INTERNATIONAL EXERGY, ENERGY AND ENVIRONMENT SYMPOSIUM

Transkript

1 8 th INTERNATIONAL EXERGY, ENERGY AND ENVIRONMENT SYMPOSIUM May 1-4, 2016 Antalya, Turkey Proceedings of the 8 th International Exergy, Energy and Environment Symposium ISBN:

2 Proceedings of the 8 th International Exergy, Energy and Environment Symposium ISBN: th INTERNATIONAL EXERGY, ENERGY AND ENVIRONMENT SYMPOSIUM May 1-4, 2016 Antalya, Turkey i

3 ORGANIZING COMMITTEE CONFERENCE CHAIR Onder KIZILKAN Suleyman Demirel University, Turkey HONORARY CHAIR T. Nejat VEZIROGLU International Association for Hydrogen Energy, USA FOUNDING CHAIR Ibrahim DINCER University of Ontario Institue of Technology, Canada ORGANIZING COMMITTEE MEMBERS Ibrahim DINCER University of Ontario Institue of Technology, Canada Ali Kemal YAKUT Suleyman Demirel University, Turkey Reşat SELBAS Suleyman Demirel University, Turkey İskender AKKURT Suleyman Demirel University, Turkey Onder KIZILKAN Suleyman Demirel University, Turkey Ahmet KABUL Suleyman Demirel University, Turkey Can Ozgur COLPAN Dokuz Eylül University, Turkey Sandro NIZETIC University of Split, Croatia Mehmet Akif EZAN Dokuz Eylül University, Turkey Melik Ziya YAKUT Suleyman Demirel University, Turkey Fatih YIĞIT Suleyman Demirel University, Turkey LOCAL ORGANIZING COMMITTEE MEMBERS Adnan OZDEN Dokuz Eylul University, Turkey Anil ERDOGAN Dokuz Eylul University, Turkey Ceren YUKSEL Dokuz Eylul University, Turkey Omer Faruk ATACAN Dokuz Eylul University, Turkey David OUELLETTE Dokuz Eylul University, Turkey Gamze YAKUT Suleyman Demirel University, Turkey Mustafa ERCELIK Dokuz Eylul University, Turkey Ozum CALLI Izmir University of Economics, Turkey Ugur GENCALP Dokuz Eylul University, Turkey ii

4 INTERNATIONAL ADVISORY COMMITTEE (In Alphabetic Order) A. Bejan, USA A. Beyene, USA A. Hepbasli, Turkey A. Kabul, Turkey A. Midilli, Turkey A. E. Ozgur, Turkey A. F. Miguel, Portugal A. H. Reis, Portugal A. K. Yakut, Turkey B. V. Reddy, Canada C. Koroneos, Greece D. Queiros-Conde, France E. Michaelides, USA E. Sciubba, Italy E. Shirani, Iran F. Aloui, France F. Hamdullahpur, Canada G. Lebon, Belgium G. Tsatsaronis, Germany G. F. Naterer, Canada H. El-Qarnia, Morocco H. Kwak, Korea H. Peerhossaini, France H. Yamaguchi, Japan H. C. Bayrakci, Turkey H. M. Sahin, Turkey I. Benko, Hungary I. Dincer, Canada I. Yildiz, Canada J. Yan, Sweden M. Feidt, France M. A. Gadalla, UAE M. A. Rosen, Canada M. B. Pate, USA O. Altuntas, Turkey O. Arnas, USA O. Kizilkan, Turkey P. Grammelis, Greece P. Lund, Finland R. El Emam, Egypt R. Selbas, Turkey S. Lorente, France S. Obara, Japan S. Nizetic, Croatia S. A. Sherif, USA T. Akiyama, Japan T. H. Karakoc, Turkey T. N. Veziroglu, USA X. Li, Canada X. R. Zhang, China V. I. Ugursal, Canada Y. Demirel, USA Y. A. Cengel, Turkey Y. Iwamoto, Japan Z. Sen, Turkey iii

5 PREFACE During the past few decades, the world has faced various critical challenges in various dimensions, ranging from energy to environment and from economy to sustainability. In all these dimensions, energy plays the most critical roles since the way that we produce, transfer, transport, convert, and use energy affects all other dimensions significantly. There is a strong need for clean energy solutions to overcome environmental, resource, efficiency, cost, energy security, and sustainability issues. When the Exergy, Energy and Environment Symposium was first launched 13 years ago in 2003 by the founding chair Dr. Ibrahim Dincer in Izmir, Turkey, this was a kind of ultimate goal.. The symposium, since then, has been running successfully under the title of International Exergy, Energy and Environment Symposium (IEEES). The conference was held in the following cities: Kos, Greece (2005); Evora, Portugal (2007); Sharjah, United Arab Emirates (2009); Luxor, Egypt (2011); Rize, Turkey (2013) and Valenciennes, France (2015). All symposium chairs, in this regard, deserve a clear recognition. Our warm thanks go to all the organizers who have contributed in the success of this conference series. This year, the 8 th International Exergy, Energy and Environment Symposium (IEEES-8) is bringing all disciplines together. IEEES-8 is a multi disciplinary international symposium, covering three main areas of exergy, energy and the environment and aims to provide a forum for researchers, scientists, engineers and practitioners from all over the world to exchange information, to present high-quality research results and new developments in the wide domain covered by exergy, energy, and the environment, and discuss the future direction and priorities in the field. The primary theme of the conference is exergy, energy and environment, not only in engineering and science but also in all other disciplines (e.g., ecology, education, social sciences, economics, management, political sciences, and information technology). Therefore, papers on related topics were solicited from all relevant disciplinary areas, ranging from current problems, projections, new concepts, modeling, experiments and measurements, to simulations. The IEEES-8 has received extraordinary international attention from every corner of the world. Here are some summary figures to share with you: Number of abstracts received: 319 Number of papers received: 169 Number of presentation scheduled for the program: 267 (with 139 oral and 128 poster presentations) IEEES-8 includes plenary sessions, keynote lectures, and several oral and poster sessions on different topics related to exergy, energy and environment. The conference aims to offer a distinctive platform for an effective and fruitful communication between the research, government and industrial communities. We iv

6 hope that everyone will find IEEES-8 both enjoyable and technically enlightening. We are sure everyone will also enjoy Antalya, which is one of the most beautiful cities in Turkey. As we are all aware, the efforts required in organizing and holding this kind of symposium are extensive. First, I would like to take this opportunity to express my sincere appreciation to Dr. T. Nejat Veziroglu, who is the Honorary Chair of the symposium. Second, I would like to express my warmest thanks to Dr. Ibrahim Dincer as the Founding Chair of the symposium. Third, my special thanks go to all the organizing committee members for their exemplary efforts. Last, but not least, I acknowledge my gratitude to the IEEES-8 keynote speakers, authors, session chairpersons and attendees, whose contributions and efforts have made us to come up with this stellar program. Dr. Onder Kizilkan IEEES-8 Symposium Chair v

7 TABLE OF CONTENTS ORGANIZING COMMITTEE... ii INTERNATIONAL ADVISORY COMMITTEE... iii PREFACE... iv TABLE OF CONTENTS... vi ENERGY, ENTROPY AND EXERGY ANALYSIS AND MANAGEMENT... 1 A Study on Exergetic Performance of Afsin Lignite Stoichiometric Combustion Process, Sefa Yalcin, Alp Er S. Konukman, Adnan Midilli... 2 Page Exergy Analysis of Nitrogen Liquefaction Process, Arif Karabuga, Resat Selbas, Ahmet Kabul Investigation of Irreversibility with CO 2 Emission Measurement in Industrial Enamel Furnace, Sedat Vatandas, Atakan Avci, M. Ziya Sogut Advanced Exergy Analysis of an Application of Waste Heat Powered Ejector Refrigeration System to Rotary Kiln, Abid Ustaoglu, Mustafa Alptekin, Mehmet Emin Akay, Resat Selbas Thermodynamic Evaluation of Absorption-Compression Cascade Refrigeration Cycles for Advanced Exergy Analysis, Mustafa Alptekin, Abid Ustaoglu, Mehmet Emin Akay, Resat Selbas Exergy Optimization of the Hybrid Compression-Absorption Industrial Refrigeration Systems, Mahmoud Afshar, Hamid Rad Energy and Exergy Analysis of a Steam Power Plant Considering Effect of Varying Plant Loads, Mehmet Bilgili, Mehmet Tontu, Besir Sahin Long Term Energy Demand and Supply Projections and Evaluations for Turkey, Esra Ozdemir, Muhsin Kilic Evaluating Exergetic Sustainability Indicators for an Electrolyte Supported SOFC Stack, Adnan Midilli, Ugur Akbulut Life Cycle Assessment of Nuclear Based Ammonia Production Options: A Comparative Study, Yusuf Bicer, Ibrahim Dincer Energy and Exergy Efficiency Evaluations of R134a Clathrates with Additives for Cooling Applications, Sayem Zafar, Ibrahim Dincer, Mohamed Gadalla Thermodynamic Performance Analysis of a Raw Mill System in Cement Plant, Mehmet Altinkaynak, Murat Ozturk, Ali Kemal Yakut THERMAL SYSTEMS AND APPLICATIONS Exergetic Assessment of PTSC Integrated Power-Refrigeration System Working with CO 2, Ahmet Kabul, Onder Kizilkan Cooling of Concentrated Photovoltaic System Using Microchannel Heat Sink, Ali Radwan, Mahmoud Ahmed, Shinichi Ookawara Thermodynamic Analysis of Parabolic Solar Collector Driven Double-Effect Absorption Cooling System, Fatih Yigit, Ahmet Kabul, Onder Kizilkan Energy and Exergy Analyses of a Biomass Fired Regenerative ORC System, Ozum Calli, Can Ozgur Colpan, Huseyin Gunerhan vi

8 Transient Analysis of an Absorption Solar Refrigerator with External and Internal Irreversibilities, Yasmina Boukhchana, Ali Fellah, Ammar Ben Brahim A Study on Adsorption Characteristics of Activated Carbon-R134a and Activated Carbon- R404a Pairs, Muhsin Kilic, Ersan Gonul Performance Investigation of a Geothermal Powered Organic Rankine Cycle for Natural Working Fluids, Mustafa Alptekin, Onder Kizilkan, Ahmet Kabul, Resat Selbas An Experimental Investigation on Exergy Analysis of an Ejector Expansion Refrigeration System, Nagihan Bilir Sag, Halil Kursad Ersoy, Arif Hepbasli Thermodynamic Assessment of Ozone Friendly Cascade Refrigeration System Using Natural Refrigerants, H. Cenk Bayrakci, Onder Kizilkan, Ahmet Kabul, Selin Cekin Thermodynamic Analysis of an Integrated System with A Concentrating Collector for Multi-Generation Purposes, Yunus Emre Yuksel, Murat Ozturk Heat Recovery Analysis of a Rotary Kiln in Cement Industry, Ahmet Yakup Cumbul, Mehmet Akif Ezan, Ismail Hakki Tavman, Arif Hepbasli, M. Ziya Sogut SOLAR ENERGY Experimental Analysis of Latent Thermal Energy Storage for Solar Heating Applications: Preliminary Results, Onder Kizilkan, Ahmet Kabul, Sefika Yildirim, Gamze Yildirim A review of Solar Energy Status in Iraq and Current Status, Ahmed Emad, Ahmet Kabul, Onder Kizilkan Effect of Solar - Geothermal Heat Exchanger Design and Fluid Type on the Thermodynamic Performance of a Power Plant, Anil Erdogan, Duygu Melek Cakici, Can Ozgur Colpan Thermal Regulation Enhancement of Concentrated Photovoltaic Systems Using Phase- Change Materials, Mohamed Emam, Mahmoud Ahmed, Shinichi Ookawara Solar Radiation Exergy and Enviroeconomic Analysis for the West Black Sea Region in Turkey, Yusuf Kurtgoz, Emrah Deniz Comparison of Regression Analysis, ANN and ANFIS Methods in the Prediction of Monthly Mean Global Solar Radiation: A Case Study, Yusuf Kurtgoz, Emrah Deniz Reduction of Entropy Production by Using of Solar Cooling Systems Based on SOLITERM Parabolic Trough Collectors Combined with Double Effect Absorption Chillers, Ahmet Lokurlu Optimal Off-Design Conditions for Solar-Driven Organic Rankine Cycles, Caglan Sevinc, Eray Uzgoren One-Dimensional Transient Thermal Model for Parabolic Trough Collectors Using Closed- Form Solution of Fluid Flow, Eray Uzgoren Design and Performance Analysis of Linear Fresnel Reflector, Melik Ziya Yakut, Arif Karabuga, Ahmet Kabul, Resat Selbas Exergy Analysis of a Solar Photovoltaic Panel within Karabük Climate Conditions, Mutlucan Bayat, Mehmet Ozalp A Comparative and Experimental Study on Different Exergetic Efficiency Methods of a Solar Panel, Mehmet Ozalp, Mutlucan Bayat Solar Assisted Multi-Generation System Using Nanofluids: A Comparative Analyzes, Muhammad Abid, Tahir A. H. Ratlamwala, Ugur Atikol New Climate Zone Definitions of Turkey by Using Typical Meteorological Year Data, Serpil Yilmaz, İsmail Ekmekci vii

9 Exergetic and Energetic Performance Evaluation of a Flat Plate Solar Collector in Dynamic Behavior, Hamed Mouna, Ben Brahim Ammar Optimization of Tilt Angles of PV Arrays for Different Seasons, Ahmet Senpınar Key Factors for the Operation of a Solar Air Collector: A Parametric Study, Ahmet Caglar, Mustafa Burak Bahadir Experimental Analysis of Solar Space Heater Performance, Guvenc Umur Alpaydin, Serhan Kucuka SUSTAINABLE AND RENEWABLE ENERGY DEVELOPMENT Load Side Management in Smart Grids using a Global Optimizer, Abdelmadjid Recioui, Mossaab Djehaiche, Abderrahim Boumezrag Performance Assessment of Various Greenhouse Heating Systems; a Case Study in Antalya, M. Tolga Balta, Fatih Yilmaz, Resat Selbas Cost Risk Modeling for Hybrid Power Generation from Geothermal, Biomass Resources and CSP in Turkey - Southeastern Anatolia and Eastern Anatolia Region, Yildirim Ismail Tosun Sustainable Re Use of Dairy Cow Manure as Bedding and Compost: Nutrients, Pathogens and Self-Heating Potential from Increased Residence Time in a Tumbling Composter, Joe Ackerman, Ehsan Khafipour, Nazim Cicek Cost Modeling for Thermal Energy Storage in Hybrid Power Generation from CSP and Biomass Resources in Turkey - Southeastern Anatolia and Eastern Anatolia Region, Yildirim Ismail Tosun Vacuum Stripping Membrane Desalination for Marmara Sea-Water, Filiz Ugur Nigiz, Nilufer Durmaz Architecture in the Net Zero Houses of the Future, Okay Gonulol, Ayca Tokuc Control System for a Novel Photobioreactor in the Building Envelope, Gulden Kokturk, Ayca Tokuc, Anil Unal Energy and Exergy Analysis of a Solar Air Heater Having Transverse Wedge Shaped Rib Roughness, Cihan Yildirim, Ismail Solmus Performance Analysis of Three Soft Computing Methods for Predicting the Heat Load of Buildings, Cihan Turhan, Tugce Kazanasmaz, Gulden Gokcen Akkurt Ventilation Strategies for the Preventive Conservation of Manuscripts in Necip Paşa Library, İzmir-Turkey, Turgay Coskun, Cem Dogan Sahin, Ozcan Gulhan, Zeynep Durmus Arsan, Gulden Gokcen Akkurt Investigation of Thermodynamic and Environmental Performance Based on Subcooling of Refrigerants in Direct Expansion System for Supermarket Applications, Onder Altuntas, M. Ziya Sogut, Enver Yalcin, T. Hikmet Karakoc Techno-Economic Assessment of Solar-Geothermal Based Multigeneration System for a Community, Farrukh Khalid, Ibrahim Dincer, Marc A. Rosen PV Array Based Smart Home Automation System, Ahmet Senpinar Energy and Exergy Analyses of a Solar Energy Driven Multigeneration System for Green Buildings, Yunus Emre Yuksel, Murat Ozturk, Ibrahim Dincer Achieving Sustainable Buildings via Energy Efficiency Retrofit: A Case Study of an Industrial Building, Muhsin Kilic, Ayse Fidan Altun Passive Thermal Management of a Photovoltaic Panel: Influence of Fin Arrangements, Ceren Yuksel, Cem Kalkan, Mustafa Aydin, Güven Nergiz, Mehmet Akif Ezan viii

10 Multi-Criteria Selection Factors for Evaluation of Intelligent Buildings; A Novel Approach for Energy Management, Elnaz Asadian, Katayoun Taghizadeh Azari, Ali Vakili Ardebili, Samira Mahmoodkelayeh HYDROGEN GENERATION, STORAGE AND TECHNOLOGY Effect of Nitrogen Doping on Hydrogen Storage of Graphene-TiO 2 Nanocomposites, Zahra Gohari Bajestani, Alp Yurum, Omid Akhlaghi, Yuda Yurum Performance Assessment of Solar-Based Hydrogen Production via H 2SO 4 Cycle, Fatih Yilmaz, M. Tolga Balta, Resat Selbas Exergoeconomic and Optimization of a Solar Based Integrated Energy System for Hydrogen Production, Shoaib Khanmohammadi, Parisa Heidarnejad, Nader Javani, Hadi Ganjehsarabi Exergy Analysis and Optimization of a Solid Oxide Electrolysis Cell for Hydrogen Production, Abdullah A. AlZahrani, Ibrahim Dincer Energy and Exergy Analyses of a Solar Based Hydrogen Production and Compression System, Hasan Ozcan, Ibrahim Dincer Energy and Exergy Analysis of a Novel Combined System Producing Power, Water and Hydrogen, Kiyan Parham, Hamed Alimoradiyan, Mohsen Assadi Energy and Exergy Analyses of a Solar, Wind and Geothermal Based Integrated System for Hydrogen Production, Abbas Alpaslan Kocer, Murat Ozturk Exergy Based Environmental Effect of PEM Electrolyser Integrated Hydrogen Gas Storage System, Mert Ozsaban, Selcuk Inac, Adnan Midilli FUEL CELLS Studying the Effect of Electrolyte Thickness on Exergetic Performance for an Electrolyte Supported SOFC Stack, Ugur Akbulut, Adnan Midilli, Ibrahim Dincer Multiphase Non-Isothermal Modeling of a Flowing Electrolyte - Direct Methanol Fuel Cell, Faruk Atacan, David Ouellette, Can Ozgur Colpan Multi-Inlet- Multi-Outlet Anode Flow Field Design for Micro Direct Methanol Fuel Cells, Radwan M. El-Zoheiry, Mahmoud Ahmed, Shinichi Ookawara Synthesizing and Testing of Nafion/SiO 2 and Nafion/TiO 2 Composite Membranes for the DMFC Applications, Mustafa Ercelik, Adnan Ozden, Yilser Devrim, Can Ozgur Colpan Evaluation of Design and Performance of Two Different Power Systems for a Small UAV, Mohamed Gadalla, Sayem Zafar Effect of Cathode Flow Field Configuration on the Performance of Flowing Electrolyte- Direct Methanol Fuel Cell, Ugur Gencalp, David Ouellette, Can Ozgur Colpan The Effects of Three Different Coating Techniques on the Performance of DMFCs, Adnan Ozden, Mustafa Ercelik, Yagmur Nalbant, Hasan Kiyik, Can Ozgur Colpan The Effects of Bio-Inspired Flow Field Design on the Performance of DMFCs, David Ouellette, Adnan Ozden, Mustafa Ercelik, Can Ozgur Colpan FLUID MECHANICS, HEAT AND MASS TRANSFER Diesel-Like Fuel from Waste Engine Oil by Thermo-Catalytic Pyrolysis, Tarabet Lyes, Maamouri Mohamed, Zouad Youcef, Khiari Karim, Mahmoud Rachid, Mohand Tazerout 558 MHD Natural Convection and Entropy Generation in a Nanofluid Filled Cavity with a Conductive Partition, Fatih Selimefendigil, Hakan F. Oztop ix

11 CFD Simulations to Optimize Flow Distribution in a FGD Wet Scrubber, Osman Gozutok, Murat Baranak, Goktug N. Ozyonum, Asli I. Kaya Determination of Flow Characteristics of Multiple Slot-Jets Impingement Cooling, Nuri Kayansanayan, Ersin Alptekin, Caner Erdogan Experimental and Numerical Investigations of Heat Transfer in Multi-Port Tubes, Kemal Ermis, H. Ibrahim Coban, Mehmet Coban Effect of Length of the Wavy Shaped Splitter Plate on Flow around a Circular Cylinder, Mustafa Sarioglu, Mehmet Seyhan, Yahya Erkan Akansu Investigation of the Effect of the Plasma Actuators Vertically Placed On the Spanwise of NACA0015 Airfoil, Hurrem Akbiyik, Hakan Yavuz, Yahya Erkan Akansu Second Law Analysis of Coupled Heat and Mass Transfer through Combined Non Gray Gas Radiation within a Cylindrical Annulus, Sakly Abir, Mazgar Akram, Ben Nejma Faycal A Numerical Study on Phase Change inside a Spherical Capsule, Ersin Alptekin, Muhammet Ozer, Murat Top, Nuriye Bozkurt, Muruvvet Zenginoglu, Fazil Erinc Yavuz, Mehmet Akif Ezan Phase Change Materials in Textile Fabrics: A Numerical Survey, Mehmet Akif Ezan, Berkant Murat Gul, Hüseyin Kurt, Atif Canberk Ezan, Ersin Alptekin Experimental Investigation of a Panel Radiator with Latent Heat Storage, Guvenc Umur Alpaydin, Serhan Kucuka Energy and Exergy Analyses of a Hybrid Solar-Geothermal Power Plant, Duygu Melek Cakici, Anil Erdogan, Can Ozgur Colpan FUELS AND COMBUSTION TECHNOLOGY MW Hybrid Power Generation in ORC Unit from Co-Incineration of Agricultural, Forestry Biomass Waste and Biogas in Stoker and Through Parabolic Solar Panel (CSP), Yildirim Ismail Tosun Optimization Methods of Radiative Transfer Calculation Applied to a Cylindrical Sodium Vapor Plasma, Soumaya Hadj Salah Waste to Energy with a Combine Membrane Technology: Biobutanol Production and Purification, Filiz Ugur Nigiz, Nilufer Durmaz Hilmioglu Biodiesel Production from High Acid Value Sunflower Oil By Using Zirconium Sulfate as a Heterogeneous Acid Catalyst, Melike Imge Senoymak, Oguzhan Ilgen Biodiesel Production over CaMgAl Hydrotalcite like Compounds from Waste Cooking Oil, Emine Emel Cakirca, Ayse Nilgün Akın Production of Upgraded Bio-Oils by Biomass Catalytic Pyrolysis Using Low Cost Food Industry Waste, Nurgul Ozbay, Adife Seyda Yargic, Rahmiye Zerrin Yarbay Sahin, Elif Yaman Numerical Investigation of Fixed Bed Downdraft Woody Biomass Gasification, Ebubekir Siddik Aydin, Ozgun Yucel, Hasan Sadikoglu Influence of Boron Loading Sequence on HDS Catalyst Activity, Yesim Dusova-Teke, Esra Yonel-Gumruk, Orhan Ozcan, M. Efgan Kibar, A. Nilgun Akin Production of a Low-Sulfur Oil from Scrap Tires Pyrolysis Using a Two-Stage Pyrolysis Process and Additives, Gyung-Goo Choi, Young-Kon Choi, Joo-Sik Kim Utilization of Kayseri-Menteş Iron Ore as Oxygen Carrier in Chemical Looping Combustion of Syngas: Deconvolution of the Gas Analysis Data, Nesibe Dilmac, Omer Faruk Dilmac x

12 Pyrolysis of Waste Polyethylene Plastics and Investigation of the Fuel Potential of Pyrolysis Products, Merve Sogancioglu, Esra Yel, Gulnare Ahmetli Pyrolysis of Washed Waste HDPE Plastics and Production of Epoxy Composite from the Pyrolysis Char, Merve Sogancioglu, Esra Yel, Gulnare Ahmetli Production of Hazelnut Oil Biodiesel through Investigating KOH-Catalyzed Transesterification Reaction Parameters, Mert Gulum, Atilla Bilgin Second Law Analysis of a CI Engine Fueled With Biodiesel-Diesel Blends, Abdulvahap Cakmak, Atilla Bilgin Coupled Diesel Engine Hazardous Emissions Fixation, and Microalgae Biomass Production Enhancement, D. O. Correa, B. Santos, J.V.C. Vargas, A.B. Mariano, W. Balmant, M.P. Rosa, D.C. Savi, V. Kava, J.C. Ordonez Regression Models for Predicting Some Important Fuel Properties of Corn Oil Biodiesel- Diesel Fuel Blends, Atilla Bilgin, Mert Gulum Rational and Hyperbolic Models to Estimate Kinematic Viscosities of Hazelnut Oil Biodiesel-Diesel Fuel Blends, Mert Gulum, Atilla Bilgin Experimental Investigation of the Effects of Water Adding to the Intake Air on Diesel Engine Performance and Heat Release Analysis, Zehra Sahin, Orhan Durgun, Mustafa Tuti Emission due to Pollution from Ships Main Engine and Auxiliary Machinery, Munir Suner POSTER PRESENTATIONS Heat and Mass Transfer in a Composite Fluid-Porous Layer, Noureddine Hadidi, Ziane Farouk, Rachid Bennacer, Yacine Ould-amer Fuzzy Control of the Compression System by the Throttle and Coupled Valves in Petroleum Companies, Razika Zamoum Boushaki, Tarik Boushaki, Farida Kessal Experimental Study and Energy Optimization of a Solar Domestic Refrigerator Incorporating a Phase Change Materials, Tetbirt Ali, Mokrane Mehdi, Abbas Mohammed, Berdja Mohand, Ferhat Yahi Removal of Ions Pb 2+ and Cd 2+ from Aqueous Solution by Containment Geomaterials, Souhila Ait Hamoudi, Boualem Hamdi, Jocelyne Brendle The Dissolution Behavior of Lead Oxide in Aqueous Organic Acid Solutions, S. Bendebane, S. Djerad, L. Tifouti Effects of Parameters on the Extraction Yield of Acid Orange10 by Elm from an Aqueous Solution: Application of Plackett-Burman Design, Farida Bendebane, Lynda Bahloul, Hazem Meradi, Mohammed Saddek Lachgar, Abbes Boukhari, Fadhel Ismail Effect of Biopolymer on the Properties of Oil-In-Water Microemulsions, Nedjhioui Mohammed, Moulai Mostefa Nadji, Tir Mohamed Taguchi Optimization Approach for Methyl Orange Removal from Aqueous Solution Using Electrochemical Process, Mohamed Tir, Mohamed Nedjhioui Multilayer Perceptron Model for Predicting Acute Toxicity of Fungicides on Rats: Validation and Domain of Application, Hamadache Mabrouk, Benkortbi Othmane, Hanini Salah, Amrane Abdeltif Thermoeconomic and Enviroeconomic analysis of ISCCS in Algeria, Tarik Boushaki, Pr. Kacem Mansouri Effect of Pb Content and Heat Treatment on Thermoelectric Properties of AgPb 18+xSbTe 20 alloys, Sheng-Long Lee, Jo-Kuang Nieh, Yu-Chih Tzeng xi

13 Thermodynamic and Kinetic Studies for the Adsorption of Amoxicillin onto Modified Wheat Grains, Othmane Benkortbi, Asmaa Boukhelkhal, Mabrouk Hamadache, Salah Hanini A Novel Technique for the Production of Fuel Bioadditive Ethyl Levulinate: Green Process by the Catalytic Membrane, Derya Unlu, Nilufer Hilmioglu Biodiesel Synthesis by Using the Smart Catalytic Membrane, Derya Unlu, Aynur Hacioglu, Nilufer Hilmioglu Characterization of Bio-Oil Obtained from a Food Industry Waste Pyrolysis, Nurgul Ozbay, Elif Yaman, Adife Seyda Yargic, Rahmiye Zerrin Yarbay Sahin Synthesis Gas Production from Tri-Reforming and Partial Oxidation of Simulated Biogas over Ni/ZrO 2-MgO-Al 2O 3, Merve Dogan, Emel Engintepe, Orhan Ozcan, Murat Efgan Kibar, Ayse Nilgun Akin Partial Oxidation of Biogas for Hydrogen Production over Ce-Promoted Ni/Mgal Hydrotalcite-Like Catalyst, Emel Engintepe, Merve Dogan, Orhan Ozcan, Murat Efgan Kibar, Ayse Nilgun Akin Desalination in Algeria, Case of Skikda Seawater Desalination Plant, Mounira Rouainia, Karima Mehri Performance Enhancement of Ni-based Oxygen Carrier by Adding Other Oxygen Carrier, Ho-Jung Ryu, Dong-Ho Lee, Chang-Keun Yi, Sung-Ho Jo, Jeom-In Baek Continuous Operation Results of 263 kwth Chemical Looping Combustor, Ho-Jung Ryu, Dong-Ho Lee, Gyoung-Tae Jin, Seung-Yong Lee, Jeom-In Baek Rehabilitation Alternatives for Flue Gas Desulfurization Units, Asli Isler Kaya, Fatih Aydin, Mustafa Malkoc, Savas Altinisik, Omer Orcun Er Effect of Zeolite Supported Iron Catalyst on Upgrading of Pyrolysis Bio-Oil, Elif Saracoglu, Esin Apaydin-Varol, Basak Burcu Uzun Air Gasification of Dried Sewage Sludge Using a Multi-Stage Gasifier: Effects of the Equivalence Ratio and Activated Carbon on Tar Removal, Young-Kon Choi, Gyung-Goo Choi, Joo-Sik Kim Life Cycle Assessment of a Maintenance Process for a Training Aircraft, Yasin Sohret, Selcuk Ekici, Onder Altuntas, T. Hikmet Karakoc Development and Multi Objective Exergy Based Optimization of a Solar Micro CHP System Based on Organic Rankine Cycle for domestic applications, Alireza Noorpoor, Parisa Heidarnejad, Shoaib Khanmohammadi, Nader Javani Evaluation of Bio oils produced from Pomegranate Pulp Catalytic Pyrolysis, Eylem Pehlivan, Nurgul Ozbay Effect of Air Exchange Rate on the Economic Outputs of Aircraft Environmental Control Systems, Ramazan Atilgan, M. Ziya Sogut, Onder Turan Interactions between Polysaccharide and Anionic Surfactant and Their Effects on the Interfacial and Rheological Behaviours, Nedjhioui Mohammed, Moulai Mostefa Nadji, Tir Mohamed, Skender Abdelhak Numerical Study of Latent Heat Thermal Energy Storage inside a Porous Matrix, Mehdi Fetiti, Amel Alidrous CD 4 effective Hamiltonian in order 6 for the Pentad (2ν 4, ν 2+ν 4, ν 1, 2ν 2, ν 3). A simultaneous Line Position analysis of GS-GS, Dyad (ν 2, ν 4)-Dyad, Dyad-GS and Pentad- GS, Ouardi Okkacha, Kaarour Abdelkrim Study of the ν 3 Fundamental Band of 12 CD 4, Kaarour Abdelkrim, Ouardi Okkacha xii

14 Exergy Analysis of Benzene Production Cycle, Masoud Taghavi, Gholamreza Salehi, Rasoul Hajibabaei xiii

15 ENERGY, ENTROPY AND EXERGY ANALYSIS AND MANAGEMENT 1

16 A Study on Exergetic Performance of Afşin Lignite Stoichiometric Combustion Process Sefa Yalcin 1*, Alp Er Ş. Konukman 2, Adnan Midilli 3 1 Energy Institute, TUBITAK Marmara Research Center, Gebze 41470, Kocaeli, Turkey 2 Gebze Technical University, Department of Mechanical Engineering, Gebze 41400, Kocaeli, Turkey 3 Recep Tayyip Erdoğan University, Department of Mechanical Engineering, Rize, Turkey * sefa.yalcin@tubitak.gov.tr, konukman@gtu.edu.tr, adnan.midilli@erdogan.edu.tr Abstract This study aims to investigate parametrically the energetic and exergetic performance of Afşin lignite combustion process. In this regard, in terms of the First law and the Second law of thermodynamics, the energy and exergy analyses have been achieved by using the complete combustion (air fuel ratio, λ=1) reaction and the proximate and ultimate analyses results of Afşin lignite samples taken from Afşin basin that is the largest lignite basin of Turkey. For these purposes, a schematic model system to produce process steam (151 o C, 5 bar) and hot water (90 o C) for practical applications has been developed, which consists of combustion chamber, steam production heat exchanger, hot utilization water production heat exchanger, recuperator for combustion air heater and chimney. The combustion chamber exit gas temperatures are taken to be 1000, 1200 and1300 K, respectively. Meanwhile, Afşin lignite adiabatic flame temperature has been calculated to be 1327 K. However, maximum reaction temperature is taken to be 1300 K for the analyses. Accordingly, it is determined that, in order to increase the energetic and exergetic performances of Afşin lignite combustion process, the heat losses resulting from the convectional, conductional and radiational irreversibilities should be minimized by adducting the practical flame temperature to the adiabatic flame temperature of Afşin lignite and increasing the useful energy from the process. Keywords: Afşin lignite, energy, exergy, efficiency, combustion. I. Introduction Afşin-Elbistan lignite basin constitutes 38% of the total lignite reserves with 4,8 billion tonnes reserves in Turkey (EÜAŞ Yıllık Rapor, 2014). Afşin lignite basin that is the largest lignite basin in Turkey has coals whose caloric values are between 900 and 1250 kcal/kg (World Energy Council Energy Report 2013). Therefore, it is inevitable to do performance improvements for lignite combustion system. In order to determine the performance of the lignite combustion system thermodynamic analysis should be done. The first law of thermodynamics does not give any information about irreversibilites and degredations, occurring in the system. Therefore, energy analysis of industrial systems is needed to be performed as well as exergy analysis (Ohijeagbon IO, 2013). Exergy analysis is a result of the second law of thermodynamics and it is a method used to determine the useful work potential of given amount of energy at some special cases. Exergy analysis is commonly used in thermal and thermo-chemical system design, simulation and performance assessment (Saidur, 2010). efficient energy system design and energy resources use on environment. In Turkey, steam boilers are used in many sectors of industry. Textiles, paper, sugar, tea and many other industries, for use in production processes, boilers for steam generation is used. Most of steam boilers used in industry utilize fossil fuels. According to the percentage use of fossil fuels for steam generation sector; food procesesing (57%), paper processing (%81), chemical production (42%), oil refining (23%) and primary metal production (10%) is realized (Saidur R, 2010). The largest share in the food processing sector, sugar and tea processing plants are located. Tea processing plants need a lot of process steam during wet tea processing. ÇAYKUR produced 65% of the dry tea in Turkey, also private sector produces 35%. ÇAYKUR has 46 fresh tea processing factory also private sector has 230. Production capacity of Çaykur is 6760 ton/day, also private sector s production capacity is 8746 ton/day (Turkey Black Tea Sector Report, 2009). According to Dincer et al. (2003) exergy analysis is an important tool to determine the impact of more 2

17 II. Main Considerations II.I. System Design and Operating Principle In this work, a facility whose daily fresh tea production capacity is 280 t/h is considered, which has five steam boilers. The desired properties of steam to be obtained from the system are given in Table 1 (Korkmaz, 2012). Tab. 1: Properties of steam needed tea processing plant Steam pressure 5 bar Steam Temperature C Steam amount 11 t/h A schematic model system to produce process steam (151 o C, 5 bar) and hot water (90 o C) for practical applications has been developed, which consists of combustion chamber, steam production heat exchanger (HeX-1), hot utilization water production heat exchanger (HeX-2), recuperator for combustion air heater (R-1) and chimney (Fig. 1). Coal properties used in the study were obtained by doing proximate and elemental analysis of samples taken from the Afşin basin. Proximate and elemental analysis of coal are conducted in accordance with ASTM D 1372 and ASTM D 3176 standards, respectively. Coal properties are presented in Table 2. Tab. 2: Proximate and elemental analysis of Afşin lignite Sym Value Unit c 15,05 %w h 1,48 %w o 4,80 %w n 0,37 %w s 1,98 %w w 46,43 %w a 29,88 %w LHV 919 kcal/kg HHV 1250 kcal/kg (c; carbon content, h; hydrogen content, o; oxygen content, n; nitrogen content, s; sulphur content, w; moisture content, a; ash content) Fuel W F1 Q 1 Q Fresh Air 25 C, 1 atm Air Preheater Hot Air 300 C, 1 atm Fuel 25 C Combustion Chamber CC Flue gas T stack, 1 atm Flue gas Tstack, 1atm Ash Stack W F3 Flue gas 150 C, 1 atm R - 1 Hot Air 300 C, 1 atm WF2 Flue gas T HeX 2,out Utilization Water 90 C HeX - 2 Feed Water-2 20 C HeX - 1 Flue gas T stack,hex-1,out Feed Water-1 20 C Process Steam 151 C Fresh Air 25 C, 1atm Fig. 1: A schematic representation of the coal-fired steam generating plan 3

18 Closed formula for Afşin lignite is determined at equation (6) by using values obtained from the results of the analysis (El-Wakil,1984, Bejan et al., 1996). n C = %c, n M H = %h n C M O = %o H M O n S = %s n M N = %n (1) S M N n top = n C + n H + n O + n N + n S (2) n C = n C n H = n H n n O = n O top n (3) top n top n N = n N n n S = n S top n top C nc H nh O no N nn S NS + γo min (O 2 + 3,762N 2 ) aco 2 + bh 2 O + cso 2 + dn 2 (4) where, γ; excess air coefficient (for stoichiometric combustion is taken 1). III.1. Energy Analysis In this work, energy analyses of each equipment individually and the whole system are performed according to the 1 st law of thermodynamics. a) Combustion Chamber Energy Analysis Energy production in coal fired steam and hot water producing system is provided by the transformation of chemical energy of coal burned with atmospheric air into heat energy. Combustion air with energy m m E air and fuel with energy E fuel enter into the combustion chamber and they perform combustion reaction and then they leave the system with the m energy E flue gas. Heat transfer occurs by the amount Q of E radiation and E ash m from the combustion chamber to the outside (Fig. 2.). E Q Radiation O min = n C + n H 4 + n S n O 2 (5) E m fuel C 0,4025 H 0,4728 O 0,0964 N 0,0085 S 0, ,4924(O 2 + 3,762N 2 ) 0,4025CO 2 + 0,2364H 2 O + 0,01985SO 2 + 1,857N 2 (6) E m air Control Volume CV E m fg The fuel fed into the system is burned with stoichiometric conditions in the combustion chamber. Process steam is obtained by transferring the the energy of the resulting combustion gas to the feed water at at 20 C in the HeX-1-exchanger. The flue gas releasing the some fraction of its energy in the HeX-1-exchanger produces the proses water at 90 C with its a portion of energy by entering the HeX- 2 that is used for production of the hot water used in the system. Then the flue gas releasing the some fraction of its energy in the HeX-2-exchanger is fed to R1 recuperators where the combustion air temperature is increased necessary for the combustion of the fuel. Finally, the flue gas that increases the entering fresh air temperature from 25 C to 300 C in the recuperator is discharged to the atmosphere from the chimney at 150 C due to the fact that condensation occurs at below temperatures. III.Analysis Assumptions The assumptions made in this energy and exergy analyses of the system; 1. The system is constantly open continuous flow 2. Heat transfers from exchanger to the environment are neglected. 3. The fans used in the system are chosen to be 1 kw. 4. Subsystems and processes are considered steady state. 5. Combustion air consist of 21 % O2 and 79 % N2 and it is assumed as an ideal gas. E m ash Fig. 2: Combustion chamber energy flow diagram Fuel and air enter into the combustion chamber as the mass. Outlet mass of the combustion chamber consists of the flue gas occurring after combustion and the unburned ash. In this way, For a general steady-state process, mass and energy balances, respectively, can be written as: m in = m out (7) m in = m coal + m air (8) m out = m fluegas + m ash (9) E in E out = 0 (10) E in = E in m + E in W Q + E in (11) E in m = m fuel LHV fuel + m air h air,573 K (12) E in W = 0 (13) E in Q = 0 (14) Mass energy input stem from only air and fuel into the combustion chamber. Any work or heat transfer is not an issue. Similarly, the energy outputs from the system are to be written as follows. 4

19 E out = E out m E out m + E out W Q + E out (15) flue = E gas ash out + E out (16) flue E out gas flue = m flue gas (h gas T CC ) (17) E out ash = ignore (18) E out W = 0 (19) E out Q Q = E Radiation (20) Energy efficiency of the combustion chamber is defined as the ratio of the output useful energy value to input energy value at equation 21. ƞ 1,CC = E flue gas out E in (21) b) HeX 1 Heat Exhanger Energy Analysis The process steam, the main useful output, is obtained by transferring the flue gas energy to the feed water in the HeX-1 heat exchanger. Hot flue gas in with E fg(t) energy entering into the heat exchanger leaves its energy to the feed water entering into the in heat exchanger with E fw(293 K) energy and then out leaves the system with E fg(t HeX1 out) energy. Cold feed water enters into the system leaves the system out as process steam with the energy E ps(424 K) by changing phase (Fig.3). Energy input to the HeX-1 heat exchanger takes place with the hot flue gas from the combustion chamber. Apart from this, there is not any work or interaction. Energy output from the HeX-1 is provided with the flue gas that have transferred a portion of its energy to the feed water for steam production and the generated process steam. The energy balance equation for the HeX-1; E in E out = 0 (24) m fg h fg(t) m fg h fg(t HeX1 out) = m ps h ps(424 K) m fw h fw(293 K) (25) Q HeX1 = m fg (h fg,1000k h fghex1 out ) (26) h fghex1 out = (Q HeX2 m fg) + h fghex2 out (27) m steam = Q HeX1 (h ps(424 K) h fw(363 K) ) (28) Efficiency expression for HeX-1 is defined in the equation 29 as the ratio of the total useful energy obtained from the heat exchanger and the total energy entering the system. ƞ 1,HeX1 = Q HeX1 E in (29) The process steam mass flow rate as defined in equation 28, is calculated as the main useful output obtained from the system. c) HeX 2 Heat Exhanger Energy Analysis E in fg (T) HeX - 1 E out fg (T-HeX1-out) CV E in fw (20 C) E out ps (151 C) Hot water to be used at the factory is also obtained from the system which is designed for steam generation. The flue gas that has leaved a portion of its energy for steam production in the HeX-1 enters to the HeX-2 heat exchanger for hot water production in with the energy E fg(t HeX1 out) and leaves the heat out exchanger with the energy E fg(t HeX2 out) by leaving a portion of its energy to the feed water that in enters to the system with the energy E fw(293 K). The cold feed water that has entered to the system leaves out the system with the energy E uw(363 K), which takes from the flue gas (Fig. 4). Fig. 3: HeX - 1 heat exhanger energy flow diagram The output hot flue gas from the combustion chamber and the feed water fed for steam production enter to the HeX-1 heat exchanger. The total mass from the system consists of prosess steam, which is obtained in the HeX-1 heat exchanger, and the flue gas. Thereby; The mass balance equation for the HeX-1; E out fg (T-HeX2-out) E out uw (90 C) HeX - 2 CV E in fw (20 C) E in fg (T-HeX1-out) m in = m out (22) m fw + m fg,in = m ps + m fg,out (23) Fig. 4: HeX-2 heat exchanger energy flow diagram 5

20 The output hot flue gas from the HeX-1 and the feed water fed for hot use water production enter to the HeX-2 heat exchanger. The total mass from the system consists of hot use water and the flue gas. Thereby; The mass balance equation for the HeX-2; m in = m out (30) m fw + m fg,(t HeX1 out) = m uw + m fg,(t HeX2 out)(31) The energy balance equation for the HeX-2; E in E out = 0 (32) Efficiency expression for HeX-2 is defined in the equation 37 as the ratio of the total useful heat obtained from the heat exchanger and the total energy entering the system. ƞ 1,HeX2 = Q hex2 E in d) R1 Recuperator Energy Analysis (37) Combustion air is fed into the combustion chamber at 573 K for the purpose of increasing the combustion efficiency in the coal combustion system. The flue gas exits from the HeX-2 which is produced hot out utilization water with the energy E fg(t HeX2 out) and enters into the R1 in order to increase the fresh air temperature to 573 K. The flue gas leaving a portion of its energy leaves from the system with the energy out E fg(t 423 K) and is thrown from the chimney into the atmosphere. In order to prevent the flue gas from condensation in the chimney, the flue gas exit temperature is kept constant at 423 K (Fig.5). E out air,300 C the system consist of hot combustion air quantity derived in the recuperator and the flue gas. Thereby; The mass balance equation for the R-1; m in = m out (38) m air,298 K + m fg,(t HeX2 out) = m air,573 K + m fg,(t 423 K) (39) The energy balance equation for the R-1; E in E out = 0 (40) m air,573 K h air,573 K m air,298 K h air,298 K = m fg h fg(t HeX2 out) m fg h fg(t 423 K) (41) Q R1 = m air (h air,573 K h air,298 K ) (42) Q R1 = m fg (h fghex2 out h fg423 K ) (43) h fghex2 out = Q R1 + h (44) m fg 423 K fg Efficiency expression for R1 is defined in the equation 45 as the ratio of the total useful heat obtained from the recuperator and the total energy entering the system ƞ 1,R1 = Q R1 E in e) Coal-Fired Steam Production System Energy Analysis (45) According to the first law of thermodynamics, energy analysis of the common system was carried out. The energy efficiency of the system is defined as the ratio of total produced useful output to the total entering energy value and is presented in the equation 46. ƞ I,Sys = Q HeX1+Q HeX2+Q R1 E in (46) III.2 Exergy Analysis E out fg (T-150 C) R1 E in air,25 C CV E in fg (T-HeX2-out) E in w,f1 Fig. 5: R1 recuperator energy flow diagram The output hot flue gas from the HeX-2 and the atmospheric air fed for combustion air production at 573 K enter to the R-1 recuperator. The total mass of Second law of thermodynamics, which describes the exergy term, defines the quality of energy as well as the quantity of energy. The final state of the system is called as dead state. This is the state that kinetic and potential energy exchanges are zero (Taner T, 2015). As the thermodynamics point of view, exergy is described as the maximum work amount that can be produced by the system at the reference ambient conditions. Exergy is not protected like energy. it is consumed due to irreversibilities or destroyed in the real prosess (Dincer I, 2003). A general exergy equation consists of different form of exergy like kinetic and potential exergy, physical and chemical exergy and exergy of radiation (Amelio A, 2016). General exergy equation consists of physical and chemical exergy at steady-state condition with negligible potential and kinetic exergy 6

21 changes for combustion process (Ohijeagbon O, 2013). Physical exergy is defined as the maximum theoretical useful work that can be obtained from a flowing stream as it is brought to the environmental state (Riverio R, 2006; Dincer I, 2003). According to Dincer and Rosen (2007), the physical exergy can be calculated as follows; E x ph = m [(h h 0 ) T 0 (s s 0 )] (47) The chemical exergy is defined as the maximum theoretical useful work that can be obtained from a flowing stream as it is brought from the environmental state to the dead state (Ohijeagbon Ol, 2013; Bilgen S, 2008; Dincer I, 2003). individually and the whole system are performed according to the 2nd law of thermodynamics. a) Combustion Chamber Exergy Analysis Combustion air enter into the combustion chamber with exergies E x ph ch air and E x air and fuel enter into ch the combustion chamber with exergy E x fuel and they perform combustion reaction and then they leave the system with the exergy E x ph fg. ph Heat transfer occurs by the amount of E x Rad from the combustion chamber to the outside as well as exergy destruction by the amount of E x D due to internal irreversibilities (Fig. 6). Chemical exergy value of pure substances is calculated via standard chemical exergy values taken from tables. Chemical exergy of gas mixtures is to be calculated by standard chemical exergy of pure substances generating the mixture as following; E x m fuel E x Q Radiation ch e x gas mix = x k e x ch k k + R T 0 k x k lnx k (48) where; xk, mole fraction of gas mixture at gas phase; e x k ch, standard chemical exergy values of gases; R, universal gas constant; T0, ambient temperature. E x m air Control Volume CV E x,d E x m fg Industrial fuels occur tens of chemical compounds that are difficult to calculate their chemical exergy. Standard chemical exergy of fuels that their chemical formulations are known can easily be found from the tables. However, some fuels are composed of multiple components and can not be read directly from the tables. Szargut and Styrylska developed a statistical method that represent the chemical exergy of industrial fuels (Midilli A, Chemical Exergy Calculations, 2014; Szargut J and Styrylska T, 1964). According to this method, φ dry colerations are expressed below according to the cases for the o/c ratio of a fuel composed of C, H, O and N: o/c 0.667; φ dry = h c o c n c (49) o/c 2.67; φ dry = h c ( h c ) n c o c (50) Standard chemical exergy value of a moist fuel is expressed as: ch e x fuel = ((NCV) fuel + w + h fg ) φ dry + (e x ch s (NCV s )) s (51) Here; w is the mass fraction of water; s is the mass fraction of sulphur and hfg is the enthalpy of vaporization of water. In this work, energy analyses of each equipment 7 E x m ash Fig. 6: Combustion chamber exergy flow diagram Exergy balance for the combustion chamber; E x in E x out E x D = E x sys (52) For steady-state condition, E x in E x out = E x D (53) Ex in = E x,in W + E Q x,in + E x,in m + E x,in,ke + E x,in,pe (54) Any work or heat transfer does not enter into the combustion chamber. Kinetic and potential energy changes are neglected. Therefore, E x,in W = 0 (55) Q E x,in = 0 (56) as written. Mass and exergy transfer into the combustion chamber takes place by fuel and combustion air. The fuel fed into the combustion chamber has chemical exergy because of being exposured to chemical reactions. Due to the fact that the fuel fed into combustion chamber at dead state, there is no physical exergy of the fuel. Combustion air fed into the combustion chamber at 573 K, so it transports physical exergy into the combustion chamber. Also combustion air reacts with fuel, so it has chemical exergy. Exergy transfer equations with mass are written as follows:

22 E x,in m = E fuel x,in + E air (57) x,in fuel = E E x,in ph E x,fuel ph E xair ch x,fuel ph + E x,fuel (58) = 0 (59) = (hair,573 K h air,298 K ) (s air,573 K s air,298 K )T 0 (60) Ex PH in,fg (T) HeX - 1 CV Ex PH in,fw1 (20 C) e x ch air = [(x O2 e x ch O2 ) + (x N2 e x ch N2 )] + R T 0 [(x O2 lnx O2 ) + (x N2 lnx N2 )] (61) Chemical exergy of combustion air is calculated with the help of values at Table 3. Tab. 3: Standard chemical exergy of air compounds at reference conditions (Bejan vd., 1996) Compounds of Air Molar Flow Rate (kmol/s) Molar Fraction (xk) Standard Molar Chemical Exergy (kj/kmol) O2 0,0153 0, N2 0,0577 0, Total 0, Exergy output from the combustion chamber takes place with the physical exergy of the flue gas and the exergy of the heat transferred by radiation and convection from the combustion chamber surfaces and discharged ash from the system. In this case, the physical exergy value carried by the flue gas which is the useful output obtained from the combustion chamber was calculated according to the equation 63. ph E x,out = E x,flue gas ph E x,flue gas Q rad + E + E x,out Q conv x,out +E x,out ash (62) = m flue gas (h T fg h 0 fg ) T 0 (s T fg s 0 fg ) (63) Exergy efficiency expression for the combustion chamber is defined as the ratio of the physical exergy value of the flue gas, which is the output useful obtained from the combustion chamber, to the total exergy value enters to the combustion chamber and shown in equation 64. ƞ II,cc = E ph x,flue gas fuel E x,in +E x,in air (64) b) HeX-1 Heat Exchanger Exergy Analysis The flue gas from the combustion chamber with the ph exergy E x in,fg(t) and the feed water with the ph E x in,fw1 enter to the heat exchanger for the production of process steam. While flue gas leaves ph the system at the exergy E x out,fg(t HeX1 out), the proses steam leaves the heat exchanger at the ph exergy E x out,ps (Fig. 8). Ex PH out,fg (T-HeX1-out) Ex PH out,ps (151 C) Fig. 8: HeX-1 heat exhanger exergy flow diagram Exergy balance for the HeX-1 heat exchanger; E x in E x out = E x D (65) ph ph ph [E x fg(t) + E x fw1(293 K) ] [E x fg(t HeX1 out) + ph E x ps(424 K) ] = E x D (66) written as that. If instead of the values in the equation are to be written; m fg [(h fg(t) h 0,fg ) T 0 (s fg(t) s 0,fg )] + m fw1 [(h fw1 h 0,fw1 ) T 0 (s fw1 s 0,fw1 )] m fg [(h fg(t HeX1 out) h 0,fg ) T 0 (s fg(t HeX1 out s 0,fg )] m ps [(h ps h 0,ps ) T 0 (s ps s 0,ps )] = E x D (67) The useful physical exergy obtained from the HeX-1 heat exchanger was calculated according to the equation 68. Q E x usefull,hex1 = Q HeX1 (1 T 0 T ps ) (68) The main useful output of the system is equal to the physical exergy value of the steam obtained from the heat exchanger. Exergy efficiency of HeX-1 heat exchanger is defined as the ratio of the useful exergy to the total exergy entering the system and shown in the equation 69. ƞ II,HeX 1 = Q HeX1 (1 T0 Tps ) ph E x fg(t) +E x fw1(293 K) ph (69) c) HeX-2 Heat Exchanger Exergy Analysis The flue gas that gets out from the HeX-1 enters into ph the HeX-2 with the exergye x out,fg(t HeX1 out), and increases the temperature of the water that enters ph into the heat exchanger with the exergy E x in,fw2, and finally provides to obtain the use water with the 8

23 ph exergy E x out,uw. The flue gas that transfers a portion of its own exergy to the feed water at HeX-2, leaves ph the system at the exergy E x out,fg(t HeX2 out) (Fig. 9). gas that transfers a portion of its own exergy to the combustion air at R-1, leaves the system at the ph exergy E x out,fg(t 423 K) (Fig. 10). Exergy balance for the R-1 recuperator; Ėx PH out,fg (T-HeX2-out) HeX - 2 CV Ėx PH out,fg (T-HeX1-out) E x in E x out = E x D (75) ph [E x fg(t HeX2 out) ph ph + E x air(298 K) ] [E x fg(t 423 K) + Ėx PH out,uw (363 K) Ėx PH in,fw2 (293 K) ph E x air(573 K) ] = E x D (76) The useful physical exergy obtained from the R-1 recuperator was calculated according to the equation 77. Fig. 9: HeX-2 heat exchanger exergy flow diagram Exergy balance for the HeX-2 exchanger; E x in E x out = E x D (70) ph [E x fg(t HeX1 out) ph [E x fg(t HeX2 out) ph + E x fw2(293 K) ] ph + E x uw(363 K) ] = E x D (71) If instead of the values in the equation are to be written; m fg [(h fg(t HeX1 out) h 0,fg ) T 0 (s fg(t HeX1 out) s 0,fg )] + m fw1 [(h fw1 h 0,fw1 ) T 0 (s fw1 s 0,fw1 )] m fg [(h fg(t HeX2 out) h 0,fg ) T 0 (s fg(t HeX2 out s 0,fg )] m uw [(h uw h 0,uw ) T 0 (s ps s 0,uw )] = E x D (72) The useful physical exergy obtained from the HeX-2 exchanger was calculated according to the equation 73. Q E x useful,hex2 = Q HeX2 (1 T 0 T uw ) (73) The main useful output of the system is equal to the physical exergy value of the hot utilization water obtained from the heat exchanger. Exergy efficiency of HeX-2 heat exchanger is defined as the ratio of the useful exergy to the total exergy entering the system and shown in the equation 74. ƞ II,HeX 2 = Q HeX2 (1 T 0 Tuw ) ph E x fg(t HeX1 out) +E x fw(293 K) ph (74) d) R-1 Recuperator Exergy Analysis The flue gas that gets out from the HeX-2 enters into ph the R-1 with the exergy E x out,fg(t HeX2 out), and increases the temperature of the fresh water that enters into the recuperator with the exergy ph E x air (298 K), and finally provides to obtain the ph combustion air with the exergy E x air (573 K). The flue 9 Q E x useful,r1 Ėx PH out,fg (T-423 K) = Q R1 (1 T 0 T air,573 K ) (77) Ėx PH air (298 K) Ėx PH air (573 K) R1 CV Ėx PH out,fg (T-HeX2-out) Ėx W f1 Fig. 10: R-1 recuperator exergy flow diagram The main useful output of the system is equal to the physical exergy value of the combustion air obtained from the recuperator. Exergy efficiency of R-1 recuperator is defined as the ratio of the useful exergy to the total exergy entering the system and shown in the equation 78. ƞ II,R1 = Q T0 R1 (1 ) T air,573 K ph E x fg(t HeX2 out) +E x air(298 K) ph (78) e) Coal-Fired Steam Production System Exergy Analysis Coal-fired steam generation system has three main useful output. The first of these is the process steam at 151 C, 5 bar pressure obtained from the HeX-1, the second is the use hot water with 2,5 t/h 90 C obtained from the HeX-2 and the third is combustion air at 1 atm to 300 C fed into the combustion chamber. In order to obtain the beneficial outcomes, coal and combustion air are supplied as the input to the system.

24 Exergy efficiency of the system is defined as the ratio of the useful exergy output to the total exergy entering the system. Thus, exergetic efficiency definition of the coal fired steam production system is expressed in the equation 79. ƞ II,sys = E Q x HeX1 E x,in Q +E x HeX2 air +E fuel +E x,in Q +E x R1 fan (79) x,in IV. Results and discussions In this study, the coal-fired system that are to be produced prosess steam and hot utilization water needed for the tea factory. As an energy resource Afşin lignite is used and analyses has been realized with assumption of stoichiometric combustion. Firstly adiabatic flame temperature is calculated according to ultimate and proximate analysis results of Afşin lignite. Because maximum flue gas temperature obtained with burn of coal related to adiabatic flame temperature. Adiabatic flame temperature of Afşin lignite has been estimated 1327 K. Hence, flue gas exit temperature has been limited to max K. Energy and exergy analysis results for the combustion chamber are reported in Tables 4 and 5. Tab. 4: Combustion chamber energy analysis results T [K] E in [kw] E out [kw] E fg [kw] E rad+conv+ash [kw] ƞ 1,CC % Tab. 5: Combustion chamber exergy analysis results A [K] E x in [kw] E x out [kw] E x D [kw] fuel [kw] E x,in air [kw] ch [kw] E x,in E x,fuel ph E xair ch E xair ph E x,flue gas [kw] [kw] [kw] Ƞ II,CC % According to combustion chamber energy and exergy analysis results has shown that as flue gas exit temperature close to adiabatic flame temperature combustion chamber efficiency increases. Maximum energy and exergy efficiency have been calculated 93% and 43%, respectively. Owing to the fact that Afşin lignite has high ash content even if heat losses stem from the system is prevented exergy efficiency is to be obtained as max. 43%. According to HeX-1 energy and exergy analysis results, is to be obtained maximum steam amount and HeX-1 heat exchanger efficiency for different flue gas exit temperature and using 1 kg/s Afşin lignite are presented in Table 6 and 7. Tab. 6: HeX-1 heat exchanger energy analysis results T [K] in [kw] E fg(t) out E fg(t HeX1 out) [kw] Q HeX1 [kw] m steam [kg/h] ƞ I,HeX1 % Because of the fact that HeX-1 exhanger is gas-fluid exchanger its energy and exergy efficiency is lower than gas-gas or fluid-fluid heat exchanger. Maximum steam production takes place approximately 3,5 t/h with 54% energy and 31% exergy efficiency at flue gas exit temperature with 1300 K in the HeX-1 heat exchanger. Accordingly, system capacity ought to be tripled by using 1 kg/s Afşin lignite for providing steam needs of the tea factory. Tab. 7: HeX-1 heat exchanger exergy analysis results T [K] E x in [kw] E x out [kw] E x D [kw] ph [kw] E x,fg(t) ph [kw] E x,fw1(20 C) ph [kw] E x,fg(t HeX 1 out) ph [kw] E x,ps(151 C) Q E x useful,hex1 [kw] ƞ II,HeX 1 % HeX-2 heat exchanger provides produce hot utilization water at 90 C for internal needs at tea factory. Mean hot water need of tea factory is assumed as 2,5 t/h. This requirement is provided by HeX-2 heat exchanger by using flue gas energy. Hot water production takes place with decrease of flue gas temperature at 23 C and with 9,4% energy and 7,36% exergy efficiency. The main reason of energy and exergy efficiency is so much low by using very high quality exergy resource is obtained to output that has low exergy value. Another reason for low exergy efficiency of HeX-2 heat exchanger is its gasfluid exchanger like HeX-1 heat exchanger. HeX-2 heat exchanger energy and exergy analysis results are presented in tables 8 and 9. 10

25 Tab. 8: HeX-2 heat exchanger energy analysis results T [K] in [kw] E fg(t HeX1 out) out E fg(t HeX2 out) [kw] Q HeX2 [kw] m water [kg/h] ƞ 1,HeX2 % Tab. 9: HeX-2 heat exchanger exergy analysis results T [K] E x in [kw] E x out [kw] E x D [kw] ph [kw] E x,fg(t HeX 1 out) ph [kw] E x,fw2(20 C) ph [kw] E x,fg(t HeX 2 out) ph [kw] E x,uw(90 C) Q E x useful,hex1 [kw] ƞ II,HeX 2 % R-1 recuperator used for the purpose that raising the temperature of the combustion air fed to the combustion chamber in a coal-fired steam generation system. Especially, when combustion air is given the high temperature plays a role in enhancing the combustion efficiency. So, the combustion air fed into the combustion chamber at 573 K in the system. Combustion air entering into R-1 recuperator at 298 K its temperature has been reached 573 K by using flue gas energy with 30% energy efficiency and 63,3% exergy efficiency. R-1 recuperator energy and exergy analysis results are presented in tables 10 and 11. Tab. 10: R1 recuperator energy analysis results T [K] E in [kw] in [kw] E fg(t HeX2 out) out E fg(t 150 C) [kw] Q R1 [kw] in E w,f1 [kw] ƞ 1,R1 % Tab. 11: R1 recuperator exergy analysis results T [K] E x in [kw] E x out [kw] E x D [kw] ph [kw] E x,fg(t HeX 2 out) ph [kw] E x,air (25 C) w [kw] E x,f1 ph [kw] E x,fg(t 150 C) ph [kw] E x,air (300 C) Q E x usefull,r1 [kw] ƞ II,R1 % As a consequence, while coal-fired steam production system energy efficiencies were calculated as 44%, 60% and 66%, respectively, its exergy efficiencies were calculated 13%, 17% and 19%, respectively at combustion chamber exit gas temperatures at 1000 K, 1200 K and 1300 K. Coal-fired steam production system energy and exergy analysis results are presented in tables 12 and 13. Tab. 12: Coal-fired steam production system energy analysis results T [K] E in [kw] Q HeX1 [kw] Q HeX2 [kw] Q R1 [kw] ƞ I,Sys % Tab. 13: Coal-fired steam production system exergy analysis results T [K] E x in [kw] fuel [kw] E x,in air [kw] E x,in fan [kw] E x,in Q E x HeX1 [kw] Q E x HeX2 [kw] Q E x R1 [kw] Q E x useful [kw] ƞ II,sys % V. Conclusion In this study is a parametric study on the energetic and exergetic performance of Afşin lignite stoichiometric combustion process. In this regard, in terms of the First law and the Second law of thermodynamics, the energy and exergy analyses have been achieved by using the stoichiometric combustion (air fuel ratio, λ=1) reaction and the proximate and ultimate analyses results of Afşin lignite samples taken from Afşin basin that is the largest lignite basin of Turkey. The following concluding remarks are made; The overall system exergy efficiency increases with an increase adiabatic flame temperature of the Afşin lignite but in that Afşin lignite has high ash and moisture content and low heating value its adiabatic temperature is to be constrictedly increase. Increasing of quality of the Afşin lignite with enrichment process like ash-free and drying process or using high quality coal requires for increase exergy efficiency of the system. 11

26 Acknowledgements The authors gratefully acknowledge the financial support from The Scientific and Technological Research Council of Turkey (TUBITAK). Nomenclature LHV HHV M Omin T E x ph E x ch E x D s T0 : low heating value (kcal/kg) : high heating value (kcal/kg) : molecular weight (kg/kmol) : minimum oxygen requirements (kcal/kg) : temperature ( C) : physical exergy (kw) : chemical exergy (kw) : exergy destruction (kw) : entropy (kj/kgk) : reference temperature (K) : enthalpy (kj/kg) h Greek letters λ : excess air coefficient (-) : efficiency (%) Superscripts m : mass Q : heat W : work rad : radiation Subscripts max : Maximum min : Minimum in : input out : output fw : output fg : flue gas ps : process steam uw : utilization water sys : system cc :combustion chamber Electricity Generation Company., Annual Report, Turkey, (2014). El-Wakil M.M., Power Plant Technology, 1st edition, Chapter 4, McGraw Hill, (1984). Korkmaz F., Current situation of Turkish tea sector and energy efficiency analysis of a tea factory, MsC Thesis, İstanbul Technical Universty, (2012). Midilli A., Chemical exergy calculations, Summer Course on Exergy and Its Applications, Ankara, June (2014). Ohijeagbon I.O., Waheed M.A., Jekayinfa S.O., Methodology for the physical and chemicial exergetic analysis of steam boilers, Energy 53, (2013). Rivero R., Garfias M., Standard chemical exergy of elements updated, Energy 31, (2006). Saidur R., Ahamed J.U., Masjuki H.H., Energy, exergy and economic analysis of industrial boilers, Energy Policy 38, (2010). Szargut J., Styrylska T., Approximate evaluation of the exergy of fuels, Breennstoff Waerme Kraft 16 (12), (1964). Turkey Black Tea Sector Report., Enterprise Europe network, Turkey (BlackSea), (2009). Taner T., Sivrioglu M., Energy-exergy analysis and optimisation of a model sugar factory in Turkey, Energy 93, (2015). World Energy Council., Energy Report, ISSN: , Ankara (2014). References Amelio A., de Voorde T.V., Creemers C., Degreve J., et. al., Comparison between exergy and energy analysis for biodiesel production, Energy 98, (2016). Bejan A., Tsatsaronis G., Moran M.J., Thermal Design and Optimization, John Wiley, , ISBN : (1996). Bilgen S., Kaygusuz K., The calculation of the chemical exergies of coal-based fuels by using the higher heating values, Applied Energy 85, (2008). Dincer I., Bejan A., Exergy : Energy, Environment and Sustainable Development, 1st Edition, Elsevier Science, (2007). Dincer I., Hussain M.M., Al-Zaharnah I., Energy and exergy use in the industrial sector of Saudi Arabia, Proc. Instn. Mech. Engrs, Vol. 217 Part A : J. Power and Energy (2003). 12

27 Exergy Analysis of Nitrogen Liquefaction Process Arif Karabuga 1 *, Resat Selbas 2, Ahmet Kabul 2 1 Suleyman Demirel University, Keciborlu Vocational School, Electrical Energy Generation, Transmission and Distribution, Isparta, 32260, Turkey 2 Süleyman Demirel Üniversitesi, Faculty of Technology, Energy Systems Engineering, West Campus, Isparta, 32260, Turkey * arif.karabuga@gmail.com Abstract Component of air that nitrogen, oxygen and argon is separated and liquefied by cryogenic method. Cryogenics is the science of very low temperature. Conventionally, the field of cryogenics is taken to start at temperatures below 120 K. Cryogenic air separation is the main method to separate air into its components. The nitrogen is used in chemical industry, food freezing, medical purpose, particle accelerators, colliders, synchrotrons, metal processing technology etc. In this study; a real nitrogen liquefaction unit has been examined. This nitrogen liquefaction unit is integrated to an air separation unit. Nitrogen provided by air separation creates the source of liquefaction unit. Energy and exergy analysis of the studied nitrogen liquefaction unit has been done. In numerical calculations and graphics EES (Equation Engineering Solver) software has been used. In results of thermodynamic calculations; exergy efficiency %36, COPactual and COPreversible 0.77 has been calculated. Furthermore, heat-exchanger block in nitrogen liquefaction unit is formed from HE-71, HE-72, HE-73. For each heat exchanger; exergy efficiency has been calculated. Exergy efficiency values are 0.55, 0.81 and 0.89 respectively. Keywords: Air separation unit, cryogenic, nitgrogen, energy and exergy. I. Introduction The earth is surrounded by air. The components of air are nitrogen, oxygen and argon. There is a virtually unlimited supply of nitrogen, oxygen and argon because of their natural occurrence within the atmosphere. Currently several methods are known in air separation. Two processes for air separation exist; cryogenic distillation and non-cryogenics distillation. Cryogenics is the science and techonolgy of very low temperature, usually below 120 K by Weisend II (1998). Non-cryogenic method to include that pressure swing adsorpsition (PSA) and membrane separation. The choice of the process to be used is based on the desired products. Cryogenic air separation is used when product high purity is needed. It is also advantageous when products are required in liquid form by Rizk et al.(2012). The cryogenic systems have the capability to deliver the largest capacities for products and for very high purities. Non-crygenics systems are employed at the lower end of production scale and generally for lower product purities by Castle (2002). In table 1 different air separation processes compared. Tab.1: Compare the process of air separation method by KLM Technology Group (2013) Process Advantages Disadvantages Cryogenic Low amount of electricity per unit nitrogen PSA Produces very high purity nitrogen Can generate liquid nitrogen for storge on site Cost-effective nitrogen production of relatively high purities Quick installation and start-up Membrane Low capital cost Large site space and utility requirements High capital cost Limited scaleability in production Long start-up and shutdown Low to moderate capital cost High maintenance equipment Production output is very flexible Quick installation and start-up Easy to vary purity and flow rate Noisy operation Limited scalability Uneconomical for high purity requirements Uneconomical for large outputs Requires relatively large amount of electricity per unit nitrogen Tab. 2: According to compare the purity values of the air separation unit by Campestrini (2014) Process Purity (%) Cryogenic N: PSA O 2: N: 99,9 up Membrane O 2: 85 99,7 N: 10 ppb * *ppb: Parts per billion 13

28 The largest markets for oxygen are in primary metals production, chemicals and gasification, clay, glass and concrete products, petroleum refineries and welding. The use of medical oxygen is an increasing market. Gaseous nitrogen is used in the chemical and petroleum industries and it is also used extensively by the electronic and metal industries for its inert properties. Liquid nitrogen is used in applications ranging from cryogenic grinding of plastics to food freezing. Argon, the third major component of air, finds uses as an inert material primarily in welding steelmaking, heat treating and in manufacturing processes for electronics by Vinson (2006). Otherwise liquid nitrogen is used in physic applications, particle accelerators, colliders, synchrotrons by Thomas et al. (2011). An exergy analysis is carried out to analyze the possibilities of fuel saving in the cryogenic distillation process. It is shownthat more than half of the exergy loss takes place in the liquefaction unit and almost one-third in the air compression unit. Exergy loss in the compressor is reduced by improving by Cornelissen and Hırs (1998). Amin et al. (2014) made simulation of nitrogen separation from air. In this study, liquefaction processes are predicated on Linde-Hampson method as a thermodynamic cycle. Liquefaction degree is taken -200 C under maintanence and affective parameters. In nitrogen from air simulation, HYSYS programme used and purity rate of nitrogen found as % as a result of simulation. Rizk et al. (2012) Made simulation of three types of cryogenic process column and calculated exergy losses of different columns. For each column accurate analyses has been defined. Exergy analyses between distillations columns have been compared double diabetic column s exergy efficiency is 23 % more efficient than traditional adyabatic double columns. Van der Ham and Kjelstrup (2010) made two different air separation units exergy analyses one of the studied units is three columns other is two columns. Three columns design has 12 % less exergy losses than two columns design. II. Experimental facility In this study an integred system is examined. Nitrogen liquefaction unit has been integrated to air separation system. Nitrogen cryogenic is used as air enters the separation unit under atmosphere pressure and ambiente temperature. Air liquid passed from air filter, passes through three staged air compressor, then enters air purity units. In this stage, separated from particles and damp air enters to cold box. Cold box consists of main heat exchanger blocks and distillation column is formed two different column and argon column. These two different columns are high and low pressure column. After that dry air enters heat exchanger and leaves it in a temperature close to liquafection degree and that dry air enters high pressure column of to cold box. There are 50 separation trays in high pressure column and 78 separation trays in low pressure column. 14 Fig. 1: Distillation column in cold box Air is separated as oxygen, nitrogen and argon through temperature differences. In distillation column, nitrogen is separated from other components and transferred into liquefaction unit. Nitrogen liquefaction unit consists of nitrogen recycle compresson CP-77, booster compressor CE-77, booster compressor last cooler HE-771, nitrogen chiller R-60 and three exchangers. While entering nitrogen liquefaction unit with a nearly 5 bar pressur, nitrogen has been risen to nearly 32 bar pressure in recycle compresson. Nitrogen leaves booster compressor, turbine with nearly 45 bar pressure. Compressor enables this pressure work from turbine through booster compressor last cooler and enters to first heat exchanger blocks HE-1. It leaves from HE-1 with a temperature of 251 K. Entering nitrogen chiller cooler leaves its heat here. After that nitrogen enters to HE-2 heat exchanger and leaves it with a temperature of 182 K. Nitrogen s 3/4 liquid mass leaving HE-2 heat exchanger is send to booster compressure turbine and enables the necessary work for pressure in compressor. Liquid from turbine combines with average pressure nitrogen and passes though heat exchanger blocks in liquefaction unit them of combines with average nitrogen from main heat exchanger of cold box and finally enters nitrogen recycle compressor ¼ of nitrogen from HE-2 heat exchangerenters to HE-3 heat exchanger and leaves HE-3 with a temperature of 112 K and 45 bar pressure in liquid phases.

29 COP actual = q L,gas w in (6) For COP vaule of liquid per mass in liquefaction unit, equation 2 is calculated as reversible work by Dinçer and Rosen (2007). COP rev = Q L,liquid w rev (7) Reversible work in equation 7 is defined in equation 8 by Dinçer and Rosen (2007). w rev = h 17 h 2 T 0 (s 17 s 2 ) (8) Fig. 2: Air separature unit by Linde Group (2009) III. Exergy analysis For making exergy analysis of nitrogen liquefaction unit, it s necessary to define two cooling effect in liquid and gas phases. The refrigeration effect per unit mass of the liquefied gas is give by Dinçer and Rosen (2007). q L,gas = h 4 h 2 (1) h 4 value defines the enthalpy leaving from compressure, h 2 value defines the enthalpy entering the compressor. From energy balance on the cycle, the refrigerant effect per unit mass of the liquefied gas is given by Dinçer and Rosen (2007). q L,liquid = h 2 h liquid (2) hliquid defines the enthalpy value of liquid nitrogen leaving cycle. If energy balance in compressor is written for gases per mass according to compression by Dinçer and Rosen (2007). w in = RT 0 ln(p 2 P 1 ) (3) Here R is nitrogen gases constant, T0 is ambient temperature, P values are entering and exit pressures. (T0=298.15) For finding exergy efficiency in cycle, COPactual value rate of COPrev values have been calculated. η ex = COP actual COP rev (9) To find exergy efficiency in heat exchanger of nitrogen liquefaction cycle by Thomas et al. (2011) η ex HE = m HP (ex heat out ex heat in ) m LP (ex cold in ex cold out ) (10) Here m is the rate of mass flow through heat exchanger, ex is defined as hot and cold exergy. Tab. 3: Enthalpy and entropy values in nitrogen liquefaction cycle Referance point values Enthalpy (kj/kg) Entropy (kj/kg) Referance point values Enthalpy Entropy (kj/kg) (kj/kg) For finding liquefaction rate of gases in cycle, fraction of equation 2 to equation 1 has been calculated by Dinçer and Rosen (2007). y = q L,gas q L,liquid (4) If actual work in cycle is written fornitrogen per mass by Dinçer and Rosen (2007). w actual = w in y (5) Fig. 3: T-s diagram of nitrogen liquefaction unit If actual coefficient of performance (COP) in cycle is written for gases per mass by Dinçer and Rosen (2007). 15

30 In table 6 ıt is observed analyses decreases if T2 temperature entering the system increases in circumstances of T0 value is 20 C and 25 C. But if ambiatte temperature is 20 C COPactual value is a little much comparing to 25 C however if ambiante temperature is 25 C exergy efficiency inreases comparing to 20 C. Fig.4: P-h diagram of nitrogen liquefaction unit IV. Results and discussions With the values given in table 3, equations solved and values in table 4 obtained. Tab. 4: Calculated result Calculated Result Calculated Result Values Values q L,gas 52.3 kj/kg COP actual q L,liquid kj/kg COP rev y η ex 0.36 w actual,gas kj/kg η ex HE w actual,liquid kj/kg η ex HE w rev kj/kg η ex HE It ambiente temperature increases COPactual and COPrev values decrease in pic 5. Fig. 7: COPactual and exergy efficiency change graphic depending on h4 value In figure 7 ıt s observed that exergy efficiency increases if enthalpy value rises leaving compressor. V. Conclusions In this study a real air separation and nitrogen liquefaction units have been examined. Energy and exergy analyses has been made and exergy efficiency has been calculated as 0.36, COPactual value and COPrev as Moreover exergy efficiencies of HE-1, HE-2 and HE-3 heat exchangers have been calculated as 0.55, 0.81 and In result of this study, compressor efficiency is low but ıf enthalpy value increses as showed in figure 7 exergy efficiency of cycle will increase. Nomenclature Fig. 5: COP values change graphics depending on T0 temperature m ex COP w h s : mass : Exergy : Coefficient of Performance : Work : Enthalpy : Entropy Fig. 6: Exergy effiency change graphics depending on T2 temperature 16 Greek letters η : Effiency Subscripts ηex : Exergy efficiency References Weisend II, J.G., Hanbook of Cryogenic Engineering. 504, Taylor & Francis, USA. Rizk, J., Nemer, M., Clodic, D., A Real Column Design Exergy Optimization of a Cryogenic Air

31 Separation Unit. Energy, 37, Castle, W. F., Air Separation and Liquefaction: Recent Developments and Prospects for the Beginning of the New Millennium. International Journal of Refrigeration, 25, Campestrini M., Thermodynamic study of solid-liquid-vapor equilibrium: application to cryogenizs and air separation unit, doctora thesis, 147. Vinson, D. R., Air separation control technology, Computers and chemical engineering, 30, Thomas, R. J., Ghosh, P., Chowdhury, K., Exergy analysis of helium liquefaction systems based on modifield Claude cycle with two-expanders. Cryogenics, 51, Amin R., Islam A., Islam R., Islam S., Simulation of N2 Gas Separation Process From Air, IOSR Journal of Applied Chemistry, Volume: 6, Issue: 5, van der Ham, L. V., Kjelstrup, S., Exergy Analysis of Two Cryogenic Air Separation Processes, Energy, 35, Dinçer, İ., Rose, M.R., Exergy: Energy, Environment and Sustainable Development, 454. Elsevier, Canada. KLM Technology group, Air Separation Units. The Linde Grup, About Air Separation Units. 17

32 Investigation of Irreversibility with CO2 Emission Measurement in Industrial Enamel Furnace Sedat Vatandas 1*, Atakan Avci 2, M. Ziya Sogut 3 1 Energy Efficiency and management Department, Enervis, Bursa Turkey 2 Mechanical Engineering Department, Engineering Faculty, Uludağ University, Bursa, Turkey 3 Mechanical Engineering Department, Engineering Faculty, Bursa Orhangazi University, Bursa, Turkey. * sedat.vatandas@enervis.com.tr, mzsogut@gmail.com Abstract The increasing use of energy in Turkey, especially in industry, the need for energy efficiency raises to the forefront every day. External dependence on sources of energy, energy costs and competitive factors, addressed the energy efficiency to take into account as a new energy source. At the same time, increasing competitiveness, reducing energy costs and decreasing environmental impacts can be achieved only through energy efficiency. In this study, primarily energy and exergy analysis of enamel oven is made which has a significant energy consumption in the facility then energy efficiency improvements are evaluated. Due to the analysis, irreversibility found approximately 88.71%. Finally, economic savings provided by emission reductions are evaluated in accordance with improvements.irreversibility due to changes in their CO2 emissions are discussed separately in this study. Exergy and environmental effects of improvements are assessed. Keywords: Enamel ovens, energy analysis, exergy, analysis, CO2 emissions, irreversibility. I. Introduction The threat posed by an increasingly global warming today, many studies are performed for solutions. Considering the results and cost of the solutions, the importance of energy efficiency increases even more. Reducing energy costs and carbon emissions can be achieved only through energy efficiency. As half the energy used in industry, ensuring energy efficiency is extremely important in terms of competitiveness and reduction of environmental impact. It is impossible to continue to compete without reducing specific energy consumption with increasing energy needs and costs. One of the significant energy users in facilities are industrial furnaces. In processes where the industrial furnace is, energy cost is the highest after the raw material cost. Significant amount of energy in high-temperature furnaces cannot be converted to useful energy. Industrial furnaces carry significant potential for energy efficiency targets desired to be achieved in the industry. Minimizing losses and energy recovery will be returned as a profit by ensuring energy efficiency in industrial plants. In this study, in order to determine the current state of modeled enamel furnace, energy and exergy analysis was performed after calculating the energy, transferred into the furnace and turned into useful energy after processing. Afterwards studies were described that was made to upgrading the furnace. And energy and exergy analyses re performed with energy efficiency improvement projects. According to the results of analyses energy recovery potential was evaluated and results are discussed. II. Industrial furnaces Industrial furnaces are used in metallurgical production to smelt at high temperatures, in heat treatment, tempering, in some areas such as the food industry, drying and for fermentation at low temperatures (Hazi et al., 2009). Metal industry which has a significant share in the total industrial energy consumption, examples of furnaces working in the high temperature zones; Enamel cooking 600 C C Heat treatment of metals 1100 C Rolling, extrusion, cooking ceramic materials, such as heat treatment and pressing 1350 C Melting and smelting of metals 1700 C According to the type of heat generation, furnaces can be divided into two main groups. Electrical furnaces and furnaces that use fuel. According to the type of fuel that combustion furnaces use, furnaces classified as solid, liquid and gaseous furnaces (Trink et al., 2004). Electric ovens, works as an arc furnace or an induction furnace. Some advantages of electric ovens are easy to operate and easy to be managed. The absence of any loss due to the flue gas is another advantage of electric ovens. On the other hand main disadvantage is the price of electricity. In addition heating can be provided by plasma arc, laser, radio frequency or a combination of heating by electromagnetic. 18

33 Combustion furnaces can be classified according to the type of heat transfer of heat, operating state and by providing the shape of the heat recovery. Industrial furnaces shall design to provide maximum amount of the product with homogeneously heat diffusion. The specific energy consumption should be kept constantly under control as well as quality. Furnaces should be operated with minimum fuel and minimum maintenance while furnace design should allow maximum heat transfer to maximum material in defined time. In order to ensure these conditions, criteria's should be considered that ordered below; How much heat will be transferred to the material Determination of the heat necessary for heating the mass and losses. Once these issues are identified, the studies such as, how much of the losses will be minimized, what will be used as a refracter material and how will the temperature of the furnace body stabilize, should be made. The model industrial furnace is used for the curing of the applied enamel of the tank and the boiler to be used in the heating sector. After the processes of metal forming, welding, degreasing, cleaning and enameling of the tank and boiler curing process is taken in model furnace. According to the sizes, adequate number of tank and boiler is processed in the furnace at 860 C. The flow line of the furnace is given in Figure 1. III. Theoretical analyses Industrial furnaces are treated as thermal processes and analyzed as a continuous flow system under the first and second laws of thermodynamics. Such systems, in order to be performed energy and exergy analysis firstly temperature, environmental conditions, specific heat capacity and mass flow of the input and output materials must be defined. In the evaluation of industrial furnaces mass balance of input and output material defined as below. (Sogut ve Oktay, 2006); m. i. m o (1) Energy is a protected property under the first law(cheng et al,. In this case, the mass flow rate for the furnace, general energy balance caused by work and heat is;. E. Q i. E o. mi h W i. m o h o (2) (3) (Sogut ve Oktay, 2006; Balkan et al., 2005). In equation (2) E. i refers input total energy amount, Ė o refers output total energy amount. In equation (3) Q is ( Q Q and W refers (. W.. W net net. i. Q Q ) total heat. W o o. W i ) amount of work. h is the enthalpy value of input and output materials. The first law of thermodynamic describes a quantitative measure of energy as independent for the direction of energy. Practically systems lose energy, with the irreversibility and changings in environmental conditions. As this situation is evaluated as production of entropy, in the second law of thermodynamic it is expressed in the concept of exergy. The exergy of a system is the maximum useful work possible according to the environmental conditions of system. The exergy balance of the prose's expressed (Szargut vd., 1998 and Utlu e al.2011); E x = Ex kin + Ex pot + Ex phy+ Ex che (4) Fig. 1: The flow line of the furnace In the equation Ekin is kinetic exergy, Epot is potential exergy, Ephy is physical exergy, and Eche is chemical exergy. When the furnaces are considered as continuous flow system, exergy balance is;. Ex i. Ex o. l (5) 19

34 (Sogut, 2010).. Ex o Ex. i refers to output exergy. refers to input exergy and,. l defines the exergetic destruction. In these systems, the potential, kinetic and chemical exergy can be neglected. Hence general exergy balance can be written as below; T..... ( 1 0 ) Qk W mi i mo o l (6) T In equation (6) 1 Q. k, Ṫ k refers heat transfer rate, W. is work amount, is flow exergy, s entropy and 0 index represents (P 0 and T 0) in terms of dead state of environment. Hence mass and material flow exergy is; ( h h0 ) T0 ( s s0) (7) Performance evaluation in the system is defined by efficiency. Efficiency is defined as the ratio of input to output of the system. Depending on the analysis conducted under the first law energy efficiency of process is found by the equation below; Here process and (8) Ė refers to the total output energy of o Ė i refers to the total input energy of process. Exergy efficiency of process is calculated with the equation of (9) (Cornelissen 1997).. Ex o ıı (9). Ex i CO2 emission in thermal systems, caused by energy losses, is dependent on the type of fuel used and the waste energy potential. It can be expressed as CO 2 where CO Q 2 W CO 2 I (10) is the unit energy CO2 emission Q coefficient and W is the amount of total waste energy. In works where such waste energy is exploited, depending on the recycled energy potential, total CO2 should be calculated taking CO2 emissions caused by furnaces. Hence, total CO2 emission (CO ) may be expressed as follows: 2 CO. E. o ı E i Q Q (1 ) (11) CO CO 2 2 BW i Wi j j 2 CO2i CO2 j Ri I i I j Q where W is the total energy of furnace, is rational exergy efficiency. is the function among environment temperatures, waste heat, and inlet and outlet temperatures of furnace(kılkış, 2004). Van Gool (1997) has noted that maximum improvement in exergy efficiency for a process or system is obviously achieved when the exergy loss or irreversibility E x E x ) is minimized. ( i o Consequently, he suggested that it is useful to employ the concept of an exergetic improvement potential when analyzing different processes or sectors of the economy. Hammond and Stapleton (2001) give this improvement potential in a rate form, denoted as IP below. IP 1 )( ) (12) ( g IV. Energy and exergy analyses Providing mass balance of enamel furnace which is taken as a model, energy and exergy analyses are conducted according to the reference environmental conditions. The energy analyses of the furnace is made and results is given in Table 1. Material Tab. 1: Energy analyses of the furnace Through energy analyses, energy efficiency of furnace was found %17.3 according to the equation (8). mo. ho ,1 1 = 0,173 mi. hi In real furnace processes environmental conditions directly affect the efficiency. In this aspect the importance of exergy analyses become important which is connected to the second law of thermodynamics. However, firstly environmental ç Input Materials M T1 Cp Δh kg/h K kj/kgk kj/h Boiler 279, , ,31 N.gas 40, , ,20 Air 35, , ,17 Enamel 2, ,35 212,21 Electricity 4,30 Total 357, ,2 Material Output Materials M T1 Cp Δh kg/h C kj/kgk kj/h Boiler 279, , ,21 Flue gas 77, , ,87 Total 356, ,1 20

35 parameters need to be defined for exergy analyses. The environmental reference pressure was taken (P0) 1 atm and temperature 25 C for conditions where the furnace is located. For exergy analyses some assumptions are made. According to this; effect of pressure on the enthalpy and entropy characteristics of input and output materials were neglected. Gases are considered as an ideal gas mixture for input and output material flow. As the furnace is continuous flow, kinetic, potential and physical exergy value of input and output materials are neglected. According to the assumptions exergy analyses were conducted and represented in Table 2. Material Tab. 2:Exergy analyses of the furnace Furnace exergy efficiency is defined as the ratio of the total exergy output to total exergy input. Accordingly exergy efficiency of the furnace is;. Ex o ıı =. 0, 1128 Ex M T1 Cp Δh Δs ψi kg/h C kj/kgk kj/h kj/k kj/h Boiler 279, , ,31 8, ,8771 N.gas 40, , ,2 5, ,72 Air 35, , ,17 2, ,75538 Enamel 2, ,35 212,21 0, , Electricity 4,3 0 4,3 Total 356, ,2 16, ,89 Material i , ,89 Input Materials Output Materials M T1 Cp Δh Δs ψi i kg/h C kj/kgk kj/h kj/k Boiler 279, , ,21 172, ,5888 Flue gas 77, , ,87 146, ,2978 Total 356, ,08 319, ,89 According to the results of analyses, efficiency of the furnace was determined too low and causes were investigated. Causes were found as heat losses in the furnace surface (Figure 2 and 3), inefficient combustion system and optimum design is not provided. Also waste heat potential was determined according to the results of flue gas analyses shown in Figure 4. Fig. 3: Heat losses in the furnace surface (side) Fig. 4: Flue Gas Analyses Based on this information by changing the furnace combustion systems instead of conventional eight burners, six recuperative burners are used. Furnace panel is renewed in accordance with the modified combustion system in order to provide control. Furnace refreacter material has been renewed in order to avoid the leakage losses. Although waste heat was used by recuperative burner (Figure 5), still waste heat potential is determined for the bath process of the boilers. In bath process hot water is needed for the degreasing process. Economizer is used for heating water of bath process(figure 6). Nearly kcal was obtained with the economizer. Fig. 5: Recuperative burner section Fig. 2: Heat losses in the furnace surface (Front) 21

36 V. Measurement of CO 2 emissions As a result of studies exergy efficiencies were respectively 11.28% and 22.73%. With the equation of 12 improvement potential (IP) of furnace was found After improvement projects IP was calculated Table 5 is prepared according to the equation 10. Tab. 5 CO2 Emissions saving after improvement CO2 CO2 Exergy Improvement CO2 Emission Efficiency Potential (kg CO2/h) Emission Factor Before Improvement 11,28% , After Improvement 22,73% , Fig. 6: Recuperative burner in furnace Exergy analyses and energy analyses have been renovated and after efficiency improvement project and the results of the energy analysis are given Table 3. Tab. 3: The balance of energy analyses Energy efficiency of the furnace according to the equation 8 was found %35. Material Material M T1 Cp Δh kg/h C kj/kgk kj/h Boiler 449, , ,89 N.gas 26, , ,80 Air 58, , ,37 Enamel 3, ,35 318,31 Electricity 4,30 Total 538, Material Input Materials Output Materials M T1 Cp Δh kg/h C kj/kgk kj/h Boiler 449, , ,79 Flue gas 88, , ,90 Total 538, Tab. 4. The balance of exergy analysis M T1 Cp Δh Δs ψi kg/h C kj/kgk kj/h kj/k kj/h Boiler 449, , ,888 13, ,80196 N.gas 26, , ,8 Air 58, , ,369 4, ,4766 Enamel 3, ,35 318, , , Electricity 4,3 0 4,3 Total 538, ,7 18, ,72 Material Input Materials Output Materials M T1 Cp Δh Δs ψi i kg/h C kj/kgk kj/h kj/k Boiler 449, , ,79 277, ,01 Flue gas 88, , ,90 166, ,58 Total 538, ,69 444, ,59 Exergy efficiency of the furnace according to the equation 9 was found % 22, With the equation (CO2b / CO2a)/ CO2b, %74 CO2 saving potential has been identified with the projects that are explained in part IV. VI. Conclusions Studies show that energy efficiency projects pay their selves in a short time when compared to energy generation projects. Therefore, facilities should think constantly energy efficiency and perform systematically energy saving projects. In order to maximize the energy efficiency, facilities should focus on large energy consumer as furnaces. In this study, enamel furnace was taken as a model which is defined as significant energy user in energy management system. Firstly looses and leakages defined and fixed, afterwards efficiency was increased in the view of technological developments. To provide the maximum amount of production per unit time, furnace height was increased. And 3 pieces of boiler has started processed which was 2 before improvement. In addition, due to technological advances, burner systems were upgraded. After improvements, energy and exergy efficiency of the enamel furnace was doubled. Likewise enamel furnace improvement potential is reduced by half. CO2 emission depending on exergy efficiency was decreased four times. Also with the design changing's in furnace, not only more production was provided per unit time but also CO2 emission per unit of production was reduced. This situation also shows that the energy efficient design is very important in the case of energy efficiency. References Balkan, F., Colak N., Hepbasli, A. (2005) Performance evaluation of a triple-effect evaporator with forward feed using exergy analysis, Int. J. Energy Res

37 Chen X., Zhang Y., Zhang S., Chen Y., Liu S.(2007) Exergy analysis of iron and steel eco-industrial systems The third international Exergy, energy and Envıronment symposıum, 1 5 July, Evora, Portugal Cornelissen R.L.(1997) Thermodynamics and sustainable development: The use of exergy analysis and the reduction of irreversibility, Ph.D thesis, University of Twente, The Netherlands. Hammond G.P., Stapleton A.J. (2001) Exergy analysis of the United Kingdom energy system, Proceedings of the Institute of Mechanical Engineers, 215 (2) Hazi A., Badea A., Hazi Gh., Necula H., Grigore R. (2009) Exergy Evaluation of Renewable Use in the Pulp and Paper Industry, IEEE Bucharest Power Tech Conference, June 28th-July 2nd, Bucharest,Romania Kılkıs Bir. (2004) An Exergy a ware optimization and control algoritma for sustainable buildings, İnternational Journey of Green Exergy, 01/2004;No 1:65-77 Sogut Z., Oktay Z.,(2006) Energy And Exergy Analyses In Thermal Process Of Production Line Of Cement Factory And Application, Igec-2 International Green Energy Conference, Ontario Institute of Technology (UOIT), Canada, June 2006 Sogut Z., Oktay Z. Karakoç H.(2010) Mathematical modeling of heat recovery from a rotary kiln Applied Thermal Engineering 30 (2010) Szargut, J., Morris, D.R., Steward, F.R.(1988) Exergy Analysis of Thermal and Metallurgical Processes, Hemisphere Publishing Corporation. Utlu Z., Hepbasli A., Turan M.(2011) Thermodynamic analyses of a Industrial dryer mill, X National Sanitary Engineering congress, 13/16 April 2011, İzmir/Turkey Trinks W., Mawhinne M. H.,(2004) Shannon R. A., Reed R. J., Garvey J. R. Industrial Furnaces Sixth Edition, ISBN: , Copyright 2004 John Wiley & Sons, Inc, New Jersey, USA W. Van Gool, (1997) Energy policy: fairy tales and factualities, in: O.D.D. Soares, A. Martins da Cruz, G. Costa Pereira, I.M.R.T. Soares, A.J.P.S. Reis (Eds.), Innovation and Technology Strategies and Policies, Kluwer, Dordrecht, pp

38 Advanced Exergy Analysis of an Application of Waste Heat Powered Ejector Refrigeration System to Rotary Kiln Abid Ustaoglu 1*, Mustafa Alptekin 2, Mehmet Emin Akay 3, Resat Selbas 2 1 Bartin University, Faculty of Engineering, Department of Mechanical Engineering, Bartin, 74100, Turkey 2 Suleyman Demirel University, Faculty of Engineering, Department of Energy Systems Engineering, Isparta, 32260, Turkey 3 Karabuk University, Mechanical Engineering Department, Karabuk, 78050, Turkey. * abidusta@gmail.com ; austaoglu@bartin.edu.tr Abstract The rotary kiln consumes the biggest share of energy in a cement factory and has great heat loss which causes significant reduction in efficiency. This reduction becomes excessive for a cement factory using the wet method. In the progress of clinker production, about 33% of total energy is exhausted from the chimney of rotary kiln. In this study, conventional and advanced exergy analyses of a waste heat powered ejector refrigeration system were performed by using the exhausted heat from the chimney of the rotary kiln. Conventional and advanced exegy analyses are carried out to the system. The exergy destructions, thermal efficiency and exergy efficiency were calculated. By means of the advanced exergy analysis, avoidable and unavoidable exergy destruction rates were found in order to determine the improvement potentials of both the components and the overall system. Moreover, the effects of condenser temperature, generator temperature and evaporator temperature on system performance were also investigated. The largest exergy destruction occurs in generator, accounting 48.9% of total exergy destruction and followed by ejector with ratio of 39.13%. The largest share of total avoidable exergy destruction occurs in ejector and that is about half of total avoidable part of overall system (82.94%). Therefore, it is important to concentrate on this component to improve overall system performance. About 31 % of the total destruction is falling into the part of unavoidable exergy destruction. Namely, the system has potential to improve by reducing the avoidable part of 40% Keywords: Advanced exergy analysis, waste heat, rotary kiln, ejector refrigeration I. Introduction Renewable energy is one of the most important solutions for a clean energy future. Another alternative to solve these problems is utilization of the waste heat. Ejector refrigeration systems are promising technologies since they can utilize renewable energy and harvest low-grade waste heat from industrial processes for cooling demand so that the problems related with greenhouse gas emission and the energy cost can be reduced. Cement industry is one of the most energy consuming industries in the world. The energy consumption rate reaches 12-15% of total energy consumption in industry. The rotary kiln has biggest share in terms of energy consumption in a cement factory and has great heat loss which causes significant reduction in efficiency. This reduction becomes excessive for a cement factory using the wet method. In the progress of clinker production, about 33% of total energy is exhausted from the chimney of rotary kiln apart from the heat loss through the wall of rotary kiln Ustaoglu et al. (2016). Therefore, reutilization of the exhausted gas becomes substantial due to the great waste heat. In order to utilize this heat, one of the preferable options is to use ejector refrigeration cycle. In the ejector refrigeration cycle, ejector is an important component. An ejector can increase the pressure without using mechanical energy directly. Therefore, it is better to use an ejector than applying mechanical devices to increase the pressure such as compressor, pump, etc. since it may be safer and simpler technology. Apart from that, the system combined with ejectors and other components are also simple. Therefore, researchers have been made endeavor to have more knowledge about the ejector behavior and improve the system performance with various methods. The ejector theory was proposed by Keenan et al (1950) and that underlies many other ejector model and designs. Since proposition of the ejector theory, many studies have been carried out to improve the performance of the refrigeration cycles by improving design or combining with other cycle, and to evaluate the first and second low efficiency of thermodynamic for more realistic and detail approach. Huang et al. (1999) predicted the ejector performance for critical conditions and validated the results with experimental data. Zhu et al. (2007) developed a two-dimensional ejector model by considering of a shock circle at the entrance of the constant area section. The second law efficiency evaluation is important tool to determine the location, magnitude and source of the exergy destruction. Pridasawas and Lundqvist (2004) carried out exergy analysis for a solar-driven 24

39 ejector refrigeration system using butane as a working fluid and obtained that the most substantial exergy destruction in the ejector refrigeration cycle occurred in the ejector. Dahmani et al. (2011) said that more than half of the total exergy destruction in the ejector refrigeration cycle by using working fluid, R134a was due to the ejector. The exergy analysis for ejector enhanced refrigeration systems has also been largely carried out to evaluate the improvement of the system performance by adapting the ejector. Yan et al. (2015) proposed a new ejector enhanced auto-cascade refrigeration cycle using R134a/R23 and compared it with a conventional auto-cascade refrigeration cycle. Their cycle achieved % better efficiency than a basic cycle at the same operation conditions as the ejector achieved pressure ratio lifts of They found that the highest exergy destruction occurs in the compressor, the condenser, cascade condenser, expansion valve, ejector and evaporator, respectively. Yang et al. (2016) proposed a novel combined power and ejector-refrigeration cycle using zeotropic mixture, and evaluated the cycle performance with different fluid composition and compared the novel cycle with conventional combined cycle. The results showed that the cycle has better performance in lower condenser temperature. Although the refrigeration cycle achieves lower evaporating temperature in higher generating temperature, the power output decreased. Zhao et al. (2015) conducted a numerical study to analyze the performance of the ejectorexpansion refrigeration cycle (EERC) using zeotropic mixtures (R134a/R143a). The results showed that the compressor and ejector have the most exergy destruction, and the cycle exergy efficiency achieves a maximum value with the mass fraction of 0.7 for R134a. Apart from the conventional analysis method, the exergy method is an important tool that can show the useful work that can be generated through the process. However, it cannot explain the interaction among the component or estimate the actual improvement potential. Conventional approach for optimizing of the system may be wrong without taking into account of the interaction between the components particularly for complicated systems in which many components having interaction with each other. In the case of the ejector refrigeration cycles, the operation parameters of the ejector depend on itself and other components Chen et al. (2014). A recent developed technique, the advanced exergy analysis by splitting the exergy destruction into avoidable-unavoidable and endogenous-exogenous part can enable us to evaluate the system in more detail and investigate the capacity of the improvement Tsatsaronis (1999). The advanced exergy analysis of different refrigeration systems, including vapor compression refrigeration systems Morosuk and Tsatsaronis (2009), Morosuk et al. 25 (2012), absorption refrigeration systems Gong and Goni (2014), Morosuk T, Tsatsaronis (2014) ejector refrigeration system Chen et al. (2015) and heat pump Erbay and Hepbasli (2014) have been carried out. In a previous study, energy and exergy analysis of a wet type rotary kiln were carried and recovery capacity of waste heat was evaluated by using an organic Rankine cycle Ustaoglu et al. (2016). The results showed that a great amount of heat energy of 30.5 MW is exhausted from the chimney of rotary kiln. In this study, the waste heat from the rotary kiln was evaluated for an ejector refrigeration system and advanced exergy analysis was carried out to determine avoidable and unavoidable exergy destruction rates in order to determine the improvement potentials of both the components and the overall system. II. Material and Method II.1. System Description Figure 1 shows a waste heat powered ejector refrigeration system.. Each state of the working fluids is represented in the point as seen in figure. The working principle of this system can be expressed as follows: the working fluid in saturated liquid phases leaving from the condenser is separated as two parts; one goes to pump where that fluid is compressed to generator pressure, and the other one goes to the expansion valve. The properties of the working fluid in points 1 and 4 are same as point 8. The compressed fluid is pumped to generator to be vaporized by using the exhausted gas from the rotary kiln chimney and leaves the generator as saturated vapor. The other fluid s pressure and temperature decreases to the evaporator level in expansion valve. The fluid is vaporized in the evaporator by using the heat of the cooling ambient. Either saturated vapors from the upper and below cycle enter to ejector and mix. When the condenser pressure below the critical value, ejector can entrain same amount of secondary fluid Huang (1985). Thus, the cooling capacity and COP are kept constant. The mixture leaves from the ejector to be superheated vapor at condenser pressure. The forking fluid entering water-cooled condenser release its heat to the water and leaves the condenser from point 8 to be saturated fluid, and then again separates. Thus, the cycle is completed. R142b is used as working fluid. II.2. Thermodynamic Evaluation The ejector refrigeration system is considered as steady flow open-system and modeled based on the first and second laws of thermodynamics, and these laws are applied to each component in the system. In the steady flow open systems, the mass and energy of control volume are stationary. General equations of mass, energy and exergy balances by ignoring the kinetic and potential energy variations can be expressed.

40 E F,tot = E F,GE + W PU (6) E P,tot = E P,EV (7) E D,tot = E D,k (8) The coefficient of the performance (COP) can be described as a ratio of the refrigerating capacity generated by evaporator to the heat supplied to generator and electricity consumption of the pump COP = Q EV (Q GE + W PU ) = m EV(h 6 h 5 ) m GE(h 3 h 1 ) = μ(h 6 h 5 )/(h 3 h 1 ) (9) where the ratio of the mass flow rate of secondary fluid coming from evaporator to the primary flow in the reversible ejector coming from the generator indicates ideal entrainment ratio McGovern et al. (2012). Table 2 shows the equations of the entropy calculation for each component in the ejector refrigeration system. Fig. 1: Waste heat powered ejector refrigeration system m in = m out (1) Q + W = m out h out m outh out (2) E heat + W + E D = E out E in (3) S F S P + S GEN = 0 (4) E D = T 0 S GEN (5) where ṁ is mass flow rate, h is enthalpy, Q is heat, and Ẇ is net work. Ėheat, ĖD, Ėin and Ėout are the exergy input through heat, exergy destruction, exergy input and output, respectively. ṠF, ṠP and ṠGEN are the entropy of fuel, product and generation, respectively. Basic thermodynamic evaluation and conventional exergy analysis can be carried out by using these equations. Table 1 shows the equations of the exergy calculation for each component in the ejector refrigeration system. By using these equations, fuel exergy, product exergy and exergy destruction for overall system can be determined by the following equations, respectively Tab. 1: Evaluation of the exergy for each component Component Exergy of fuel (Ė F,XX) Exergy of product (Ė P, XX) Exergy destruction Generator (GE) ṁ 9(e 9- e 10) ṁ 1(e 2- e 3) Ė F,GE- Ė P,GE Condenser (CO) ṁ 7(e 7- e 8) ṁ 11(e 12- e 11) Ė F,CO- Ė P,CO Evaporator (EV) ṁ 4(e 6- e 5) ṁ 13(e 13- e 14) Ė F,EV- Ė P,EV Ejector (EJ) ṁ 3(e 3- e 7) ṁ 6(e 7- e 6) Ė F,EJ- Ė P,EJ Pump (PU) W PU ṁ 1(e 2- e 1) Ė F,PU- Ė P,PU Expansion valve (EXV) ṁ 4e 4 ṁ 4 e 5 Ė F, e 4, EXV - Ė P, e 4, EXV 26 Tab. 2: Evaluation of the entropy for each component Component Entropy of fuel (Ṡ F,XX) Entropy of product (Ṡ P,XX) Exergy destruction Generator (GE) ṁ 3s 3+ṁ 10s 10 ṁ 2s 2+ṁ 9s 9 (Ṡ P,GE - Ṡ F,GE)T 0 Condenser (CO) ṁ 8s 8+ṁ 12s 12 ṁ 7s 7+ṁ 11s 11 (Ṡ P,CO - Ṡ F,CO)T 0 Evaporator (EV) ṁ 6s 6+ṁ 14s 14 ṁ 5s 5+ṁ 13s 13 (Ṡ P,EV - Ṡ F,EV)T 0 Ejector (EJ) ṁ 7s 7 ṁ 3s 3+ṁ 6s 6 (Ṡ P,EJ - Ṡ F,EJ)T 0 Pump (PU) ṁ 2s 2 ṁ 1s 1 (Ṡ P,PU - Ṡ F,PU)T 0 Expansion valve (EXV) II.3. Ejector Model ṁ 5s 5 ṁ 4s 4 (Ṡ P,EXV - Ṡ F,EXV)T 0 An ejector can increase the pressure without using mechanical energy directly. Therefore, it is better to use an ejector than applying mechanical devices to increase the pressure such as compressor, pump, etc. since it may be safer and simpler technology. An ejector system is composed of four section including nozzle, mixing chamber, throat and supersonic diffuser. The high pressure steam known to be primary fluid coming from the generator expands and accelerates along with primary nozzle. The steam blows out with supersonic speed to achieve very low pressure at the end of the nozzle and in mixing chamber. Due to the pressure different, a lower-pressure vapor, secondary fluid, can be entrained within the mixing chamber.they are mixed in this stage. In the diffuser stage, the pressure is recovered to the condenser pressure. In order to facilitate evaluating of ejector model, the following assumption were made; 1. The ejector flow was assumed to be one dimensional and steady state. 2. The velocity variation of the fluid at the ejector inlet and outlet was neglected. 3. The pressure drop of the working fluid and the heat loss were neglected. 4. The pressure in mixing process in the ejector was considered to be constant with mass,

41 momentum and energy conservation. 5. The losses of the flow inside the ejector are all considered using isentropic efficiencies in the motive nozzle (n), in the mixing chamber (m) and in the diffuser (d). The isentropic efficiency of motive nozzle can be determined by following equation n = (h P,3 h P,3a ) (h P,3 h P,3s ) (10) where hp,3, hp,3a and hp,3s are the enthalpy of inlet flow, enthalpy of the exit flow, and enthalpy of the exit flow with isentropic expansion process. For a given nozzle efficiency, hp,3a can be determined. By applying the energy equations to the motive nozzle, the speed of nozzle at the exit can be calculated by using following equations 2 u p,3a 2 = h P,3 h P,3a (11) When Eqs 9 and 10 are solved, the following equation can be obtained u p,3a = 2 n (h P,3 h P,3s ) (12) In a similar way, the velocity of the exit flow can be determined for the secondary fluid u p,3a = 2 n (h P,6 h P,3s ) (13) However, this value can be neglected since it is very small compare to the primary flow velocity. After mixing the fluid, the mass conservation can be defined by m p + m s = m t (14) where mp, ms and mt represents the mass flow rates of primary fluid, secondary fluid and total mass flow rate of primary and secondary mixture. Although the actual results deviate slightly in ejector component, in order to facilitate the evaluation, the mass flow rate of the primary fluid is assumed to be 0.15 kg/s as a first approximation and that is very close the value of the primary fluids in literature. However, it should be defined by calculating of entrainment ratio. According to the momentum conservation, the following equation can be obtained m p u p,3a + m s u s,3a = m t u t,3m (15) The efficiency of the fluid mixture can be calculated by 2 m = u t,3m 2 u t,3s (16) when the energy conservation equation is applied to the mixture of the fluid. 2 m p (h P,3a + u p,3a 2 2) + m s (h s,3a + u s,3a 2) 2 = m t (h t,3m + u t,3m 2) (17) 27 Thus, the value of the enthalpy ht,3m and velocity ut,3m of the mixture can be determined. The first term can be neglected due to its comparatively low value. After the throat section, the mixture comes to supersonic diffuser section. The isentropic efficiency of the diffuser section can be expressed by d = (h 7s h t,3m ) (h 7 h t,3m ) (18) where h7 and h7s are the enthalpy of the exit flow and the enthalpy of the exit flow for isentropic compression process. II.4. Advanced Exergy Analysis All actual operations are irreversible due chemical reactions, heat transfer at finite temperature difference, mixture of matters, infinite expansion and frictions Petrakopoulou (2011). Conventional exergy analysis can describe components having high exergy destruction and its reasons, which component has irreversibility and its magnitude. However, it cannot explain the interaction among the component or estimate the actual improvement potential. Conventional approach to optimize the system may be wrong without taking into account of the interaction between the components particularly for complicated systems in which many components having interaction with each other. Therefore, advanced exergy analysis has been carried out. A detail exergy analysis, where the exergy destruction is split into several parts, is called to be advance exergy analysis. These parts are avoidable-unavoidable and endogenous-exogenous exergy destructions Morosuk and Tsatsaronis (2009). The value of the total exergy destruction can be calculated through an exergy balance for this component as follows E D,k = E F,k E P,k = T 0 S P,k = T 0 m ks P,k (19) where T 0, E F,k and E P,k shows reference temperature, exergetic fuel and exergetic product, respectively. Thus, the exergy efficiency can be found by following equations for kth component ε k = E P,k E F,k = 1 E D,k E F,k (20) where the exergy destruction E D,k can be split into AV UN avoidable E D,k and unavoidable E D,k parts, Tsatsaronis and Park (2002), Cziesla et al. (2006), Tsatsaronis and Morosuk (2008a, 2008b); Petrakopoulou et al. (2012). E D,k = E D,k UN AV + E (21) D,k This can provide more realistic approach to measure the potential improvement of the thermodynamic efficiency of the components. The unavoidable part of the exergy destruction cannot be shrunk due to the

42 technological limitations including material and manufacturing cost and availability. The other part of indicates avoidable destruction. EN The other approach is split into endogenous E D,k EX and exogenous E D,k parts as follows, Tsatsaronis (1999), Morosuk and Tsatsaronis (2006) E D,k = E D,k EN EX + E (22) D,k The endogenous and exogenous part of the exergy destruction provides the reason of the exergy destruction in the component caused by the EN component itself or by the other components. E D,k part can be obtained when the all irreversibilities occurs in the kth component while the other component is assumed to be ideal and has no irreversibility with having its current efficiency. On the EX other hand, E D,k part occurs within the kth component due to the irreversibilities in the other components. For each case, the power output of the overall system is kept constant and equal to actual case. Thus, the exogenous destruction is the remaining part of the total exergy destruction in kth component and can be calculated by subtracting the endogenous exergy destruction from the total exergy destruction Petrakopoulou (2011). In order to calculate the unavoidable exergy destruction, each component should be considered isolated and separated from the system. The exergy UN destruction rate per unit product exergy (E D E P ) k can be calculated by assuming the system operating with high efficiency and low losses. Thus, the unavoidable exergy destruction for kth compoentn by using the real case product exergy rate can be expressed by Petrakopoulou (2011) UN = E E D,k P,k real (E D,k E P,k ) UN (23) when the unavoidable exergy destruction is known, the avoidable part can be calculated by Eq 20. The unavoidable endogenous and exogenous, and avoidable endogenous and exogenous part of the exergy destructions are expressed respectively, by Tsatsaronis and Morosuk (2008a, 2008b) UN,EN = E E D,k E D,k P,k EN (E D,k E P,k UN,EX = E UN E E D,k D,k AV,EN = E EN E E D,k D,k AV,EX = E AV E D,k ) UN (24) UN,EN D,k UN,EN D,k AV,EN D,k (25) (26) (27) In order to evaluate the advanced exergy analysis, modified exergy efficiency can be described by Tsatsaronis and Morosuk (2008a, 2008b) In this study, the exergy destruction rate split into avoidable-unavoidable parts were evaluated. The endogenous- exogenous exergy destructions and combined two splitting approaches were not considered. In a further study, these parts will be evaluated. In order to facilitate the analyses some assumptions are made; 1. Each process in the system is assumed to be steady state. 2. Potential and kinetic energy variations are neglected. 3. The heat transfer to/from ambient and pressure drops in the pipes are neglected. 4. The working fluid at the inlet of pump is assumed as saturated liquid. 5. The dead state pressure P0 and temperature T0 are considered to be kpa and 20 C, respectively. Tab. 3: Input values to the system Parameters Values Pump isentropic efficiency 85% Evaporator temperature 4 C Condenser temperature 35 C Generator temperature 95 C Cooling capacity 20 kw Mass flow rate of exhausted gas 44.5 kg/s Mass flow rate of refrigerant at the inlet of 0.15 kg/s pump Inlet temperature to generator of exhausted 277 C gas Inlet temperature to condenser of cooling 27 C water Outlet temperature from condenser of 32 C cooling water Inlet temperature to evaporator of water 15 C Outlet temperature from evaporator of water 10 C Ambient pressure kpa Ambient temperature 25 C III. Results and Discussions The calculation program was written in Engineering Equation Solver (EES). Table 4 expresses the parameters of the components in the ejector system for the real, ideal and unavoidable conditions. The calculated thermodynamic data of the ejector refrigeration cycle at real operation case, at ideal operation case and at the unavoidable operation conditions for the working fluid R142b are shown in Tables 5-7, respectively. Tab. 4: The parameters used for real, ideal cycles, and the cycle for the unavoidable exergy destruction Component Parameter Real Ideal Unavoidable Pump (PU) ɳ P Generator (GE) ΔT GE 7.5 C C Ejector (EJ) ɳ n Ejector (EJ) ɳ m Ejector (EJ) ɳ d Condenser (CO) ΔT CO 3 C C Expansion valve (EXV) - Isenthalpic Isentropic Isenthalpic Evaporator (EV) ΔT EV 2 C C ε modified = E P,k (E F,k E D,k UN E AV,EX ) (28) D,k 28

43 Tab. 5: Data of the ejector refrigeration cycle at real operation case for the working fluid R142b m P T h s e Loc. Subs. (kg/s) (kpa) ( C) (kj/kg) (kj/kg.k) (kj/kg) 1 R142b R142b R142b R142b R142b R142b R142b R142b R142b Air Air Air Water Water Water Water Water Tab. 6: Data of the ejector refrigeration cycle at ideal operation case for the working fluid R142b m P T h s e Loc. (kg/s) (kpa) ( C) (kj/kg) (kj/kg.k) (kj/kg) Tab. 7: Data of the ejector refrigeration at the unavoidable operation conditionsfor the working fluid R142b Loc. Loc. m (kg/s) P (kpa) T ( C) h (kj/kg) s (kj/kg.k) e (kj/kg) 1UN UN UN UN UN UN UN UN III.1. Evaluation of Convensional Exergy Analysis In recognition of the fuel exergy of the overall system, it was provided by generator and pump. The fuel exergy of generator is provided by the exhausted gas from the rotary kiln and that of pump is provided from the electrical work. 29 Table 8 indicates the results of the conventional exergy analysis of ejector refrigeration system in terms of entropy balance and destruction. The major part of the exergy destruction is arisen from the generator system (48.9%) and followed by the ejector system (39.1%). The pump, expansion valve and evaporator systems have the lowest exergy destruction. The total exergy destruction rate of these components generates about 4.73% of the overall system exergy destruction. In terms of the exergetic efficiency, the expansion valve and pump show the best performance to be about 91 and 90.4 %, respectively. On the other hand, the lowest efficiency was obtained in the generator (34.2%) and followed by the condenser with an exergy efficiency of 38.3%. The remaining components show relatively better efficiency. Although the exergy destruction rate of the ejector forms the major part, its exergetic efficiency is quite preferable. It is observed that the main attention to improve the overall system performance should be paid to the generator system since the exergy destruction is mostly resulted from this system. The overall system shows a low exergetic efficiency of 5.58%. This is arisen from that most of the fuel exergy is destroyed due to limited use of the ejector refrigeration system. Tab. 8: The conventional exergy analysis results S F Component (kw) (kw) (kw) ε (%) Pump (PU) Generator (GE) Ejector (EJ) Condenser (CO) Expansion valve (EXV) Evaporator (EVA) Overall System III.2. Evaluation of Advanced Exergy Analysis Table 8 shows the advanced exergy analysis results where the exegy destruction rates are split into unavoidable-avoidable parts for each component. The second column indicates the exergy destruction for each component. The third and fourth columns show the unavoidable and avoidable parts which can provide the potential improvement of the system components. Tab. 9: The advanced exergy analysis results Splitting the exergy destruction Component Ė D (kw) Ė UN (kw) Ė AV (kw) Pump (PU) Generator (GE) Ejector (EJ) Condenser (CO) Expansion valve (EXV) Evaporator (EVA) Overall System Figure 2 shows these exergy destructions as percentages in the total destruction rate for each S P Ė D

44 component. Although the generator has the largest value of the exergy destruction; most of the destruction is composed of unavoidable part. Namely, the improvement potential for itself and for overall system is very weak. The expansion valve and evaporator show very similar characteristic and have unavoidable exergy destruction rates of about 70%. Moreover, they have very low exergy destruction. On the other hand, the ejector is the most promising component to improve the overall system performance since it has the largest avoidable exergy destruction. Moreover, about 65% of the exergy destruction can be avoided. About 31% of the total exergy destruction is in the part of avoidable exergy destruction for overall system and the ejector has the largest share of overall system. Although the conventional exergy analysis results indicate that the largest exergy destruction is in the generator and may be considered for the system improvement, the advanced exergy analysis results show that the main attention to improve the overall system performance should be paid to the ejector system since the avoidable part of the exergy destruction is mostly resulted from this system. Fig. 2: Exergy destruction rate for avoidable and unavoidable parts IV. Conclusions Conventional and advanced exergy analyses are carried out for ejector refrigeration cycles. Moreover, the exergy destruction was split into avoidable-unavoidable parts in order to determine the improvement potential of each component in the system. The following part summarizes the study. 1. The largest exergy destruction occurs in generator, accounting 48.9% of total exergy destruction and followed by ejector with ratio of 39.13%. 2. The exergetic efficiency of the overall system is about 5.58%. 3. The largest share of total avoidable exergy destruction occurs in ejector and that is about half of total avoidable part of overall system (82.94%). Therefore, it is important to concentrate on this component to improve overall system performance. 4. About 69 % of the total destruction is falling into the part of unavoidable exergy destruction. Namely, the system has potential to improve by reducing the avoidable part of 31%. Nomenclature COP : The coefficient of the performance (-) 30 E : Exergy (kw) e : Specific exergy (kj.kg-1) h : Specific enthalpy (kj.kg-1) ṁ : Mass flow rate (kg.s-1) P : Pressure (kpa) Q : Heat load (kw) s : Specific entropy (kj.kg-1 K-1) T : Temperature (C) W : Work (kw) Greek letters ε : Exergetic efficiency Superscripts AV : Avoidable EN : Endogenous EX : Exogenous UN : Unavoidable Subscripts a : Actual CO : Condenser D : Destruction EJ : Ejector EXV : Expansion valve EV : Evaporator F : Fuel GE : Generator In : Inlet k : The k-th component n : nozzle out : Outlet P : Product PU : Pump s : Isentropic tot : Total XX : Component representative References Cai D.H., He G.G., Tian Q.Q., Tang W.E., Thermodynamic analysis of a novel aircooled non-adiabatic absorption refrigeration cycle driven by low grade energy, Energy Convers. Manage., 86, (2014). Chen J., Havtun H., Palm B., Conventional and advanced exergy analysis of an ejector refrigeration system, Applied Energy, 144, (2015). Chen J., Havtun H., Palm B., Parametric analysis of ejector working characteristics in the refrigeration system, Appl. Therm. Eng., 69, (2014). Cimsit C., Ozturk I.T., Analysis of compression-absorption cascade refrigraiton cycles, Applied Thermal Engineering, 40, (2012). Cziesla F., Tsatsaronis G., Gao Z., Avoidable thermodynamic inefficiencies and costs in an externally fired combined cycle power plant, Energy, 31, (2006). Dahmani A., Aidoun Z., Galanis N., Optimum design of ejector refrigeration systems with environmentally benign fluids., Int. J. Therm. Sci., 50, (2011).

45 Erbay Z., Hepbasli A., Application of conventional and advanced exergy analyses to evaluate the performance of a ground-source heat pump (GSHP) dryer used in food drying, Energy Conversion and Management, 78, (2014). Gong S., Goni B.K., Parametric study of an absorption refrigeration machine using advanced exergy analysis, Energy, 76, (2014). Huang B.J., Chang J.M., Wang C.P., Petrenko V.A., A 1-D analysis of ejector performance, Int. J. Refrig.,22, (1999). Huang B.J., Jiang C.B., Hu F.L., Ejector performance characteristics and design analysis of jet refrigeration system, Trans ASME, 107, (1985). Kairouani L., Nehdi E., Cooling performance and energy saving of compressioneabsorption refrigeration system assisted by geothermal energy, Applied Thermal Engineering, 26, (2006). Kaska O., Energy and exergy analysis of an organic Rankine for power generation from waste heat recovery in steel industry, Energy Conversion and Management, 77, (2014). Keenan J.H., Neumann E.P., Lustwerk F., An investigation of ejector design by analysis and experiment, J. Appl. Mech. Trans. ASME, 72, (1950). McGovern R.K., Narayan G.P., Lienhard V.J.H., Analysis of reversible ejectors and definition of an ejector efficiency, Int. J. Therm. Sci., 54, (2012). Morosuk T., Tsatsaronis G., A new approach to the exergy analysis of absorption refrigeration machines., Energy, 33, (2008). Morosuk T., Tsatsaronis G., Advanced exergetic evaluation of refrigenration machines using different working fluids, Energy, 34, (2009). Morosuk T., Tsatsaronis G., Splitting the exergy destruction into endogenous and exogenous parts application to refrigeration machines. In: Frangopoulos C, Rakopoulos C, Tsatsaronis G, editors. Proceedings of the 19th international conference on efficiency, cost, optimization, simulation and environmental impact of energy systems, July 12 14; 2006, National Technical University of Athens, Greece, 1, (2006). Morosuk T., Tsatsaronis G., Zhang C.. Conventional thermodynamic and advanced exergetic analysis of a refrigeration machine using a Voorhees compression process, Energy Convers. Manage., 60, (2012). Petrakopoulou F., Tsatsaronis G., Morosuk T., Carassai A., Conventional and advanced exergetic 31 analyses applied to a combined cycle power plant, Energy, 41, (2012). Pridasawas W., Lundqvist P., An exergy analysis of a solar-driven ejector refrigeration system, Sol. Energy, 76, (2004). Tarique S., Siddiqui M.A., Performance and economic study of the combined absorption/compression heat pump, Energy Conversion Management, 40, (1999). Tsatsaronis G., Recent developments in exergy analysis and exergoeconomics, Int. J. Exergy, 5, (2014). Tsatsaronis G., Strengths and limitations of exergy analysis, In: Bejan A, Mamut E, editors. Thermodynamic optimization of complex energy systems. Dordrecht: Kluwer Academic Publishers, (1999). Tsatsaronis G., Morosuk T., A general exergy-based method for combining a cost analysis with an environmental impact analysis, Part I. Theoretical development, in: Proceedings of the ASME IMECE, Boston, Massachusetts, USA (2008a). Tsatsaronis G., Morosuk T., A general exergy-based method for combining a cost analysis with an environmental impact analysis. Part II. Application to a cogeneration system, in: Proceedings of the ASME IMECE, Boston, Massachusetts, USA (2008b). Tsatsaronis G., Park M.H., On avoidable and unavoidable exergy destructions and investment costs in thermal systems, Energy Convers. Manage., 43, (2002). Ustaoglu A., Alptekin M., Akay M.E., Energy and exergy analysis of a wet type rotary kiln and utilization of waste heat powered ORC, Applied Thermal Engineering, Submitted (2016). Yan G., Chen J., Yu J., Energy and exergy analysis of a new ejector enhanced auto-cascade refrigeration cycle, Energy Conversion and Management, 105, (2015). Yang X., Zheng N., Zhao L., Deng S., Li H., Yu Z., Analysis of a novel combined power and ejector-refrigeration cycle, Energy Conversion and Management, 108, (2016). Zhao L., Yang X., Deng S., Li H., Yu Z., Performance analysis of the ejector-expansion refrigeration cycle using zeotropic mixtures, International journal of refrigeration, 57, (2015). Zhu Y., Cai W., Wen C., Li Y., Shock circle model for ejector performance evaluation., Energy Convers. Manage., 48, (2014).

46 Thermodynamic Evaluation of Absorption-Compression Cascade Refrigeration Cycles for Advanced Exergy Analysis Mustafa Alptekin 1*, Abid Ustaoglu 2, Mehmet Emin Akay 3, Resat Selbas 1 1 Suleyman Demirel University, Faculty of Engineering, Department of Energy Systems Engineering, Isparta, 32260, Turkey 2 Bartin University, Faculty of Engineering, Mechanical Engineering Department, Bartin, 74100, Turkey 3 Karabuk University, Mechanical Engineering Department, Karabuk, 78050, Turkey. * mustafaalptekin@sdu.edu.tr Abstract In this paper, thermodynamic assessment of waste heat powered absorption-compression cascade refrigeration system is performed using advanced exergy method by using the exhausted heat from the chimney of rotary kiln. NH3-H2O fluid pairs and R134a are selected as working fluids in the absorption-compression cascade refrigeration cycles. The heat energy of generator is provided from waste heat of rotary kiln. Conventional and advanced exegy analyses are carried out to the system. The exergy destruction was split into endogenous-exogenous and avoidable-unavoidable parts in order to reveal interdependency within the components and determine the improvement potential of each component in the system. The largest exergy destruction occurs in the generator, accounting 63% of total exergy destruction. Therefore, it is important to concentrate on this component to improve overall system performance. In respect of overall system destruction, about 53.1% of the total destruction is caused by the system components themselves. Furthermore, about 60% of the total destruction is falling into the part of unavoidable exergy destruction. Namely, the system has potential to improve by reducing the avoidable part of 40%. Keywords: Advanced exergy analysis, absorption-compression cascade refrigeration system, refrigerant pairs I. Introduction Cement industry is one of the most energy consuming industries in the world. The energy consumption rate reaches 12-15% of total energy consumption in industry. The rotary kiln has biggest share in terms of energy consumption in a cement factory and has great heat loss which causes significant reduction in efficiency. This reduction becomes excessive for a cement factory using the wet method. In the progress of clinker production, about 33% of total energy is exhausted from the chimney of rotary kiln apart from the heat loss through the wall of rotary kiln Ustaoglu et al. (2016). Therefore, reutilization of the exhausted gas becomes substantial due to the great waste heat. Research and development studies for cooling systems have been gradually increased recently. The conventional refrigeration systems consume a great amount of energy through the compressor. Absorption systems are promising technologies in the refrigeration applications they can be operated by providing of a low level of thermal energy, including solar energy, biomass, geothermal and waste heat energy in the industrial process, compared to compression refrigeration systems Cai et al. (2014). Thus, the utilization of waste heat through production process and renewable energy even if low level can be provided in a useful form of energy thereby reducing amount of carbon emission and environmental pollutions. The compression-absorptions refrigeration cycles are another alternative among the refrigeration systems. Even tough more complex structure, these cycles can provide electrical energy save compering to vapor compression cycles and utilize advantages of absorption and vapor compression cycles by providing the use of electricity and heat for refrigeration. These type cycles also can be categorized into two group including combined refrigeration cycles and cascade refrigeration cycles. In the case of combined cycles, the compression rates of vapour compression and absorption sections are same with each other and that of the combined cycles. However, absorption and vapour compression cycles are connected in serial form for cascade systems Cimsit and Ozturk. (2014). There have been done many studies about the compression-absorption refrigeration cycles. Tarique and Siddiqui (1999) compared a conventional vapor compression cycle using ammonia with a combined refrigeration cycle using NH3 and NaSCN in the identical conditions in terms of performance and economical aspect. İt is found that the capital and operation costs were reduced significantly for the case of combined cycle. Kairouani et al. (2011) carried out a performance evaluation of compression-absorption refrigeration cascade cycle for NH3-H2O fluid pair for absorption section as the R717, R22 and R134a were used for the vapour compression section. The results showed that performance coefficient of the cycle have improvement of %37-54 comparing to a conventional 32

47 vapour compression cycle in which the same working fluids were used in the same operation conditions. Cimsit and Ozturk (2012) used LiBr-H2O and NH3-H2O pairs in the absorption section as R134a, R-410A and NH3 in the vapour compression section of a cascade refrigeration cycle. They showed that the cycle using LiBr-H2O fluid pair could show better performance coefficient compare to the cycle using NH3-H2O fluid pair for all working fluid of the vapour compression section. The second law efficiency evaluation is important tool to determine the location, magnitude and source of the exergy destruction. Rezayan and Behbahaninia (2011) carried out a thermoeconomic optimization of the system and applied exergy analysis to a CO2/NH3 cascade refrigeration cycle. The results showed that regarding to the exergy analysis on the optimized system, the highest exergy destruction occurred in the condenser (33.5%) as the lowest exergy destruction became in the expansion valve of the carbon dioxide circuit (5.2%). Yan et al. (2015) proposed a new ejector enhanced auto-cascade refrigeration cycle using R134a/R23 and compared it with a conventional auto-cascade refrigeration cycle. Their cycle achieved % better efficiency than a basic cycle at the same operation conditions as the ejector achieved pressure ratio lifts of They found that the highest exergy destruction occurs in the compressor, the condenser, cascade condenser, expansion valve, ejector and evaporator, respectively. Cimsit et al. (2015) carried out thermoeconomic optimization of LiBr/H2O-R134a compression-absorption cascade refrigeration cycle. The analysis results showed that the evaporator equipage and solution heat exchanger should be designed carefully according to the exergoeconomic factor values. Apart from the conventional analysis method, the exergy method is an important tool that can show the useful work that can be generated through the process. However, it cannot explain the interaction among the component or estimate the actual improvement potential. Conventional approach for optimizing of the system may be wrong without taking into account of the interaction between the components particularly for complicated systems in which many components having interaction with each other. In the case of the ejector refrigeration cycles, the operation parameters of the ejector depend on itself and other components Chen et al. (2014). A recent developed technique, the advanced exergy analysis by splitting the exergy destruction into avoidable-unavoidable and endogenous-exogenous part can enable us to evaluate the system in more detail and investigate the capacity of the improvement Tsatsaronis (1999). The advanced exergy analysis of different refrigeration systems, including vapor compression refrigeration systems Morosuk and Tsatsaronis (2009), Morosuk et al. (2012), absorption refrigeration systems Gong and Goni (2014), Morosuk T, Tsatsaronis (2008) ejector refrigeration system Chen et al. (2015) and heat 33 pump Erbay and Hepbasli (2014) have been carried out. To the best of author s knowledge, there is no study about evaluation of an absorption-compression cascade refrigeration cycle in terms of advanced exergy analysis. In a previous study, energy and exergy analysis of a wet type rotary kiln were carried and recovery capacity of waste heat was evaluated by using an organic Rankine cycle Ustaoglu et al. (2016). The results showed that a great amount of heat energy of 30.5 MW is exhausted from the chimney of rotary kiln. In this study, the waste heat from the rotary kiln was evaluated for an absorption-compression cascade refrigeration cycle and advanced exergy analysis was carried out to determine endogenous, exogenous, avoidable and unavoidable exergy destruction rates in order to determine the improvement potentials of both the components and the overall system along with the interactions within components. II. Material and method II.1. System description Figure 1 shows the schematic view of the absorption compression cascade refrigeration cycle. The below and upper cycles are vapor compression and absorption refrigeration cycles, respectively. R134a is used as working fluid in the case of vapor compression cycle. LiBr-H2O and NH3-H2O pairs are selected for absorption refrigeration cycle. Each state of the working fluids is represented in the point as seen in figure. The working principle of this system can be expressed as follows: Vaporized working fluid in evaporator enters to the compressor to be compressed to point 2. The compressed and slightly heated fluid releases its heat to the other working fluid having lower temperature in the upper cycle, and leaves the heat exchanger. In the expansion valve, the temperature and pressure of the working fluid decreases to evaporator level. The working fluid entering evaporator absorbs the heat to be vaporized and again enter to the compressor. Thus, the below cycle is completed. As the refrigerant is circulated in points 11, 12, 13 and 14, fluid pair is circulated in points 5, 6, 7, 8, 9 and 10. The vaporized refrigerant by receiving heat from the below cycle enters from point 14 as weak solution enters from point 10 to the absorber where they dissolves and reacts with each other to form strong solution, fluid pairs. This strong solution is pumped to the solution heat exchanger and receives some of the heat of the fluid pairs coming from the generator and enters to generator where the heat of the exhausted gas is transfer to the fluid pair. The fluid having low boiling point in the strong solution is vaporized and enters to the condenser. However, all of that fluid cannot be vaporized. The weak solution leaving from the generator releases some of its heat in the heat exchanger and enters to the expansion

48 valve where its heat and pressure decreases to the absorber level. The fluid leaving generator from point 11 is condensed in condenser and enters to the expansion valve where its pressure and temperature decreases then the fluid enters to the heat exchanger to be vaporized by heat from the below cycle. Thus, all cycle is completed. frictions Petrakopoulou (2011). Conventional exergy analysis can describe components having high exergy destruction and its reasons, which component has irreversibility and its magnitude. However, it cannot explain the interaction among the component or estimate the actual improvement potential. Conventional approach to optimize the system may be wrong without taking into account of the interaction between the components particularly for complicated systems in which many components having interaction with each other. Therefore, advanced exergy analysis has been carried out. A detail exergy analysis, where the exergy destruction is split into several parts, is called to be advance exergy analysis. These parts are avoidable and unavoidable, or endogenous and exogenous exergy destructions Morosuk and Tsatsaronis (2009). The value of the total exergy destruction can be calculated through an exergy balance for this component as follows E D,k = E F,k E P,k = T 0 S P,k (4) where T 0, E F,k and E P,k shows reference temperature, exergetic fuel and exergetic product, respectively. Thus, the exergy efficiency can be found by following equations for kth component ε k = E P,k E F,k = 1 E D,k E F,k (5) Fig. 1: Schematic view of absorption-compression cascade refrigeration cycle II.2. Thermodynamic Evaluation The absorption- compression cascade refrigeration cycle is considered as continuous flow open-system and modeled based on the first and second laws of thermodynamics, and these laws are applied to each component in the system. In the continuous flow open systems, the mass and energy of control volume are stable. General equations of mass, energy and exergy balances by ignoring the kinetic and potential energy variations can be expressed m in = m out (1) Q + W = m out h out m outh out (2) E heat + W + E D = E out E in (3) where ṁ is mass flow rate, h is enthalpy, Q is heat, and Ẇ is net work. Basic thermodynamic evaluation and conventional exergy analysis can be carried out by using these equations. All actual operations are irreversible due chemical reactions, heat transfer at finite temperature difference, mixture of matters, infinite expansion and Exergy efficiency of the overall system can be determined by ε overall = COP COP carnot (6) where the coefficient of performance COP can be calculated by COP = Q EV/(Q GE + W CO + W PU) (7) where Q GE = m 8h 8 + m 11h 11 m 7h 7 (8) W CO = m R(h 2 h 1 ) (9) W PU = m 5(h 6 h 5 ) (10) ṁ R indicates the mass flow rate of the R134a. The coefficient of performance for Carnot can be obtained by COP carnot = ((T GE T AB )T EV ) ((T CO T EV )T GE ) (11) where T indicates the temperature, the subscripts GE, AB and EV represent generator, absorber and evaporator. In the evaluation, the temperature should be in Kelvin. The exergy destruction E D,k can be split into 34

49 AV UN avoidable E D,k and unavoidable E D,k parts, Tsatsaronis and Park (2002), Cziesla et al. (2006), Tsatsaronis and Morosuk (2008a, 2008b), Petrakopoulou et al. (2012). E D,k = E D,k UN AV + E (12) D,k This can provide more realistic approach to measure the potential improvement of the thermodynamic efficiency of the components. The unavoidable part of the exergy destruction cannot be shrunk due to the technological limitations including material and manufacturing cost and availability. The other part of indicates avoidable destruction. EN The other approach is split into endogenous E D,k EX and exogenous E D,k parts as follows, Tsatsaronis (1999), Morosuk and Tsatsaronis (2006) E D,k = E D,k EN EX + E (13) D,k The endogenous and exogenous parts of the exergy destruction provide the reason of the exergy destruction in the component caused by the EN component itself or by the other components. E D,k part can be obtained when the all irreversibilities occurs in the kth component while the other component is assumed to be ideal and has no irreversibility with having its current efficiency. On the EX other hand, E D,k part occurs within the kth component due to the irreversibilities in the other components. For each case, the power output of the overall system is kept constant and equal to actual case. Thus, the exogenous destruction is the remaining part of the total exergy destruction in kth component and can be calculated by subtracting the endogenous exergy destruction from the total exergy destruction Petrakopoulou (2011). In order to calculate the unavoidable exergy destruction, each component should be considered isolated and separated from the system. The exergy UN destruction rate per unit product exergy (E D E P ) k can be calculated by assuming the system operating with high efficiency and low losses. Thus, the unavoidable exergy destruction for kth component by using the real case product exergy rate can be expressed by Petrakopoulou (2011) UN = E E D,k P,k real (E D,k E P,k ) UN (14) When the unavoidable exergy destruction is known, the avoidable part can be calculated by Eq 12. The unavoidable endogenous and exogenous, and avoidable endogenous and exogenous part of the exergy destructions are expressed respectively, by Tsatsaronis and Morosuk (2008a, 2008b) AV,EN = E EN E E D,k E D,k D,k AV,EX = E AV E D,k UN,EN D,k AV,EN D,k (17) (18) In order to evaluate the advanced exergy analysis, modified exergy efficiency can be described by Tsatsaronis and Morosuk (2008a, 2008b) ε modified = E P,k (E F,k E D,k UN E AV,EX ) (19) In order to facilitate the analyses some assumptions are made; 1. Each process in the system is assumed to be steady state. 2. Potential and kinetic energy variations are neglected. 3. The heat transfer to/from ambient and pressure drops in the pipes are neglected. 4. The fluid pairs in absorption refrigeration system at the inlet of pump is assumed as saturated liquid. III. Results and Discussions The calculation program was written in Engineering Equation Solver (EES). Table 1 shows the input parameters in order to carry out the energetic and exergetic analysis of the cycle. Table 2 expresses the parameters of the components in the absorption-compression cascade refrigeration system for the real, ideal and unavoidable conditions. The calculated thermodynamic data of the absorption-compression cascade refrigeration at real operation case, at ideal operation case and at the unavoidable operation conditions for the working fluid R142b are shown in Tables 3-5, respectively. Tab. 1: Input parameters to the system Parameters Values Pump isentropic efficiency 85% Compressor isentropic efficiency 85% Evaporator temperature -20 C Condenser temperature 45 C Generator temperature 100 C Absorber temperature 43 C Heat exchanger temperature for bottom cycle 5 C Heat exchanger temperature for upper cycle 20 C Inlet temperature to generator of exhausted gas 277 C Inlet temperature to condenser of cooling water 25 C Outlet temperature from condenser of cooling 30 C water Inlet temperature to evaporator of water 30 C Outlet temperature from evaporator of water 25 C Inlet temperature to absorber of cooling water 30 C Outlet temperature from absorber of cooling water 35 C Subcooling and superheating temperature 5 C Mass flow rate of exhausted gas 44.5 kg/s Effectiveness of solution heat exchanger 50% Cooling capacity 20 kw Ambient pressure kPa Ambient temperature 25 C D,k UN,EN = E E D,k P,k EN (E D,k E P,k ) UN (15) UN,EX = E UN E E D,k D,k UN,EN D,k (16) 35

50 Tab. 2: The parameters used for real, ideal cycles, and the cycle for the unavoidable exergy destruction Component Parameter Real Ideal Unavoidable Evaporator ΔT EV 10 C C Compressor η C Heat exchanger ΔT HE 5 C C Expansion valves - Isenthalpic Isentropic Isenthalpic Condenser ΔT CO 10 C C Absorber ΔT AB 3 C C Pump Solution heat Exchanger β Generator ΔT GE 7 C C Tab. 3: Data of the ejector refrigeration cycle at real operation case Loc. Subs. m P h s e T ( C) (kg/s) (kpa) (kj/kg) (kj/kg.k) (kj/kg) 1 R134a R134a R134a R134a R134a NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH NH NH NH Water Air Air Air Water Water Water Water Water Water Water Tab. 4: Data of the ejector refrigeration cycle at ideal operation case Loc. Subs. m P h s e T ( C) (kg/s) (kpa) (kj/kg) (kj/kg.k) (kj/kg) 1 R134a R134a R134a R134a NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH NH NH NH Tab. 5: Data of the ejector refrigeration at the unavoidable operation conditions Loc. Substance m P T ( C) h s e (kg/s) (kpa) (kj/kg) (kj/kg.k) (kj/kg) 1 R134a R134a R134a R134a NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH 3 H 2O NH NH NH NH III.1. Evaluation of Conventional Exergy Analysis Table 6 shows the results of the conventional exergy analysis of absorption-compression cascade refrigeration system. The generator shows the largest exergy destruction on the cycle which is more than one half of the overall exergy destruction rate of system (%63.86), and followed by evaporator (10.8%) and solution heat exchanger (6%). The lowest exergy destructions are observed in pump, and followed by the expansion valve 2, 1 and 3, respectively. The exergy efficiency of pump is the highest in the system and followed by Expansion valve 3, 1 and 2. These are arisen from the relatively low exergy destructions of these systems compare to their fuel and product exergy. The lowest exergy efficiency is obtained in evaporator with about 23% and followed by condenser, generator and heat exchanger with 28%, 32% and 42%, respectively. The other components show quite preferable performances. The system COP and COPcarnot were calculated to be and , respectively. As for the overall system, the efficiency, which can be calculated by Eq 6, is about 46%. The fuel exergy of overall system is composed of the fuel exergy of generator, compressor and pump as the exergy product of overall system indicates that of evaporator. Tab. 6: The conventional exergy analysis results Component Ė F (kw) Ė P (kw) Ė D (kw) ε (%) Compressor (COMP) Heat exchanger (HE) Expansion valve Evaporator (EV) Condenser (CO) Expansion valve Absorber (AB) Pump (PU) E Solution heat exchanger(she) Generator (GE) Expansion valve Overall System III.2. Evaluation of Advanced Exergy Analysis Table 7 shows the advanced exergy analysis results where the exegy destruction rates are split into endogenous-exogenous and unavailable-available 36

51 parts for each component. The second column indicates the exergy destruction for each component. The third and fourth columns show the endogenous EN and exogenous parts, E D,k ande D,k EX, those state how much exergy destruction caused by components itself and how much exergy destruction caused by the other components in the cycle. Tab. 7: The advanced exergy analysis results Component Compressor (COMP) Heat exchanger (HE) Expansion valve 1 Evaporator (EV) Condenser (CO) Expansion valve 2 Ė D (kw Splitting the exergy destruction Ė EN (kw) Ė EX (kw) Ė UN (kw) ĖAV (kw) Absorber (AB) Pump (PU) Solution heat exchanger (SHE) Generator (GE) Expansion valve 3 Overall System 1.02E E E E E In Figure 2, the endogenous exogenous parts were shown in terms of percentages in the total exergy destruction. In the case of evaporator, the exergy destruction is mostly caused by the component itself. The endogenous exergy destruction is more dominant for the compressor and generator compare to the exogenous. The remaining components exergy destructions are mostly caused by the rest of the components. For the case of expansion valve 3, pump and absorber, the exogenous exergy destruction rates are over the 90% and followed by compressor and solution heat exchanger with about 82% and 77%, respectively. The heat exchanger, expansion valve 1 and expansion valve 2 show similar behaviors and have exogenous exergy destruction rates of about 65%. As regards to the overall system, the endogenous part of the exergy destruction is comparable with the exogenous exergy destruction. Figure 3 shows the avoidable and unavoidable parts of the total exergy destruction for each component. These results state the improvement potential of each system with more realistic approach. Although 42.3% of total exergy destruction is avoidable part in generator, it shows the largest avoidable part of the total exergy destruction and that is more than two third of avoidable part of overall system as seen in Table 7. Therefore, it is important to concentrate on this component to improve overall system performance. The second biggest share of the avoidable part is observed in the solution heat exchanger and 75% of the exergy destruction can be avoided. The third largest share is the absorber and 81% of the total exergy destruction is avoidable part for that component. Although the second largest exergy destruction occurs in the evaporator, all exergy destruction is unavoidable part. In the case of condenser, it is 93.2%. The highest rates of avoidable parts are observed expansion valve 3, absorber, and pump and solution heat exchanger, respectively. Together, they contribute 25% of the avoidable part of overall system. However, some components including expansion valves and pump have very small share. As for overall system, about 60% of the total exergy destruction is unavoidable part while the 40% of the exergy destruction can be eliminated. Fig. 2: Exergy destruction rate for exogenous and endogenous parts Fig. 3: Exergy destruction rate for avoidable and unavoidable parts 37

52 IV. Conclusions Conventional and advanced exegy analyses are carried out for absorption-compression cascade refrigeration cycles. Moreover, the exergy destruction was split into endogenous-exogenous and avoidable-unavoidable parts in order to reveal interdependency within the components and determine the improvement potential of each component in the system. The following part summarizes the study. 1. The largest exergy destruction occurs in generator, accounting 63% of total exergy destruction and followed by evaporator and solution heat exchanger with ratios of 10.8% and 6.6%, respectively. 2. The exergetic efficiency of the overall system is about 46%. 3. The endogenous exergy destruction is more dominant for the compressor and generator compare to the exogenous part of the exergy destruction. 4. In the case of the expansion valve 3, pump and absorber, the exergy destruction can be largely reduced by improvement of other remaining components. 5. The largest share of total avoidable exergy destruction occurs in generator and that is more than two third of avoidable part of overall system (68.23%). Therefore, it is important to concentrate on this component to improve overall system performance. 6. Although the second largest exergy destruction occurs in the evaporator, the 100% of exergy destruction is caused by unavoidable exergy destruction part. 7. In respect of overall system destruction, about 53.1% of the total destruction is caused by the system components themselves. Furthermore, about 60% of the total destruction is falling into the part of unavoidable exergy destruction. Namely, the system has potential to improve by reducing the avoidable part of 40%. Nomenclature E : Exergy (m.s-1) h : Specific enthalpy (kj.kg-1) ṁ : Mass flow rate (kg.s-1) P : Pressure (kpa) Q : Heat load (kw) s : Specific entropy (kj.kg-1 K-1) T : Temperature (C) W : Work (kw) Greek letters ε : Exergetic efficiency Superscripts AV : Avoidable EN : Endogenous EX : Exogenous UN : Unavoidable Subscripts D : Destruction F : Fuel 38 k : The k-th component P : Product Abreviations AB : Absorber EV : Evaporator COMP : Compressor HE : Heat exchanger EXV : Expansion valve EXP V : Expansion valve PU : Pump SHE : Solution heat exchanger GE : Generator References Cai D.H., He G.G., Tian Q.Q., Tang W.E., Thermodynamic analysis of a novel aircooled non-adiabatic absorption refrigeration cycle driven by low grade energy, Energy Convers. Manage., 86, (2014). Cimsit C., Ozturk I.T., Analysis of compression-absorption cascade refrigraiton cycles, Applied Thermal Engineering, 40, (2012). Cimsit C., Ozturk I.T., Hosoz M., Second law based thermodynamic analysis of compression-absorption cascade refrigeration cycles, J. of Thermal Science and Technology, 34, 9-18 (2014). Chen J., Havtun H., Palm B., Parametric analysis of ejector working characteristics in the refrigeration system, Appl. Therm. Eng., 69, (2014). Cziesla F., Tsatsaronis G., Gao Z., Avoidable thermodynamic inefficiencies and costs in an externally fired combined cycle power plant, Energy, 31, (2006). Erbay Z., Hepbasli A., Application of conventional and advanced exergy analyses to evaluate the performance of a ground-source heat pump (GSHP) dryer used in food drying, Energy Conversion and Management, 78, (2014). Gong S., Goni B.K., Parametric study of an absorption refrigeration machine using advanced exergy analysis, Energy, 76, (2014). Kairouani L., Nehdi E., Cooling performance and energy saving of compressioneabsorption refrigeration system assisted by geothermal energy, Applied Thermal Engineering, 26, (2006). Kaska O., Energy and exergy analysis of an organic Rankine for power generation from waste heat recovery in steel industry, Energy Conversion and Management, 77, (2014). Morosuk T., Tsatsaronis G., Advanced exergetic evaluation of refrigenration machines using different working fluids, Energy, 34, (2009). Morosuk T., Tsatsaronis G., Splitting the exergy

53 destruction into endogenous and exogenous parts application to refrigeration machines. In: Frangopoulos C, Rakopoulos C, Tsatsaronis G, editors. Proceedings of the 19th international conference on efficiency, cost, optimization, simulation and environmental impact of energy systems, July 12 14; 2006, National Technical University of Athens, Greece, 1, (2006). cycle, Energy Conversion and Management, 105, (2015). Morosuk T., Tsatsaronis G., A new approach to the exergy analysis of absorption refrigeration machines., Energy, 33, (2008). Morosuk T., Tsatsaronis G., Zhang C.. Conventional thermodynamic and advanced exergetic analysis of a refrigeration machine using a Voorhees compression process, Energy Convers. Manage., 60, (2012). Petrakopoulou F., Tsatsaronis G., Morosuk T., Carassai A., Conventional and advanced exergetic analyses applied to a combined cycle power plant, Energy, 41, (2012). Rezayan O., Behbahaninia A., Thermoeconomic optimization and exergy analysis of CO2/NH3 cascade refrigeration systems, Energy, 36, (2011). Tarique S., Siddiqui M.A., Performance and economic study of the combined absorption/compression heat pump, Energy Conversion Management, 40, (1999). Tsatsaronis G., Strengths and limitations of exergy analysis, In: Bejan A, Mamut E, editors. Thermodynamic optimization of complex energy systems. Dordrecht: Kluwer Academic Publishers, (1999). Tsatsaronis G., Morosuk T., A general exergy-based method for combining a cost analysis with an environmental impact analysis, Part I. Theoretical development, in: Proceedings of the ASME IMECE, Boston, Massachusetts, USA (2008a). Tsatsaronis G., Morosuk T., A general exergy-based method for combining a cost analysis with an environmental impact analysis. Part II. Application to a cogeneration system, in: Proceedings of the ASME IMECE, Boston, Massachusetts, USA (2008b). Tsatsaronis G., Park M.H., On avoidable and unavoidable exergy destructions and investment costs in thermal systems, Energy Convers. Manage., 43, (2002). Ustaoglu A., Alptekin M., Akay M.E., Energy and exergy analysis of a wet type rotary kiln and utilization of waste heat powered ORC, Applied Thermal Engineering, Submitted (2016). Yan G., Chen J., Yu J., Energy and exergy analysis of a new ejector enhanced auto-cascade refrigeration 39

54 Exergy Optimization of the Hybrid Compression-Absorption Industrial Refrigeration Systems Mahmoud Afshar*, Hamid Rad, Petroleum University of Technology, MahmoudAbad, Energy, , Iran * Rad@put.ac.ir Abstract To prevent the creation of condensate liquid in gas trunk lines, the heavy molecules of the supply natural gas have to be removed. Proper natural gas dew point adjustment process is crucial in the separation of its heavy molecules: It is typically performed by the compression refrigeration system with a vast power demand for gas compressors, usually provided by gas turbines. In this paper a Hybrid Compression-Absorption Refrigeration System (HCARS) is presented and optimized in terms of exergy losses. The main idea in the proposed hybrid system is to utilize the recovered compression refrigeration system gas turbines hot exhaust gases (with a temperature of 540 ) as the heat source required for the Absorption Refrigeration System (ARS) absorber regeneration process which would reduce energy consumption and greenhouse gases generation. The proposed system is worked out for the dew point unit of a gas refinery which results in 63% additional cooling capacity of the hybrid system (12550 KW) in comparison o the current compression refrigeration system (7670 KW) for the same fuel gas consumption. The coefficient of performance of the hybrid system is , and totally about SCMD of fuel gas is saved by the possibility of shutting off one of three running turbo-compressors of the unit. Keywords: Gas Refinery Dew Point Unit, Hybrid Refrigeration System, Exergy Destruction Optimization I. Introduction Energy efficiency has received heightened interest with the increasing attention paid to climate change and environmental pollution. It has been a major topic of discussions on natural resources preservation and costs reduction. Environment preservation must also be considered through energy optimization studies. Oil and gas industry consumes a large amount of energy and gas turbines are typically used to drive mechanical equipment and to generate electricity. High temperature exhaust gas from the gas turbine is discharged to the surroundings. Energy efficiency can be improved by using an absorption refrigeration system (ARS) to convert waste heat into useful cooling energy. The application of ARSs reduces the energy consumption of conventional vapor compression refrigerators. Hydrocarbon liquid dropout can cause a number of problems in gas transmission lines, including increased pressure drop, reduced line capacity, and equipment problems such as compressor damage. To avoid liquid dropout, most current operating specifications for gas transmission lines require that the lines be operated above the hydrocarbon dew point. As noted, natural gas dew point temperature is regulated by refrigeration units and needs a lot of cooling load. Benjamin Brant et al (1998) has Installed the first kind Waste Heat Absorption Refrigeration Plant. The refrigeration unit was designed to provide refrigeration for two process units at the refinery while utilizing waste heat as the energy source. The ARS enables to cool the gas to C, thereby recovering 45% of high molecular weight hydrocarbons by kalinowski et al (2009). Priedeman and Christensen (1999) presented a general ammonia-water absorption heat pump cycle that was modeled and tested. The experimental findings were found to be in close agreement with the simulation results by Jianbo et al (2013). Studies on the absorption and compression refrigeration systems in recent years were continued. A.K. Pratihar et al. (2011) simulated an ammoniawater compression-absorption system by incorporating detailed heat and mass transfer calculations in the absorber and desorber of the system. They studied the effect of relative solution heat exchanger area on the COP. With increase in the relative solution heat exchanger area from 17% to 50%, COP increased initially, got maximum value at 39% and then decreased. N.A. Darwish et al. (2013) analyzed the absorption refrigeration water ammonia system using Aspen Plus flow sheet simulator. A very good agreement between the simulator s results and the experimental measurements was found. Based on the above Description, in addition to simulation of absorption refrigeration cycles, analysis of them requires thermodynamic equations to 40

55 knowledge of simulation accuracy and compliance the design assumptions with the physics laws. Morethrefore simulations that have been associated with experimental models, have been matched with simulation software, especially sequential software. In the dew point units of natural gas, compression refrigeration systems are mainly used to achieve the desired temperature. Absorption refrigeration cycles can be used alongside conventional compression cycle to achieve the desired temperature. In other words, the hybrid refrigeration system used. A hybrid absorption compression refrigeration system consists of a heat-driven compression refrigeration subsystem and an absorption refrigeration subsystem, which are integrated by an absorber and a rectifier. High-temperature flue gas successively goes into a heat recovery generator to heat the working fluids of the compression refrigeration subsystem. Exist of restrictions in the unit and keeping constant defines the basic structure of the propane cycles cause going away from the original definition of hybrid refrigeration cycle and intersections of absorptioncompression refrigeration limited to nature of the system does not change. II. Energy Analysis of Absorption System Absorption systems have been modeled in the past in a variety programs, such as the one by Lazzarin et al (1996). Modern modeling is usually done by one of two software, developed by Oak Ridge National Laboratory (2014) and Engineering Equation Solver (EES), developed at the University of Wisconsin allows the user to compute thermophysical properties of working fluids, providing results with very good accuracy when compared to experimental results by Liao et al (2004). Selecting the correct property method is crucial for getting meaningful results. As in the previous paper written on modeling ammonia/water systems in ASPEN HYSYS program, it was found that Peng- Robinson was the best available method by Darwish (2013). This section explores the possibility of using absorption chillers to utilize waste heat. To accomplish this, models are created in ASPEN HYSYS and a variety of cycle options are considered. The waste heat source is the exhaust stream from a gas turbine, and since gas turbine models are already available, the bulk of the modeling work reported here focuses on developing absorption cycle models by Rad et al (2015). In fact, many calculations of thermodynamics are based on the assumption of ideal conditions such as reversible processes, real processes are nevertheless amenable to thermodynamic analysis. The object of such an analysis is to determine the overall efficiency of the use of energy and to calculate 41 the inefficiencies of the various steps of a certain process. The principle of mass conservation and the First and Second Laws of Thermodynamics were applied to each component of the system for the analysis. Every component was considered as a control volume, taking into account the heat transfer, work interaction and inlet and outlet streams. The gas turbine was not modeled at the same level of detail as the absorption chiller, since it is not the focal point of this work by Rad and Afshar (2015). The governing equations for mass conservation are: Mass balance: m i m o = 0 (1) Mass balance for ammonia: m ix i m ox o = 0 (2) Where x i and x o correspond to the inlet and outlet ammonia mass fractions. Energy balance: The First Law of Thermodynamics yields the energy balance of each component of the whole system in following form Q W = m oh o m ih i (3) Entropy balance: S gen,k = m es e m is i Q K T K (4) The known data of the model are the pump s flow rate, the condition of some streams, maximum and minimum pressure of cycle, ammonia concentration in all streams and the values of other points must therefore be calculated. II.1. Exergy Analysis In the absence of nuclear, magnetic, electrical and surface tension effect, the total exergy of the system E can be divided into four components: physical exergy E PH, kinetic exergy E K, potential exergy E PT and chemical exergy E CH. If the kinetic and potential exergy are neglected: E = E PH + E CH (5) E PH = (h h 0 ) T 0. (s s 0 ) (6) E CH = x n M a. e a ch + 1 x n M w. e w ch (7) Exergy destruction at the individual component level could also be calculated using the following equation: E D,k = in m (h T 0 s) m (h T 0 s) out (8)

56 Where T 0 is a reference temperature maintained at K here. It is noted that the definition of the exergy destruction rate in (Eq. 8) accounts for both the physical and chemical exergy of the fluid streams, as suggested by Palacios Bereche et al. and is also consistent with the approach of Morosuk et al (2005). The calculation procedure of the chemical exergy of various substances based on standard chemical exergy values of respective. Exergy balance at the steady state for a control volume reads: i m i. e i + (1 T 0 j ). Q T j = e m j e. e e + W CV + E D (9) A detailed exergy analysis includes calculation of exergy destruction, exergy loss, exergetic efficiency, two exergy destruction ratios in each component of the system along with the overall system and second law efficiency for the cycle. Mathematically these are expressed as follows: E D = E F E P E L (11) ε = E P E F = 1 [ E D+E L ] (12) E F II.2. COP and exergetic efficiency The COP is defined as the useful heat rate from the Q E divided by the required heat rate to the generator Q G by Lostec et al (2013). COP Abs = Q E = [m (h e h i )] ammonia (12) Q G [m (h i h e )] steam The COP of compression refrigeration system is defined as: COP Comp = Q Evap W Comp (13) Which can be defined COP for compression refrigeration operating units that are using gas turbines to run compressors of compression refrigeration system as: COP Comp = Q Evap Input Heating Energy of NG (14) Therefore, with the help of the equations 12 and 14, the COP of Hybrid refrigeration System defined by Rad (2015) as: COP Hybrid = (Q Evap) Abs +(Q Evap) Com Input Heating Energy of NG (15) The purpose of this section is to compare of the absorption cycle COP with compression refrigeration cycle COP and achieve to definition of hybrid refrigeration system COP that uses gas turbine as power source. III. Hybrid Refrigeration System The Dew Point Unit is already working utilizing standard mechanical compression refrigeration systems, and the proposed Hybrid Refrigeration System design is explained based on this refinery design specifications as well as real operational working conditions. Gas turbine exhaust operational specifications are shown in Table 1. Tab.1: Exhaust flue gas operational specifications Exhaust flue gas temperature 540 Exhaust flue gas mass flow rate 75.5 kg s Exhaust flue gas pressure kpa The cooling of the main cycle evaporator, i.e MW, is utilized for cooling of natural gas parallel to the available compression refrigeration cycle as shown in Figure 1. Considering the 7.26 MW cooling of natural gas in operational compression refrigeration cycle, the hybrid system natural gas cooling capacity will be =11.75 MW, which is 1.62 times the original cooling capacity. The original design is for cooling of 19.95MMSCMD natural gas, therefore the proposed hybrid system can cool MMSCMD of natural gas. Table.2 Shows the COP of Compression System, Absorption Refrigeration System that added to available system and Hybrid Absorption-Compression Refrigeration System COP in design condition and real condition. Tab.2: COP of System in Design and Real Condition Design COP Real COP Compression system Absorption system Hybrid system The results presented indicate the feasibility of the proposed hybrid refrigeration system for natural gas cooling under current real working condition of the compression refrigeration cycle. IV. Results of Advanced Analysis Exergy analysis of the process is performed and the process is optimized to minimize the exergy destruction of the process. The effect of partial load operation of the available compression refrigeration system on the performance of ARS is shown in Figure 3. This is done by extrapolation of the real field data collected in compression refrigeration cycle in Fajr-e-jam refinery by Ashar and Rad (2015). The exhaust temperature and mass flow rate of the flue gas is obtained with second order regression. As shown in Figure 3. The absorption cycle cooling load changes are within 5% of full load operation. Figure 4 also shows COP of Hybrid Refrigeration System in partial cooling load of compression refrigeration system. 42

57 Fig. 1: Proposed Hybrid Refrigeration System Layout Fig. 2: Schematic View of Suggested Cascade Absorption Cycle for Dew Point Unit 43

58 In the field data collected from propane compression refrigeration system by Afshar and Rad (2015), reduced cooling load of compression refrigeration system results in the exhaust flue gas temperature and mass flow rate changes: mass flow rate reduces in direct ratio with cooling load of compression system, while the temperature increases. Therefore, the increase in flue gas temperature compensates to some degree the drop in its mass flow rate. Indeed, despite reduction in compression refrigeration cooling load, the absorption refrigeration cycle cooling load remains almost unchanged and works near full load conditions. Tab. 3: Stream Constraints used in optimization of the proposed hybrid refrigeration system Stream Constrains Strea Constrains m 1A Quality=0 1B Quality=0 3A 4A Quality=0 Quality=0 X 3 X 5 3B X 3 X 5 T 5 T T 5 T X NH X NH T B T 4 40 P 5 P 4 P 5 P 4 5A Quality = 0 5B Quality = 0 P 5 P 3 &P 4 P 5 P 3 &P 4 6A T 6 T 2 6B T 6 T 2 8A Quality = 0 P 8 P 4 P 8 P 4, 8B T 8 T 4 T 8 T 4 P A P B T A T B Quality A Quality B P 12 P 1 12A P 12 P 1 12B X 14 = X 1 14A X 14 = X 1 T 14 = T 1 T 14 = T 1 14B P 14 = P 1 P 14 = P 1 P 8 P 4 Fig. 3: Cooling Load of Absorption Cycle Based on Partial Cooling Load of Compression Cycle Fig. 4: COP of Hybrid Refrigeration System in Partial Cooling Load of Compression Cycle The proposed hybrid refrigeration cycle is optimized by using optimization methods (such as Sequential Linear Programming Method) in HYSYS (version7.2). The objective function is: The total optimized Exergy destruction rate is 4005 KW, and the refrigeration rate increased to 4.68 MW from previous 4.49 MW, and also the COP of the cascade absorption refrigeration cycle increased to from its previous value of The exergetic efficiencies of all components are calculated. The rate of fuel exergy ( E Fuel ), rate of product exergy (E P), rate of destruction exergy (E D), rate of loss exergy ( E L ) for all components are calculated first. Four other parameters that are first exergy destruction ratio ( Y D ), second exergy destruction ratio (Y D ), exergy loss ratio (Y L ) and finally the exegetic efficiency are calculated based on the above four values. All of theses parameters are defined in section 3.2 of chapter three. The values of these parameters for all components of hybrid refrigeration cycle are shown in Table 4. Only % of the input exergy is converted to cooling as the system useful output, while the remaining exergy is either lost to the environment or destructed due to irreversibilities in the various components of the system. The total exergy supplied to the system is kw, out of which % is converted to useful product which is equivalent to kw, % is destroyed which is equivalent to kw, and the remaining 31.1 % is lost to environment which is equivalent to kw. n Minimize k=1 (E D,k) (k is number of cascade cycle components) Subject to constrains shown in table3. (16) 44

59 Tab. 4: Exergetic Destruction, Loss and Efficiency Using EDM Component E Fuel E P E D E L Y D Y L Y D ε kw kw kw kw % % % % Pump(A) Generator(A) Rectifier(A) SHX(A) Valve1(A) Cond.Evap.assly(A) L/V HX(A) Valve2(A) Absorber(A) Pump(B) Generator(B) Rectifier(B) SHX(B) Valve1(B) Cond.Evap.assly(B) Valve2(B) L/V HX(B) Absorber(B) Overall System The highest total exergy destruction is found in SHX- A which is equal to 9.08 % of the total exergy destruction of the system. The second highest exergy destruction is found to be in the SHX-B which amounts to kw, equivalent to 7.89 % of the total exergy destruction. The exergetic efficiency for condenser-evaporator assembly and L/V HX of main and cascade cycles are almost low. The reason for low exergetic efficiency of condenser-evaporator assembly may be attributed to the large temperature difference between the working fluid and brine in the evaporator and the working fluid and cooling air in the condenser. Similarly the high temperature difference between both the fluids and relatively very less mass flow rate of gaseous ammonia from the evaporator compared to the condensed ammonia from the condenser may be the reason for the poor exergetic efficiency of L/V HX. V. Conclusion Energy efficiency can be improved by using an absorption refrigeration system to convert waste heat into useful cooling energy and one example is the dew point unit of gas refineries. Since the source of the heat is the flue gas of compression refrigeration systems, a hybrid compression-absorption refrigeration system is proposed in this paper for full achievement of possible energy savings. By conducting exergy destruction optimization, the total minimum exergy destruction rate is 4005 KW, while the ARS cooling rate and COP increased to 4.68 MW and from their its previous values of 4.49 MW and , respectively. The exergetic efficiencies of the rectifiers and absorbers are the highest, and the exergetic 45 efficiencies of condenser evaporator assemblies and liquid/vapor heat exchangers are the least. The highest total exergy destruction is found in SHX-A which is equal to 9.08 % of the total exergy destruction of the system. The second highest exergy destruction is found to be in the SHX-B, equivalent to 7.89 % of the total exergy destruction. Low exergetic efficiency of condenser evaporator assembly is attributed to the large temperature difference between the working fluid and NG in the evaporator and the working fluid and cooling air in the condenser. Similarly the high temperature difference between both the fluids and relatively very low mass flow rate of ammonia from the evaporator compared to the condensed ammonia from the condenser may be the reason for the poor exergetic efficiency of L/V HX. References Rad, H Feasibility Study of Hybrid Cooling Systems in Separation of High Molecular Weight Hydrocarbons in Natural Gas. Thesis Submitted to the University of Petroleum University of Technology In Energy Systems Engineering. Afshar, M. Rad, H Simulation and Exergy Analysis of Cascade Absorption Refrigeration System with Heat Recovery of Dew Point Units Waste Heat. 4 th Conference of Emerging Trends in Conservation Energy. Wei, H., Liuli, S., Danxing, Z., Hongguang, J., Sijun, M., and Xuye, J., New hybrid absorption compression refrigeration system based on cascade use of mid-temperature waste heat. International Journal of Applied Energy, Vol.106, pp Andre, M., Sergio, M., Luben, G., and Ricardo, S.,

60 2010. Using engine exhaust gas as energy source for an absorption refrigeration system. International Journal of Applied Energy, Vol.87, pp Kalinowski, P., Yunho, H., Reinhard, R., Hashimi, S., and Rodgers, P., Application of waste heat powered absorption refrigeration system to the LNG recovery process. International Journal of Refrigeration, Vol. 32, pp Manzela, A., Hanriot, S., Cabezas, G., and Sodre, J., Using engine exhaust gas as energy source for an absorption refrigeration system. International Journal of Applied Energy, Vol.87, pp Erickson, D., Anand, G., and Kyung, I., 2004, Heatactivated dual-function absorption cycle. International Journal of ASHRAE Trans, Vol.110, Part1. Fan, Y., Luo, L., and Souyri, B., Review of solar sorption refrigeration technologies: development and applications. Journal of Renew Sustain Energy Rev, Vol.11, pp Srikhirin, P., Aphornratana, S., and Chungpai, S., A review of absorption refrigeration technologies. Journal of Renew Sustain Energy Rev, Vol.5, pp Gas technology. Technical information about natural gas cleaning and treatment. See also URL Priscilla, B., Machado, J., Monterio, J., Medeiros, H., and Epsom, O., Supersonic separation in onshore natural gas dew point plant. International Journal of Natural Gas and Engineering, Vol.6, pp Pongsid, S., Satha, A., and Supachart, C., A review of absorption refrigeration technologies. International Journal of Renewable and Sustainable Energy Reviews, Vol.5, pp Jianbo, L., and Shiming X., The performance of absorption-compression hybrid refrigeration driven by waste heat and power from coach engine. International Journal of Applied Thermal Engineering, Vol.61, pp Brant, B., Brueske, S., Erickson, D., and Papar, R., Refinery Waste Heat Ammonia Absorption Refrigeration Plant. Journal of ESL-IE Pratihar, A., Kaushik, S., and Agarwal, R., Performance evaluation of a small capacity compression-absorption refrigeration system. International Journal of Applied Thermal Engineering, Vol. 42, pp Darwish, N., Hashimi, S., and Mansoori, A., A Performance Analysis and Evaluation of a 46 Commercial Absorption-Refrigeration Water- Ammonia (ARWA) System. International Journal of Refrigeration, Vol. 31, No. 7, pp Kececiler, A., Acar. H., and Dogan A, Thermodynamic analysis of absorption refrigeration system with geothermal energy: an experimental study. Journal of Energy Conversion and Management, Vol. 41, pp Lazzarin, R.M., Gasparella, A., and Longo, G.A., 1996, Ammonia-Water Absorption Machines for Refrigeration: Theoretical and Real Performances, International Journal of Refrigeration, Vol. 19, No. 4, pp Ruiz, E., Ferro, V.R., Riva, J., Moreno, D., and Palomar, J., 2014, Evaluation of ionic liquids as absorbents for ammonia absorption refrigeration cycles using COSMO-based process simulations. International Journal of Applied Energy, Vol. 123, pp Grossman, G., Zaltash, A., Modular Simulation of Absorption Systems. International Journal of Refrigeration, Vol. 24, No. 6, pp Liao, X., 2004, The Integration of Air-Cooled Absorption Chiller in CHP Systems. Ph.D. Thesis, University of Maryland, College Park, MD, USA. See also URL Darwish, N., Al-Hashimi, S., Al-Mansoori, A., Performance Analysis and Evaluation of a Commercial Absorption-Refrigeration Water- Ammonia (ARWA) System. International Journal of Refrigeration, Vol. 31, No. 7, pp Peng, D., Robinson, D., A New Two-Constant Equation of State. Industrial & Engineering Chemistry Fundamentals. International Journal of Chemistry Fundumental, Vol.15, pp Janilson, A.R., Edson, B., Thermodynamic Modeling of an Ammonia-Water Absorption System Associated with a Micro-turbine. International Journal of Thermodynamics, Vol. 12, pp Garousi, L.F., Infante F., Mahmoudi, S.M.S., and Rosen, M.A., First and second law analysis of ammonia/salt absorption refrigeration systems. International Journal of Refrigeration, vol. 40, pp Lostec, B.L., Galanis, N., and Millette, J., Simulation of an ammonia-water absorption chiller. International Journal of Renewable Energy, Vol.60, pp Aghniaea, S., Mahmoudi, S.M, Exergy Analysis of a Novel Absorption Refrigeration Cycle with Expander and Compressor. Indian Journal of Sci.Res, Vol.1, pp

61 Zare, V., Mahmoudi, S.M., Yari, M., and Amidpour, M., Thermoeconomic analysis and optimization of an ammonia-water power/cooling cogeneration cycle. International Journal of Energy, vol.47, pp Takeshita, K., Amano, Y., and Hashizume, T., Experimental study of advanced cogeneration system with ammonia-water mixture cycles at bottoming. International Journal of Energy, Vol. 30, pp

62 Energy and Exergy Analysis of a Steam Power Plant Considering Effect of Varying Plant Loads Mehmet Bilgili 1, Mehmet Tontu 2, Besir Sahin 2* 1 Cukurova University, Ceyhan Engineering Faculty, Mechanical Engineering Department, Adana, 01950, Turkey 2 Cukurova University, Engineering and Architecture Faculty, Mechanical Engineering Department, Adana, 01330, Turkey * bsahin@cu.edu.tr Abstract In this study, the energy and exergy analyses of the existing steam power plant have been performed with three different operating loads (100%, 70% and 40%). This steam power plant is designed to operate at a subcritical steam conditions and has a power capacity of 660 MW at full load. The primary objectives of this study are to analyze the system components separately and to identify and quantify the sites having largest energy and exergy losses. Influences of three different loads on the useful power, reversible power and irreversibility have been investigated for each component. In addition, the second law efficiency of system components and the overall efficiency of the power plant have been computed. Energy losses mainly occur in the condenser and that exergy losses mainly take place in the boiler. According to these results, the exergy analysis is more significant compared to the energy analysis. It is found that if the exergy losses are reduced, the power plant efficiencies are positively affected. It is concluded that notable amount of input exergy can t be used as a useful work due to friction, mixing and irreversibility in the cycle. Keywords: Efficiency, energy analysis, exergy analysis, operating load I. Introduction Energy is one of the basic necessities in modern life. There is not any field of activity in everyday life that is the energy not being used. Nowadays, energy consumption of societies is considered as an indicator of development. Especially, electrical and heat energies play an important role in human lives. These energies which are produced from limited natural resources have become more valuable due to an increase of demand. Energy is considered to be a vitally important tool in economic, social and industrial developments (Yazıcı and Selbaş, 2011). Erkin (2006) reported that Turkey is one of the fastest growing energy markets in the world for nearly two decades because of a young and growing population, an increase of electricity consumption per person, rapid urbanization and strong economic growth. Unfortunately, the growth in electricity generation in recent years has been lower comparing to the growth of electricity demand. According to the report of IEA, (2013), the population growth, industrialization, expansion of technology and rising of welfare are proportional to the increase of energy consumption. In order to reach the level of developed countries, developing countries must involve in research and development and produce more advanced technological products. Although the development rate of renewable energy technologies increases rapidly, the world energy needs are still heavily depended on fossil fuels for electricity production. That is to say, the majority of the world s power generation is met by fossil fuels, particularly coal and natural gas. Despite the growth of renewable energy installations like wind and solar power, a heavy dependence on fossil fuels is expected to continue for decades as stated by Regulagadda et al. (2010). They also reported that despite the depletion of fossil fuel reserves and environmental concerns such as climate changes, the growth in the fossil oil demand is expected to be 47.5% (For example, 91.6% from the natural gas and 94.7% from the coal, until 2030). On the other hand, cleaner renewable energy sources are being rapidly developed. But, their relative cost and current state of renewable energy technologies have not advanced to a stage where they can significantly reduce our dependence on the fossil fuel. Environmental concern, climate changes, and continued reliance on the fossil fuel forces the fossil fuel plants to reduce their environmental impact by operating thermal power stations more efficiently and improving their technologies. In this study, the energy and exergy analyses of the existing steam power plant were performed with three different operating loads (100%, 70% and 40%). Influences of three different loads on the useful power, reversible power and irreversibility were investigated for each component. II. Power Plant Description Steam power plant considered in the present study is the coal fired plant located in Turkey. It operates with Rankine cycle supplying competitive power to the national grid under acceptable environmental protection. Power plant generally consists of pumps, 48

63 turbines, boiler, condensers and heaters also includes flue gas desulfurization unit and electrostatic filters. The present power plant has one HP turbine, one IP turbine and two LP turbines. HP turbine is a single flow type and has 13 stages. IP is double flow and has three uncontrolled extraction points and 14 stages. LP turbines are double flow and each LP turbines have three uncontrolled extraction points and 14 stages. All turbines directly coupled with the generator which rotates at 3000 rpm (50Hz). Gross outlet power is 660 MW and net power is 605 MW at a full load for each unit. In coal fired thermal power plant, steam is obtained in very high pressure inside the steam boiler by burning the pulverized coal. This steam is then super-heated in the super heater to extreme high temperature. This super-heated steam is then allowed to enter into the turbine, as the turbine blades are rotated by the pressure of the steam. After entering into the turbine, the steam pressure suddenly falls leading to corresponding increase in the steam volume. After having imparted energy into the turbine rotors, the steam is made to pass out of the turbine blades into the steam condenser of turbine. In the condenser, cold water at ambient temperature is circulated with the help of pump which leads to the condensation of the low pressure wet steam. Then this condensed water is further supplied to low pressure water heater where the low pressure steam increases the temperature of this feed water, it is again heated in high pressure. This outlines the basic working methodology of a thermal power plant. The water-steam cycle of power plant is shown in Fig. 1. III. Analysis III.1. Energy Analysis In a steady state control volume: m in = m out (1) The total energy content remains constant as the mass balance during steady state operation: E in = E out (2) Q W = m out h out m in h in (3) Here, subscripts in and out show inlet and outlet states, Q is the heat transfer rate, W is the work rate, m is the mass flow rate and h is the specific enthalpy. Thermal efficiency (η ı ), may also be defined as the 1st law efficiency, which can be expressed as the ratio of the work rate to the fuel energy input rate: η I = W net Q in (4) III.2. Exergy analysis Regulagadda et al. (2010) indicated that exergy analysis will characterize the work potential of a system. In the exergy analysis of this study, the properties at the dead state were denoted by subscript zero. For instance P0 and T0 refer to the dead-state pressure and temperature, respectively. Here, T0 was assumed to be 25 o C (298 K) and P0 was assumed to be 1 bar. The method of exergy analysis overcomes the limitations of the first law of thermodynamics. The concept of exergy is based on both the FLT and the SLT. Exergy analysis clearly indicates the locations of energy waste in a process and can therefore lead to improved operation or technology. Exergy analysis can also quantify the quality of heat in a waste stream. A main aim of exergy analysis is to identify exergy efficiencies and true magnitudes of exergy losses. Exergy analysis provides those tools and it helps in locating weak spots in the process. (Dinçer and Rosen, 2011). The exergy analysis for present system is carried out for each component in the subsystems, to evaluate the exergy losses in the individual component and then the analysis is performed on the overall individual subsystems to find out the exergy losses in each subsystem. Finally the exergy analysis for the overall plant is carried out and the second law efficiency is computed. Quantitative measure of disorder is called entropy of the system at the microscopic level. Entropy generation, Sgen of the system is presented as follows: Ql S ms ms gen (5) out in Exergy balance can be obtained by using following equations: X heat X X mass, out T T in 1 0 k m work X Ql out X loss W mass, in X m X T 0 out loss X 0 in (6) (7) Where Tk is the surrounding temperature and the change in the flow exergy is; out in h out h in T ( s 2 2 Vout V in (8) g( zout z in) 2 X m( in ) (9) out 0 out s in ) 49

64 Fig.1: The water-steam cycle of power plant 50

65 Reversible work is obtained with following equations. out V W 2 out rev 2 in V h 2 in out h g( z in out T ( s 0 z in ) out m( in ) out s in ) (10) (11) Irreversibility can be calculated as: I T0S gen or, I W rev W in (12) (13) Considering no kinetic and potential energy, the expression for exergy becomes; ( h h0 ) T0 ( s s0) (14) Where h, s are specific enthalpy and entropy values at a given state, h0, s0 are specific enthalpy and entropy values at a dead state. III.3. Fuel and Combustion Analysis Boilers need a heat source which has an enough temperature to generate steam with a high pressure and temperature using fossil fuel in the combustion chamber. It is known that the main constituent of coal is carbon. Coal also contains varying amounts of oxygen, hydrogen, nitrogen, sulfur, moisture, and ash as shown in Table.1. In the present thermal power plant, imported coal with high calorific value is used and coal components are indicated below. Lower calorific value of coal is defined as 25,800 kj/kg. (Unal, 2009). Pure oxygen is used only in special application but in this study air is used for combustion. Air is considered to be % 21 oxygen and % 79 nitrogen on a molar basis. With this idealization the molar ratio of the nitrogen to the oxygen is 0.79/0.21 = 3.76, when air supplies is accompanied by 3.76 moles of nitrogen (Moran and Shapiro, 1995). Tab. 1: Chemical compositions of presently used Coal Coal Components (% ratio) C H N S O Ash Water Combustion process in this study is: (cc+hh+nn+ss+oo+wh2o+)+1.5a(o2+3,76n2) (x CO2+yN2+zH2O+tSO2) Required minimum amount of oxygen for completing combustion is calculated using this equation 17 (Geredelioğlu, 2011); Amount of oxygen = %C %H %S %O2 (15) Enthalpy of flue gases can be calculated at desired temperature using these equations: h h fg xco. hco xn. h N xso. h SO xh O. h H O (16) fg h M fg fg Entropy of components of flue gases, s s CO N s s ln 2 K CO R x, 2 CO2 (17) (18) (19) s ln 2 K R x SO, N 2 N 2 (20) s ln 2 K R x, SO2 SO2 s s R x 2 H O 2 K, H2O ln H O (21) fg xco. sco xn. sn xso. sso xh O s H O (22) s s fg s fg (23) M fg Chemical exergy of flue gas and combustion air formula are represented below respectively. ch fg x RT ( x CO2. ln x CO2 x N2 x. N2 ln x x SO2 x. SO2 ln x H2O x. H2O 0 CO2 CO2 N2 N2 SO2 SO2 H2O H2O ln (24) ) ch ch fg fg (25) M fg X ch ch fg m (26) fg ch fg air x x. RT ( x ln x x ln x ) (27) O. 2 O2 N2 N2 0 O2 O2 N2 N2 ch ch air air (28) M air X ch air ch mair air (29) Coal chemical exergy, h o n (30) c c c w f h f (31) 4.18 X f m f f (32) 51

66 Standard chemical exergy of components is given Table 2 (Kotas,1995). Tab. 2: Standard chemical exergy of components Components ψ ch kj/kmol O CO N2 720 SO H2O 9500 IV. Results and Discussions Energy losses of thermal power plant equipment at different loads are shown in Fig. 2. As it can be seen from the figure energy loss of condenser is always higher than the other equipment. For example, 48.1%, 48.8% and 49.6% of total energy are rejected to the environment via condenser at 100%, 70% and 40% loads respectively. Because enthalpy of exhaust steam which comes from LP turbines is higher compared to the condense water. Reason of condenser losses is the latent heat of exhaust steam is transferred to the cooling water. Also 5.5 %, 6.3% and 6.6% of total energy rejected to the environment via boiler s outer surface and flue gases at 100%, 70% and 40% loads, respectively. Fig. 3 shows exergy losses of all thermal power plant equipment. As it is observed from the figure, exergy losses of the boiler is much higher than the other equipment. Despite the fact that major energy loss occurs in the condenser, but, major exergy loss is taken place in the boiler. That is to say, 51.9 %, 53.6% and 56.1% of input exergy are destroyed through exhaust gases at 100%, 70% and 40% loads respectively. The high level of exergy loss in the boiler is due to finite differences between flue gas and working fluid (water and steam). A combustion process usually occurs simultaneously with heat transfer. Both chemical reaction and heat transfer are irreversible processes. The losses in the boiler are due to increase in the entropy generation of flue gases. Additionally substantial amount of heat loss is conveyed by the flue gas to the environment with high temperature. The exergy losses take place in pumps due to compression ratio and flow control method. Pump losses are usually very small compared with the other equipment. Exergy loss in the condenser is due to heat transfer between exhaust steam and cooling water but it s thermodynamically insignificant because temperature differences between them is about 5 K. So that the rate of exergy loss is small. Fig. 2: Energy losses of power plant equipment Fig. 3: Exergy losses of thermal power plant equipment 52

67 Fig. 4: Irreversibilities of turbines Fig. 5: Irreversibility of heaters Fig. 4 shows irreversibility of turbines at three different loads. In turbines, 5.34 %, 5.43% and 5.55% of total exergy are destructed at loading capacities of 100%, 70% and 40% respectively. The exergy losses in turbines are due to the pressure drop and expansion process. Fluid friction happens when the fluid expands through the steam turbine blades. These friction losses result in the dissipation of part of its energy into heat itself at the expense of useful work. The fluid then does less work and leaves through the exhausts with a higher temperature. The more irreversible process, the higher turbine exit temperature and the less the work. As the load increases, amount of irreversibility increases in turbines but percentage of irreversibility decreases. The exergy loss in LP turbines is higher than IP and HP turbines because LP turbines work at a vacuum pressure so that condensation process begins and steam leaves the LP turbines as a water-steam mixture therefore the rate of entropy and irreversibility increases substantially with condensation process. But the exergy loss in HP turbines is lower, because pressure ratio between inlet and outlet is small compared to the IP and LP turbines. The value of exergy losses in the LP 1 and LP 2 turbines are close to each other, because the operating conditions are almost the same. Irreversibility in the HP and LP heaters at different load are shown in Fig. 5. It can be seen from this figure that irreversibility in the HP heaters is always higher than LP heaters due to the extraction steam of HP heaters at a high pressure and high temperature. Besides temperature differences between extraction steam and working fluid is higher and also as the load increases, although irreversibility in the HP heaters raises, irreversibility in the LP heaters drops down because while the load increases, temperature differences between extraction steam and working fluid diminishes in LP heaters but in HP heaters temperature differences between them are increases. In addition aim of using HP and LP heaters is to recover latent heat of exhaust steam just before conveying to the condenser. The second law efficiency which is belonged to power plant equipment is shown in Fig. 6. At three different loads, boiler feed pump is very efficient but boiler is least efficient. And also according to the figure the second law efficiency of equipment enhances with increasing load. Besides boiler feed pump efficiency is higher compared to the condenser extraction pump because flow rate of boiler feed pump is controlled by frequency converter but flow rate of condenser extraction pump is controlled by a valve (throttling). Pump exergy loss is related with the 53

68 compression ratios which are 21.4, 21.6 and 22.2 in the boiler feed pump and 595.2, and 913,4 in the condenser extraction pump at different loading capacities such as100%, 70% and 40% respectively. Overall efficiency of thermal power plant is shown in Fig. 7. According to energy analysis based on the first law analysis the major energy losses are due to heat rejection in the condenser and heat losses with flue gases but according to exergy analysis based on the second law analysis the major losses are due to steam generation where the fuel exergy is destroyed. It is seen that the first and second law efficiencies enhance with increasing load capacity. Overall power plant efficiencies are affected by its component efficiencies and its raises with increasing load. Although the amount of energy loss and exergy loss of power plant is increased, percentage of these losses is decreased while load is raised so that overall efficiencies are increased. As a consequence, the power plant is not suitable for partial load, it is suitable for full load. Net output powers of thermal power plant are measured 605 MW, 420 MW and 235 MW at a loading capacity of 100%, 70% and 40% respectively. Second law efficiency BFP Turbines CEP Boiler Fig. 6: The second law efficiency of equipments 100% 70% 40% First Law Second Law Efficiency V. Conclusions Plant Load % 40 Fig. 7: Overall efficiencies of power plant thermal power plant. In this study, energy and exergy analyses of the active thermal power plant were performed using thermodynamic principles for three different loads. Energy and exergy losses and also second law efficiencies of each equipment of the thermal power plant were calculated. From these analyses, the vulnerable spots can be seen in power production processes clearly. In actual coal fired thermal power plant operation, abnormal exergy parameters of the corresponding equipment s can be detected and used to define the faulty locations. So, the exergy efficiency and energy loss analysis of the power plant are helpful to malfunctions identification and diagnosis of From the present results, the following conclusions have been drawn: In the present coal fired thermal power plant, the maximum energy loss was found in the condenser at three different loads, but, exergy losses are lower because these losses thermodynamically are insignificant which are at low quality. Heat losses can be used in heating greenhouses if it is possible to set it near the thermal power plant. In terms of the exergy loss or irreversibility, maximum losses were found in the boiler at three 54

69 different loads. Because flue gas temperature is very high according to the ambient temperature. These losses may be used in the absorption cooling system or it may be used for coal drying purposes. Also, the losses may be used for heating sites. But it is considered that in the case of flue gas temperature which is lower than 120 C, acid formation can occur with condensation of flue gas, this situation cause defect the channel and any other equipment. As the load increases, irreversibility or the exergy loss increases at each equipment but percentage of irreversibility decreases. Irreversibility in pumps is directly proportional to the compression ratio. And the type of flow control method can also affect the irreversibility. As the load increases, compression ratio falls down but second law efficiency increases. Irreversibility in the HP heaters is always higher than LP heaters at three different operating loads. Irreversibility in heaters is directly proportional to the temperature differences between heating medium and working fluid. If the differences between them are increased, heat transfer rate is enhanced as a result, entropy and irreversibility rise. The causes of exergy losses in turbines are mainly related to their design, frictional losses and pressure ratio between inlet and outlet. Also condensation process should be considered in the last blades of LP turbines. It may cause wear and tear in the last blades of LP turbines because of water droplets and hence the entropy and irreversibility increase rapidly. The first law efficiency of thermal power plant was found to be 41.5%, 39.7% and 36.4% % at the loading capacities of 100%, 70% and 40% respectively. The second law efficiency of thermal power plant was found to be 39.1%, 37.4% and 34.3% at loading capacities of 100%, 70% and 40% respectively. Finally, it can be concluded that the overall efficiency of the thermal power plant is best at a full operating load. analysis of a thermal power plant with measured boiler and turbine losses, Applied Thermal Engineering, 30: (2010). Yazıcı H., Selbaş R., Energy and exergy analysis of steam power plant, Selçuk University, Journal of Technical Online, 10(1): (2011). Dinçer, I., Rosen, M.A., Thermal energy storage; systems and applications, John Wiley and Sons, 599p (2011). Geredelioğlu, Ç., Energy and exergy analysis of Çayırhan thermal power plant: Msc Thesis, Süleyman Demirel University, The Institute of Science, Isparta, 122p (2011). Unal, F., Exergy analysis of a thermal power plant: Msc Thesis, Yıldız Teknik University, The Institute of Science, İstanbul, 91p. (2009) References Erkin T., Energy and exergy analysis of thermal power plant, MSc Thesis, Denizli: Pamukkale University (2006). IEA, Energy Efficiency Policy Recommendations, (Date of Access: ). Kotas. T.J., The exergy method of thermal plant analysis, Krieger Publishing Company, Malabar, FL (1995). Moran, M.J., Shapiro, H.N., Fundamental of Engineering thermodynamics, Third Edition. John Wiley & Sons, New York (2000). Regulagadda P., Dincer, I., Naterer G.F., Exergy 55

70 Long Term Energy Demand and Supply Projections and Evaluations for Turkey Esra Ozdemir 1, Muhsin Kilic 2* 1 Uludag University, Engineering Faculty, Department of Mechanical Engineering, Bursa, Turkey 2 Uludag University, Yenisehir Ibrahim Orhan of Vocational School, Department of Machine, Bursa, Turkey * mkilic@uludag.edu.tr Abstract Energy demand of countries changes depending on many socia-economic factors like their population, the level of social and economic development, industrialization, urbanization and technological development. In the past decade, economic growth and social development in our country has been led to increase in energy consumption. Accordingly, the amount of energy needed must be provided that so as to realize the economic growth and social development in time, satisfactory, uninterrupted and taking into account the environmental impact. For this reason, it is necessary to determine and prioritize the alternative energy strategies. In this study, used energy datas from Turkey s Energy-Balance Tables and developed a long-term energy demand projections for Turkey in Long Range Energy Alternatives Planning System Program. Two scenarios for Turkey's energy demand are created and their results are evaluated. In addition, scenarios are created that the renewable energy resources' ratios which are solar, wind, hydro and geothermal energy etc. is increased in total energy supply and their results are evaluated for energy demand and for electricity generation. Keywords: Energy, energy demand, energy supply, environmantal effects, long term projection. I. Introduction Energy has become one of the most significant factors in ensuring that countries provide a competitive advantage since the beginning of the 20th century. In the 21st century, the technological innovations, increasing the permeability of international borders, capital mobility and development of communication cause increasing the amount and speed of energy use by Kavak (2005). For this reason, generating policies for the future and energy management should be done by planning energy from today. The Republic of Turkey with population 78 million and area km 2, forms a natural bridge between Europe and Asia by TUIK (2016). Turkey is a rapidly growing economy and over the past decade, its Gross Domestic Product (GDP) has increased at an significant rate compared to other OECD countries. Turkey is the 17th largest economy of the world by IEA (2001). Turkey has experienced considerable changes in its electricity market in the past decades. Rapid growth in the electricity demand has led to considerable transformation in the electricity sector with large increases in the generation capacity to accompany it. According to the energy balance sheets of Turkey, which are published by World Energy Council-Turkish National Committee (2016), between the years 2000 and 2013, the electricity demand of Turkey almost doubled and reached GWh. Energy has a strategic importance for developing Turkey. According to Turkey Statistical Institute's datas (2016), the energy imports of Turkey increased to 60.1 billion dolars in However this value decreased in the last two years and that was billion dolars and 54.91billion dolars in 2013 and 2014 respectively. Accounting for these values, it is considered that Turkey doesn t have substantial reserves of conventional fuels. Hence, Turkey should make energy planning and energy policy. Due to the lack and poor quality of primary resources, Turkey is highly dependent on imported energy. According to the Ministry of Energy, import dependency was above 72% in This is underpinned by the dependency on natural gas imports which account for nearly 43% of total electricity production by Republic of Turkey Ministry of Energy (2016). There are several challenges that the Turkish energy market faces. The high level of dependence on imported energy sources and the negative externalities caused by the utilization of fossil fuels stand as the main problems the policy makers will strive to solve for the immediate future. The primary aim of Turkey is to realize its own energy security. To this end, Turkey has for objective to -diversify its energy supply routes and source countries, -increase the share of renewables and include the nuclear in its energy mix, -take significant steps to increase energy efficiency, -contribute to Europe s energy security by Republic of Turkey Ministry of Foreign Affairs (2016). As a potential candidate country for Europe Union accession, Turkey is under a pressure to reduce its CO2 emissions. Thus Turkey should focus on building up its capacity for mitigation of greenhouse gas 56

71 (GHG) emissions and adaptation to climate change. One of the common approaches in the world for mitigation of GHGs is the sector-based emission mitigation policy. Accordingly, Turkey has a national policy on increasing share of renewable energy sources in the electricity generation by Ozer at al. (2013). Energy demand and supply projections working which lead to the energy planning and energy policy, has gained importance in recent years. Long term energy supply and demand projections constitute the basis of long term energy planning and investment. A combination of an optimal system to meet the energy demand can be determined in the lowest cost with energy projections. Thus, it is possible that both energy and financial resources are used more efficient and possible scenarios can be tested and revised. In the light of this working, decision-makers are able to have an idea about needed policies and practises for future from today. Some energy projections in the literature are as follows; Zhang et al. (2007) calculated the external costs of electricity generation in China under different energy scenarios by using Long Range Energy Alternatives Planning (LEAP) system. They estimated the energy demand in electricity generation of China from 2003 to Lin and Ouyang (2014) evaluated how demand of fossil fuels and carbon dioxide and sulphur dioxide in China are affected by the reforms in the country and macro-economic developments with general equilibrium models they created. Cai et al. (2013) rated between by creating different scenarios of electricity production models for transition to clean and more efficient use of energy in China by using LEAP software. Ozer et al. (2012) predicted electricity demand until 2030 in Turkey by using the LEAP program and have planned electricity production improvement scenarios for reduction of greenhouse gas emissions. Hotunoglu and Karakaya (2011) evaluated the final results by scenarizing how energy demand will develop in case of economic growth is stabilized, energy densities decrease in the future and economic growth changes in each five years by using artificial neural Networks technique. Dilaver and Hunt (2011) estimated Turkish aggregate electricity demand depending on GDP, electricity price and Underlying Energy Demand Trend (UEDT). Shin et al. (2005) have made planning electricity production by creating projection of being used waste gas (landfill gas) to produce electricity more environmentally friendly and economically in Korea by using LEAP model, within the context of the Kyoto Protocol. Egelioglu et al. (2001) investigated the influence of economic variables on the annual electricity consumption in Northern Cyprus and they found that a model using number of customers, number of tourists and electricity prices has a strong predictive ability. Song et al. (2007) accomplished environmental and economic assessents of chemical absorption processes in Korea using the LEAP model. They analyzed the scenario based on the data of a pilot plant (2 ton/day) that is installed in the Seoul coal steam power plant, and an alternative scenario 57 is set according to energy policy chang by climate change egreement and development of CO2 mitigation technology. Ozturk et al. (2005) used the genetic algorithm approach to investigate the relationship between total electricity consumption, gross national product (GNP), population, imports and exports for the period in Turkey with annual data. Total electricity demand of Turkeywas estimated as 220 TWh and 300 TWhin 2020 with exponential and quadratic forms of the genetic algorithm electricity demand models respectively. However, Ozturk and Ceylan (2005) concluded that aggregate electricity demand for Turkey would be between about 462 TWh and 500 TWh in 2020 for the low and high growth scenario respectively as the results of genetic algorithm electricity demand (GAED) quadratic model. The goal of this paper is to evaluate the future energy demand, energy supply and CO2 emission potential of Turkey s energy sector. For this purpose, firstly, the total energy demand was estimated depending on the six sectors: industrial, residential and services, transport, agriculture and non energy use. The estimation was based on the population, gross domestic product (GDP) and the proportion of each demand sector in total consumption with the annual growth rates. Correspondingly electricity generation scenarios were built. In this paper, two scenarios for energy demand and four scenarios for electricity generation based on Long-range Energy Alternatives Planning system (LEAP) model were employed to simulate the current energy situation and to develop forecasts under certain assumptions. The demand scenarios include Business As Usual (BAU) and Mitigation Scenario options. The electricity generation scenarios are created that the renewable energy resources' ratios which are solar, wind, hydro and geothermal energy etc. is increased in total energy supply and their results are evaluated for energy demand and for electricity generation. II. Energy Consumption and Supply in Turkey II.1. The structure of the energy sector in Turkey Importance of Turkey increases as a regional energy transit hub and growing consumer in the energy market. According to Energy Information Administration (2013), Turkey's energy demand has increased in recent years in very quickly and it is predicted that this increase will continue in the next years. According to Energy Report (2013), overall distribution of energy resources in Turkey is that, Turkey has become one of the fastest growing energy markets in the world and has been experiencing rapid demand growth in all segments of the energy sector for decades. Turkey comes in possession of the most dynamic energy economies of the world in terms of increase in energy demand.

72 Having a substantial potential for the renewable energy resources, Turkey ranks seventh in the world and first in Europe in terms of geothermal energy. Turkey aims at further increasing its use of hydro, wind and solar energy resources and Turkey has potential producing %30 of its electricity need from the renewable by Turkey is geographically located in close proximity to more than %70 of the world s oil and gas reserves Annual electricity generation is approximately 179, 5 billion kwh in Turkey. Renewable energy and energy efficiency projects are assisting to reduce CO2 emissions in Turkey by more than 3 million tons annually. Turkey has different kinds of energy sources which Turkish energy sector is becoming more active, competitive and attracting the attention of investors. According to Energy Balance Sheets of Turkey (EBST), thousand tons of oil equilavent (TOE) primary energy supply occurred with thousand TOE domestic production and thousand TOE the value of imported energy in Fig. 1 shows the distribution of sources in total primary energy supply. The highest energy resource was natural gas with the rate of 32%. This value was followed by coal and oil sources with 29% rates. According to Fig.1, Turkey's energy supply comprises of fossil fuels by 90%. Wood 2% Hydro 4% Oil 29% Geo., wind, solar 2% Others 2% Coal 29% Naturel gas 32% Fig. 1: The distribution of sources in Turkey's primary energy supply On a sectoral basis, thousand TOE of the energy-supply were consumed by the sector of cycle and energy, thousand TOE were consumed by the sectors of industrial, residential and services, transport, agriculture and non energy use. The highest energy demand took place in the industrial sector and the residential and service sector (Tab. 1). Generally, the industrial sector in the energy consumption have increased since 2002, although in 2008 and 2009 the production decreased due to the global economic crisis in Turkey. However, the value of energy consumption of industrial sector has continued to rise since 2010 (WEC-TNC). 58 Tab. 1: Sectoral energy consumption of Turkey in 2013 (EBST 2013) Sectors Energy consumption Rate (Thousand TOE) (%) Industry 30, Transportation 22, Housing (Residence) 31, and Services Agriculture Non-Energy Usage Cycle and Energy 30, II.2. The structure of the electricity sector in Turkey According to Energy Balance Sheets of Turkey (EBST), electricity generation and consumption increased more than threefold since The gross electricity demand of Turkey increased by 6.8% annually from GWh (Gigawatt-hours) in 1990 to GWh in The total installed capacity of the power industry is approximately 49.5 GW (Gigawatt) at the end of 2013, while it was 16.3 GW in The annual growth rate is about 5.8% (Fig. 2) Electricity Energy Generation (GWh) Installed Capacity (MW) Fig. 2: Turkey's installed capacity and electricity energy generation Production and consumption in 2013, compared to 2012, increased by 0.3% and 1.6% respectively. As seen in Tab. 2., generation and consumption increased by 0.3% and 1.6% respectively. When we look at the past five years, the change in consumption and peak demand was the average annual level of 5.6%. Compared with the previous five years, the decline of increase rate is realized significantly. However, growth in the installed capacity has continued and reached MW with an increase 12%. Electricity Generation Company (EGC) with its subsidiaries in the production has a share by 34%. As of 2013, the total share of public sector in the market was 60% and the share of free market production was 40%. The weight of natural gas continues in electricity generation. Consumption of natural gas increased significantly from 1990 to 2013 while the share of natural gas in electricity generation increased from

73 about 18% in 1990 to 43.8% in Hydro, lignite and imported coal power plants had 25%, 13% and 12% respectively. Tab. 2: General electricity generation and consumption of Turkey (EBST 2013) Unit change (%) change (%) Installed MW capacity Peak Demand MW Generation GWh İmportation GWh Exportation GWh Consumption GWh According to Energy Balance Sheets of Turkey (EBST), electricity consumption per capita increased more than threefold since 1990, as shown in Fig. 3. Net electricity consumption per capita was 2577 kwh/person and 2568 kwh/person in 2012 and Gross electricity consumption per capita was 3205 kwh/person and 3132 kwh/person in 2012 and 2013 respectively. LEAP based on a comprehensive accounting such as, energy production, conversion and consumption in a particular region or economy under conditions of alternative assumptions based on population, economic development, technology, price and so on. With the program, possible future problems are identified; perspective is created for the energy supply and demand in the future years by evaluating possible effects of energy policy and it allows energy planning and policy from today. In addition, an environmental assessment of greenhouse gas emissions arising from energy use can be done. LEAP contains technology and environment database. It allows extensive information indicated the impacts of the environment, cost, technical characteristics of energy technology and also it allows to do projections of energy supply and demand for long-term planning. Fig. 4: The structure of LEAP (Song et al. 2007). Fig. 3: Electricity consumption per capita by the years When the final electricity demand by sector of Turkey for 2013 is analyzed, Turkey s energy end-use is dominated by the industrial sector, which takes up about 45% of total end-use electricity consumption while it was 62.4% in 1990 and has an average annual growth rate of about 5%. The residential sector accounts for about 25%, the commercial and services sectors follow by a cumulative 27% in III. Methodology This study uses an accounting and scenario-based modeling platform called LEAP to assess the impact of energy consumption and energy supply processes in Turkey. Long Range Energy Alternatives Planning (LEAP) system is an energy environment modeling tool developed at Stockholm Environment Institute (SEI), Boston, to assess the effects physical, economic and environmental of alternative energy programs, technologies and other energy initiatives by Song et al. (2007). 59 The structure of LEAP is presented in Fig. 4. In this approach, the LEAP software tool is used to analyze the current energy patterns and to simulate alternative energy futures along with environmental emissions under a range of user-defined assumptions. LEAP consists of four modules: energy scenarios, aggregation, environmental data base and fuel chain. Each module makes it possible to analyze extending impact by technology and policy change through a description of natural resources structured as energy sector, conversion course, final energy form and final energy demand. Moreover, it is possible to analyze for resource, conversion, demand and environmental emission through scenario analysis by (Song et al. 2007). Gross domestic product (GDP), which expresses the income level of countries is an entry data in the creation of projections. In LEAP model, the GDP PPP and GDP MER datas for the value of GDP which are gotten from the World Bank, were used. Turkey Statistical Institute population value for population, which is another factor in the incresing of energy demand is used in LEAP model. For this data,

74 scenarios was created using the growth rate of population in the newsletter which is "the Demographic Structure and Future of Turkey, " published by TUIK. All of the datas were gotten form Energy Balance Sheets of Turkey (EBST), and for which were entered for years in LEAP model. Fig. 6: The fuel ratios of enery demand According to the BAU scenario, the energy consumption of housing and services sector will be million GJ at the end of 2023, while it was million GJ in Compared with the 2011 value, the total growth rate is about 45% in The energy consumption of industry sector will reach million GJ at the end of 2023, while it was million GJ in Compared with the 2011 value, the total growth rate is about 40.5% in In the agricultural sector, energy demand will be million GJ with a 117% increase at the end of In the transportation sector, energy demand will be million GJ with a 28.5% increase at the end of In the non energy use sector, energy demand wil be million GJ with a 103% increase at the end of 2023 (Tab. 3). Fig. 5: Energy system diagram of LEAP model IV. Results and discussions IV.1. Energy Demand Scenarios IV.1.1. Business As Usual (BAU) Scenario The Business As Usual represents the energy pathway that is implied if current energy policies, supply and demand trends in Turkey persist. This includes basically economic growth and energy conversion. Current trends in the Turkish economy and the power sector continue in the BAU Scenario by Ozer et al. (2013). For the growth rate of Turkey's population and the growth rate of GDP MER, 1.25% and 2.9% are used in scenario. Accordingly, it is estimated that Turkey's population will reached to 86.7 million people and GDP MER will reached to billion $ in Scenarios were developed in based the historical development of fuels from 1985 until 2012 for energy demand sectors. According to the scenario, Turkey's total energy demand in 2023 will be million GJ (Gigajoule). Electricty, natural gas and oil energy demand will reached to million GJ, million GJ and million GJ respectively in 2023 (Fig. 6). Tab. 3: Sectoral energy consumption of Turkey in next years Sectors (MGJ) Housing (Residence) and Services 1254, , , , ,93 Agriculture 240,97 292,76 355,68 432,13 525,00 Industry 1348, , , , ,61 Transportation 667,80 708,98 754,07 803,58 858,12 Non-Energy Usage 185,99 222,14 265,32 316,89 378,49 Total 3698, , , , ,15 According to Tab.4, it was predicted that the electricity demand of housing and services sector will reached to TWh with about 54% growth, the electricity demand of industry sector will reached to TWh with about 49.5% growth in The electricity demand of agriculture sector and transportation sector will reached to TWh and 0.42 TWh respectively at the and of Tab. 4: Sectoral electricity consumption of Turkey in next years Sectors (TWh) Housing (Residence) and Services 88,07 97,06 107,35 120,75 136,17 Agriculture 4,36 6,10 8,51 11,86 16,49 Industry 93,90 104,26 114,52 126,56 140,24 Transportation 0,53 0,34 0,37 0,39 0,42 Total 186,86 207,76 230,75 259,56 293,32 60

75 In Turkey's energy demand, the most destructive gas is CO2 emissions. According to BAU scenario,as a result of energy consumption, it is estimated that greenhouse gas affect will reach to million metric tons of CO2 at the end of Tab. 5: Greenhouse gas affect IV.1.2. Mitigation Scenario Economic and demographic situation are jumping the shark in the mitigation scenario. It is assumed that the develepment of Turkey's economy slowed towards Hence, the growth rate of Turkey's population has been 1%. The growth rate of GDP MER has been as "Growth (2,9%; 2015; 2,5%; 2018; 1,5%; 2020; 0,5%; 2023; -0,5%)" in projection. The energy demand of housing and services sector is decreased 1% annually, the energy demand of industry sector remains constant and the enerdy demand of non-energy use sector is increased by 0.5% growth. Accordingly, it is estimated that Turkey's population will reached to 84.2 million people and GDP MER will reached to billion $ in According to the mitigation scenario, the energy consumption of housing and services sector will be million GJ at the end of 2023, while it was million GJ in Compared with the 2011 value, the total decrease rate is about 4.5% in The energy consumption of industry sector will be million GJ at the end of 2023, while it was million GJ in Compared with the 2011 value, the total decrease rate is about 0.3% in In the agricultural sector, energy demand wil be million GJ with a 29.6% increase at the end of In the transportation sector, energy demand wil be million GJ with a 11.3% increase at the end of In the non energy use sector, energy demand wil be million GJ with a 29.6% increase at the end of 2023 (Tab. 6). According to Tab.7, it is estimated that the electricity demand of housing and services sector will reached to TWh, the electricity demand of industry sector will reached to TWh in The electricity demand of agriculture sector and transportation sector will reached to 9.81 TWh and 0.37 TWh respectively at the end of Tab. 7: Sectoral electricity consumption of Turkey in next years Sectors (TWh) Housing (Residence) and Services 88,07 87,10 86,45 87,26 88,31 Agriculture 4,36 5,55 6,97 8,40 9,81 Industry 93,90 95,69 96,47 97,85 99,52 Transportation 0,53 0,34 0,36 0,37 0,37 Total 186,86 188,68 190,25 193,88 198,01 According to mitigation scenario, it is estimated that greenhouse gas affect will reach to million metric tons of CO2 at the end of Tab. 8: Greenhouse gas affect IV.2. Electricity Generation Scenarios IV.2.1. Hydro Scenario In this scenario, it is aimed that the installed capacity of hydro-based electricity generation will be MW in Accordingly, it can be seen from the Fig.7 that the electricity generation with hydro-power growth will be TWh at the end of 2023, while it was TWh in Tab. 6: Sectoral energy consumption of Turkey in next years Sectors (MGJ) Housing (Residence) and Services 1254, , , , ,55 Agriculture 240,97 266,51 291,33 306,18 312,34 Industry 1348, , , , ,42 Transportation 667,80 708,98 745,31 754,75 743,42 Non-Energy Usage 185,99 205,70 224,87 236,32 241,08 Total 3698, , , , ,31 61 Fig. 7: The fuel ratios of hydro scenario IV.2.2. Nuclear Scenario In addition to the increase of hydro power generation capacity, in the nuclear scenario, it is aimed that Akkuyu Nuclear Power Plant with 4800 MW capacity will be activated in 2019 and Sinop Nuclear Power

76 Plant with 1200 MW capacity will be activated in For these data, step function was used in modelling. was used in modelling. Results were shown at the Fig.10. According to nuclear scenario as seen in Fig.8, it is predicted that electricity generation will reach to TWh and the genaration from nuclear energy will be TWh. Fig. 10: The fuel ratios of total scenario Fig. 8: The fuel ratios of nuclear scenario According to total scenario, it is estimated as given in the Tab.8 that electricity generation will reach to TWh. The genaration from hydro, geothermal, wind, natural gas, solar and nuclear energy will be ; 3.68; 87.6; 12.67; 9.2; and TWh, respectively. Tab. 8: A comparison of the value of electricity generation scenarios Scenario (TWh) Hydro 240, , , ,625 Geo and wind 246, , , ,243 Nuclear 240, , , ,465 Total 247, , , ,340 Fig. 9: The fuel ratios of geo and wind scenario IV.2.3. Geo and Wind Scenario In addition to the increase of hydro power generation capacity, in the geo and wind scenario, it is targetted that the value of geothermal energy in electricity generation in 2016 and 2023 will reached to 300 MW and 600 MW respectively. In addition that, the value of wind energy in electricity generation in 2016 and 2023 will reach to MW and MW respectively. According to the scenarios, the highest energy generation values were obtained with the total scenario which combined hydro, geothermal, wind, solar, natural gas and nuclear energy. When looking at the environmental emission values, the most environmentally friendly scenario was expected to take place with total scenario as shown in the Fig.11. According to this scenario, it is calculated as shown in the Fig.9 that electricity generation will reach to TWh. The generation from geothermal energy will be 3.68 TWh and the generation from wind energy will be 87.6 TWh. IV.2.4. Total Scenario Total scenario consists of increase of hydro-power, nuclear, geo and wind generation capacity. Also, it is aimed that the installed capacity of solar energy will be 3000 MW and natural gas installed capacity will be MW in For these data, step function 62 Fig. 11: The greenhouse effect of scenarios V. Conclusions Energy supply and demand forecast from today is significant for investors and energy planners to

77 evaluate energy policies and environmental policies. The importance of energy demand and electricity generation is dealt with in this study. The energy demand and electricity generation of Turkey were estimated according to population and GDP increasing rate up to Affect the energy consumption are analyzed. The modelling for demand of energy and electricity generation has been done in Long Range Energy Alternatives Planning system program. Two energy demand scenarios and three electricity generation scenarios were built up for energy consumption, power technology and environmental policies. In the scenarios generated, the results of energy demand were obtained primarily by making BAU scenario in which everything continues in conditions of today and mitigation scenario. In the electricity generation scenarios, hydro, nuclear, geo and wind and total, energy consumption and emission values were calculated by creating possible future plans with various policies. In this line of study, generating policies for the future and energy management will be possible by planning energy today. References Başkan, Ö., Haldenbilen, S., Ceylan, H,. The modeling of energy demand in transportation sector and sustainable policies (in Turkish). World Energy Council-Turkish National Committee, 10. Energy Congress of Turkey (2005). Cai, L., Guo, J., Zhu, L., China's future power structure analysis based on LEAP. Energy Sources, Part A: Recovery, Utilization and Environmental Effects, 35:22, (2013). Canyurt, O.E., Öztürk, H.K., Hepbaşlı, A., Utlu, Z., Genetic Algorithm (GA) approaches for the transport energy demand estimation: model development and application. Energy Sources, Part A: Recovery, Utilization and Environmental Effects, 28: (2006). Dilaver Z., Hunt L.C., Turkish aggregate electricity demand. An outlook to 2020, Energy, 36: (2011). Egelioglu F., Mohamad A.A., Güven H. Economic variables and electricity consumption in Northern Cyprus. Energy, 26: (2001). EIA (Energy Information Administration), Online at: U, [accessed: ] IEA (International Energy Agency), Online at: [accessed: ] Kavak K., Energy Efficiency in the world and in Turkey and the analysis of energy efficiency in the Turkish industrial, Dissertation Thesis, State Planning Organisation (in Turkish), publication number: SPO:2689, Ankara (2005). 63 Lin B., Ouyang X., A revisit of fossil-fuel subsidies in China: challenges and opportunities for energy price reform, Energy Conversion and Management, 82, (2014). MEANR, Republic of Turkey Ministry of Energy and Natural Resources, Ankara. Online at: US/Mainpage, [accessed: ] Ozturk H.K., Ceylan H., Canyurt O.E., Hepbaslı A., Electricity estimation using genetic algorithm approach: a case study of Turkey, Energy, 30: (2005). Ozturk H.K., Ceylan H., Forecasting total and industrial sector electricity demand based on genetic algorithm approach: Turkey case study, International Journal of Energy Research, 29: (2005). SEI. LEAP long-range energy alternatives planning system; user guide for leap version Online at: 08UserGuideEnglish.pdf; [accessed: ]. Shin H., Park J., Kim H. ve Shin E., Environmental and economic assessment of landfill gas electricity generation in Korea using LEAP model, Energy Policy, 33, (2005). Song H.J., Lee S., Maken S., Ahn S.W., Park J.W., Min B., Koh W., Enviranmental and economic assessment of the hemical absorption process in Korea using the LEAP model, Energy Policy, 35: (2007). Zhang Q., Weili T., Yumei W., Yingxu C., External costs from electricity generation of China up to 2030 in energy and abatement scenarios, Energy Policy, 35: (2007). TUIK (Turkish Statistical Institute), TUIK Datas, Ankara. Online at: [accessed: ] WB (The World Bank), Datas and statistics for Turkey. Online at: [accessed: ] WEC-TNC, The Energy Balance Sheets of Turkey, Ankara. Online at: [accessed: ] WEC-TNC (World Energy Council-Turkish National Committee), Energy Report 2013, Ankara. Online at: ]

78 Evaluating Exergetic Sustainability Indicators for an Electrolyte Supported SOFC Stack Adnan Midilli*, Ugur Akbulut Department of Mechanical Engineering, Faculty of Engineering, Recep Tayyip Erdogan University, Rize, Turkey * amidilli@gmail.com Abstract In terms of exergy, this paper introduces and evaluates exergetic sustainability indicators for an Yttria Stabilized Zirconia (YSZ) electrolyte supported SOFC stack. For this purpose, the perfectly insulated SOFC stack is considered, having 300 μm-thickness-ysz electrolyte, 50 μm-thickness Ni-YSZ anode, 50 μm-thickness LSM-YSZ cathode, and consuming fuel as 97% H2+3% H2O and oxygen of the air, and operating at temperatures ranging from 1073 to 1273 K. In order to introduce exergetic sustainability indicators of such a stack, the following parameters are taken into account, which are total exergy input, desired exergy output, exergy destruction, exergy outputs by unused hydrogen and air, and exergy output by water. For better understanding and evaluating the critical points of the exergetic sustainability indicators as a function of current density, the critical exergetic efficiency is assumed to be 0.30 Accordingly, for making an evaluation of such a SOFC stack in terms of exergy based environmental and sustainability assessment, the following indicators should be taken into consideration, which are exergetic efficiency, waste exergy ratio, exergy destruction ratio, environmental effect factor and exergetic sustainability index. Under the selected operating conditions, it is noticed that waste exergy ratio increases with the rise of temperature. However, exergy destruction ratio and environmental effect factor decrease with temperature elevation at a constant current density while increasing with the rise of current density at a constant temperature. Moreover, exergetic sustainability index decreases with the increment of current density at a constant temperature while increasing with the rise of temperature at a constant current density. Considering the critical value of exergetic efficiency (=0.30), the critical environmental effect factor and the critical exergetic sustainability index are respectively determined to be and while the critical current density is A/cm 2 for 1073 K, A/cm 2 for 1173 K and A/cm 2 for 1273 K. Thus, it is suggested that, in order to exergetically operate the YSZ electrolyte supported SOFC stack under the selected design and operating conditions in accordance with the required electricity generation and fuel consumption, i) environmental effect factor should not be higher than 2.320, ii) exergetic sustainability index should not be lower than 0.428, iii) current density should be selected lower than A/cm 2 at 1073 K, A/cm 2 at 1173 K and A/cm 2 at 1273 K. Keywords: YSZ electrolyte supported SOFC, exergetic efficiency, waste exergy ratio, exergy destruction ratio, environmental effect factor, exergetic sustainability index I. Introduction Solid oxide fuel cells, commonly manufactured to be tubular and planar SOFCs, has recently attracted considerable interest for domestic and industrial applications. The most important features of these cells can be compiled as higher energy efficiency, lower pollutant emissions, fuel flexibility, and high operating temperature which allows a variety of cogeneration possibilities (Singhal, 2002; Khaleel et al., 2004; Xue et al., 2005; Ni et al., 2007; Hussain et al., 2009; Ahn et al., 2009; Xu et al., 2014). In terms of the energy conversion management, it is generally said that there are three types of the SOFCs which are anode supported SOFCs, cathode supported SOFCs and electrolyte supported SOFCs. In electrolyte-supported SOFCs, the thickness of YSZ electrolyte is generally between 150 ~ 300 μm, which operates during 800~1000 C (Singhal and Kendall, 2004; Han et al., 2010). In anode-supported SOFCs, the thickness of YSZ electrolyte is generally 15~30μm, which operates during 600~800 C (Singhal and Kendall, 2004; Han et al., 2010). In an electrolyte-supported SOFCs, the ohmic effect of the electrolyte layer should be also taken into account for energetic or exergetic performance improvement because electrolyte is the thickest component. Moreover, exergy based environmental and sustainability assessments of these fuel cells should be investigated. In this regard, it can be said that in order to ensure the exergetic sustainability, a clean and abundant fuel such as hydrogen should be used in the SOFCs. Of the most promising energy carriers for the future, hydrogen is more versatile, an energy-efficient, low polluting, environmentally benign fuel that will meet most of our energy needs in near future (Dincer, 2002). If so, it can be said that use of renewable hydrogen in the SOFCs can help reduce environmental effect of the SOFCs and achieve exergetic sustainability. Under these important considerations, in order to introduce and evaluate the exergy based environmental and sustainability assessment of an YSZ electrolyte supported SOFC stack, the following exergetic sustainability indicators, which are derived based on the operating principle of the YSZ electrolyte supported SOFC stack, should be taken into account; i) exergetic efficiency, ii) waste exergy ratio, iii) exergy recoverability ratio, iv) exergy 64

79 destruction ratio, v) environmental effect factor, vi) exergetic sustainability. Moreover, some effective parameters are considered, which are i) operating temperature (ranging from 1073 K to 1273 K), ii) operating pressure (=1 atm), iii) anode and cathode thicknesses (= 50 μm each) and iv) current density (ranging from 0 to 3 A/cm 2 ), v) electrolyte thickness (ranging from 100 to 500 μm). For this purpose, a detailed literature review has been carried out by considering the evaluation of exergetic sustainability indicators for an YSZ electrolyte supported SOFC stack. However, it is noticed that no studies are conducted on this subject. Actually, this lack of information indicating the originality of this paper is the motivation behind this work. However, some works have been found on exergetic sustainability indicators (e.g. Midilli and Dincer, 2009; 2010; Midilli et al., 2010; Kucuk and Midilli, 2015; Ozsaban and Midilli, 2016). Considering such important facts, as the scientific and industrial benefits, this study, which includes all details on exergy-based sustainability of a high pressure hydrogen production and storage system, aims to contribute to find out new exergetic dimensions and the exergy based-environmental and sustainability aspects of an YSZ electrolyte supported SOFC stack. II. Main Considerations In order to introduce the exergetic sustainability indicators of an YSZ electrolyte supported SOFC stack, the following required assumptions and parameters have been taken into consideration. II.1. Assumptions The following assumptions have been considered for this evaluation: An electrolyte supported SOFC stack is considered. Hydrogen is used as a fuel and only hydrogen is electrochemically reacted. Fuel consists of 97% H2 and 3% H2O and air as oxidant consists of 79% N2, 21% O2 (Costamagna et al., 2004; Ni et al., 2007; 2009; Yonekura et al., 2011; Ranjbar et al., 2014). The chemical exergy of nitrogen is not taken into account because it is inert gas and not a function of temperature. The fuel cell is insulated perfectly, so there is no heat interaction with environment. Temperatures at channel inlets and exits are the same (Colpan et al., 2007; Ranjbar et al., 2014). Radiation heat transfer between gas channels and solid structures is neglected (Ranjbar et al., 2014). Contact resistances are ignored. Radiation transfer between solid structure and gas channels is ignored. The electrodes and electrolyte materials of the SOFC are taken to be Ni-YSZ/YSZ/LSM-YSZ (Chan et al., 2002; Costamagna et al., 2004; Singhal and Kendall, 2004; Colpan et al., 2007; Han et al., 2010; Zheng et al., 2014). SOFC stack operates under steady-state conditions (Hussain et al., 2006; Colpan et al., 2007). Kinetic and potential exergies are neglected. SOFC operating pressure is taken to be 1 atm (Costamagna et al., 2004, Colpan et al., 2007; Mirahmadi et al., 2011; Verma et al., 2013). Dead state pressure is 1 atm and dead state temperature is 298 K. SOFC operating temperatures are taken between 1073 and 1273 K, and membrane thickness is selected between 100 and 500 μm considering the values in the literature (Singhal and Kendall, 2004; Han et al., 2010). Each reactant in SOFC is an ideal gas (Chan et al., 2002; Larminie and Dicks, 2003; Hussain et al., 2006; Colpan et al., 2007; Tanim et al., 2014). Flow of reactants is steady, incompressible and laminar. Pressure drops along the fuel cell are neglected (Hussain et al., 2006; Colpan et al., 2007; Tanim et al., 2014). The product water is in vapor phase. Current density is taken from 0 to 3 A/cm 2. The utilization ratios of hydrogen and oxygen are taken to be 80% and 50%, respectively (Ishak et al., 2012; Tanim et al., 2014) Active area of a single cell is selected to be 100 cm 2 and a stack is composed of 100 cells. All activation, ohmic, and concentration polarizations are considered. All values for hydrogen, oxygen, nitrogen and water vapor are taken from the NIST (Website) Tab. 1: The parameters taken for the calculations Parameter Value Ref. Operating temperature K (Singhal and Kendall, 2004; (T) Han et al., 2010). Operating pressure (P) 1 atm (Mirahmadi and Valefi, 2013) Current density (J) 0-3 A/cm 2 Assumed. Faraday constant (F) A.s/mol (Verma et al.,2013) Number of electrons per 2 (Costamagna et al., 2004) mol (ne) Anode thickness (La) 50 μm (Ni et al., 2007; Verma et al, 2013) Electrolyte thickness μm (Singhal and Kendall, 2004; (Le) Han et al., 2010). Cathode thickness (Lc) 50 μm (Ni et al., 2007) Porosity (ε ) 30% Tortuosity (ξ) 6 (Ishak et al., 2012) Pore radius (r p ) 0.5 μm Diffusion volume of 6.12x10-6 m 3 /mol hydrogen (v H2 ) Diffusion volume of 13.1x10-6 m 3 /mol (Zheng et al., 2014) water (v H2O ) Diffusion volume of 16.3x10-6 m 3 /mol oxygen (v O2 ) Diffusion volume of 18.5x10-6 m 3 /mol nitrogen (v N2 ) Molar fraction of %97 hydrogen (Yonekura et al., 2011) Molar fraction of water %3 Molar fraction of oxygen %21 II.2. Main Calculations In order to perform this work, the required equations for the specific calculations related to the electrolyte supported SOFC stack are taken from the literature, which are not presented in this study. 65

80 1. The values of ohmic, activation and concentration overpotentials, total cell overpotential (irreversible cell voltage) and net cell voltage can be calculated from the literature (Colpan et al. 2008; Kazempoor et al., 2010; Costamagna et al, 2004; Tanim et al., 2014). 2. Assuming that a single cell has 100 cm 2 active area and a stack is composed of 100 cells, total active area of the SOFC stack can be calculated from the literature (Trendewicz and Braun, 2013). 3. In order to estimate the molar flow rate of hydrogen, the fuel utilization factor (UF) can be calculated from the literature (Campanari, 2001; Ishak et al., 2012). 4. The molar flow rates of hydrogen consumed in the SOFC, and the molar flow rates of water output, oxygen and nitrogen at the air channel inlet and exit can be calculated from the literature (Colpan et al., 2007). 5. The heat generated from the SOFC stack can be calculated by considering the general energy balance equation for a SOFC stack (Chan et al., 2001). II.3. System Description The control volume of an YSZ electrolyte supported SOFC stack can be illustrated as below. H assumptions), v) E x 2 O out includes chemical and W physical exergies of water vapor, vi) E x d,out includes exergy of electricity generated by the stack. Fig. 2: Exergy balance diagram of an YSZ electrolyte supported SOFC stack II.4. Exergy Balance of the Stack Considering Fig. 2, the general exergy balance equation can be written as below in terms of the Second Law of Thermodynamics. Fig. 1: Control volume of an YSZ electrolyte supported SOFC stack In this figure, fuel is hydrogen; desired output is electricity generated by the stack; air includes mostly oxygen and nitrogen; and H2O is in vapor phase while crossing the boundary of control volume; heat loss will be equal to zero because it is assumed that the stack is perfectly insulated. However, the heat loss should be taken into account if it occurs in case of practical applications. Considering the general control volume of an YSZ electrolyte supported SOFC stack, general exergy balance diagram is shown in Fig. 2. In Fig. 2, i) E x in fuel includes chemical and physical fuel exergies of hydrogen gas, ii) E x uu includes chemical and physical exergies of unused hydrogen air gas, iii) E x in includes chemical and physical air exergies of air (see assumptions), iv) E x uu includes chemical and physical exergies of unused air (see 66 E x in - E x out = E x loss (1) In this equation, total exergy output ( E x out ) includes desired output (E x W d,out ) from the stack while total exergy loss ( E x loss ) can sometimes include waste exergy and exergy destruction based on the operating principle of the system (Midilli and Dincer, 2009). In this regard, Eq. (1) can be written as E x in - E x d,out = E x w + E x D (2) In Eq. (2), E x w can consist of recoverable waste exergy ( E x rw ) and unrecoverable waste exergy (E x urw ) (Midilli and Dincer, 2009; 2010). If so, Eq. (2) can be written as E x in - E x d,out = E x rw + E x urw + E x D (3) Considering Eq. (3) and assumptions, general exergy balance of the stack can be derived as ch E x in,h2 ch (Ėx uu,h2 ph + E x in,h2 ph + E x uu,h2 ch + E x in,o2 ph E x out,h2o ) + (E x uu,o2 ph + E x in,o2 ph + E x in,n2 W = ph ph + E x uu,o2 + E x out,n2 + E x out,h2o ch ) + E x D (4) ch + Under these considerations, exergetic sustainability parameters depending on the exergy parameters in Eq. (4) can be derived as below. III. Exergetic Sustainability Parameters How to derive the exergetic sustainability indicators

81 for a system or process? In order to derive the exergetic sustainability indicators for a system or a process, the steps summarized below should be basically followed, i) detailed operating principle is determined, ii) detailed control volume including all inputs and outputs is drawn, iii) operating parameters and assumptions are determined, iv) mass balance equation is written, v) energy analysis is achieved in terms of the First Law of thermodynamics, vi) exergy analysis is achieved in terms of the Second Law of thermodynamics, vii) Exergy efficiency is defined as the ratio of exergy of desired output to total exergy input, viii) Waste exergy ratio is defined as the ratio of total waste exergy output to total exergy input. Here, total waste exergy output should contain all exergy outputs from the system or process to reference environment, which should not include exergy destruction, ix) Exergetic recoverability ratio is defined as the ratio of the recoverable exergy from total waste exergies to total exergy input. Here, recoverable exergy from total waste exergy output should cover the exergies that are possible to be reused for the same system or same process, and/or for any system or any process, x) Exergy destruction ratio is defined as the ratio of exergy destruction in the system or process to total exergy input. Here, exergy destruction is determined through the exergy analysis of the system or process, xi) Environmental effect factor is defined as the ratio of total waste exergy output to exergetic efficiency, xii) Exergetic sustainability index is defined as the reverse of environmental effect factor. More details on derivation, conceptual and physical meanings and mathematical formulations of these indicators were presented in the literature (Midilli and Dincer, 2009; 2010; Midilli et al., 2012). Therefore, mathematical derivation procedure of these parameters will not be presented in this study. However, main definitions of these parameters will be introduced here as below. i) Exergetic efficiency (ee) Total exergy of useful output ee (5a) Total exergy input In this regard, exergetic efficiency of the SOFC is mainly based desired output exergy, total exergies of hydrogen and air. E x (ee) = W = E x in,h2 +E x in,air W E x ch ph in,h2 +E x in,h2 +E x ch ph in,o2 +E x in,o2 +E x in,n2 ph (5b) where E x W indicates exergy of SOFC stack power ch ph ch ph output while E x in,h2, E x in,h2, E x in,o2, E x in,o2 and ph E x in,n2 represents chemical and physical exergy of hydrogen, oxygen and nitrogen gas entering the SOFC stack, respectively. ii) Waste exergy ratio (wer) Total waste exergy output wer (6a) Total exergy input Waste exergy consists of physical and chemical exergies of unused hydrogen, unused oxygen, nitrogen, water vapor leaving the SOFC is taken into consideration. In this regard, waste exergy ratio (ranging from 0 to 1) can be written as wer = ch ph ch ph ph ch ph E x uu,h2 +E x uu,h2 +E x uu,o2 +E x uu,o2 +E x out,n2 +E x out,h2o +E x out,h2o E x ch ph in,h2 +E x in,h2 +E x ch ph ph in,o2 +E x in,o2 +E x out,n2 (6b) ph ch where E x out,h2o and E x out,h2o indicates physical and chemical exergies of water vapor, respectively. iii) Exergy recoverability ratio (err) Total re coverable exergy err (7a) Total exergy input Exergy recoverability ratio (ranging from 0 to 1) indicates the exergy potential that is possible to be recovered in the system. If so, it includes i) physical exergies of unused hydrogen, unused oxygen, nitrogen, water vapor, and ii) the chemical exergies of unused hydrogen and water vapor leaving the SOFC. In this regard, exergy recoverability ratio can be written (err) = E ch x uu,h2 (7b) ch +Ėx uu,h2o E x ch in,h2 ph ph ph ph +E x uu,o +E x 2 out,n2 +E x uu,h2 +E x out,h2o ph +E x in,h2 +E x ch ph ph in,o2 +E x in,o2 +E x out,n2 ch ch where E x uu,h2 and E x uu,o2, indicates chemical exergy of unused hydrogen and oxygen output; ph ph ph E x uu,h2, E x uu,o2, E x out,h2o, represents the physical exergy of unused hydrogen, oxygen and water outputs respectively. iv) Exergy destruction ratio (edr) Total exergy destruction edr (8a) Total exergy input Exergy destruction ratio (ranging from 0 to 1) is a function of exergy destruction and total exergy input. (edr) = E x d E x ch ph in,h2 +E x in,h2 +E x ch ph in,o2 +E x in,o2 +E x out,n2 v) Environmental effect factor (eef) eef ph (8b) Waste exergy ratio Exergy destruction ratio (9a) Exergetic efficiency Environmental effect factor (ranging from 0 to + ) indicates whether or not SOFC has some damage potential on the environment due to waste exergy output and exergy destruction. 67

82 (eef) = wer+edr = ch E x uu,h2 ee ph ch ph ph ch ph +E x uu,h2 +E x uu,o2 +E x uu,o2 +E x out,n2 +E x out,h2o +E x out,h2o +E x d W (9b) vi) Exergetic sustainability index (esi) Exergy of useful output esi (10a) Waste exergy output Exergy destruction Exergetic sustainability index (ranging from 0 to + ), which is defined as a function of environmental effect factor, reveals exergy-based sustainability of system or process in terms of the second-law of thermodynamics. (esi) = 1 eef = E x ch uu,h2 (10b) W ph ch ph ph ch ph +E x uu,h2 +E x uu,o2 +E x uu,o2 +E x out,n2 +E x out,h2o +E x out,h2o +E x d IV. Result and Discussion In this study, the main aim is to introduce and evaluate the exergetic sustainability indicators for an Yttria Stabilized Zirconia (YSZ) electrolyte supported SOFC stack. In this regard, the following investigations have been performed; i) variation of exergetic efficiency as a function of current density under various operating temperatures (see Fig. 3), ii) variation of environmental effect factor as a function of current density under various operating temperatures (see Fig. 4), iii) variation of exergetic sustainability index as a function of current density under various operating temperatures (see Fig. 5). Figure 3 introduces the variation of exergetic efficiency as a function of current density under various operating temperatures which are taken to be 1073, 1173 and 1273 K. It is assumed that anode, cathode and electrolyte thicknesses are taken to be 50, 50 and 300 μm, respectively while operating pressure is equal to 1 atm. As noticed from this figure, exergetic efficiency (ranging from 0 to 0.497) decreases with the rise of current density from 0.1 to 3 A/cm 2. At a constant current density, for example, for 0.3 A/cm 2, exergetic efficiency goes up from to with the increase of operating temperature from 1073 to 1273 K. Accordingly, in order to increase the exergetic efficiency of the YSZ electrolyte supported SOFC stack, lower current densities are targeted under the given design and operating conditions. The exergetic efficiency can be commercially assumed to be higher than at least 0.3. Considering this assumption and the critical point of current densities (0.271 A/cm 2 at 1073 K, 0.61 A/cm 2 at 1173 K, and 1.17 A/cm 2 at 1273 K) resulting from the operating temperatures, the graph can be separated in two region which are efficient operating region (exergetic efficiency is higher than 0.3) and inefficient operating region (exergetic efficiency is lower than 0.3). On the other word, it can be said that, considering the assumed reference line, commercially available operating region is upper side of the reference line. 68 Figure 3. Variation of exergetic efficiency as a function of current density under various operating temperatures. Figure 4 display the variation of environmental effect factor as a function of current density under various operating temperatures. It is assumed that anode, cathode and electrolyte thicknesses are taken to be 50, 50 and 300 μm, respectively while operating pressure is equal to 1 atm. Moreover, operating temperatures are taken to be 1073, 1173 and 1273 K. Environmental effect factor (eef) La=50 μm, Lc=50 μm, Le=300 μm, P=1 atm T=1073 K T=1173 K T=1273 K Current density (A/cm 2 ) efficient operating region critical environmental effect factor line Figure 4. Variation of environmental effect factor as a function of current density under various operating temperatures. As shown in this figure, environmental effect factor (ranging from to ) goes up with the rise of current density from 0.1 to 2.2 A/cm 2. At a constant current density, for example, for 0.3 A/cm 2, rising operating temperature (from 1073 to 1273 K) decreases environmental effect factor (from to 1.201). Considering the reference line of exergetic efficiency (as in Fig. 3) and the critical points of the current densities at the selected operating temperatures, the critical values of environmental effect factor are determined to be at 1073 K, at 1173 K, at 1273 K in case La=Lc=50 μm and Le=300 μm. Thus, under the selected design and operating conditions, in order to minimize exergy based environmental effect of the YSZ electrolyte supported SOFC stack, it should be operated at the lower values than almost of environmental effect factor.

83 Figure 5 displays the variation of exergetic sustainability index as a function of current density under various operating temperatures. It is assumed that anode, cathode and electrolyte thicknesses are taken to be 50, 50 and 300 μm, respectively while operating pressure is equal to 1 atm. Moreover, operating temperatures are taken to be 1073, 1173 and 1273 K. As shown in this figure, exergetic sustainability index (ranging from 0.04 to 0.988) decreases with the increase of current density (from 0.1 to 2.2 A/cm 2 ). At a constant current density, for example, for 0.3 A/cm 2, exergetic sustainability index increases (from to 0.833) with the rise of operating temperature from 1073 to 1273 K. In order to increase exergetic sustainability index the current density should be decreased while operating temperature increases. Considering the reference line of exergetic efficiency (as in Fig. 3) and the critical points of the current densities at the selected operating temperatures, the critical values of exergetic sustainability index are estimated to be at 1073 K, at 1173 K, at 1273 K in case La=Lc=50 μm and Le=300 μm. Thus, under the selected design and operating conditions, in order to maximize exergetic sustainability index of the YSZ electrolyte supported SOFC stack, it should be operated at the higher values than almost of the exergetic sustainability index. Exergetic sustainability index (esi) La=50 μm, Lc=50 μm, Le=300 μm, P=1 atm Efficient operating region Current density (A/cm 2 ) citical exergetic sustainability index line T=1073 K T=1173 K T=1273 K Figure 5. Variation of exergetic sustainability index as a function of current density under various operating temperatures. V. Conclusion In this paper, exergetic sustainability indicators were introduced and evaluated for an Yttria Stabilized Zirconia (YSZ) electrolyte supported SOFC stack. In this regard, the following concluding remarks can be drawn: Exergetic sustainability index decreases with the rise of current density at a constant temperature while going up with the increase of operating temperature at a constant current density. Considering the critical value of exergetic efficiency (=0.30), the critical current densities are determined to be A/cm 2 for 1073 K, A/cm 2 for 1173 K and A/cm 2 for 1273 K. 69 Considering the reference line of exergetic efficiency, the critical values of environmental effect factor are determined to be at 1073 K, at 1173 K, at 1273 K in case La=Lc=50 μm and Le=300 μm. Considering the reference line of exergetic efficiency, the critical values of exergetic sustainability index are estimated to be at 1073 K, at 1173 K, at 1273 K in case La=Lc=50 μm and Le=300 μm. Consequently, it can be said that, under the selected design and operating conditions, in order to minimize environmental effect factor of the YSZ electrolyte supported SOFC stack, it should be operated at the lower values than of the critical environmental effect factor. Moreover, under the selected design and operating conditions, in order to maximize exergetic sustainability index of the YSZ electrolyte supported SOFC stack, it should be operated at the higher values than of the critical exergetic sustainability index. Thus, it is suggested that, in order to exergetically operate the YSZ electrolyte supported SOFC stack under the selected design and operating conditions in accordance with the required electricity generation and fuel consumption, i) environmental effect factor should not be higher than 2.320, ii) exergetic sustainability index should not be lower than 0.428, iii) current density should be selected lower than A/cm 2 at 1073 K, A/cm 2 at 1173 K and A/cm 2 at 1273 K. Nomenclature Symbols E x : Exergy rate (kw) W : Power (kw) Subscripts D : Destruction d,out : desired output in : Input w : waste References Ahn J.S., Pergolesi D., Camaratta M.A., Yoon H., Lee B.W., Lee K.T., Jung D.W., Traversa E., Wachsman E.D., High-performance bilayered electrolyte intermediate temperature solid oxide fuel cells, Electrochemistry Communications 11, (2009). Campanari S., Thermodynamic model and parametric analysis of a tubular SOFC module, Journal of Power Sources, 92, (2001). Chan S.H., Khor K.A., Xia Z.T., A complete polarization model of a solid oxide fuel cell and its sensitivity to the change of cell component thickness, Journal of Power Sources, 93, (2001). Chan S.H., Low C.F., Ding O.L., Energy and exergy analysis of simple solid-oxide fuel cell, power

84 systems, Journal of Power Sources, 103, (2002). Colpan C.O., Dincer I., Hamdullahpur F., Thermodynamic modeling of direct internal reforming solid oxide fuel cells operating with syngas, International Journal of Hydrogen Energy, 32, (2007). Colpan C.O., Dincer I., Hamdullahpur F., A review on macro-level modeling of planar solid oxide fuel cells. International Journal of Energy Research, 32, (2008). Costamagna P., Selimovic A., Borghi M.D., Agnew G., Electrochemical model of the integrated planar solid oxide fuel cell (IP-SOFC), Chemical Engineering Journal, 102, (2004). Dincer I., Technical, environmental and exergetic aspects of hydrogen energy systems, International Journal of Hydrogen Energy, 27, (2002). Han M.F., Liu Z., Zheng Z., Liu M., High performance solid oxide fuel cells based on tri-layer yttria-stabilized zirconia by low temperature sintering process, Journal of Power Sources, 195, , (2010). Hussain M.M, Li X., Dincer I., Mathematical modeling of planar solid oxide fuel cells, Journal of Power Sources, 161, , (2006). Hussain M.M, Li X., Dincer I., A general electrolyte electrode-assembly model for the performance characteristics of planar anode-supported solid oxide fuel cells, Journal of Power Sources, 189, (2009). Ishak F., Dincer I., Zamfirescu C., Energy and exergy analyses of direct ammonia solid oxide fuel cell integrated with gas turbine power cycle, Journal of Power Sources, 212, (2012). Kazempoor P., Dorer V., Ommi F., Modelling and Performance Evaluation of Solid Oxide Fuel Cell for Building Integrated Co- and Polygeneration, Fuel Cells, 10(6), (2010). Khaleel M. A., Lin Z., Singh P., Surdoval W., Collin D., A finite element analysis modeling tool for solid oxide fuel cell development: coupled electrochemistry, thermal and flow analysis in MARC, Journal of Power Sources, 130, (2004). Kucuk H., Midilli A., Assessment of exergetic sustainability indicators for a single layer solar drying system, International Journal of Exergy, 16(3), (2015). Larminie J.E., Dicks A., Fuel Cell Systems Explained, John Wiley and Sons, West Sussex (2003). Midilli A., Dincer I., Development of some exergetic parameters for PEM fuel cells for measuring 70 environmental impact and sustainability, International Journal of Hydrogen Energy, 34(9), (2009). Midilli A., Dincer I., Effects of some micro-level exergetic parameters of a PEMFC on the environment and sustainability, International Journal of global Warming, 2(1), (2010). Midilli A., Kucuk H., Dincer I., Environmental and sustainability aspects of a recirculating aquaculture system, Environ Prog Sustain, 31, (2012). Mirahmadi A., Valefi K., Study of thermal effects on the performance of micro-tubular solid-oxide fuel cells, Ionics, 17, (2011). Ni M., Leung M.K.H., Leung D.J.C., Parametric study of solid oxide fuel cell performance, Energy Conversion and Management, 48, (2007). Ni M., Leung D.J.C., Leung M.K.H., Electrochemical modeling and parametric study of methane fed solid oxide fuel cells Energy Conversion and Management, 50, (2009). Ozsaban M., Midilli A., A parametric study on exergetic sustainability aspects of high-pressure hydrogen gas compression, International Journal of Hydrogen Energy, 41(11), (2016). Ranjbar F., Chitsaz A., Mahmoudi S.M.S., Khalilarya S., Rosen M., Energy and exergy assessments of a novel trigeneration system based on a solid oxide fuel cell, Energy Conversion and Management, 87, (2014). Singhal S. C., Solid oxide fuel cells for stationary, mobile, and military applications, Solid State Ionics, , (2002). Singhal S.C., Kendall K., High Temperature Solid Oxide Fuel Cells: Fundamentals, Design and Applications, Elsevier Ltd, Oxford, (2004). Tanim T., Bayless D.J., Trembley J. P., Modeling a 5 kwe planar solid oxide fuel cell based system operating on JP-8 fuel and a comparison with tubular cell based system for auxiliary and mobile power applications, Journal of Power Sources, 245, (2014). Trendewicz A.A., Braun R.J., Techno-economic analysis of solid oxide fuel cell-based combined heat and power systems for biogas utilization at wastewater treatment facilities, Journal of Power Sources, 233, (2013). Verma J.K., Verma A., Ghoshal A.K., Performance analysis of solid oxide fuel cell using reformed fuel, International Journal of Hydrogen Energy, 38, (2013)

85 Website, accessed at March 20, Xue X., Tang J., Sammes N., Du Y., Dynamic modeling of single tubular SOFC combining heat/mass transfer and electrochemical reaction effects, Journal of Power Sources, 142, (2005). Xu H., Dang Z., Bai B.F., Electrochemical performance study of solid oxide fuel cell using lattice Boltzmann method, Energy, 67, (2014). Yonekura T., Tachikawa Y., Yoshizumi T., Shiratori Y., Ito K., Sasaki K., Exchange current density of solid oxide fuel cell electrodes, ECS Transactions, 35(1), (2011). Zheng K., Li L., Ni M., Investigation of the electrochemical active thickness of solid oxide fuel cell anode, International Journal of Hydrogen Energy, 39, (2014). 71

86 Life Cycle Assessment of Nuclear Based Ammonia Production Options: A Comparative Study Yusuf Bicer*, Ibrahim Dincer Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario L1H 7K4, Canada * yusuf.bicer@uoit.ca Abstract In this study, nuclear power based ammonia production options ranging from thermochemical cycles to high temperature electrolysis are comparatively evaluated using life cycle assessment (LCA) tool. Ammonia is produced by extracting nitrogen from air and hydrogen from water and combining them with the help of nuclear energy. Since production of ammonia contributes about 1% of global greenhouse gas (GHG) emissions, new methods with less environmental impact are under close investigation. Nuclear, as a sustainable energy source compared to conventional fossil fuels, emerges as an alternative option for ammonia synthesis. Within the current study, the selected ammonia production systems are (i) three step nuclear Cu-Cl thermochemical cycle, (ii) four step nuclear Cu-Cl thermochemical cycle, (iii) five step nuclear Cu-Cl thermochemical cycle, (iv) nuclear power based electrolysis integrated to Haber-Bosch process and (v) nuclear high temperature electrolysis integrated to Haber-Bosch process. Electrolysis units for hydrogen production and a Haber-Bosch process for ammonia synthesis are utilized for the electrolysis based options while hydrogen is produced thermochemically using excess heat in nuclear power plants for thermochemical based systems. The waste heat in nuclear power plant is also utilized for high temperature electrolysis in order to decrease the amount of required power for electrolysis process. Using the LCA methodology, the environmental impacts of selected ammonia synthesis methods are comparatively identified and quantified from cradle to gate. The life cycle assessment includes fuel elements, chemicals, and diesel requirements as well as the relevant transport requirements. Cryogenic air separation is mostly used method for massive amount of nitrogen production. In the life cycle assessment of nitrogen production, electricity for process, cooling water, waste heat and infrastructure for air separation plant are included. The LCA results for the selected ammonia production methods show that nuclear electrolysis based ammonia production method yields lower global warming and climate change impacts while thermochemical based options yield higher abiotic depletion and acidification values. Keywords: Ammonia production, nuclear, life cycle assessment, environmental impact. I. Introduction Ammonia is potentially treated as a significant hydrogen carrier with a much higher hydrogen content. In recent years, expectations are rising for hydrogen and hydrogen carriers as a medium for storage and transportation of energy and use of renewable energy. Transportation and storage issues of hydrogen are important as hydrogen is in gas form at ambient temperature and pressure. Ammonia is one of the major synthesized industrial chemicals in the world. Ammonia synthesis consumes almost 1.2% of total primary energy and contributes about 1% of global GHGs emissions (Gilbert et. al., 2010). Approximately 1.5 tonnes of CO2 is released to the environment during the production of 1 tonne of ammonia with the current technology (Anderson et. al., 2008). Natural gas is the primary feedstock used for producing ammonia worldwide via steam methane reforming. The delivery and storage infrastructure of ammonia is similar to liquefied petroleum gas (LPG) process. Under medium pressures, both of the substances are in liquid form which brings significant benefit because of storage options. Today, vehicles running with propane are mostly accepted and used by the public since their on-board storage is possible and it is a good example for ammonia fueled vehicle opportunities since the storage and risk characteristics of both substances are similar to each other. Zamfirescu and Dincer (2009) examined the use of ammonia as a clean fuel in evaluation with further conventional fuels. They defined the possible technical benefits of ammonia usage as a sustainable fuel aimed at power production on vehicles based on some efficiency indicators containing the system efficiency, the driving distance, fuel tank compactness and the price of driving. Verma and Kumar (2015) offered a model to evaluate life cycle GHG emissions in hydrogen production from underground coal gasification with and without carbon capturing. Utilization of carbon capturing technology permits a substantial reduction in total life cycle emissions in hydrogen production from underground coal gasification. Kalinci et al. (2012) performed a life cycle assessment of hydrogen production from CFBG/DG biomass production in 72

87 order to use the generated hydrogen in PEM fuel cell vehicles by investigating the costs of GHG emissions reduction. The extreme energy consumption rates were observed in the compression and transportation of hydrogen steps for the CFBG based system. Zamfirescu and Dincer (2008) reported a few possible occasions and benefits of using ammonia as a sustainable fuel in transportation vehicles. They have compared ammonia with other conventional fuels in different aspects. Moreover, using ammonia both as a refrigerant and a fuel, they calculated refrigeration effect with respect to refrigeration power vs engine s power ending up with that ammonia is the cheapest fuel on $/GJ basis. Cryogenic air separation is usually used method for massive amount of nitrogen production which is used in this study. In the life cycle assessment of nitrogen production, electricity for process, cooling water, waste heat and infrastructure for air separation plant are taken into account. The allocation factors were obtained from the heat of vaporization and the specific heat capacity multiplied with the temperature difference from 20 C to the boiling point. The utilized software, SimaPro, has the values of nitrogen production from cryogenic air separation process in the database. Makhlouf et al. (2015) presented the results of a life cycle assessment of 1 tonne of ammonia produced in Algeria considering anhydrous liquid ammonia. They specified that Algerian ammonia plant consumes more energy than world average. Reformer processes are the main reasons of overconsumption of energy and GHG emissions. This was because of the low effectiveness of the catalytic reaction in which the catalysts were used more than 10 years. A few of the available ammonia utilization pathways can be listed as follows (Dincer et. al., 2011): Direct feed of ammonia into an internal combustion engine Ammonia thermal cracking and feed of the products (H2 and N2) all together in the internal combustion engine cylinder for combustion Separation of N2 and H2 streams simultaneously with the decomposition such that only pure H2 is combusted; and the nitrogen is expanded for work production Direct ammonia high-temperature fuel cell systems, Ammonia thermal cracking and separation and further using the hydrogen into high temperature fuel cells Ammonia electrolysis and hydrogen used in proton exchange fuel-cells with additional exploitation of ammonia s refrigeration effect Fig. 1. Selected nuclear based ammonia production options In nuclear-based high-temperature ammonia production, the system consists of a nuclear power plant, high temperature electrolyzer, cryogenic air separation unit and a Haber-Bosch synthesis plant as shown in Fig. 2. The required electricity is utilized from nuclear power plant and the required heat for high temperature electrolysis is supplied from nuclear waste heat. Nuclear power plant electricity is assumed as a mixture of 66.5% from pressure water reactor (PWR) and 33.5% from boiling water reactor (BWR) type reactors since SimaPro software database does not include CANDU type reactors. Note that ammonia is also a suitable fuel for spark-ignition engines because of its high opposition to auto ignition. On the other hand, it is of great interest to utilize ammonia in compression-ignition engines due to the popularity of compression ignition engine-driven electricity generators. For internal combustion engines, service network is already available and ready in addition to mature manufacturing technology. II. Systems description In the present study, five different ammonia production methods are selected for comparative assessment purposes as illustrated in Fig. 1 where Haber-Bosch process is utilized for ammonia synthesis. 73 Fig. 2. Nuclear high temperature electrolysis and Haber-Bosch process for ammonia production

88 Nuclear based electricity yields lower cost and reliable supply. Combining nuclear power plant with ammonia production plant is an encouraging method. In high temperature electrolysis, the excess heat in the nuclear power plant is utilized to decrease the required amount of electricity for electrolysis as seen in Fig. 2. Nitrogen (N2) Water Uranium Nuclear Electrolysis & Haber-Bosch Ammonia Synthesis Ammonia (NH3) Nuclear Power Plant Electricity Fig. 3. Energy and material nuclear electrolysis based ammonia production On the other hand, in nuclear electrolysis based option, electricity is produced in nuclear power plant and directly utilized in electrolysis coupled with Haber-Bosch ammonia synthesis loop. There is no heat assisting in this method. Hence, more electrical energy is required to split water into hydrogen and oxygen. The schematic diagram of nuclear electrolysis based ammonia production option can be seen in Fig. 3. The copper-chlorine (CuCl) cycle is a multiple step thermochemical cycle for the production of hydrogen. The CuCl cycle is a combined process that employs both thermochemical and electrolysis steps. The CuCl cycle involves four chemical reactions for water splitting, whose net reaction decomposes water into hydrogen and oxygen. Both heat and electricity are provided at the same time for hydrogen generation and then hydrogen reacts with nitrogen to produce ammonia. Input of water and energy for the production of steam are included but other infrastructure is not included, as the heating infrastructure is already a part of the respective heating modules used in the plant. Nuclear power plant electricity is assumed as a mixture of 66.5% from PWR and 33.5% from BWR type reactors for this method, too. The life cycle assessment includes fuel elements, chemicals, and diesel requirements as well as the relevant transport necessities. Water use for cooling is also taken into account. Considered radioactive waste streams are: spent fuel to reprocessing and conditioning; operational low active waste for conditioning in the intermediate repository; and, contaminated waste from dismantling. Non-radioactive wastes are taken into account. The average burnup relates to an average enrichment of 3.8% U235 for fresh uranium fuel elements in BWR type reactor. The average burnup corresponds to an average enrichment of 4.2% U235 for fresh uranium 74 fuel elements in PWR type reactor. The diesel requirements for the yearly test of diesel emergency generators are accounted for. The transport requirements are calculated with the standard distances for chemical and diesel requirements and specific distances for fuel recharge and radioactive waste. Schematic diagram of energy and material flows of nuclear thermochemical CuCl cycle based ammonia production options are shown in Fig. 4. Nitrogen (N2) Water Uranium Nuclear CuCl Thermochemical Cycle & Haber-Bosch Ammonia Synthesis Ammonia (NH3) Nuclear Power Plant Heat Electricity Fig. 4. Energy and material flows of nuclear CuCl cycle based ammonia production III. Life Cycle Assessment (LCA) LCA is a methodology from cradle to grave. This tool helps to make effective decision by analyzing the system systematically. LCA analyses the environmental impact of a product or process over the length of its entire life, beginning from raw material extraction to final disposal. LCA deliberates all the life periods of product or process to assess the overall environmental impact. There are a number of assessment methods progressed over the time to categorize and characterize the environmental flows of system. In this study, LCA is performed using CML 2001 method which was proposed by a set of scientists under the principal of CML (Center of Environmental Science of Leiden University) including a group of impact classes and characterization procedures for the impact assessment phase in The environmental impact categories considered in this study are explained as follows: III.1 Depletion of abiotic resources The key concern of this category is the human and ecosystem health that is affected by the extraction of minerals and fossil as inputs to the system. For each extraction of minerals and fossil fuels, the Abiotic Depletion Factor (ADF) is defined. This indicator has globe scale where it is related with concentration reserves and rate of de-accumulation. III.2. Human toxicity Toxic substances on the human environment are the

89 core concerns for this category. In the working environment, the health risks are not included in this category. Characterization factors, Human Toxicity Potentials (HTP), are determined with (The Uniform System for the Evaluation of Substances) USES-LCA, describing fate, exposure and effects of toxic substances for an infinite time horizon. 1,4-dichlorobenzene equivalents/kg emissions is used to express each toxic substance. Depending on the substance, the geographical scale differs between local and global indicator [9]. III.3. Fresh water aquatic eco-toxicity This indicator considers the effect of the emissions of toxic substances to air, water, and soil on fresh water and ecosystems. USES-LCA is used to calculate the Eco-toxicity Potential by describing fate, exposure and effects of toxic substances. 1,4-dichlorobenzene equivalents/kg emissions is used to express infinite Characterization factors which is the time horizon. The scale of this indicator can be applied to global/continental/regional and local scale. Marine eco-toxicity is related to effects of toxic substances on marine ecosystems [9]. III.4. Ozone depletion Due to stratospheric ozone depletion, a bigger portion of UV-B radiation spreads the world surface. It may have damaging properties upon human health, animal health, terrestrial and aquatic ecosystems, biochemical cycles and on materials. The category is output related and it is at global scale. The model of characterization is advanced by the World Meteorological Organization (WMO) and describes ozone depletion potential of various gasses in unit of kg CFC-11 equivalent/kg emission. The geographic scope of this indicator is at global scale and the span of time is infinity [9]. III.5. Acidification potential Acidifying substances causes a wide range of impacts on soil, groundwater, surface water, organisms, ecosystems and materials. RAINS 10 model is used to calculate the Acidification Potential (AP) for emissions to air, describing the fate and deposition of acidifying substances. The Regional Air Pollution Information and Simulation (RAINS) model is a European-scale integrated assessment model dealing with air quality and associated effects. SO2 equivalents/kg emission is utilized to state the AP. This category has a different geographical scale that can be local and global. Depending on the availability the Characterization aspects containing fate were used. But, when not available, the aspects used without fate (In the CML baseline version only factors including fate were used). The method was stretched for nitric acid, water, soil, and air; sulphuric acid, water; sulphur trioxide, air; hydrogen chloride, water, soil; hydrogen fluoride, water, soil; phosphoric acid, water, soil; hydrogen sulphide, soil, all not including 75 fate. Nitric oxide, air (is nitrogen monoxide) was added containing fate [9]. III.6. Global warming The greenhouse gases to air are associated with the climate change. Adversative effects upon ecosystem health, human health and material welfare can result from climate change. The Intergovernmental Panel on Climate Change (IPCC) developed the characterization model which is selected for the development of characterization factors. A kg carbon dioxide/kg emission is used to express the Global Warming Potential for time horizon 500 years (GWP500). This indicator has a global scale [9]. III.7. Eutrophication This category reflects the impacts of to excessive levels of macro-nutrients in the environment caused by emissions of nutrients to air, water and soil. The stoichiometric procedure of Heijungs is the base of the Nutrification potential (NP) which is expressed as kg PO4 equivalents per kg emission and the geographical scale varies between local and continental scale, time span is infinity, and fate and exposure are not involved [9]. Overall environmental impact of any process is not complete if only operation is considered, all the life steps from resource extraction to disposal during the lifetime of a product or process should be considered. Mass and energy flows and environmental impacts related to plant construction, utilization, and dismantling stages are taken into account in LCA analysis [10, 11]. Using SimaPro software for life cycle analysis, cradle to grave considerations of various nuclear based ammonia production methods are investigated and comparatively assessed. IV. Results and discussion Various nuclear resources based ammonia production pathways are determined, and the energy and material requirement for each route are identified and calculated. The values are used in SimaPro software for the calculations of life cycle assessment. The calculations are based on one kg of ammonia end product. The environmetanl impact results are presented herein. The impact on human health due to human toxicity is maximum for the ammonia production from nuclear electrolysis method where it corresponds to 1.41 kg 1,4-DB-eq per kg of ammonia. Ammonia from nuclear thermochemical based methods yield lower human toxicity values where the lowest is 5 step cycle with a value of 0.80 kg 1,4-DB-eq as seen in Fig. 5.

90 Fig. 5. Human toxicity values of nuclear based ammonia production methods Fig. 7. Acidification values of nuclear based ammonia production methods Fig. 8. Eutrophication values of nuclear based ammonia production methods Fig. 6. Abiotic depletion values of nuclear based ammonia production methods The abiotic resources are natural resources including energy resources, such as iron ore and crude oil, which are considered as non-living. The abiotic depletion is highest for nuclear 5 step CuCl cycle method with a value of kg Sb eq. as it is illustrated in Fig. 6. Besides, nuclear electrolysis based option has the lowest abiotic depletion corresponding to kg Sb eq. In terms of global warming potential, nuclear based electrolysis option yields the lowest environmental impact with a value of 0.48 kg CO2 eq. GHG emission. However, it is very high (3.70 kg CO2 eq.) for the thermochemical based ammonia production as shown in Fig. 9. The nuclear-based high temperature electrolysis has higher global warming potential (0.84 kg CO2 eq.) than nuclear electrolysis. Nuclear electrolysis options have approximately same values with many renewable based ammonia production options found in the literature. The acidification values are lowest for nuclear electrolysis based option (0.002 kg SO2 eq.) followed by nuclear high temperature electrolysis method (0.003 kg SO2 eq.) as shown in Fig. 7. It is higher in thermochemical cycles because of used chemical substances. Eutrophication values are in parallel with acidification values in which nuclear 5 step CuCl cycle has about kg PO4 eq. as illustrated in Fig. 8. Fig. 9. Global warming values of nuclear based ammonia production methods 76

91 The ozone layer depletion is currently an important issue which needs to be decreased. In terms of ozone depletion, nuclear electrolysis yields highest environmental impact corresponding to 7.26E-07 kg CFC-11 eq. In terms of human toxicity, nuclear 5 step CuCl cycle has the lowest environmental impact. Nuclear electrolysis based options yield lower acidification, eutrophication and Abiotic depletion values among all methods. Global warming potentials of nuclear electrolysis and nuclear high temperature electrolysis are relatively lower compared to thermochemical cycles. Nuclear based ammonia production is a promising option especially when utilizing excess heat and electricity. Acknowledgement The authors acknowledge the support by the Natural Sciences and Engineering Research Council of Canada and the Mitacs. Fig. 10. Ozone layer depletion values of nuclear based ammonia production methods The other methods have approximately same values where nuclear 5 step CuCl cycle represents a value of 5.12E-07 kg CFC-11 eq. Nomenclature BWR Boiling water reactor CCS Carbon capture storage CFBG Circulating fluidized bed gasifier DG Downdraft gasifier GHG Greenhouse gas HHV Higher heating value IPCC Intergovernmental panel on climate change LCA Life cycle analysis LPG Liquefied petroleum gas PV Photovoltaic PWR Pressurized water reactor SMR Steam methane reforming UCG Underground coal gasification USES The Uniform System for the Evaluation of Substances Fig. 11. Marine sediment ecotoxicity values of nuclear based ammonia production methods The marine eco-toxicity refers to impacts of toxic substances on marine ecosystems. Highest environmental impact is observed in 3 step CuCl thermochemical cycle which corresponds to 9.11 kg 1,4-DB eq. On the other hand, nuclear high temperature electrolysis occurs to be most environmentally benign method (7.28 kg 1,4-DB eq.) in terms of marine sediment ecotoxicity. V. Conclusions A life cycle assessment of the nuclear ammonia production methods is conducted and environmental impacts are comparatively assessed. The following concluding remarks can be written for this study: 77 References Gilbert P, Thornley P. Energy and carbon balance of ammonia production from biomass gasification. Poster at Bio-Ten Conference, Birmingham Anderson K, Bows A, Mander S. From long-term targets to cumulative emission pathways: Reframing UK climate policy. Energy Policy. 2008;36(10): Zamfirescu C, Dincer I. Ammonia as a green fuel and hydrogen source for vehicular applications. Fuel Processing Technology. 2009;90(5): Verma A, Kumar A. Life cycle assessment of hydrogen production from underground coal gasification. Applied Energy. 2015;147(0): Kalinci Y, Hepbasli A, Dincer I. Life cycle assessment of hydrogen production from biomass gasification systems. International Journal of Hydrogen Energy. 2012;37(19): Zamfirescu C, Dincer I. Using ammonia as a sustainable fuel. Journal of Power Sources. 2008;185(1): ) Makhlouf A, Serradj T, Cheniti H. Life cycle impact

92 assessment of ammonia production in Algeria: A comparison with previous studies. Environmental Impact Assessment Review. 2015;50: Dincer I, Zamfirescu C. Apparatus for using ammonia as a sustainable fuel, refrigerant and NOx reduction agent. Google Patents; Consultants P. SimaPro Life Cycle Analysis Database version 7.3 (software). International Organization for Standardization (ISO) ISO Environmental Management - Life Cycle Assessment e Requirements and Guidelines, Dincer I, Rosen MA, Exergy: Energy, Environment and Sustainable Development, 2nd ed., Elsevier, NY

93 Energy and Exergy Efficiency Evaluations of R134a Clathrates with Additives for Cooling Applications Sayem Zafar* 1, Ibrahim Dincer 1, Mohamed Gadalla 2 1 Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario L1H 7K4, Canada 2 Department of Mechanical Engineering, American University of Sharjah, Sharjah, UAE * s: sayem55@hotmail.com Abstract An experimental investigation is performed to evaluate the energetic and exergetic efficiencies of R134a with additives, being treated as phase change materials (PCMs). PCMs charging and discharging characteristics are analysed and evaluated for cooling applications. PCMs are formed using R134a clathrate and distilled water with five different additives. The used additives are sodium chloride, magnesium nitrate hexahydrate, aluminum, copper and ethanol. The refrigerant mass percentage is maintained at 35% while the additive mass percentage at 1% is selected for all the PCMs. The discharge tests are also conducted in which the PCMs are used to cool the air. The main objective of using additives is, then, to study their potential for enhancing the R134a clathrate and exergy contents. The tests are also conducted to determine the energetic and exergetic efficiencies. A comparative assessment study is implemented to compare both energy and exergy efficiencies of different PCMs made up of suggested additives. The exergy destruction evaluations and thermoeconomic analyses are also performed. The present results indicate that the sodium chloride based PCM has the highest and magnesium nitrate hexahydrate based PCM has the lowest exergy destruction. The thermoeconomic analyses include the evaluation of thermoeconomic factor and cost-benefit analyses for the PCMs and the related parameters are studied for each PCM which included its energy, containment and PCM components costs. The ethanol additive is found to have the best overall efficiency when compared with other PCMs. It can be safely concluded that liquid additives are more feasible than other tested additives as they dissolve homogeneously and improve the phase change heat absorption. Finally, the thermoeconomic results show that magnesium nitrate hexahydrate based PCM has the highest thermoeconomic factor while sodium chloride based PCM has the lowest thermoeconomic factor. Keywords: R134a clathrate, phase change materials, cooling, energy, exergy, efficiency, thermoeconomics. I. Introduction The energy management is a challenge that needs to be dealt with in order to achieve the goal of sustainable development. In order to improve the performance of energy systems, more effective tools need to be utilized (Rosen et al. 1997). Exergy is defined as the maximum obtainable work producing ability by a system or a flow of matter as it comes to equilibrium. The exergy of an energy form or a substance is a measure of its usefulness or quality or potential to cause change (Rosen and Dincer, 1997). Exergy analysis is an effective thermodynamic method for design and analysis of thermal systems while it is an efficient technique for revealing the improvement capacity of thermal systems. With energy and exergy precisely known, it becomes easier to evaluate their efficiency for any given system. In thermodynamics, efficiency specifies the effectiveness of the energy conversion process. Efficiency is sometimes misunderstood and defined incorrectly. This is due to the fact that efficiency is often used without being properly defined first (Cengel and Boles, 2015). Basically, efficiency can be described as the ratio of output against the input. This definition of efficiency holds true for all thermodynamic systems and is clearly understood. However the output and input parameters are specific to a system or a component which needs to be specified clearly for efficiency assessment (Zafar et. al. 2014). Research has shown that refrigerant clathrates can be used for cooling applications where phase change is desired above freezing (Mori et al. 1991). Clathrates tend to form when gas molecules get trapped in the water molecule cage under low temperature and high pressure (George. 1989, Sloan. 1990). Refrigerant clathrates can be used for active as well as passive cooling applications hence are considered more effective compared with other type of PCMs (Bi et al. 2004, Inba. 2000). Refrigerant clathrates have high heat of fusion and high density which allows them to store more energy per unit volume. Refrigerant clathrates are no more toxic than the base refrigerant so the existing systems can be utilized to contain them. Many refrigerants form clathrates but only handful are in commercial use. Several chlorofluorocarbons 79

94 (CFCs), hydro- chlorofluorocarbons (HCFCs), and hydrofluorocarbons (HFCs) can form clathrates of refrigerant (Eslamimanesh. 2011). For use in cold thermal energy storage system, the refrigerant clathrate should form between temperature range from 278 K to 285 K (Guo et al. 1996). CFCs are forbidden due to stratospheric ozone layer depletion concerns which leave the hydro-chlorofluorocarbon and hydrofluorocarbons to be used for PCM. Refrigerant clathrates of R-134a show they can be an effective in their role as cold thermal energy storage through phase change (Guo et al. 1996). PCMs based on refrigerant clathrates have poor thermal transport properties. To make refrigerant clathrates as effective PCMs, additives of different materials have been studied. For instance, adding calcium hypochlorite or benzenesulfonic acid sodium salt improved the cold energy storage capacity and the cold energy transfer rate of R141b based clathrate (Bi et al. 2006). Adding alcohol in R134a based clathrate has also been studied which shows it accelerates the cool storage rate and eliminates the floating clathrate during the hydration process (Wua et al. 2012). Adding ethanol as an additive in R134a clathrate has also shown to improve the charging and discharging perforamce of the PCM (Zafar et al. 2015). This paper studies the energy and exergy of the R134a clathrates with and without additives. The paper also studies the thermoeconomic analyses of R134a clathrate with and without additives. The refrigerant clathrate and additives were studied as phase change materials (PCMs) for cooling applications. The PCMs are studied for their charging and discharging energy and exergy values and efficiencies. The cost-benefit analyses are also presented in this paper for each PCM. II. Experimental Setup For the experiments, a cold constant temperature bath from The Clifton Range is used as a constant temperature source (Clifton, NE7). The refrigerant clathrate with additive, named PCM, are formed in glass tubes from ACE Glass Incorporated (ACE Glass Inc.). The tubes are submerged in the constant temperature water bath for which the temperature is set at 276 K and 278 K. The constant temperature bath works by providing cold energy and heat simultaneously to the distilled water in the bath to maintain its temperature at a set value. The graphic illustration of the experimental system is shown in Fig. 1. the amount of energy. A stirrer is also used which circulated the water in the bath. Without the stirrer, the water near the hot or cold source would change its temperature while the water away from the source would see its effect later. Fig. 1: A schematic diagram of the proposed PCM testing system The PCM is formed in the glass tubes. First the glass tube is filled with distilled water and the desired additive. The exact mass of the tube with its constituents is measured using a high accuracy digital weighing scale. The tube is sealed and then vacuumed to get rid of excess air. The last step is to fill the desired refrigerant using a needle valve that allows one way flow. The glass tube is then submerged into the cold temperature water bath for charging. The tubes are visually observed after regular interval to observe the onset and end set of freezing. The freezing times, PCM temperatures and pressures are recorded for each test. Onset of freezing is usually east to detect as the top layer starts freezing. The end set is challenging to pin point so it is important to continue observing the PCM until after the last observed changes in the PCM structure. PCM usually rises as it freezes so height is observed for the end set. The K-type thermocouples are attached to a reader to read the temperatures. For initial charging test, only one temperature reading is taken. For thermal property tests, temperatures are taken at two different locations. The tube is comprehensively tested for leaks and provisions are made to make sure there are no leaks. It is important to use a glass tube since the onset of phase change needs to be observed visually. The illustrative Fig. of the glass tube, its connections and used systems are shown in Fig. 2. A refrigeration system with cooling coils around the water bath pumps out the heat. A controller constantly monitors the water temperature in the bath while continues to provide the desired heat to maintain the desired temperature. The bath is converted into constant energy bath for thermal properties experiments. A constant cold and hot energy is provided to the water in the bath to maintain 80

95 Q The thermal exergy absorbed by the PCM Ex in,pcm is described as: Q (m PCM ex PCM )f - (m PCM ex PCM )i = Ex in,pcm (7) The thermal exergy released by the discharging fluid Q Ex out,c or the thermal exergy released by the stationary solid is described as: Q (mcexc)i - (mcexc)f =Ex out,c = E x Q supply Δt (8) The overall system efficiencies can now be described since useful output and required inputs have been established. The required input is the energy/exergy released by the charging material to change the phase or charge the PCM. The useful output is the energy/exergy absorbed by the discharging material which in turn is absorbed by the PCM. The overall system s energy efficiency can be described as ηoa = Q in Q out (9) The overall system s exergy efficiency can be described as Fig. 2: Instruments for experimental measurements III. Analysis In order to determine the efficiencies, it is first important to describe the useful input and required output of the system. For the charging process, the heat absorbed by the charging fluid Qin,c is described as: [(m chc)δt]out - [(m chc)δt]in = Qin,c (1) While heat given out by the PCM Qout,PCM is (m PCM h PCM )i - (m PCM h PCM )f = Qout,PCM (2) The thermal exergy absorbed by the charging fluid Q is described as: Ex in,c Q [(m cexc)δt]out - [(m cexc)δt]in = Ex in,c (3) Q While thermal exergy given out by the PCM Ex out,pcm is defined as: Q (m PCM ex PCM )i - (m PCM ex PCM )f = Ex out,pcm (4) For the discharging process, the heat absorbed by the PCM Qin,PCM is described as: (m PCM h PCM )f - (m PCM h PCM )i = Qin,PCM (5) The heat released by the discharging fluid Qout,c or the heat emitted by the stationary solid is described as (mchc)i - (mchc)f = Qout,c = Q Δt (6) 81 Ψoa = Ex in Ex out (10) The thermoeconomic analysis on the PCMs is also conducted with the following equation: f TE = Z k Z k +ξ Ex dst (11) where Z k is the total cost of the items used in the PCM in dollars, ξ is the energy cost in $/J and and f TE is the thermoeconomic factor. IV. Results and discussion Fig. 3 shows the average energy utilizations for charging R134a clathrate using five tested additives. The graph shows that magnesium nitrate hexahydrate has the lowest overall energy utilization followed by copper, ethanol, aluminum and then sodium chloride. For 0.01 additive mass fraction, the energy decreased by 55% for magnesium nitrate hexahydrate and 22% for copper. Ethanol and aluminum maintained the energy utilization relatively the same as required by the base R134a clathrate. Sodium chloride increased the energy utilization by 13%. At high additive concentrations, the energy utilization decreased by 27% for magnesium nitrate hexahydrate. Copper, ethanol, aluminum and sodium chloride increased the energy utilization by 27%, 26%, 23% and 60% respectively. Fig. 4 presents the average exergy utilizations during charging process of R134a clathrate for the five tested additives. Similar to the energy utilization trend, magnesium nitrate hexahydrate has the lowest overall energy utilization followed by copper, ethanol, aluminum and then sodium chloride. For 0.01 additive mass fraction, the exergy decreased by 55%

96 for magnesium nitrate hexahydrate and 33% for copper. Ethanol and aluminum maintained the energy utilization relatively the same as required by the base R134a clathrate. Sodium chloride increased the energy utilization by 13%. At high additive concentrations, the energy utilization decreased by 27% for magnesium nitrate hexahydrate. Copper, ethanol, aluminum and sodium chloride increased the energy utilization by 27%, 26%, 23% and 60% respectively. Fig. 5: Energy comparison between PCMs during discharging phase Another factor that hampers the energy absorption is non-homogenous mixing of additives. The metal additives would absorb more energy if they mix well in the clathrate. Ethanol on the other hand, makes hard solid clathrate which allows it to provide cool energy longer than the others. Fig. 3: Energy comparison between PCMs during charging phase Fig. 6 presents the average cool exergy released by each PCM during the discharge phase. Exergy follows the same trend as energy. Magnesium nitrate hexahydrate, again, has the lowest exergy while ethanol additive has the highest. The reason for magnesium nitrate hexahydrate additive s low energy release is its soft clathrate structure. Ethanol, on the other hand, makes hard solid clathrate which allows it to provide cool energy longer than the others. Fig. 4: Exergy comparison between PCMs during charging phase Fig. 5 shows the average cool energy released by each PCM during the discharge phase. The graph shows that ethanol additive has the highest overall energy release followed by aluminum, copper, base clathrate and magnesium nitrate hexahydrate. The reason for magnesium nitrate hexahydrate additive s low energy release is its soft clathrate structure. Additives that make soft small clathrate structures tend to absorb low energy. 82 Fig. 6: Exergy comparison between PCMs during discharging phase Fig. 7 presents the overall energy and exergy efficiencies of each PCM. The solid bar represents the exergy values while bars with pattern fill represent the energy. The input and output values are utilized to evaluate these efficiencies. The efficacy graph shows the true picture of which additive results in most gain. Ethanol additive shows the highest efficiency while sodium chloride shows the least. Ethanol, in spite taking long to freeze, yields the most during discharge hence proved to be the most useful. Sodium chloride additive talks long

97 to charge while does not last very long during discharge which makes it the least efficient additive. Fig. 8 shows the exergy destruction for each PCM used in the experiments. Sodium chloride based PCM has the highest exergy destruction of 12 kj. Magnesium nitrate hexahydrate based PCM has the lowest exergy destruction of 4 kj. Sodium chloride based PCM has high charging time and a comparative low discharge time which results in high exergy destruction. Magnesium nitrate hexahydrate based PCM has low charging time which results in relatively low exergy destruction. It is to be noted that even though magnesium nitrate hexahydrate based PCM has the lowest exergy destruction, it may not necessarily be the most useful PCM, overall. IV.2. Results of Thermoeconomic Analysis Using the equations described in the Analyses section, thermoeconomic analyses are conducted on the PCMs used in the experiments. Thermoeconomic analyses include the evaluation of thermoeconomic factor and cost-benefit analyses for the PCMs. Fig. 9 shows the variation of thermoeconomic variable, f TE as it changes with respect to each PCM. Thermoeconomic variable, f TE, is studied for each PCM including its energy, containment and PCM components costs. For thermoeconomic factor, higher the value, more feasible it is. The results show that magnesium nitrate hexahydrate based PCM has the highest thermoeconomic factor while sodium chloride based PCM has the lowest thermoeconomic factor. The low thermoeconomic factor for magnesium nitrate hexahydrate based PCM is due to its low exergy destruction. Similarly, high thermoeconomic factor for sodium chloride based PCM is due to its high exergy destruction. Fig. 7: Overall energy and exergy efficiencies of PCMs Fig. 9: Thermoeconomic variable values of each PCM Fig. 8: Exergy destruction values of each PCM using different additives 83 Fig. 10: Energy costs of producing and using PCMs

98 Fig. 10 shows the energy cost of producing the PCM and amount saved using the PCM. The energy cost is calculated using the electricity unit rate of $0.32. Ethanol, having the highest efficiency, gives the highest return in terms of discharge energy. Sodium chloride has the greatest difference between charging and discharging price due to its low efficiency. As it can be seen from the graph that the energy cost of charging 100 units is not very high and it remains below $5. Magnesium nitrate hexahydrate shows to have the lowest energy cost but it also has the lowest return. aluminum and then sodium chloride. For discharging process, PCM with ethanol additive has the highest overall energy and exergy release followed by aluminum, copper, base clathrate and magnesium nitrate hexahydrate. The efficacy graph shows the true picture of which additive yields most gain. Ethanol additive shows the highest efficiency while sodium chloride shows the least. Sodium chloride based PCM has the highest exergy destruction of 12 kj. Magnesium nitrate hexahydrate based PCM has the lowest exergy destruction of 4 kj. The results show that magnesium nitrate hexahydrate based PCM has the highest thermoeconomic factor while sodium chloride based PCM has the lowest thermoeconomic factor. Ethanol, having the highest efficiency, gives the highest return in terms of discharge energy. Ethanol additive makes the most economical PCM since it costs relatively low to produce it yet it gives the highest efficiency. Fig. 11: Cost of producing 100 units of each PCMs Fig. 11 shows the costs of producing 100 units of each PCM with additives compared to the base PCM without any additive. This Fig. includes the price for 80 gram additive and the costs of energy to charge it. The energy cost is calculated using the electricity unit rate of $0.32. Copper additive proves to be the most expensive primarily due to the price of copper particles. Magnesium nitrate hexahydrate has low cost since it does not take long to get charged. It should be noted that magnesium nitrate hexahydrate does not produce a lot either so its low price is somewhat deceiving. Ethanol has the second lowest cost due to its low price. Ethanol additive makes the most economical PCM since it costs relatively low to produce it yet it gives the highest efficiency. V. Conclusions Experimental studies are undertaken on refrigerant clathrates for use in cooling applications. Refrigerant R134a is used to form the clathrate. Sodium chloride, magnesium nitrate hexahydrate, aluminum particles, copper particles and ethanol is used as an additive to determine their impact on the refrigerant clathrate. Energy, exergy, efficiencies and thermoeconomics are evaluated for R134a with and without additives. Based on the obtained results, the following findings may be drawn: For charging process, PCM with magnesium nitrate hexahydrate has the lowest overall energy and exergy utilization followed by copper, ethanol, 84 Nomenclature Ex : Exergy (J) ex : Specific exergy (J/kg) fte : Thermoeconomic factor h : Specific enthalpy (J/kg) m : Mass (kg) m : Mass flow rate (kg/s) PCM : Phase change material Q : Heat (J) Q : Heat flow rate (W) t : Time (sec) Zk : Total cost ($) Greek letters ξ : Energy cost ($/J) η : Energy efficiency ψ : Exergy efficiency Subscripts c : Charging d : Discharging i : initial f : Final Superscript Q : Thermal References ACE Glass Incorporated. Pressure Tube 185 ml. Item Number Bi YH, Guo TW, Zhu TY, Zhang L, Chen L., Influences of additives on the gas hydrate cool storage process in a new gas hydrate cool storage system. Energy Conversion and Management 47: (2006) Bi YH, Guo TW, Zhu TY, Fan SS, Liang DQ, Zhang L., Influence of volumetric-flow rate in the crystallizer on the gas-hydrate cool-storage process in a new gas-hydrate cool-storage system. Applied Energy 78:

99 (2004) Cengel Y A, M. A. Boles Thermodynamics: An Engineering Approach, 8th ed. McGraw-Hill: New York. Eslamimanesh A, Mohammadi AH, Richon D., Thermodynamic model for predicting phase equilibria of simple clathrate hydrates of refrigerants. Chemical Engineering Science 66: (2011) George A., Hand book of thermal design. Guyer C, editor. Phase change thermal storage materials. McGraw Hill Book Co.; chapter 1 (1989) Guo KH, Shu BF, Zhang Y., Transient behavior of energy charge discharge and solid liquid phase change in mixed gas-hydrate formation. In: Wang, B.X. (Ed.), Heat Transfer Science and Technology. Higher Education Press, Beijing (China): (1996) Inaba H., New challenge in advanced thermal energy transportation using functionally thermal fluids. International Journal of Thermal Science 39: (2000) Mori YH, Isobe F., A model for gas hydrate formation accompanying direct-contact evaporation of refrigerant drops in water. International Communications in Heat Mass Transfer 18: (1991) Rosen MA., Dincer I., On exergy and environmental impact, International Journal of Energy Research (1997) Rosen M.A. and I. Dincer On exergy and environmental impact. International Journal of Energy Research 21: Sloan ED., Clathrate hydrates of natural gases. New York: Marcel. (1990) The Clifton Range. Chillo Baths NE7 Series 8 igerated-water-baths/ Wua J, Wangb S., Research on cool storage and release characteristics of R134a gas hydrate with additive. Energy and Buildings 45: (2012) Zafar S, Dincer I., Efficiency Assessment of Crude Oil Distillation System. In Progress in Exergy, Energy and Environment, edited by Ibrahim Dincer, Adnan Midilli and Haydar Kucuk, Chapter 19, pp United States, ISBN : Springer International Publishing Switzerland, (2014). Zafar S, Dincer I, Gadalla. M., Experimental testing and analysis of R134a clathrates based PCMs for cooling applications. International Journal of Heat and Mass Transfer 91: (2015). 85

100 Thermodynamic Performance Analysis of a Raw Mill System in Cement Plant Mehmet Altinkaynak 1*, Murat Ozturk 2, Ali Kemal Yakut 1, 1* Süleyman Demirel University, Faculty of Technology, Department of Energy Systems Engineering, 32260, Isparta, Turkey 2 Süleyman Demirel University, Faculty of Technology, Department of Mechatronic Engineering, 32260, Isparta, Turkey * mehmetaltinkaynak@sdu.edu.tr Abstract The cement production process is one of the most power-intensive and higher harmful gas emitting process in the world. The energy policies of many development and under developing countries focus on increasing energy efficiency in the industry, which in turn, causes to decrease harmful gas emissions. In this paper, the energy and exergy analyses of a raw mill in cement plant are investigated for better understanding of the system design dynamics. Exergy destruction rate and exergy efficiency are obtained using by the thermodynamic analysis. The system design parameters, which effect the process performance, such as the ambient temperature, the mass flow rate and component temperature are analyzed. Keywords: Thermodynamic analysis, cement plant, raw mill. I. Introduction The cement production facility is the energy extensive plant which makes every effort related to power consumption, performance and generation. The cement production process contains many indicators, such as i-) grinding and blending raw materials (limestone, shells or chalk, and shale, clay, sand, or iron ore), ii-) heating those materials to very high temperatures in a kiln, iii-) cooling and mixing those materials with gypsum, and iv-) finally, grinding down the mixture to form cement powder. The exergy analysis presents the system design as closely as allowable to the maximum theoretical limit. The development of design technique for the raw mill component with increasing performance is a vital task. The transfer of heat between inlet layer of the raw mill and environment is the most generally encountered operation in component design process. Numerous theoretical and experimental analyses to investigate the raw mill system in cement industry have reported in the literature. Sogut et al. (2010), have analyzed the heat recovery modelling from a rotary kiln process in cement industry to the environment using by the energy and exergy analysis viewpoints. Also, the authors have investigated the energetic and exergetic efficiencies, and exergy destruction rates of the rotary kiln for cement production process. Madlool et al. (2012), have given the energy and exergy analyses equations, the exergy balance equations, and also energy and exergy efficiencies for the components of a cement industry for investigating of the cement generation processes. In addition, Ahamed et al. (2012), have investigated for increasing the energetic efficiency, the exergetic efficiency and the recovery performance of a cooling process through the optimization of its working indicators, such as i-) the mass flow rate of working fluid and clinker, ii-) the cooling working fluid temperature, and iii-) the grate speed. Also, they have analyzed the thermodynamic performance analysis to investigate how the working indicators of grate clinger cooling process and the heat recovery from the hot exhaust gasses. Atmaca and Kanoglu (2012), have studied the energetic and exergetic analyses of a raw mill in cement plant and specific measures in order to decrease the quantity of energy consumption in grinding system. also, they have found the energetic and exergetic efficiency as 61.5% and 16.4%, respectively. Gutierrez et al. (2012), have investigated the power consumption and the exergy destruction rate of the calcination system in the vertical shaft kilns, in order to identify the indicators affecting energy consumption. They have given that the most exergy destruction rate, due to fuel combustion, internal heat and momentum transfer taking place in the kiln process. In this paper, the methods for determining the magnitudes and causes of exergy destruction in the raw mill system in cement industry is detailed investigated by using thermodynamic analysis. Furthermore, the impact of design parameters on the system performance are evaluated under different operating conditions. II. System description The raw mill system is an important component among other parts of the cement plant. Because, the raw mill system is used to grain the crude inputs into the farine output which is the semi-product of clinker output. The schematic diagram of the raw mill system in a cement plant is illustrated in Fig. 1. In this process, the raw materials, such as CaCO2, SiO2, Al2O3, Fe2O3, MgO, K2O, SO3 and Na2O at reference temperature and pressure enter the raw mill system to produce farine. The producing farine, which is consisting of CaO, CO2, SiO2, Al2O3, Fe2O3, MgO, 86

101 K2O, SO3 and Na2O enters to the farine silo at point 2. Also, the producing farine dust goes to the filter at point 3. The additional of stack gases for the farine production process has significant indicators for the heat needs during farine generation. Therefore, the heated gaseous, such as N2, O2, CO2, CO and SO2 input to the raw mill system at point 4 to give its heat to the raw materials. m i = m e (2) where m is the mass flow rate, subscript i and e are the inlet and outlet flows, respectively. III.2. Energy balance The energy analysis of the control volume deals with all energy parts of the chosen control volume. The energy balance equation, which is given as the first law of the thermodynamics, can be given as follows; Q + m inh in = W + m out h out (3) where Q, W and h are the heat transfer rate, power and specific enthalpy, respectively. III.3. Exergy balance The exergy can be described as the maximum work that should be provided from the process at a chosen state. To evaluate the exergy analysis, firstly the reversible work must be defined. The reversible work can be defined as the maximum useful work that can be provided as the system goes through a process between two given states. The general exergy balance rate can be written as follows; E x Q + m in ex in = E x W + m out ex out + E x D (4) Fig. 1. Schematic diagram of a raw mill system III. Thermodynamic analysis Thermodynamic assessments based on the energy and exergy analyses are used to examined the performance, energy loss rate and exergy destruction rate in order to increase the efficiency of the investigated process and its components. The mass, energy and exergy balance equations are used to investigate the exergy destruction rate, the energy and exergy content of any stream, the energy efficiency and exergy efficiency of the process for detailed information. Generally, based on the usual principle, the thermodynamic balance equation for a quantity in investigated system can be defined as (Dincer, 2012) Input + Generation Output Consumption = Accumulation (1) Also, in the steady state condition, the accumulated indicator in the Eq. (1) is equal to zero, because whole properties in the system are unchanging with time. III.1. Mass balance The conservation of mass is a fundamental procedure in investigating any thermodynamic process. The mass balance equation can be written as follows; where ex is the specific exergy and E x D is the rate of exergy destruction. E x Q and E x W are the rate of exergy transfer by heat, and work, respectively, and can be calculated as follows; E x Q = (1 T o T ) Q (5) E x W = W (6) where T_o is the reference temperature, and T is the temperature at which heat transfer takes place. The specific exergy can be given as follows; ex = ex ke + ex pe + ex ph + ex ch (7) where ex ke, ex pe, ex ph and ex ch are the kinetic, the potential, the physical and the chemical exergies, respectively. In this paper, kinetic and physical exergy are accepted negligible. The physical exergy or specific flow exergy can be written as ex ph = (h h o ) T o (s s o ) (8) where s is the specific entropy. The chemical exergy of gas mixture can be given as ex ch = x i ex o ch + RT o x i ln(x i ) (9) o where ex ch is the standard chemical exergy of an element in kj mol and x i is the mass fraction of in element i and subscript o stands for dead state. Total exergy rate can be given as follows; 87

102 E x = m ex (10) III.4. Thermodynamic analysis of raw mill In this section, the energy and exergy analyses of raw mill system in cement plant are investigated. To reach this aim, the mass, energy and exergy balance equations for input and output flows of the raw mill process are evaluated. The mass balance equation of the raw mill sub-system in the cement production process can be written as follows: m 1 + m 4 = m 2 + m 3 (11) Based on the general energy balance equation, which given in the Eq. (3), the energy balance equation for the raw mill can be defined as follows: m 1h 1 + m 4h 4 + Q loss RM = m 2h 2 + m 3h 3 (12) The exergy balance equation of the raw mill is given as m 1ex 1 + m 4ex 4 + E x RM Q,loss = m 2ex 2 + m 3ex 3 + E x RM D (13) The Raw materials enters the Raw mill at point 1. The chemical compositions of Raw materials are given in Table 1. Tab 1. Chemical compositions of Raw materials at point 1. Mass Molar Mass Mass Flow Raw Concentration (kg/kmol) Rate Materials (wt. %) M (kg/s) Y CaO 56,077 CaCO 2 CO 2 44, SiO 2 60, Al 2O 3 101, Fe 2O 3 159, MgO 40, K 2O 94, SO 3 80, Na 2O 61, Total The specific exergy of the flow at state1 is given as follows; ex 1 = Y CaO M CaO ex ch CaO ch + Y CO2 M CO2 ex ch CO2 + ch + Y SiO2 M SiO2 ex SiO2 + Y Al2 O 3 M Al2 O 3 ex Al2 O 3 ch ch Y Fe2 O 3 M Fe2 O 3 ex Fe2 O 3 + Y MgO M MgO ex MgO + ch Y K2 OM K2 Oex K2 O + Y SO3 M SO3 ex ch ch SO3 + Y Na2 OM Na2 Oex Na2 O (14) where chemical exergy of CaO, CO 2, SiO 2, Al 2 O 3, Fe 2 O 3, MgO, K 2 O, SO 3 and Na 2 O are given as follows, respectively; ch ex CaO 0 = h CaO (h Ca + 0.5hO2 ) T0 [s CaO (s 0 Ca + 0.5s 0 O2 )] + ex ch ch Ca + 0.5ex O2 ex ch CO2 0 = h CO2 ex ch ch C + ex O2 (15) (h C + ho2 ) T0 [s CO2 (s 0 C + s 0 O2 )] + (16) 88 ch ex SiO2 0 = h SiO2 s 0 O2 )] + ex ch ch Si + ex O2 ch ex Al2 O 3 0 (s Al2 ch ex Fe2 O 3 0 (s Fe2 ch ex MgO 0 (s Mg ex ch K2 O 0 = h Al2 O (h Si + ho2 ) T0 [s SiO2 (s 0 Si + 0 (h Al s 0 O2 )] + ex ch ch Al ex O2 0 = h Fe2 O 3 0 (h Fe s 0 O2 )] + ex ch ch Fe ex O2 0 = h MgO 0 (h Mg + 0.5s 0 O2 )] + ex ch ch Mg + 0.5ex O2 0 = h K2 O 0.5s 0 O2 )] + ex ch ch K ex O2 ex ch SO3 0 = h SO3 s 0 O3 )] + ex ch ch S + 1.5ex O2 ch ex Na2 O 0 (s Na2 0 = h Na2 O h O2 ) T0 [s Al2 O h O2 ) T0 [s Fe2 O h O2 ) T0 [s MgO (17) (18) (19) (20) (h K hO2 ) T0 [s K2 O (s 0 K (h S + 1.5hO2 ) T0 [s SO3 (s 0 S + 0 (h Na s 0 O2 )] + ex ch ch Na ex O h O2 ) T0 [s Na2 O (21) (22) (23) where Mi is the molar mass (kg/kmol) of ith substance, Yi is the mass concentartion (wt. %) of the ith substance. Also, Mi and Yi are given in Table 1. The stack gases enters the Raw mill at point 19. The chemical compositions of stack gases are written in Table 2. Tab 2. Chemical compositons of stack gases at point 4 Mass Molar Mass Mass Stack Concentration (kg/kmol) Flow Rate Gases (wt. %) M (kg/s) Y N 2 28, O 2 31, CO 2 44, CO 28, SO 2 64, Total The spesific exergy of the flow at state 4 can be calculated as follows; ex 4 = Y N2 M N2 ex ch N2 + Y O2 M O2 ex ch O2 + Y CO2 M CO2 ex ch CO2 + Y CO M CO ex ch ch CO + Y SO2 M SO2 ex SO2 (24) where the chemical exergy equations of CO and SO 2 are given as ex ch CO = h CO (hc + 0.5hO2 ) T0 [s 0 CO (s 0 C + 0.5s 0 O2 )] + ex ch ch C + 0.5ex O2 ex ch SO2 0 = h SO2 ex ch ch S + ex O2 (25) (h S + ho2 ) T0 [s SO2 (s 0 S + s 0 O2 )] + (26)

103 The Raw materials exit from the raw mill, is called as farine stored in the farine silo at point 2. The chemical compositions of farine at point 2 are given in Table 3. Tab 3. Chemical compositions of farine at point 2 Farine Molar Mass Mass Mass Flow (kg/kmol) Concentration Rate M (wt. %) (kg/s) Y CaO 56, CO 2 44, SiO 2 60, Al 2O 3 101, Fe 2O 3 159, MgO 40, K 2O 94, SO 3 80, Na 2O 61, Total The specific exergy of flow at point 2 is given as follows; ex 2 = Y CaO M CaO ex ch CaO ch + Y CO2 M CO2 ex ch CO2 + ch + Y SiO2 M SiO2 ex SiO2 + Y Al2 O 3 M Al2 O 3 ex Al2 O 3 ch ch Y Fe2 O 3 M Fe2 O 3 ex Fe2 O 3 + Y MgO M MgO ex MgO + ch Y K2 OM K2 Oex K2 O + Y SO3 M SO3 ex ch ch SO3 + Y Na2 OM Na2 Oex Na2 O (27) The stack gases plus farine dust exist from the raw mill, and goes to the electro filter at point 3. The chemicalcomposition at flowing materials at point 3 are given in Table 4. Tab 4.Chemical compositions of flowing materials at point 3. Molar Mass Mass Mass Flow Flowing (kg/kmol) Concentration Rate Materials M (wt. %) (kg/s) Y CaO 56, CO 2 44, SiO 2 60, Al 2O 3 101, Fe 2O 3 159, MgO 40, K 2O 94, SO 3 80,064 < < Na 2O 61, N 2 28, O 2 31, CO 28, SO 2 64, Total The specific exergy of flow at point 3 is given as follows; ex 3 = Y CaO M CaO ex ch CaO ch + Y CO2 M CO2 ex ch CO2 + ch + Y SiO2 M SiO2 ex SiO2 + Y Al2 O 3 M Al2 O 3 ex Al2 O 3 ch ch Y Fe2 O 3 M Fe2 O 3 ex Fe2 O 3 + Y MgO M MgO ex MgO + ch Y K2 OM K2 Oex K2 O + Y SO3 M SO3 ex ch SO3 + ch Y Na2 OM Na2 Oex Na2 O + Y N2 M N2 ex ch N2 + Y O2 M O2 ex ch O2 + Y CO M CO ex ch ch CO + Y SO2 M SO2 ex SO2 (28) The heat loss rate from the raw mill system to the environment, which given in Eq. (12), can be calculated as follows (Atmaca and Kanoglu, 2012); Q loss RM = T RM T o R tot (29) where T RM is the raw mill temperature R tot is the net thermal resistance of the raw mill, and can be evaluated as follows; R tot = R conv,1 + R cond + R conv,2xr rad R conv,2 +R rad (30) The convection 1 and 2, conduction and radiation thermal resistance can be calculated as follows; R conv,1 = 1 2πr 1 hl R conv,2 = 1 2πr 2 hl ) R cond = ln(r 2 r 1 2πkL R rad = 1 h rad A (31) (32) (33) (34) Where h and k are the convection coefficient and thermal conductivity, respectively, h rad the radiation heat transfer coefficient, and can be defined as follows; 2 h rad = εσ(t out,surf + T 2 out )(T out,surf + T out ) (35) where ε is the emissivity of the rwa mill system surface and σ is the Stefan-Boltzman constant as5.67x10 8 W m 2 K 4. III.5. Energy efficiency The efficiency of any system should be written in terms of useful outputs from the system boundary divided by the total inputs to the system. According to this description, for a general system, the energy efficiency can be defined as (Dincer, 2011) η = useful energy in outputs total energy inputs (36) The energy efficiency equation of the raw mill system in cement plant is defined as follows; η RM = m 2h 2 +m 3h 3 m 1h 1 +m 4h 4 (37) III.6. Exergy efficiency The exergetic efficiency analysis gives some very significant indicators about the process and its parts for increasing and efficiently use. For a general process, an exergy efficiency equation is given as follows (Dincer, 2011); ψ = exergy in outputs total exergy inputs (38) The exergy efficiency of the raw mill system in cement plant, which is presented in Fig. 1, is written as 89

104 ψ RM = m 2ex 2 +m 3ex 3 m 1ex 1 +m 4ex 4 (39) IV. Results and discussions In this paper, the reference temperature and pressure are taken as 25 C and kpa, respectively. The thermodynamic properties of the material flows in the raw mill system for the investigating cement production process are determined using by the EES software program (Klein, 2010). The EES code is also developed to investigate the performance of the raw mill system and its parts. The heat loss rate from the raw mill system to the ambient is calculated as 14, MJ/h. The exergy efficiency of the raw mill system is calculated as 34.45%. In addition, the drying part and grinding part have the maximum heat loss rate for the process. The heat loss rate for these components are calculated as and 64.74%, respectively. The design parameters of the raw mill system used in this study, such as the mass flow rate, temperature, pressure, specific exergy and exergy rate are given in Table 5. Tab 5. Thermodynamic properties of materials with each state point for raw mill system State point Mass flow rate (kg/s) Temperature ( C) Pressure (kpa) Exergy rate (MJ) , , , ,160 Generally, the exergy efficiency of any process is lower than energy efficiency, and that important development possibility exist. The cement producing sectors are the energy and exergy intensive system. There are significant opportunities to identify parts where energy and exergy savings measurements should be performed so that energy and exergy should be saved along with the increase of harmful gaseous. The effects of input material temperature on the exergy destruction rate and exergy efficiency of the raw mill system are illustrated in Fig. 2. As seen in this figure, increasing temperature of the input materials has positive effect on the system exergy efficiency. Also, the exergy destruction rate of the raw mill system is decreased with increasing input material temperature. The impacts of input materials mass flow rate on the exergy destruction rate and exergy efficiency of the raw mill system are given in Fig. 3. As illustrated in this figure, increasing input material mass flow rate increase the exergy destruction rate and exergy efficiency of the raw mill system. The heated gaseous temperature at point 4 is very important for efficiently system design. Therefore, the effects of heated gaseous temperature on the system energy efficiency and exergy efficiency is given in Fig. 4. The energy efficiency of the raw mill system is increased from 52.25% to 54.25% with increasing 90 heated gaseous temperature from 350 C to 750 C. Also, the exergy efficiency of the raw mill system is increased from 31.45% to % with increasing heated gaseous temperature. Ex D (MJ) Ex D (MJ) Temperature of input materials Fig. 2. Effect of input materials temperature on exergy destruction rate and exergy efficiency Fig. 3. Effect of mass flow rate on exergy destruction rate and exergy efficiency Energy efficiency Fig. 4. Effect of flow 4 temperature rate on energy and exergy efficiency V. Conclusions Ex D (MJ) y In this paper, the energy and exergy analyses of the raw mill system in cement plant are given to investigating cement generation process using with the actual facility data. Because energetic viewpoint cannot give adequate information about the energy losses, exergetic viewpoint is performed in order to investigated real efficiencies and destructions of the raw mill system in cement production process. Also, the parametric studies are conducted in order to find out how the input material temperature, mass flow rate of input materials and temperature of flow at Mass flow rate of input materials (kg/s) Ex D y h y Temperature of flow Exergy efficiency Exergy efficiency Exergy efficiency

105 point 4 affect the efficiency of the raw mill system in cement plant. References Sogut, Z., Oktay, Z., Karakoc, H. Mathematical modeling of heat recovery from a rotary kiln, Applied Thermal Engineering 30 (2010) N.A. Madlool, R. Saidura, N.A. Rahim, M.R. Islam, M.S. Hossian, An exergy analysis for cement industries: An overview, Renew Sust Energ Rev 16 (2012) U Ahamed, N A Madlool, R Saidur, M I Shahinuddin, A Kamyar, H H Masjuki, Assessment of energy and exergy efficiencies of a grate clinker cooling system through the optimization of its operational parameters, Energ 46 (2012) A Atmaca, MKanoglu, Reducing energy consumption of a raw mill in cement industry, Energ 42 (2012) A S Gutiérrez, J B Cogollos Martínez, C Vandecasteele, Energy and exergy assessments of a lime shaft kiln, Appl Therm Eng 51 (2013) Dincer, I., Rosen, M. A. Exergy: Energy, Environment and Sustainable Development, Elsevier, 225 Wyman Street, Waltham, MA 02451, USA, Second edition Dincer, C. Zamfirescu, Sustainable energy systems and applications, New York, NY: Springer, Klein, S.A., Engineering equation solver. Academic Professional, Version 8,

106 THERMAL SYSTEMS AND APPLICATIONS 92

107 Exergetic Assessment of PTSC Integrated Power-Refrigeration System Working with CO2 Ahmet Kabul 1, Onder Kizilkan 2* 1,2 Süleyman Demirel University, Faculty of Technology, Department of Energy Systems Engineering, 32260, Isparta, Turkey * onderkizilkan@sdu.edu.tr Abstract This study deals with energy and exergy analysis of a solar driven combined power-refrigeration system. The system comprises a supercritical Brayton cycle, a transcritical organic Rankine cycle and a subcritical vapor compression refrigeration cycle. The three systems operates with carbondioxide (CO2) as working fluid because of its zero ozone depleting potential and with negligible global warming potential. Also it is a sustainable working fluid. The combined process includes parabolic trough solar collector system for providing the heat demand of the supercritical Brayton Cycle. The rejected heat from supercritical Brayton Cycle is used for heat energy demand of organic Rankine cycle. The refrigeration cycle is driven by the power generated from the organic Rankine cycle. With the results, all the irreversibility rates of the combined system are determined. Additionally, a parametric study is carried out to examine the variation of energy and exergy efficiency rates of the three systems. Keywords: Thermal energy storage, solar energy, phase change material, latent heat I. Introduction The demand for energy is continually increasing while conventional fossil fuel energy resources are being consumed at an alarming rate. It became very important that reliable and more sustainable energy resources are required to compensate for the uncertainty surrounding the supply of fossil fuels. Renewable energy sources, such as solar, biomass, geothermal, wind, and hydro, can be good alternatives to conventional fuel sources. These sustainable energy sources are available in sufficient quantities and have minimal impact on the environment (Al-Sulaiman and Atif 2015). Solar assisted power systems have the potential to generate electricity particularly in places with high insolation levels. In order to be competitive with conventional power plants, further development of this renewable technology is necessary, such as the use of more efficient power cycles and solar components, increasing deployment, reducing manufacturing cost and integration of thermal energy storage to improve dispatchable power on demand (Padilla et al., 2015). It is very important to develop a highly efficient, relatively low-cost power conversion system with the minimal environmental impacts by the development of the new concept and advanced power cycles as well as by the refinement of the existing conventional power cycles. Carbon dioxide, due to its non-toxicity, non-flammability, abundance and low cost, is an economic competitive and environmental favorable working fluid. Supercritical Carbon dioxide Brayton cycle (SCO2-BC), in view of its simple system layout, superior cycle efficiency, compact plant components, 93 and friendly environmental influence, have currently received significant attention and being studied for widely applications in nuclear, fossil, concentrating solar power (CSP), waste heat recovery and ship propulsion systems (Li et al., 2016) Research effort to find alternative methods of generating electricity is not new and the advantages of the supercritical Brayton cycle with carbon dioxide has been exposed and discussed during the 60's (Mecheri and Le Moullec, 2016). For the last decades, there is a pretty high interest on such power cycles those utilize low-moderate temperature heat sources. Akbari and Mahmoudi (2014) reported exergoeconomic analysis for a new combined supercritical CO2 recompression Brayton/organic Rankine cycle in which the waste heat from supercritical CO2 recompression Brayton cycle was utilized by an organic Rankine cycle for generating electricity. Al-Sulaiman and Atif (2015), conducted a thermodynamic comparison of five supercritical carbon dioxide Brayton cycles integrated with a solar power tower. Padilla et al. (2015) performed a detailed energy and exergy analysis of four different supercritical CO2 Brayton cycle configurations with and without reheat. Kim et al. (2016), investigated heat transfer performance and pressure drop of PCHE with CO2 as a working fluid with wide Reynolds number ranges using CFD analysis. Li et al. (2016), examined the modeling of the forced convection heat transfer of carbon dioxide at supercritical pressures within the PCHE for SCO2-BC. Mecheri and Moullec (2016), investigated the supercritical CO2 cycles performance from thermodynamic consideration for coal power plant application. Kouta et al. (2016), conducted the performance and cost analyses of a solar power tower integrated with supercritical CO2

108 Brayton cycles for power production and a multiple effect evaporation with a thermal vapor compression desalination system for water production. They performed analyses for two configurations based on two different supercritical cycles. Linares et al. (2016), presented an exploratory analysis of the suitability of supercritical CO2 Brayton power cycles as alternative energy conversion systems for a future fusion reactor based on a dual coolant lithium-lead blanket. Hu et al. (2015), studied the performance of a supercritical gas Brayton cycle using CO2-based binary mixtures as the working fluids have been studied. Liu et al. (2016), presented the theoretical analysis and on-site testing on the thermal performance of the waste heat recovery system for offshore oil production facilities. They used the ideal air standard Brayton cycle to analyze the thermal performance. Garg et al. (2015), introduced a quantitative methodology for load regulation of a CO2 based Brayton cycle power plant using the thermal efficiency and specific work output coordinate system. Padilla et al. (2016), proposed three S-CO2 Brayton cycle configurations without reheat by introducing an ejector prior the heater, which reduced the pressure at the solar receiver. They performed a comprehensive thermodynamic analysis and a multi-objective optimization. Serrano et al. (2014) proposed a new layout of the classical recompression supercritical CO2 Brayton cycle which replaces one of the recuperators by another which bypasses the low temperature blanket source. Manente and Lazzaretto (2014) analyzed the supercritical closed CO2 Brayton cycles. They made the analyses to explore the thermodynamic performance and the technical feasibility of such systems. Baroncia et al., (2015) investigated the integration of a molten carbonate fuel cell and a Brayton cycle that uses supercritical carbon dioxide as working fluid. Rovense (2015), performed a numerical analysis performed by SAM for the solar tower for the assessment of the performance of a solar closed air Brayton cycle. Rovira et al. (2015), proposed a configuration named Hybrid Rankine Brayton cycle with balanced recuperator, as well as its operating conditions and potential working fluids, for low to moderate temperature solar applications. Iverson et al. (2013), investigated the response of a prototype sco2 Brayton cycle under transient operating conditions similar to that experienced in a typical solar plant with a direct receiver. In this study, thermodynamic assessment of a solar driven combined power-refrigeration system is conducted. The system includes a supercritical CO2 Brayton cycle (sco2-bc), a transcritical CO2 Organic Rankine Cycle (tco2-orc) and a subcritical CO2 vapor compression refrigeration cycle (CO2-VCRC). A parabolic trough solar collector system (PTSC) provides the necessary heat demand of the supercritical CO2 Brayton Cycle. The ORC system uses the rejected heat from Brayton cycle. The CO2 refrigeration cycle is driven by the power generated from the ORC which are integrated together. II. Combined CO 2 System with PTSC The schematic representation of solar driven combined power-refrigeration system is shown in Fig. 1. The combined system consists of PTSC system, sco2-bc, and tco2-orc, for power generation and CO2-VCRC for refrigeration. Since the use of CO2 as a working-fluid of power and refrigeration cycles has been growing in recent years due to associated benefits (Singh et al. 2013), it has been selected as working fluid for all cycles. The P-h diagram of the three cycles which are being investigated are given in Fig. 2. Fig. 1: Schematic representation of proposed system 94

109 valve where it becomes wet vapor at low pressure. After expansion valve, the refrigerant passes through evaporator where it absorbs necessary heat energy to become saturated vapor while it refrigerates the cold room. Fig. 2: The P-h diagram of the three cycles Solar energy is collected using a PTSC system for supplying heat demand of the cycles. For PTSC system, Therminol-VP1 is selected as the heat transfer fluid (HTF) for its good heat transfer properties and good temperature control (Therminol 2014). Because of its good properties, it is being used in many high temperature applications driven by PTSC such as power plants (Kumar and Reddy 2009; Vogel et al. 2014; Cheng et al. 2012; Al-Sulaiman 2013; 2014). In sco2-bc, a compressor is used to increase the pressure of the gas and after compression process; the compressed gas enters to the boiler. In the boiler the gas is heated up to about 350 C by means of the absorbed solar energy using HTF. The high pressure CO2 then expands in the turbine and enters to the heat exchanger (HEX) where it gives the rest of its heat energy to the tco2-orc. In gas cooler, the gas is cooled to 32 C before the inlet of the compressor. The tco2-orc comprises of four compounds: a turbine, an evaporator, a condenser and a pump. The required heat energy for the evaporator of the tco2- ORC is supplied from the sco2-bc. The liquid CO2 from the condenser is pumped by means of liquid pump and fed to the HEX, where it is heated by the heat energy delivered from BC, and becomes superheated vapor. The superheated vapor then enters to the turbine and expands to a low pressure. At the exit of the tco2-orc turbine, the CO2 vapor enters to recuperator for preheating of the other fluid after pumping process. Subsequently, the turbine exhaust is intensified to liquid in the condenser by extracting heat to the environment by means of a cooling tower. The tco2-orc and the CO2-VCRC are coupled together by the turbine-compressor unit. They also use the same condenser and CO2 as working fluid. The compressor of the CO2-VCRC is driven by the turbine of tco2-orc system and the CO2 is compressed to the condenser as superheated vapor. The CO2-VCRC is subcritical cycle and after the condenser, the refrigerant enters to the expansion 95 For the refrigeration processes, the coolant is 23 % ethylene glycol water (EG-water) mixture with a freezing temperature of C. Also water is used in the cooling tower for absorbing heat energy from gas cooler and condenser. The general design parameters for modelling of the power-refrigeration system are given in Table 1. It must be noted that data for PTSC system is adapted from the reported data in the references Kalogirou (2009), Singh et al. (2013) and Al-Sulaiman (2014). III. Mathematical Modelling For the thermodynamic assessment of the proposed system, energy and exergy balance equations are employed. For the modelling of PTSC system, the equations in ref Kalogirou (2009) are used. Also the given assumptions are made for the analyses: All processes are of steady state and steady flow. The changes in potential and kinetic energies are negligible. The heat transfer to/from ambient and pressure drops in the pipes are neglected. The pump and compressor operations are adiabatic and isentropic. The dead state temperature and pressure are taken to be 25 C and kpa, respectively. Tab. 1: General design parameters of the combined system Pipe receiver inner diameter 0.08 m Pipe receiver outer diameter 0.09 m Glass cover diameter 0.15 m Total length of PTSC m Mass flow rate of HTF kg/s Receiver emissivity 0.92 Glass cover emissivity 0.87 Temperature of the sun 5739 K Absorbed solar radiation 850 W/m2 Wind velocity 5 m/s Turbine isentropic efficiency 0.93 Pump isentropic efficiency 0.92 Turbine inlet temperature 350 C Compressor inlet temperature 32 C Turbine inlet pressure kpa Turbine outlet pressure 8000 kpa Net power generation 120 kw Turbine isentropic efficiency 0.88 Pump isentropic efficiency 0.96 Turbine inlet temperature 85 C Condenser temperature 28 C Turbine inlet pressure 8000 kpa Turbine outlet pressure 6892 kpa Net power generation 120 kw Evaporator capacity kw Evaporator temperature -10 C Condenser temperature 40 C Entering EG-water temperature -4 C Exiting EG-water temperature -9 C PTSC sco2-bc tco2-orc VCRC

110 For steady-state and steady-flow processes a general mass balance can be expressed in rate form as (Dincer and Rosen, 2007) m in = m out (1) where m is the mass flow rate, and the subscripts in and out stand for inlet and outlet respectively. The general energy balance can be written as: E in = E out (2) The energy balance for a general steady-flow system can also be written more explicitly as Q + m inh in = W + m outh out (3) where E in is the rate of net energy transfer to the system, E out is the rate of net energy transfer from the system, Q is the rate of net heat, W is the rate of net work, and h is the specific enthalpy. The general exergy balance equation can be written as (Dincer and Rosen, 2007) Eẋ in = Eẋ out + Eẋ dest (4) where FR is the heat removal factor, S is the is the absorbed solar energy, Aa is the unshaded collector aperture, Ar is the receiver area, and UL is the overall heat loss coefficient of the solar collector. The useful collected energy can be also calculated from entering and exiting fluid properties: Q u = m c p (T out T in ) (11) The heat removal factor can be calculated from F R = m C p A r U L [1 exp ( A ru L F m C p )] (12) where F' is the collector efficiency factor and given by F = U 0 U L (13) Here, UL is the loss coefficient of the receiver and U0 is the overall heat transfer coefficient. Since the receiver is surrounded by glass cover and the inside space is evacuated, it is assumed that there is no heat transfer by convection. Therefore, based on the receiver area Ar and glass cover area Ag, the overall collector heat loss coefficient is given by For a fixed control volume the exergy balance in steady state can be expressed as U L = [ A r ] (h c,c a +h r,c a )A g h r,r c (14) Eẋ Q Eẋ W = m ine in m oute out + T 0 S gen (5) Here, Eẋ Q and Eẋ W terms are the exergies of heat and work, respectively, e is the specific exergy, T0 is the dead state temperature and S gen is the rate of entropy generation. In Eq. 5, the terms T 0 S gen, Eẋ Q and Eẋ W are given below: Eẋ dest = T 0 S gen (6) Eẋ Q = Q ( T T 0 T ) (7) Eẋ W = W (8) The specific exergy (thermomechanical exergy or flow exergy) is defined relative to the environment (T0, P0): e = (h h 0 ) T 0 (s s 0 ) (9) where h is enthalpy, s is entropy and the subscript 0 indicates properties at the reference (dead) state. For the thermodynamic modelling of the PTSC, mathematical equations given in reference Kalogirou (2009), is used. The useful collected energy rate is defined as Q u = F R [SA a A r U L (T in T 0 )] (10) where hc,c-a is the convection heat loss coefficient between ambient and the cover, hr,c-a is the radiation heat transfer coefficient for the glass cover to the ambient and hr,r-c is the radiation heat transfer coefficient between the receiver tube and the glass cover. These three heat transfer coefficients are defined below: h c,c a = Nu air k air D g (15) Where Nu air = Re 0.52 for 0.1 < Re < 1000 (16) Nu air = 0.3 Re 0.6 for 1000 < Re < (17) Here, k is the thermal conductivity of the air, Nu is the Nusselt number and Re is the Reynolds number. h r,c a = ε g σ (T g + T a )(T g 2 + T a 2 ) (18) where is Stefan Boltzmann constant and g is the emittance of the glass cover. h r,r c = σ(t r+t g )(T r 2 +T g 2 ) 1 εr +A r Ag ( 1 εg 1) (19) Here, the subscript g refers to glass cover and r is the emittance of the receiver. The glass cover 96

111 temperature Tg can be calculated using equation below: T g = A r h r,r c T r +A g (h r,c a +h w ) T a A r h r,r c + A g (h r,c a +h w ) (20) The overall heat transfer coefficient from the surroundings to the fluid in the tube is U 0 = [ 1 + D o + ( D o ln( U L h fi D i 2k Do ) D i )] 1 (21) where Di and Do are the inside and the outside tube diameters, hfi is the heat transfer coefficient inside the tube, and k is the thermal conductivity of the tube. The equation for hfi is given below: h fi = Nu fi k fi D i (22) where Nu fi = Re 0.8 Pr 04 for Re > 2300 (23) Nu fi = (constant) for Re < 2300 (24) The exergy from the solar radiation in terms of reference and sun s temperature given by given by Petela (2005) can be expressed as Eẋ solar = S A a ( ( T 0 T sun ) ( T 0 T sun )) (25) where Eẋ solar is the function of the outer surface temperature of sun where Tsun = 5739 K (Tiwari, 2002). Finally, the exergy efficiency can be expressed as the ratio of total exergy output to total exergy input (Dincer and Rosen, 2007): η ex = Eẋ out Eẋ in = 1 Eẋ dest Eẋ in (26) IV. Results and Discussion Solar assisted CO2 power-refrigeration system was analyzed based on the model and assumptions described previously. For determining the thermodynamic performance of the systems, energy and exergy analysis are applied to the PTSC integrated power-refrigeration system. Under the assumptions made and using the solar data of Isparta, Turkey, the calculated properties of the system are given in Table 2, according to reference points illustrated in Fig. 1. Tab. 2: Calculated properties of the PTSC integrated CO2 power-refrigeration system Reference point Fluid type T ( C) P (kpa) m (kg/s) h (kj/kg) s (kj/kgk) e (kj/kg) E x (kw) 1 CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO Water Water Water Water Water Water Water EG-water EG-water Therminol-VP Therminol-VP Therminol-VP

112 During the calculations, the net power generation of sco2-bc was taken as 865 kw, the net power generation of tco2-orc was taken as 120 kw and the refrigeration capacity of CO2-VCRC was taken as kw. According to the analyses, the energy efficiency the overall system was found to be while the overall exergy efficiency was found to be %. The exergy destruction rates of the all parts of the integrated system were determined according to the exergy analyses. The results showed that, the total irreversibility of the system was calculated as 4891 kw where the PTSC system leads with the exergy destruction of 2719 kw. In Table 3, the exergy destruction rates of the all elements are given with relative irreversibility rates. Tab. 3: Exergy destruction rate and relative irreversibility s of the system components System component E x dest RI (kw) (%) PTSC Brayton comp Boiler Brayton turbine HEX Gas cooler ORC pump Rectifier ORC Turbine Condenser VCRC Compressor VCRC Expansion valve Evaporator Cooling Tower Cooling Tower pump PTSC Pump Overall cooling system In order to identify the effect of different working conditions, some parametric studies were also carried out. For these analyses, the variable parameters were selected to be solar radiation intensity, PTSC length, and net power generated. Figure 3 shows the variation of solar radiation intensity with the total exergy destruction rate. As seen from the figure, while the solar irradiation intensity increases, the exergy destruction rate decreases. Also in Figure 4, the variation of exergy efficiency with solar radiation intensity is given. With the increase of solar radiation, the exergy efficiency increases as expected. This decrement is expected, This is because the exergetic efficiency is an inverse function of exergy destruction rate as long as the exergy loss is less than the exergy destructed. Ex dest, kw hex S, kw/m 2 Fig. 3: Variation of solar radiation intensity with exergy destruction rate S, kw/m 2 Fig. 4: Variation of solar radiation intensity with exergy efficiency In Figure 5, the variation of total power generation with PTDC length is given. With the increase of collector length, the power generation increases, as well. This is because, the more collector length results in increase of area thus the heat gain from solar energy increases. This means, the heat transfer absorbs more energy with higher temperature and transfers it to the sco2-bc. A detailed information about energy analyses can be found from the reference Kizilkan and Kabul (2015). W net, kw PTSC lenght, m Fig. 5: Variation of PTSC length with power generation 98

113 It can be considered the effect of varying the net power generation on the exergy destruction rate of the overall systems considered in Figure 6. As can be seen from the figure that, as the net power generation increases, the total exergy destruction rate increases linearly. This increase in exergy destruction rate is expected because of the increment in the area of the PTSC. Ex dest, kw W net, kw Fig. 6: Variation of power generation with exergy destruction rate V. Conclusions Exergetic assessment of PTSC integrated combined power-refrigeration system was investigated comparatively using CO2. The system was consisted of sco2-bc, tco2-orc and CO2-VCRC. For the design parameters of the cycles, the net power generation of the sco2-bc was taken as 865 kw, power generation of tco2-orc was taken as 120 kw and the refrigeration capacity of CO2-VCRC was kw. According to the second law analyses, the exergy efficiency of the overall system was found to be % while the exergy destruction rate of the whole system was found to be 4891 kw. The major contributor the exergy destruction rate was determined as PTSC system because of its huge area. Also, the effects of solar radiation intensity on the exergy efficiency and exergy destruction were investigated. It was found that with the increase of the solar irradiation intensity, the exergetic efficiency of the system increased. Additionally, it was observed that the main source of exergy destruction in the integrated cycle was the PTSC system, therefore it is very important illustrate carefully the exergy destruction of the collectors. References Akbari A.D., Mahmoudi S.M.S., Thermoeconomic analysis & optimization of the combined supercritical CO2 (carbon dioxide) recompression Brayton/organic Rankine cycle, Energy, 78, , Al-Sulaiman F.A., Atif M., Performance comparison of different supercritical carbon dioxide Brayton cycles integrated with a solar power tower, Energy, 82, 61 71, Al-Sulaiman F.A. (2013), Energy and sizing analyses of parabolic trough solar collector integrated with steam and binary vapor cycles, Energy, 58, Al-Sulaiman F.A. (2014), Exergy analysis of parabolic trough solar collectors integrated with combined steam and organic Rankine cycles, Energy Conversion and Management, 77, Baronci A., Messina G., McPhail S.J., Moreno A., Numerical investigation of a MCFC (Molten Carbonate Fuel Cell) system hybridized with a supercritical CO2 Brayton cycle and compared with a bottoming Organic Rankine Cycle, Energy, 93, , Cheng, Z.D., He, Y.L, Cui, F.Q., Xu, R.J. and Tao, Y.B. (2012), Numerical simulation of a parabolic trough solar collector with nonuniform solar flux conditions by coupling FVM and MCRT method, Solar Energy, 86, Dincer I, Rosen MA. Exergy: Energy, Environment and Sustainable Development. 1st ed. Oxford: Elsevier Science; Garg P., Kumar P., Srinivasan K., A trade-off between maxima in efficiency and specific work output of super- and trans-critical CO2 Brayton cycles, The Journal of Supercritical Fluids, 98, , Hu L., Chen D., Huang Y., Li L., Cao Y., Yuan D., Wang J., Pan L., Investigation on the performance of the supercritical Brayton cycle with CO2-based binary mixture as working fluid for an energy transportation system of a nuclear reactor, Energy, 89, , Iverson B.D., Conboy T.M., Pasch J.J., Kruizenga A.M, Supercritical CO2 Brayton cycles for solarthermal energy, Applied Energy, 111, , Kim S.G., Lee Y., Ahn Y., Lee J.I., CFD aided approach to design printed circuit heat exchangers for supercritical CO2 Brayton cycle application, Annals of Nuclear Energy, 92, , Li H., Zhang Y., Zhang L., Yao M., Kruizenga A., Anderson M., PDF-based modeling on the turbulent convection heat transfer of supercritical CO2 in the printed circuit heat exchangers for the supercritical CO2 Brayton cycle, International Journal of Heat and Mass Transfer, 98, , Kalogirou S.A., Solar Energy Engineering: Processes and Systems, Academic Press, Oxford, UK, Kizilkan O., Kabul A., Design and Energy Modelling of a Solar Driven Combined Power-Refrigeration System with Super-Trans-Sub Critical Cycles using CO2, The 2015 World Congress on Advances in Aeronautics, Nano, Bio, Robotics, and Energy (ANBRE15), Incheon, Korea, August 2015.

114 Kouta A., Al-Sulaiman F., Atif M., Marshad S.B., Entropy, exergy, and cost analyses of solar driven cogeneration systems using supercritical CO2 Brayton cycles and MEE-TVC desalination system, Energy Conversion and Management, 115, , Kumar K.R. and Reddy K.S. (2009). Thermal analysis of solar parabolic trough with porous disc receiver, Applied Energy, 86, Linares J.I., Cantizano A., Moratilla B.Y., Palacios V.M., Batet L., Supercritical CO2 Brayton power cycles for DEMO (demonstration power plant) fusion reactor based on dual coolant lithium lead blanket, Energy, 98, , Singh, R., Kearney, M.P. and Manzie, C. (2013), Extremum-seeking control of a supercritical carbondioxide closed Brayton cycle in a direct-heated solar thermal power plant, Energy, 60, Therminol (2014), Therminol VP-1, Heat transfer fluids by Eastman, accessed on Tiwari GN. Solar Energy: Fundamentals, Design, Modelling and Applications. Pangbourne: Alpha Science International Ltd.;2002. Vogel, T., Oeljeklaus, G., Görner, K., Dersch, J. and Polklas, T. (2014), Hybridization of parabolic trough power plants with natural gas, Energy Procedia, 49, Liu X., Gong G., Wu, Y., Li H., Thermal performance analysis of Brayton cycle with waste heat recovery boiler for diesel engines of offshore oil production facilities, Applied Thermal Engineering, In Press, Mecheri M., Moullec, Y.L., Supercritical CO2 Brayton cycles for coal-fired power plants, Energy, 103, , Manente G., Lazzaretto A., Innovative biomass to power conversion systems based on cascaded supercritical CO2 Brayton cycles, Biomass and Bioenergy, 69, , Padilla R.V., Too Y.C.S., Benito R., Stein W., Exergetic analysis of supercritical CO2 Brayton cycles integrated with solar central receivers, Applied Energy, 148, , Padilla R.V., Too Y.C.S., Benito R., McNaughton R., Stein W., Thermodynamic feasibility of alternative supercritical CO2Brayton cycles integrated with an ejector, Applied Energy, 169, 49 62, Petela R. Exergy analysis of the solar cylindricalparabolic cooker. Solar Energy 2005;79: Rovense F., A Case of Study of a Concentrating Solar Power Plant with Unfired Joule-Brayton Cycle, Energy Procedia, 82, , Rovira A., Muñoz M., Sánchez C., Val J.M.M., Proposal and study of a balanced hybrid Rankine Brayton cycle for low-to-moderate temperature solar power plants, Energy, 89, , Serrano I.P., Linares J.I., Cantizano A., Moratilla B.Y., Enhanced arrangement for recuperators in supercritical CO2Brayton power cycle for energy conversion in fusion reactors, Fusion Engineering and Design, 89, ,

115 Cooling of Concentrated Photovoltaic System Using Microchannel Heat Sink Ali Radwan 1*, Mahmoud Ahmed 1, Shinichi Ookawara 1,2 1 Department of Energy Recourses Engineering, Egypt-Japan University of Science and Technology (E-JUST), Egypt. 2 Tokyo Institute of Technology, Tokyo, Japan. * Ali.radwan@ejust.edu.eg Abstract The high incident heat flux on the concentrated photovoltaic (CPV) system causes a significant increase in the cell temperature and thus reduces the system efficiency. Therefore, using an efficient cooling technique is a vital issue for those systems. In the present study, a new cooling method for CPV systems is introduced. A wide microchannel and a manifold microchannel are introduced as heat sinks for those of high thermal generated systems. A comprehensive thermo-fluid model is developed that includes the whole layers of photovoltaic cell integrated with the introduced microchannel domain. The developed model is numerically simulated and validated using various sets of the previous experimental, numerical and analytical results. In the manifold microchannel heat sink, the effect of the manifold pitch is investigated. Based on the simulation results, it is found that decreasing the manifold pitch distance enhances the solar cell temperature uniformity and avoids the hot spot formation. The manifold microchannel heat sink achieves a better solar cell temperature uniformity compared with the conventional wide one at lower mass flowrates lower than 8.3 g/s. In addition, the manifold microchannel heat sink consumes less pumping power than the wide microchannel heat sink. At CR=40, the wide microchannel heat sink is suitable to achieve the highest gained net electric power at cooling flow rate from 8.3g/s to 71.1 g/s. However, if the cooling mass flow rate islower than 8.3 g/s or greater than 71.1 g/s, the manifold microchannel heat sink is producing a higher net electrical power than the wide microchannel heat sink. The differences between the maximum and minimum solar cell temperature are about15 o C and 65 o C at CR=10 and 40 respectively for the wide MCHS, while these differences are about 0.3 o C and 0.8 o C at CR=10 and 40 respectively in the case of using manifold microchannel heat sink. Keywords: Heat transfer, concentrated photovoltaic, manifold-microchannel, solar cell efficiency I. Introduction In sunlight concentration onto photovoltaic (PV) cells, the replacement of expensive solar cells area with low-priced concentrating lenses or mirrors is one method to lower the cost of solar electricity. Because of the reduction in the solar cell area, higher PV cells efficiency may be used (Royne et al., 2005; Zelin Xu and Kleinstreuer, 2014). In the meantime, it is mentioned that PV cells efficiency is less than 20% for silicon solar cells and around 40% for multi-junction solar cells (Zelin Xu and Kleinstreuer, 2014). The remainder part of solar energy is converted into the heat causing temperature rise in PV cells. The generated thermal energy in the PV systems might cause junction damages and lead to a significant decrease in its electrical efficiency (Rejeb et al., 2015; Zelin Xu and Kleinstreuer, 2014; Zhao et al., 2011). Therefore, using an efficient cooling technique in concentrated photovoltaic (CPV) systems will achieve a high electrical efficiency and allow designing the high concentration ratio (CR) systems. In addition, the extracted thermal energy could be used for any domestic or industrial application in photovoltaic/ thermal (PV/T) system. The major design considerations for cooling of concentrated photovoltaic cells are the solar cell temperature, temperature uniformity, and the consumed pumping power (Rejeb et al., 2015; Zelin Xu and Kleinstreuer, 2014; Zhao et al., 2011). The rapid development of the Micro-Electro-Mechanical Systems (MEMS) makes it possible to construct very small scale cooling devices. These small-scale devices can dissipate a large amount of heat flux from hot surfaces(kalteh et al., 2011). So the use of microchannel heat sink (MCHS) is an effective technique to limit the temperature of those high generated heat flux areas such as CPV systems (Rosell et al., 2011). Long parallel microchannels and wide microchannel heat sinks (WMCHS) are described as the conventional microchannel heat sinks. Such heat sinks have been successfully examined for the use in electronic devices cooling applications. Numerous numerical and analytical models for predicting the heat transfer characteristics and pressure drop through such cooling systems have been investigated. Although the conventional microchannel heat sinks offer a significant heat transfer augmentation, they are associated with a dramatic pressure loss. Alternative configurations had been proposed to decrease the incurred pressure loss and simultaneously increase the heat transfer. One of those configurations is the manifold microchannel 101

116 heat sink (MMCHS) (Sarangi et al., 2014). The MMCHS consists of a manifold system which distributes the cooling fluid via multiple inlet-outlet ports. By reducing the flow length through the microchannels, a significant reduction in the pressure drop was attained. Additionally, a decrease in the thermal resistance was achieved by interrupting the thermal boundary layers growth. This design was originally suggested by (Harpole and Eninger, 1991), who confirmed a significant enhancement in the heat transfer coefficient relative to the conventional MCHS at a constant pumping power. A complete twodimensional thermo-fluid model of the MMCHS has been developed [8]. They concluded that distribution of manifold channel spacing should be 333µm, i.e. (30 channels/cm). Also, they found that by using the mentioned spacing, a small rise in the lateral surface temperature non-uniformities is achieved. Rahimi (Rahimi et al., 2013) experimentally studied the performance of the combination of micro-channels and a photovoltaic module as a hybrid PV/T system using water as a coolant. In their experiments, the microchannel hydraulic diameter is mm, and the Reynolds number (Re) varies up to 70. They reported that approximately 30% increased in the output power compared to uncooled conditions. Reddy et al. (Reddy et al., 2014) concluded, based on numerical simulation, that the optimum dimensions of the microchannel were 0.5 mm width and aspect ratio of 8. Moreover, the pressure drop was found to be low in straight flow channels. Bladimir et al. (Ramos- Alvarado et al., 2011) numerically calculated the pressure loss and temperature uniformity of the heated walls of different proposed microchannel configurations. They suggested a new design to achieve a smaller pressure drop and a better flow and temperature uniformity. They recommended using microchannel distributors for cooling the concentrated PV cells, fuel cells, and electronics. The main findings of the recent literature in the field of cooling CPV or PV systems can be summarized as follows: (1) until now the minimum studied channel depth was 500 µm, 5000µm in (Agrawal and Tiwari, 2011),1200µm in(ramos-alvarado et al., 2011),500 µm in (Rahimi et al., 2013), and 220 µm in (Yang and Zuo, 2015). These dimensions are not in agreement with the basis of microchannel dimensions which should be in the range from 10 to 200 µm (Kandlikar, 2014); (2) In the most recent theoretical modeling of the PV/T systems, one-dimensional model is the most popular one (3) most of the recent microchannel studies are investigated in cooling method of the integrated circuits and electronic devices, while a few studies was performed in the concentrated photovoltaic systems. Based on the recent summary, the objective of the current work is to compare the performance of CPV system integrated with two different designs of microchannels such as a manifold microchannel (MMCHS) and a wide microchannel (WMCHS) with channel height of 100 µm. A comprehensive 102 conjugate two-dimensional thermo-fluid model including the PV layers and the microchannel is developed. The developed model is more suitable for the new investigated MMCHS with unknown thermal performance. Also, the proposed numerical model is able to present the variation of temperature which can t be obtained by the conventional energy balance model. The developed model is numerically solved and validated with recent available numerical, analytical, and experimental data. II. Physical model In the present study, a concentrated photovoltaic cell with concentration ratio up to 40 is investigated. The concentration ratio is reached by using the Fresnel lens solar concentrator which has the minimum capital cost (Whitfield et al., 1999). The typical dimensions of the solar cell unit are 12.5 cm by 12.5 cm. In the current work, the effective area of the photovoltaic cell is divided into four equal areas of 6.25 cm by 6.25 cm. Consequently, four microchannel heat sinks of 6.25 x 6.25 cm 2 are used for a single solar cell unit to ensure better temperature uniformity and lower friction power. A schematic diagram of the WMCHS and MMCHS considered in the present work is shown in Fig.1. The manifold distribution system is placed on the top of the flat microchannels, in a direction transverse to the main flow direction. The coolant is pumped in through a common inlet header, which branches out into parallel manifold inlet channels. Upon entering the microchannel, the fluid undergoes a 90 o turn and passes through a distance of the microchannel midpitch distance P/2, removing the generated heat from the concentrated solar cell, and subsequently flows through another 90 o turn then exits upward through the outlet manifold channels. Another common outlet header is used to collect the outlet flow rate. In MMCHS, the pitch (P) is defined as the distance between two consecutive inlets or outlet ports. The inlet and the outlet ports of the manifold have the same cross-sectional area and the same number for the same pitch value. The pitch value is set to have different values of 2547, , 837.6, and µm to achieve a specific number of inlet ports of 25, 50, 75, and 100 respectively. The variation of the pitch is investigated to achieve the best effect on the solar cell temperature, reduce the friction power, and attain a temperature uniformity of the solar cell. In fig. 2-A and B, the PV layers rest on the designed microchannel heat sink. Water is flowing through the channel to maintain a high efficiency and avoid excessive cell temperatures. A constant sun radiation is assumed to be 1000W/m 2, and the incident solar heat flux increases according to the value of geometrical CR.

117 (A) Vw Ta Concentrated thermal radiation Glass cover Solar Cell Tedlar Cooling Channel configurations, the microchannel wall thickness, channel height, channel material, and cooling mass flow rate are selected to be the same. The thermophysical properties of the solar cell layers and microchannel wall materials are presented in Table 1. δg δsc δt δins Lsc H Insulation Table. 1: Thermophysical properties of the microchannel material and solar cell layers Material ρ (kg/m 3 ) Cp (J/kg.K) K (W/m.K) Glass Silicon Tedlar Aluminium (B) Concentrated thermal radiation II.1. The governing equations and numerical simulation δg δsc δt H (A) (B) Vw Ta p outlet port Lsc Glass cover Solar Cell Tedlar Cooling Channel Fig.1 a neat sketch of the proposed PV layers integrated with (A) WMCHS, (B) MMCHS Inlet y x q" w H q" w Solar Cell Tedlar Channel wall δ sc δ t δ w Outlet Two-dimensional solid fluid conjugate heat transfer model is developed to estimate the electrical and thermal performance of the CPV system. In the current study, the width of the MCHS (Wsc) is very large compared to the microchannel height (H). Hence, the two dimensions assumption is appropriate. The developed model adopts the following assumptions: 1. The flow in microchannel heat sink is laminar, incompressible, and steady. 2. The fluid properties are temperature dependent due to the effect of high concentrated solar radiation, while the properties of the solar cell, tedlar, and microchannel walls are assumed to be temperature independent. 3. The effect of viscous dissipation term in the energy equation is neglected. 4. For the investigated MCHS configurations, the back side is perfectly insulated. 5. The thermal contact resistances among each layer of the solar cell and microchannel heat sink are neglected. The comprehensive thermal model includes energy balance, and thermo-fluid equations. Applying energy balance equations on the PV module, the total absorbed energy by the PV cell can be written as: Solar Cell δ sc E g G(t) (1) sc sc sc y P Tedlar Channel wall H δ t δ w The total absorbed energy by the tedlar can be written as: 1 ( t) E g G (2) T sc T x Outlet Inlet Fig. 2 computational domain for (A) WMCHS-PV system, (B) MMCHS-PV system A comparison of the CPV system performance integrated with the proposed MMCHS and conventional WMCHS is presented. In the presented 103 Consequently, the total absorbed energy by the solar cell and its tedlar can be written as follows (G. N. Tiwari and Swapnil Dubey, 2010; Zelin Xu and Kleinstreuer, 2014): 1 G( t) G( t) E (3) sc T g sc sc The electrical energy generated by solar cell can be written as: g

118 E g G(t) (4) el sc sc Where sc is the solar cell efficiency (G. N. Tiwari and Swapnil Dubey, 2010; Zelin Xu and Kleinstreuer, 2014) and can be written as follows: 1 ( T T )) (5) sc ref ( ref sc ref Where: the ref and β ref are the solar cell efficiency and cell temperature coefficient at a reference temperature (T ref =25 o C) respectively. The reference solar radiation is G=1000W/m 2 (Hedayatizadeh et al., 2013). As reported earlier, part of the total absorbed solar energy (E) is converted into electricity (Eel) in the solar cell according to its efficiency. The second part of absorbed solar energy is lost from the top (Et) of the solar cell to the environment by the effect of the wind and radiation loss. The rest is conducting through the solar cell to the microchannel heat sink. This part of solar energy causes a temperature rise in the solar cell and reduces the electrical efficiency. Hence, the amount of thermal energy passes through the solar cell can be estimated according to the following relation: q' ' E E E (6) w el t Where the top loss (Et) of thermal energy due to the effect of the wind speed and radiation can be calculated by (Zelin Xu and Kleinstreuer, 2014). 1 1 Et Ut ( Tsc Ta) (7) g U t K h g conv (8) 2008). The fluid flow and heat transfer governing equations are described for the fluid and solid domains as follows (Lelea and Laza, 2014): Continuity equation: ( u) ( v ) 0.0 x y Momentum equations in x and y-direction u ( u) P u u u v ( ) ( ) x y x x x y y v u x ( v) P v v v ( ) ( ) y y x x y y Energy equation (Dehghan et al., 2015) CuT CvT T T ( Kl ) ( Kl ) x y x x y y (12) (13) (14) (15) Energy equation for solids T T ( K s ) ( K s ) 0 (16) x x y y Where: K l and K s are the liquid and solid material thermal conductivity. The solid energy equation is applied to the solar cell, tedlar, and the microchannel heat sink material. In the current study, the solar concentration ratio is varied from 10 to 40. While the cooling fluid mass flow rate ranged from 4.47g/s to g/s which is equivalent to Re=10 to 400 for the wide microchannel configuration respectively where Reynolds number based on twice of microchannel height as a hydraulic diameter(kandlikar, 2014). In addition, the wind speed and ambient temperatures are 1m/s and 30 o C respectively through the all simulation results. The top convection heat transfer coefficient from the glass surface to the ambient,(hconv), is depending on wind speed (Vw) (Agrawal and Tiwari, 2011) as follows: h (9) conv V w The useful thermal energy (Eu), friction power, and Reynolds number are calculated according to the following equations: E P u m. C ( T (10) friction f P. m f f T f, out f, in f ) (11) Fluid temperature significantly changes inside the microchannel, especially at high CR values. Hence, the variation in the fluid thermophysical properties is substantial. It is worth mentioning that in the present calculations, the variation of the thermophysical properties of cooling water is considered using the polynomial equations reported in (Jayakumar et al., 104 II.2. Boundary conditions At the inlet the fluid velocity is identified and assumed to be uniform and it is varied according to the value of mass flow rate. In addition, a uniform inlet temperature is assumed. In the meantime, the outlet flow boundary condition is identified at the microchannel outlets. Water molecular mean free path is about 0.25 nm so, Knudsen number falls in the no-slip regime (Kn< 0.001) (Dehghan et al., 2015). Consequently, no-slip and no temperature jump boundary conditions are taken at the interface between the solid fluid domains. The thermal boundary condition for the upper wall of the solar cell is a constant heat flux which is calculated by Eq. 6. Here, it is clear that the upper wall heat flux is a function of the solar cell efficiency which is dependent on the solar cell temperature. Therefore, an iterative technique is applied to calculate the actual incident heat flux. Finally, the lower and side walls of the computational domain are assumed to be adiabatic. In more details, for a flat microchannel domain, the applied boundary conditions are indicated as follows:-

119 1. At the channel inlet (x=0) 1.1. For fluid domain at 0 y H u = Uin, v =0, T = Tin=303 K 1.2. For solid domains (Microchannel walls, tedlar, and solar cell layer) T k 0 s x 2. At the channel outlet x= Lsc 2.1. For fluid domain 0 y H v u T 0, 0, and 0 x x x 2.2. For solid domains (Microchannel walls, tedlar, and solar cell layer) T k s 0 x 3. Upper wall y= (δw +H +δsc+δt), and 0 x Lsc T k '' sc q f ( T sc ) y w 4. Lower wall at y=0 and 0 x Lsc T k ch 0 y 5. For fluid-solid interface: u= v = 0, and k f Tf kchtch 6. For solid-solid interfaces: (a) Solar cell-tedlar interface K sc T sc = K T T T (b) the interface between the tedlar and the microchannel K ch T ch = K T T T Table 2: Optical and physical characteristics of PV cell and MCHS design parameters. Parameter Value, µm Parameter Value L sc δ g H 100 erf δ Sc β ref β sc τ g δ ch 200 α sc δ w 200 α T δ T ε g Pitch (A) P= 2547 µm, N= 4/cm (B) P=1260 µm, N= 8 /cm (C) P= 837 µm, N=12/cm (D) P= 627 µm, N=16/cm 1, (Hedayatizadeh et al., 2013); 2, (Zhou et al., 2015); 3, (Zelin Xu and Kleinstreuer, 2014); 4, (Z. Xu and Kleinstreuer, 2014) II.3. Solution methods and convergence criteria The governing equations along with the described boundary conditions are solved using the commercial CFD package ANSYS FLUENT 14.5 ( ANSYS FLUENT 14.5 Theory Guide) based on discretization using the finite volume method. Pressure-velocity coupling is addressed through the SIMPLE algorithm, along with an algebraic multigrid algorithm (AMG) for solving the linearized system of discretised equations. In addition, the average pressure at the inlet and the average temperature of the solar cell are also monitored to check for convergence of the flow and energy equations. A non-uniform grid arrangement with a large number of grid points near the channel walls is arranged to resolve fluid flow and heat transfer with a consideration of the effect of the boundary layer flow.the resulting system of algebraic equations is solved using the Gauss Seidel iterative technique. The computational domain dimensions, the solar cell optical properties, and layers dimensions are reported in Table 2. The numerical code is verified in some ways to ensure the validity. For every microchannel heat sink, a grid independence test is conducted using several different mesh sizes. II.4. Mesh independence test The mesh independence study is performed for each dimension of the computational domain. The first one is for a cooled CPV system using WMCHS. A different number of cells of , , , and are tested. It is found that there no significant change in the solar cell temperature and the outlet water temperature with further increase of cell numbers after cells. Accordingly, the cell numbers of are selected for the simulation of the WMCHS. The test is established four times for the MMCHS with the studied four-pitch value. It is found that the suitable number of cells are , , , and respectively II.5. Numerical results validation. The results of the present numerical simulation of the microchannel fluid domain are validated using different sets of the available experimental, numerical and theoretical data. The comparison between the predicted fully developed Nu number values with the analytical results (Rohsenow and Hartnett, 1998) is established at different applied Boundary conditions. Table 3 illustrates the comparison between predicted fully developed Nu number and the corresponding analytical values at different boundary conditions in the case of using water as a working fluid. In this comparison, the following three types of boundary conditions are applied: (i) both walls at constant heat flux; (ii) both walls at constant wall temperature; (iii) one wall is adiabatic and the other wall is heated at constant heat flux. Based on Table 3, the maximum error between numerically predicted values and analytical results of Nu number does not exceed 0.65%. Tab. 3. Fully developed Nu number validation Nu fully developed on the heated wall, water, Re=100 B.C CFD Analytical* %Error q w1,2=const % T w1,2=const % q w1=const., q w2= % *(Kandlikar, 2014) 105

120 Where: the numbers 1 and 2 refer to the upper and the lower wall of the wide microchannel. qw and Tw are the constant wall heat flux and constant wall temperature boundary condition respectively. Figure 3 presents the comparison between the current predicted Nusselt number and those measured [30] and predicted using Lattice Boltzmann method, (LBM) [28] at different values of Reynolds numbers. A good agreement between the predicted and measured Nusselt number is found. Also, the friction factor is validated with the analytical results of the fully developed laminar flow friction characteristics presented in (Kandlikar,2014; Rohsenow and Hartnett, 1998) in the case of wide microchannels. An excellent agreement between the predicted friction factor and the analytical results is obtained. inlet ports is 2547µm and the computational domain length is µm. So the total numbers of inlet ports are four per one cm. Similarly, in the case (B), (C), and (D), the pitch values are 1260, 837, and 624 µm which gives a 8, 12, and 16 inlet ports per cm respectively. The same total mass flow rate is used for each case, and CR is selected to be 40. The comparison between the four investigated pitches is implemented based on the average solar cell temperature, cell temperature uniformity, and the consumed pumping power. Figure 4 indicates the variation of the average solar cell temperature under the influence of the cooling fluid mass flow rate. Generally, it is found that increasing the cooling mass flow rate leads to reduce the solar cell temperature. This trend was observed by several researchers (Baloch et al., 2015) and (Du et al., 2012) using water and (Z. Xu and Kleinstreuer, 2014) using 5% vol. Al2O3-water nanofluid. There are different points of view to interpret the reason for this trend. It is reported that at a lower Re range, the heat transfer mechanism between the upper wall and cooling fluid is dominated by convection, while at a higher Re range, the heat transfer mechanism is dominated by conduction within the thin layer of the laminar wall region (Z. Xu and Kleinstreuer, 2014). Another point of view relates this trend to the reduction of contact time between the fluid and the upper wall due to higher velocity associated with higher flow rate (Baloch et al., 2015). The last interpretation is that at high Re, the heat extracted by the cooling water reaches the saturated level and therefore, the cell temperature slightly increases (Du et al., 2012). Fig. 3 Comparison between the predicted Nu number and those measured results of (Kalteh et al., 2012) and previous predicted of the author using twophase Lattice Boltzmann method (Ahmed and Eslamian, 2015). Also, the predicted friction factor and the analytical results (Rohsenow and Hartnett, 1998, Kandlikar, 2014) for flow between two parallel plates. III. Results and discussion In this section local and average solar cell temperature, electrical efficiency, pumping power and temperature uniformity will be discussed for the investigated configurations. In the first part, the effect of the manifold pitch on the average solar cell temperature, temperature uniformity, and the friction power will be discussed. Then the best manifold pitch design is compared with the conventional. The comparison is presented at a mass flow rate ranged from 4.47g/s to g/s and the solar concentration ratio (CR) from 10 to 40. III.1. Effect of the manifold pitch Different values of the manifold pitch are investigated. In case (A), the distance between the two consecutive 106 Fig.4. Variation of the average solar cell temperature with the cooling fluid mass flow rate at the various manifold pitch Regarding to the pitch effect, it is found that decreasing the pitch reduces the solar cell temperature especially at lower flow rates as it is clear in the figure. This is may be attributed to the reduction of the fluid pass under the heated CPV cell. Consequently, the flow might be in the entrance

121 region. The heat transfer in the entrance region is significantly higher than that of fully developed regime. This will cause a better reduction in the solar cell temperature as shown in the case in the case of the lowest pitch (D). The solar cell temperature reduces from o C in the case of pitch (A) to 86 o C for the case of using pitch (D) at the same cooling mass flow rate and solar concentration ratio of 4.47g/s and 40 respectively. In addition, further increase of the mass flow rate beyond 50g/s leads to no significant effect on the solar cell temperature. The reason might be due to the fully developed flow regime. Figure 5 shows the variation of the local solar cell temperature with the lateral distance for the studied pitch values. It is clear that at a lower mass flow rate and the highest solar concentration ratio, increasing the manifold pitch achieves non-uniformity in the solar cell. Moreover, the case (D) achieves better solar cell temperature uniformity. The temperature uniformity is measured by the difference between the minimum and maximum local solar cell temperature MMCHS associated with different number of inlet ports per cm. Results indicated that the optimum number of inlet and outlet ports per cm is found to be 30 per cm. which is in agreement with the results obtained in (Harpole and Eninger, 1991). Figures 7-a and b show the variation of the average solar cell temperature versus the cooling fluid mass flow rate for both WMCHS and MMCH. Generally, it is found that increasing the cooling fluid mass flow rate leads to enhance the average solar cell temperature until a certain limit. Further increasing of mass flow rate attains a slight enhancement in the solar cell temperature. Additionally, increasing the solar concentration ratio leads to the increase of the solar cell temperature due to the rise in the incident heat flux. Figure 6 presents the variation of the pumping power versus the cooling rate at different values of the manifold pitches. It s found that the configuration (A) consumed the highest pumping power. The reason for such trend is probably due to two factors. The first one is that increasing the pitch tends to increase the fluid path length and hence increase the pressure drop. The second one is that reducing the pitch decreases the flow velocity due to increasing the number of inlet ports at the same mass flow rate. To conclude, from the figs 4, 5 and 6, it is clear that increasing the manifold pitch number enhances the solar cell temperature and the temperature uniformity and reduces the consumed pumping power. Fig.6. Variation of the friction power per unit width with the cooling flow rate at various manifold pitches By comparing between 7-a, and 7-b, MMCHS achieve lower average solar cell temperature especially at lower mass flow rates, while increasing the mass flow rates the WMCHS achieve a relatively lower solar cell temperature. The reason for such trend is that at a lower mass flow rate, the heat transfer coefficient for MMCHS is higher than that of WMCHS. At a higher mass flow rate, the heat transfer coefficient is approximately close for both configurations. However, there are stagnation flow points that cause hot spots on the lower side of the microchannel walls due to the flow turning in the MMCHS. Accordingly, it leads to the increase of the cell temperature. Fig.5. Variation of the solar cell temperature with the solar cell lateral length at various manifold pitches III.2.Comparison between the MMCHS and WMCHS The conventional WMCHS is compared with the 107 Comparison of the solar cell temperature uniformity is presented in Fig.8 A and B at a mass flow rates of 4.74g/s and g/s respectively. In both figures, the local solar cell temperature increases with the increase of the WMCHS axial distance while for the MMCHS, the local solar cell temperature is nearly constant along the solar cell length. For instance, the difference between the maximum local solar cell temperature and the lower local solar cell temperature is about 15 o C and 65 o C at CR=10 and 40 respectively

122 for the WMCHS. However, in the case of MMCHS, the maximum local solar cell temperature difference is about 0.3 o C and 0.8 o C at CR=10 and 40 respectively. Regarding the gained electric power, it is found that the solar cell power integrated with the WMCHS is greater than that for MMCHS especially at a higher mass flow rate as shown in Fig.9. This is due to the fact that the solar cell power is directly proportional to its efficiency which is proportional to the solar cell temperature. Moreover, the solar cell temperature is lower in the case of WMCHS than MMCHS at a higher mass flow rate as discussed in Fig. 7. q"w Solar Cell δsc also until it reaches a nearly constant solar cell power after 50g/s. Further increase in the cooling fluid mass flow rate leads to a significant rise in the friction power while remaining the solar cell power as a constant value. So the net power which is defined as the difference between the solar cell electric power and the friction power will increase and then it decreases again. The same trend also observed in (Z. Xu and Kleinstreuer, 2014). Thus in the case of WMCHS, further increase in the cooling fluid mass flow rate beyond 20g/s is not effective at CR=10. Finally, there is an optimum value of mass flow rate which will attain a maximum net output power. This value is dependent on the solar concentration ratio. It is found that the mass flow rate of 25, 35, and 45 g/s are the best suitable for the CPV systems operating with a solar concentration ratio of 20, 30, and 40 respectively. Tedlar δt Inlet y x H Channel wall δw Outlet q"w Solar Cell δsc Tedlar δt Channel wall δw y P H x Inlet Outlet Fig. 7 Variation of the average solar cell temperature with the cooling fluid mass flow rate at various CR values for (A) WMCHS and (B) MMCHS with P=627.2µm It is well known that using microchannel heat sink will significantly increase the friction power loss due to its very small size. Figure 10 shows the variation of the solar cell electric power and both the consumed friction power and the net gained electric power versus the mass flow rate at CR=10. It is clear that for WMCHS, the friction power increase with an increase in the flow rates. While the solar cell power increase 108 Fig.8 the Variation of the local solar cell temperature with the axial distance at cooling fluid mass flow rate of (A) 4.74g/s and (b) 189.6g/s

123 The comparison of the net gained electric power for the CPV systems integrated with the MMCHS and WMCHS is presented in Figs.11 A and B for CR=10 and 40 respectively. It is found that for MMCHS, the net gained power increases with the increase of the cooling fluid mass flow rate and takes the expected trend of the solar cell efficiency. This is due to that the consumed friction power is the same with the gained power. However, in the case of WMCHS, the variation of the net gained electric power is varied as discussed in Fig.10. technique for the solar concentration ratio of 10. On the other hand, at CR=40 as shown in Fig 11 -B, the WMCHS is suitable to achieve highest gained net electric power for a cooling fluid mass flow rate ranged from 8.3 to71.1 g/s. Fig.9. Variation of the solar cell electric power versus the cooling fluid mass flow rate for MMCHS and WMCHS at CR=10 Fig. 10. Variation of the friction power, net power and the solar cell power with the cooling fluid mass flowrate at CR=10 In Fig. 11-A, when the cooling fluid mass flow rate is greater than 27.7g/s, MMCHS is preferable to achieve a higher net gained power than WMCHS. If the pumping system is not capable of delivering a cooling mass flow rate greater than 27.7g/s per meter width of channel, hence the WMCHS is the suitable cooling 109 Fig.11 variation of the net gained electric power with the cooling fluid mass flow rate at (A) CR=10 and (B) CR=40 IV. Conclusions A comprehensive 2D- mathematical model has been developed to predict the performance of the CPV systems using two different designs of manifold microchannel heat sinks (MMCHS) and conventional wide microchannel heat sink (WMCHS). For the MMCHS, the manifold pitch effect is numerically investigated. It is found that the manifold pitch of µm is most appropriate for the cooling of CPV system. Such proposed manifold pitch achieves the lowest solar cell temperature, best temperature uniformity of the solar cell and the minimum friction power. Furthermore, using MMCHS attains better

124 temperature uniformity than the WMCHS. The difference between the maximum local solar cell temperature and the lower local solar cell temperature is about 15 o C and 65 o C at CR=10 and 40 respectively for the WMCHS. However, when using MMCHS, the maximum local solar cell temperature difference is about 0.3 o C and 0.8 o C at CR=10 and 40 respectively. Furthermore, at the cooling fluid mass flow rate greater than 27.7g/s, MMCHS is achieving a higher net gained power than WMCHS. If the cooling mass flow rate is greater than 27.7g/s per meter, the WMCHS is providing a hiher net gain power at solar concentration ratio of 10. Acknowledgement The author would like to thank the Egyptian government especially Ministry of Higher Education (MoHE). We also would like to express our gratitude to Egypt-Japan University of Science and Technology (E-JUST) for offering the facilities and tools. Nomenclature Cp : Specific heat of cooling fluid (J.kg -1.K) G(t) : Incident solar radiation (W.m -2 ) h : Heat transfer coefficient (W.m -2.K) H : microchannel height (m) k : Thermal conductivity (W.m -1.K) L : Microchannel length, solar cell length (m) m : Unit cooling fluid mass flow rate (Kg.s -1 ) N : Number of inlet or outlet ports Nu : Nusselt number Nu=h.Dh /Kf P : Pressure (Pa), electrical power (W) T : Temperature ( o C) T : Temperature ( o C). Ut : Overall heat transfer coefficient from the top surface of solar cell to ambient (W.m -2.K) V : Velocity vector (m.s -1 ). Vw : Wind velocity (m.s -1 ). Greek symbols α : Absorptivity β : Backing factor and solar cell temperature coefficient (K -1 ) ε : Emissivity τ : Transmittivity µ : Fluid viscosity (Pa.s) σ : Stephan-Boltzmann constant 5.67*10-8 (W. m - 2.K 4 ) ρ : Fluid density (kg.m -3 ) δ : Thickness (m) : Solar cell and thermal efficiency Subscripts a : Ambient conv. : convection eff : Effective el : Electrical f : fluid g : Glass in : Inlet out : Outlet ref : Reference condition, G=10 3 W.m -2,T=25 o C 110 Sc Sc, x T th w w References : Solar cell : Local solar cell : Tedlar : Thermal : Wall : water Agrawal, S., Tiwari, A., Experimental validation of glazed hybrid micro-channel solar cell thermal tile. Sol. Energy 85, (2011). Ahmed, M., Eslamian, M. Laminar forced convection of a nanofluid in a microchannel: Effect of flow inertia and external forces on heat transfer and fluid flow characteristics. Appl. Therm. Eng. 78, (2015). ANSYS FLUENT 14.5 Theory Guide [WWW Document], n.d. URL project/neptunius/docs/fluent/html/th/node322.htm (accessed ). Baloch, A. a. B., Bahaidarah, H.M.S., Gandhidasan, P., Al-Sulaiman, F. a.. Experimental and numerical performance analysis of a converging channel heat exchanger for PV cooling. Energy Convers. Manag. 103, 14 27(2015). Dehghan, M., Daneshipour, M., Valipour, M.S., Rafee, R., Saedodin, S.. Enhancing heat transfer in microchannel heat sinks using converging flow passages. Energy Convers. Manag. 92, (2015). Du, B., Hu, E., Kolhe, M.. Performance analysis of water cooled concentrated photovoltaic (CPV) system. Renew. Sustain. Energy Rev. 16, (2012). G. N. Tiwari and Swapnil Dubey, Fundamentals of Photovoltaic Modules and Their Applications. The Royal Society of Chemistry, (2010). Harpole, G.M., Eninger, J.E.. Micro-channel heat exchanger optimization, in: 1991 Proceedings, Seventh IEEE Semiconductor Thermal Measurement and Management Symposium. IEEE, pp (1991). Hedayatizadeh, M., Ajabshirchi, Y., Sarhaddi, F., Safavinejad, A., Farahat, S., Chaji, H., Thermal and Electrical Assessment of an Integrated Solar Photovoltaic Thermal (PV/T) Water Collector Equipped with a Compound Parabolic Concentrator (CPC). Int. J. Green Energy 10, (2013). Jayakumar, J.S., Mahajani, S.M., Mandal, J.C., Vijayan, P.K., Bhoi, R., Experimental and CFD estimation of heat transfer in helically coiled heat exchangers. Chem. Eng. Res. Des. 86, (2008).

125 Kalteh, M., Abbassi, A., Saffar-avval, M., Frijns, A., Darhuber, A., Experimental and numerical investigation of nano fl uid forced convection inside a wide microchannel heat sink. Appl. Therm. Eng. 36, (2012). Kalteh, M., Abbassi, A., Saffar-Avval, M., Harting, J., Eulerian Eulerian two-phase numerical simulation of nanofluid laminar forced convection in a microchannel. Int. J. Heat Fluid Flow 32, (2011). Kandlikar, S.G., Heat Transfer and Fluid Flow in Minichannels and Microchannels,(2014). Lelea, D., Laza, I., The water based Al2O3 nanofluid flow and heat transfer in tangential microtube heat sink with multiple inlets. Int. J. Heat Mass Transf. 69, ,( 2014). Rahimi, M., Karimi, E., Asadi, M., Valeh-e-Sheyda, P. Heat transfer augmentation in a hybrid microchannel solar cell. Int. Commun. Heat Mass Transf. 43, ,(2013). Xu, Z., Kleinstreuer, C., Concentration photovoltaicthermal energy co-generation system using nanofluids for cooling and heating. Energy Convers. Manag. 87, (2014). Xu, Z., Kleinstreuer, C., Computational Analysis of Nanofluid Cooling of High Concentration Photovoltaic Cells. J. Therm. Sci. Eng. Appl. 6, (2014). Yang, K., Zuo, C., A novel multi-layer manifold microchannel cooling system for concentrating photovoltaic cells. Energy Convers. Manag. 89, (2014). Zhao, J., Song, Y., Lam, W.-H., Liu, W., Liu, Y., Zhang, Y., Wang, D. Solar radiation transfer and performance analysis of an optimum photovoltaic/thermal system. Energy Convers. Manag. 52, (2011). Zhou, J., Yi, Q., Wang, Y., Ye, Z., Temperature distribution of photovoltaic module based on finite element simulation. Sol. Energy 111, (2015). Ramos-Alvarado, B., Li, P., Liu, H., Hernandez- Guerrero, A. CFD study of liquid-cooled heat sinks with microchannel flow field configurations for electronics, fuel cells, and concentrated solar cells. Appl. Therm. Eng. 31, (2011). Reddy, K.S., Lokeswaran, S., Agarwal, P., Mallick, T.K., Numerical Investigation of Micro-channel based Active Module Cooling for Solar CPV System. Energy Procedia 54, ,(2014). Rejeb, O., Dhaou, H., Jemni, A., A numerical investigation of a photovoltaic thermal (PV/T) collector. Renew. Energy 77, (2015). Rohsenow, W.M., Hartnett, J.R., Forced convection internal flow in ducts. McGraw-Hill, New York (1998). Rosell, J., Chemisana, D., Tadrist, L., Iban, M., Effect of a hybrid jet impingement / micro-channel cooling device on the performance of densely packed PV cells under high concentration (2011). Royne, A., Dey, C.J., Mills, D.R., Cooling of photovoltaic cells under concentrated illumination: A critical review. Sol. Energy Mater. Sol. Cells 86, , (2005). Sarangi, S., Bodla, K.K., Garimella, S. V., Murthy, J.Y., Manifold microchannel heat sink design using optimization under uncertainty. Int. J. Heat Mass Transf. 69, (2014). Whitfield, G., Bentley, R., Weatherby, C., Hunt, A., Mohring, H.-D., Klotz, F., Keuber, P., Miñano, J., Alarte-Garvi, E., The development and testing of small concentrating PV systems. Sol. Energy 67, 23 34(1999). 111

126 Thermodynamic Analysis of Parabolic Solar Collector Driven Double-Effect Absorption Cooling System Fatih Yigit 1*, Ahmet Kabul 2, Onder Kizilkan 3 1,2,3 Suleyman Demirel University, Faculty of Technology, Energy System Engineering Department, Isparta, 32000, Turkey * fatihyigit@sdu.edu.tr Abstract This paper presents an analysis of thermodynamic performance of a double effect absorption cooling system, LiBr- H2O is used as fluid couple, driven by solar energy is supplied trough parabolic collector. The analysis is performed by using a software program Engineering Equation Solve (EES). First, cooling demand of an isolated supermarket with 1000 m 2 area is assumed 100 kw based on the location of the supermarket. To meet the cooling need, a double effect absorption cooling is considered and the double effect absorption cooling system is analyzed thermodynamically. Also some parametric studies are carried out by the variation of required thermal energy for generator, evoparator temperature, solar radiation intensity and parabolic solar collector s technical properties. Keywords: thermodynamic analysis, parabolic collector, solar energy, absorption cooling I. Introduction The rapid population growth and enhancement on the technological field for the last two decades, and people s demands for higher life standards and comfort level give rise to increasingly excessive energy consumption. One of the necessaries to comfort level and health is air-conditioning due to all people need fresh air (Evangelos et al., 2016). According to the International Institute of Refrigeration in Paris, the amount of electricity consumption of airconditioning process is approximately 15% of all the electricity produced in the world. On the other hand, electricity consumption for air conditioning systems has been estimated around 45% of the whole residential and commercial buildings (Kalkan et al., 2012). There are a few system using for air-conditioning however among these systems, especially vapor compression cooling systems consume too much electricity (Yilmazoglu, 2010). So these systems support to climatic change and global warning by using fluid (CFC) and excessive electricity consumption. CFC (Chloro Fluoro Carbon) and HCFC (hydrochlorofluorocarbon) gasses which are used in conventional cooling application utilization was band due to increase global warning by weakening ozone layer (Bejan, 1997). Additionally, it is uncertain that what kind of damages will emerge because of utilization of inorganic gasses which are used in conventional cooling systems. When this situation is considered, using double effect absorption cooling system which is an environmentally friendly system compared with conventional refrigeration systems is very beneficial to our world and country s feature. Double effect absorption cooling systems have relatively very low electricity consumption thanks to using pump rather than compressor for compression process (Onan et al., 2010). Thus, double effect absorption cooling systems with low electrical energy consumption and using renewable energy such as geothermal, solar, and waste heat is very important for sustainable energy and environment. Advantage of Absorption Cooling System (Sencan, 1999); Quiet operation because of pump quite quiet when compare compressor. It requires little maintenance. They can provide a complete productivity for variable cooling loads. Reduces energy costs thanks to use renewable energy sources such as sun, geothermal and waste heat. The absorption cooling system with H2O-LiBr might not work under 75 C because of crystallization of the LiBr. Therefore, generator temperature must be on the 75 C. To reach high generator temperature must be use like evacuated or parabolic trough solar collectors. Assumed that heat loss of the parabolic trough solar collectors only via radiation heat transfer so this type solar collectors have great advantages for solar cooling applications (Ozturk, 2008; Sencan, 1999). Turkey, which has got an advantageous location has a great potential of solar energy that can be used for numerous applications such as electricity generation, heating, cooling, etc. Solar energy is generally being used for water heating especially in southern of part of Turkey. The utilization of the solar energy can lead reduction of fossil fuel consumption and thus a reduction of carbon dioxide emissions. On the other hand, the high potential of solar energy can be utilized for cooling applications in summer times, by using double effect absorption cooling systems driven by 112

127 solar energy instead of electricity (Ozturk, 2008). There are a number of studies made by different researchers in the literature about absorption refrigeration systems. Onan et al. (2010) designed a solar assisted absorption refrigeration system (SAARS) for acclimatizing of villas in Mardin and they also analyzed the performance of the system under different temperatures by using MATLAB. F. Assilzadeh et al. (2005) studied a solar cooling system that has been designed for Malaysia and similar tropical regions using evacuated tube solar collectors and they carried out LiBr absorption unit by using TRNSYS program. Aman et al. (2014) developed a thermodynamic model which is based on a 10 kw air cooled ammonia water absorption chiller driven by solar thermal energy. They conducted both energy and exergy analyses to evaluate the performance of this residential scale cooling system. Praene et al. (2011) designed a solar-driven 30 kw LiBr/H2O single-effect absorption cooling system and installed at Institut Universitaire Technologique of Saint Pierre. Ratlamwala et al. (2012) investigated a parametric study which are undertaken and the effects of some operating conditions such as geothermal temperature, geothermal mass flow rate, concentration of ammonia water vapor, temperature of inlet stream to the very high temperature generator (VHTG), and pressure of the first turbine on the outputs of the system. Solum et al. (2011) examined effect of thermodynamic quantities of any doubleeffect absorption system operating by means of double fluid, LiBr-water on system performance by using an engineering program which name is EES. This paper presents thermodynamic performance analysis of a solar assisted double effect absorption cooling system using LiBr-H2O fluid couple. In the analysis, parabolic trough solar collectors are used for utilizing of solar energy. The analysis is performed by using the Engineering Equation Solve (EES) (F-Chart, 2016) software program. Parametric studies also are carried out for some variables such as thermal energy demand for generator, solar radiation intensity, evaporator temperature and PTSC technical properties. II. System Description The investigated supermarket which is to be cooled has got the dimensions of 1000 m 2 in area, 3.7 m in height and 3700 m 3 in volume, respectively. It is assumed that the supermarket is isolated with 0.04 m insolation material which has a got a heat conduction coefficient of W/mK. The solar assisted proposed double effect absorption cooling system consists of a solar parabolic trough collector, two generators, two heat-exchangers, a condenser, an evaporator, an absorber and four expansion valves. The proposed system is given in Figure 1. Technical properties of the parabolic trough solar collector used 113 the solar assisted double effect absorption cooling system given table 1. Tab. 1: Properties of the parabolic trough solar collector Single collector width 3.5 m Receiver inner diameter 0.04 m Receiver outer diameter 0.05 m Cover diameter 0.09 m Receiver emissivity 0.92 Glass cover emissivity 0.87 Temperature of the sun 5739 K Solar radiation intensity 500 W/m 2 System design parameters and assumptions are described below. The fluid couple used in the system is LiBr- H2O Evaporator temperature is considered to be 12 C Condenser and absorber temperature is considered to be 42 C. The effectiveness of heat exchanger for a counter flow heat exchanger is 0.6 on average Pressure losses in the system was neglected. Absorber, generator, condenser and evaporator are isolated for heat loss and gain III. Thermodynamic Analysis The performance of the parabolic solar collector driven absorption cooling system is mathematically modeled using mass, energy and exergy balance equations. In order to carry out the thermodynamic analysis for the system, some assumptions are made: The system processes are steady state. Kinetic and potential energies of the changes are ignored. The fluid couple at pump inlet is saturated liquid. The pumps are adiabatic. Potential and kinetic energies are neglected. The dead state pressure and temperature are taken P0= kpa and T0= 5 C General mass balance equation can be written as (Cengel and Boles, 2006): m in = m out (1) The general energy balance can also be written as: E in = E out (2) where E in is the ratio of net energy transfer to the system, E out is the ratio of net energy transfer from the system. For steady-flow processes the general exergy balance is defined as: Q + m inh in = W + m outh out (3)

128 Fig 1. The double effect absorption cooling system Where Q is the ratio of net heat, W is the ratio of net work, and h is the specific enthalpy. The rate of useful energy delivered by solar collector is defined as (Tiwari, 2003; Kalogirou, 2009) Q u = F R [SA a A r U L (T i T a )] (4) Q u = m C p (T o T i ) (5) where F R is the heat removal factor, S is the heat absorbed by the receiver, A a is the aperture area, A r is the receiver area, and U L is the solar collector overall heat loss coefficient. The general exergy balance equation can be defined as (Dincer and Rosen, 2007): Eẋ in = Eẋ out + Eẋ dest (6) The exergy balance equation can be expressed more explicitly as: Eẋ Q Eẋ W = m ine in m oute out + T 0 S gen (7) 114 where, Eẋ Q and Eẋ W terms are the exergies of heat and work, e is specific exergy, T0 is the state temperature and S gen is the rate of entropy generation. The terms in Equation 7 are described below (Dincer and Rosen, 2007): Eẋ dest = T 0 S gen (8) Eẋ Q = Q ( T T 0 T ) (9) Eẋ W = W (10) The specific exergy can be expressed as (Cengel and Boles, 2006; Bejan, 1997): e = (h h 0 ) T 0 (s s 0 ) (11) where, h is enthalpy, s is entropy, T is the temperature and subscript 0 stands for reference state properties. The exergy efficiency can be expressed as (Dincer and Rosen, 2007): η ex = Eẋ out Eẋ in = 1 Eẋ dest Eẋ in (12)

129 where, Eẋ out is the rate of total energy output, Eẋ in is the rate of the total energy input. The solar exergy is defined as (Petela, 2005): Eẋ solar = S A a ( ( T 0 T sun ) ( T 0 T sun )) (13) where Eẋ solar is the function of the outer surface temperature of the sun. IV. Result and Discussion In this paragraph, all the simulation results are presented. Firstly, coefficient of performance curve of system are given as a function of solar radiation. The next results are related to the exergy efficiency for each component of the double effect absorption cooling system. In this section some parametric studies are given for understanding performance change of the system. Table 2 shows the calculated thermodynamic properties of the solar assisted double effect absorption cooling system. Tab. 2: Thermodynamic properties of each point of the solar assisted double effect absorption cooling system Points Fluid Phase T P Cp h m s x Ex [ᵒC] [kpa] [kj/kg.k] [kj/kg] [kg/s] [kj/kg] [%] [kw] 0 H2O H2O-LiBr Strong mix H2O-LiBr Strong mix H2O-LiBr Strong mix H2O-LiBr Strong mix H2O-LiBr Mean mix H2O-LiBr Mean mix H2O-LiBr Mean mix H2O-LiBr Weak mix H2O-LiBr Weak mix H2O-LiBr Weak mix H2O Vapor H2O Liquid &Vapor H2O Liquid &Vapor H2O Vapor H2O Liquid H2O Liquid &Vapor H2O Vapor H2O Comp. Liquid H2O Comp. Liquid H2O Comp. Liquid H2O Comp. Liquid H2O Comp. Liquid H2O Comp. Liquid H2O Comp. Liquid H2O Comp. Liquid Figure 2 shows how the Coefficient of Performance (COP) and Exergy destruction of the system change with generator I temperature. COP COP Ex DestTotal T GI ( C) Fig. 2: COP according to the change in the Generator I temperature Ex DestTotal (kw) Figure 3 presents the COP of the absorption cooling system with increase absorber temperature. It shows clearly that COP increase with absorber temperature but exergy destruction of the system goes down. COP COP Ex DestTotal T E ( C) Fig. 3: COP according to the change in the Evaporator temperature Ex DestTotal (kw) 115

130 In Figure 4 seen that with the increase of the condenser temperature exergy destruction of the system decrease and COP goes up. According to Figure 5 as increase the absorber temperature COP is decrease on the other hand exergy destruction of the system first goes up than sharply goes down. The cause of the sharply decrease of exergy destruction might be crystallization of the LiBr. COP COP Ex DestTotal Ex DestTotal (kw) Fig. 7: Exergy Efficiency for each component according to the change in the Absorber temperature T C ( C) Fig. 4: COP according to the change in the Condenser temperature COP COP Ex DestTotal Ex DestTotal (kw) Fig. 8: Exergy Efficiency for each component according to the change in the Condenser temperature T A ( C) Fig. 5: COP according to the change in the Absorber temperature In Figure 6 to 9 the variation of exergy efficiency with different system parameters are given. Each figure show the effects of evaporator temperature, absorber temperature, condenser temperature and generator temperature on the exergy efficiency of each system components. Fig. 9: Exergy Efficiency for each component according to the change in the Generator I temperature Fig. 10: Exergy Efficiency for each component according to the change in the Evaporator temperature 116 V. Conclusions In this study, solar assisted double effect absorption refrigeration system analyzed using first and second laws of thermodynamics. According to the calculations, absorption refrigeration system s COP value was found to be 0.87 when the generator I temperature was 125 C. Additionally, some parametric studies have been performed to see the variation of exergy destruction rates and exergy efficiencies of the system components. It was found that with the increase of the first generator temperature, system performance increased. Consequently, parabolic trough collector integrated systems are becoming attractive for mid-temperature

131 thermal applications such as absorption refrigeration and also power generation systems. Additionally, it can be expected that these systems can redeem itself in a short period of time related to the technological improvements in PTSC systems. Nomenclature F R A a A r U L E x h m P PTSC Q s T W x References : heat removal factor, : aperture area, : receiver area, : solar collector overall heat loss coefficient. : Exergy (kw) : Specific enthalpy (kj/kg) : Mass flow rate (kg/s) : Pressure (kpa) : Parabolic Trough Solar Collector : Heat load (kw) : Specific entropy (kj/kg.k) : Temperature (C) : Work (kw) : Concentration Aman J., Ting D.S.K., Henshaw P., Residential Solar Air Conditioning: Energy and Exergy Analyses of an Ammonia-water Absorption Cooling System, Applied Thermal Engineering, 62, , Assilzadeh F., Kalogirou S.A., Ali Y., Sopian K., Simulation and Optimization of a Libr Solar Absorption Cooling System with Evacuated Tube Collectors, Renewable Energy, 30, , Bejan, A., Advanced Engineering Thermodynamics, John Wiley and Sons, New York, USA, Cengel Y.A., Boles M.A., Thermodynamics an Engineering Approach Eighth Edition, Published by McGraw-Hill Education, New York, USA, Dincer I., Rosen M.A., Exergy: Energy, Environment and Sustainable Development, Elsevier Science; 1st ed., Oxford, UK, Onan C., Ozkan D.B., Erdem S., Exergy Analysis of a Solar Assisted Absorption Cooling System on an Hourly Basis in Villa Applications, Energy, 35, , Ozturk, H.H., Gunes Enerjisi Ve Uygulamalari, Birsen Publishing, 2008, (In Turkish). Petela R., Exergy analysis of the solar cylindricalparabolic cooker, Solar Energy, 79, , Praene J.P., Marc O., Lucas F., Miranville F., Simulation and experimental investigation of solar absorption cooling system in Reunion Island, Applied Energy, 88, , Ratlamwala T.A.H., Dincer I., Gadalla M.A., Thermodynamic analysis of a novel integrated geothermal based power generation-quadruple effect absorption cooling-hydrogen liquefaction system, International Journal of Hydrogen Energy, 37, , Sencan A., Absorpsiyonlu Sogutma Sisteminin Tasarimi ve S.D.U Oditoryumunda Uygulanabilirliginin Arastirilmasi, MSc Thesis, Süleyman Demirel University, Isparta, Turkey, 1999 (In Turkish). Solum C., Koc I., Altuntas Y., Cift Etkili Libr-H2O Akiskanli Absorpsiyonlu Sogutma Sisteminde Termodinamiksel Buyukluklerin Sistem Performansina Etkileri, Journal of Aeronautics and Space Technologies, 1, 19-26, 2011 (In Turkish). Tiwari G.N., Solar Energy: Fundamentals, Design, Modelling and Applications, Alpha Science International Ltd., Pangbourne UK, Yilmazoglu M.Z., Tek Etkili Bir Absorpsiyonlu Sogutma Sisteminin Termodinamik Analizi, Journal of Faculty of Engineering and Architecture Gazi University, 25, , 2010 (In Turkis). Evangelos B., Christos T., Kimon A. A., Exergetic, energetic and financial evaluation of a solar driven absorption cooling system with various collector types, Applied Thermal Engineering, 102, , F-Chart Software. Engineering Equation Solver (EES). accessed on Kalkan N., Young E.A., Celiktas A., Solar Thermal Air Conditioning Technology Reducing the Footprint of Solar Thermal Air Conditioning, Renewable & Sustainable Energy Reviews, 16, , Kalogirou S.A., Solar energy engineering: processes and systems, 1st ed., Academic Press, Oxford, UK,

132 Energy and Exergy Analyses of a Biomass Fired Regenerative ORC System Ozum Calli 1*, Can Ozgur Colpan 2, Huseyin Gunerhan 3 1 Izmir University of Economics, Vocational School, Ventilation-Air Conditioning Technology Department, Balcova, Izmir, 35330, Turkey 2 Dokuz Eylul University, Faculty of Engineering, Mechanical Engineering Department, Tinaztepe, Buca, Izmir, 35397, Turkey 3 Ege University, Faculty of Engineering, Mechanical Engineering Department, Bornova, Izmir, 35040, Turkey * ozum.calli@ieu.edu.tr Abstract Heat generated from the combustion of biomass can be used as an energy source in an Organic Rankine Cycles (ORC). In this paper, an integrated biomass fired regenerative ORC system is examined using energy and exergy analyses. For this purpose, several control volumes enclosing the components of the system are formed. Applying exergy balances, exergy destruction in each control volumes are calculated. Various parameters including turbine inlet temperature, mass flow rate of dry biomass, fuel-air ratio, and type of biomass are investigated and to what extent which parameter affects the electrical efficiency and exergetic efficiency is determined. Some suggestions are given for increasing the electrical and exergetic efficiencies. Keywords: Organic Rankine Cycle, ORC, energy, exergy, biomass, regenerator I. Introduction As the human population grows and technological devices advance, dependence on energy increases significantly. Organic Rankine Cycles (ORCs) represent an attractive solution for the energy requirement, where traditional applications are technologically and economically unfeasible. The principle of electricity generation by means of an ORC is similar to the conventional Rankine cycle. The difference between these two cycles is that an organic working fluid (e.g. R134a, R245fa, R123, and hydrocarbons such as iso-pentane and iso-octane) with favourable thermodynamic properties at lower temperatures and pressures is used instead of water in the ORC (Qiu et al., 2011). For instance in geothermal ORC applications boiling point of ORC fluids chosen as working fluids are less than water (e.g. Quoilin et al., 2013). This fluid is selected such that it can utilize the heat from lower temperature sources. The fluid that takes the heat from this low temperature source is used to drive a turbine to generate electricity. There are various types of energy sources that can be used in an ORC system such as geothermal, solar, waste heat from industry, and biomass. This fluid is selected such that it can utilize the heat from lower temperature sources. The fluid that takes the heat from this low temperature source is usedtodrive a turbine to generate electricity. There are various types of energy sources that can be used in an ORC system such as geothermal, solar, waste heat from industry, and biomass. Organic Rankine cycle was investigated in terms of many parameters in the studies found in the literature. The most investigated parameteristhe working fluid and the most preferred fluid is R134a (Guo et al., 2010; Chen et al., 2010; Maizza and Maizza, 1996; Marion et al., 2012; Saleh et al., 2007). On the other hand, the selection of the most suitable working fluid depends on the operating conditions of the cycle. Lakew et al. (2010) investigated different working fluids for power production at different operating conditions and heat sources with different temperatures for a subcritical Rankine cycle. Their studies showed that R227ea gives the highest power for a heat source temperature range of C and R245fa produces the highest power in the range of C. Assessment of different configurations of ORC is another topic that has been widely investigated in the literature. Forinstance, Al-Sulaiman (2014) conducted exergy analysis of a combined steam Rankine cycle and organic Rankine cycle, which are both integrated with parabolic trough solar collectors. As a result of this study, it is found that the main source of exergy destruction is the solar collector. In addition, it is shown that as the solar irradiation increases, the exergetic efficiency increases. The highest and lowest exergetic efficiencies are obtained when R134a (26%) and R600 (20%) are used as the working fluid in the combined cycle, respectively. Biomass fired ORC systems can efficiently convert the chemical energy of biomass into electricity. Biomass is an important renewable energy source, available nearly everywhere and has the advantage of continuity by contrast solar and wind is intermittent. As the sunlight is presenced, biomass stored carbon continiously. In addition, biomassisoften economically favorable. There are a few studies on the biomass fired ORC in the literature. For example, Liu et al. (2011) investigated 2 kwe biomass-fired combined heat and 118

133 power (CHP) system based on anorganic Rankine cycle (ORC) by using three different environmentally friendly refrigerants, namely HFE7000, HFE7100 and n-pentane, as the ORC working fluids. They found that the electrical efficiency of the CHP system mainly depends on the temperature of the hot water entering the biomass boiler and the temperature of the cooling water entering and exiting the ORC condenseras well as the type of the ORC fluid. In another study by the same research group (Liu et al., 2012), a 0.8 kwe biomass-fired CHP system was investigated experimentally and it was found that the electricity generation efficiency is 1.41%, which islower than that predicted by the thermodynamic modelling. With an evaporator temperature of 120 C, the thermodynamic modelling study gives the electrical efficiency in the range of 8 9%. There are two main factors responsible for the apparent difference in the electrical generation efficiency between the experiments and thethermodynamic modelling: The first one is the expander efficiency. In the model, the value of this efficiency is assumed as 85%; whereas the experimental results show that this efficiency is only 53.92%. The second one is the alternator efficiency. In themodel, it is assumed that this efficiency is 90%; but the experimental results show that it is only 50.94%. Huang et al. (2013) investigated regenerativeandnon-regenerativebiomass-orc with dry and wet working fluids and found that the highest electric power was obtained for there generative system with methylcyclohexane applied as the dry working fluid. The aim of this paper is to analyze the energetic and exergetic performances of a regenerative biomass fired ORC by investigating to what extent which parameters affect. II. System description Fig. 1 shows the schematic diagram of the regenerative biomass-fired ORC system studied in this study. The system can be divided into two sections as the biomass side and the ORC cycle. In the biomass side, biomass fuel and air enter the burner and as a result of the combustion process, heat is generated. This heat is transferred to the biomass side working fluid. The working fluid in the biomass side then enters the heat exchanger that connects the biomass side and the ORC. In the ORC, the organic fluid first gets the heat from the biomass side through the heat exchanger. This fluid then expands in theturbine, producing mechanical energy, further transformed into electric energy through a generator. The fluid expanded in the turbine enters the regenerator, which is used to increase the electrical efficiency of the ORC. The exit of the regenerator enters the condenser and pump consecutively before entering the regenerator again. The regenerator increases the heat exchanger inlet temperature and decreases the heat gained from the burner. The fluid leaving the regenerator enters the heat exchanger completing the cycle. 119 Fig. 1: Schematic diagram of a regenerative biomass-fired ORC system. III. Mathematical Model The mathematical model of the integrated biomass fired ORC system is developed using energy and exergy analyses, which are discussed in thefollowingsubsections. A commercially available software, Engineering Equation Solver (EES), is used for the solution of the equations. The main assumptions made to carry out the energy and exergy analyses of the ORC system are listed below. The system is assumedtowork at steady state. The pressure drops along the components and the lines connecting the components are neglected except in the pump and ORC turbine. The heat loss from the components to the surroundings is neglected. The kinetic and potential energy and exergy changes are neglected. Air and the combustion gases are assumed to be ideal gases; and the biomass side fluid (Thermal oil) is assumed to be incompressible fluid. III.1. Energy analysis In this subsection, the energy analysis of the system considered is presented. Energy balance for steady state control volumes can be shown as follows: Q cv W cv = n o (h + v2 g. z) i 2 + g. z) o n i (h + v2 2 + (1) where Q cv and W cv are heat transfer rate and power rate of the control volume, respectively, and n i and n o are the molar flow rate of the working fluid at the inlet and outlet of the control volume, respectively. In this study molar unit system is selected for convenience as some of the modeling equations are written as a function of the molar compositions of the

134 chemical species. Hence, most of the equations given in this study are indicated according to molar unit system. v, g and z denote the velocity, gravitational acceleration and elevation according to a reference point, respectively. In the burner, the combustion process occurs in which air and biomass fuel react. The combustion gases are emitted to the atmosphere. The chemical reaction for the combustion of biomass, which mainly consists of C, H, and O atoms, can be shown as follows: C x H y O z + (λ γ )(O N 2 ) x CO 2 + (y/2) H 2 O + (α γ )O 2 + (3.76 λ γ)n 2 (2) In this equation, λ and α denote the theoretical air and excess air coefficient, respectively. γ is the stoichiometric air coefficient for the complete burning reaction of C x H y O z when there is no excess air. The relation between the excess air coefficient and the theoretical air can be shown as: α = λ 1 = λ + 2z 4x y 4γ Energy balance for the burner can be shown as: (3) n air h air + n biomass h biomass = n excessgases h excessgases + n biomassfluid (h outlet h inlet ) (4) Energy balance around the control volume enclosing the heat exchanger that connects the biomass side with the ORC is: n biomassfluid (h 1 h 2) = n ORCfluid (h 4 h 3) (5) Energy balance for the regenerator can be written in a similar way using Eq. (1). Using the energy balance for the condenser, heat transfer rate from the condenser to the cooling water can be shown as follows. Q condenser = n ORCfluid (h 6 h 7) = n cw (h 10 h 9) (6) Applying the energy balance for the control volumes enclosing the turbine and the pump, the power output of the turbine (Eq. (7)) and the power input to the pump (Eq.(8)) can be found, respectively. In these equations, η s,t denote the isentropic efficiency of the turbine, which can be defined as the actual work output of the turbine to the work output of the turbine if the turbine undergoes an isentropic process. η s,p is the isentropic efficiency of the pump, which is the ratio of the work input for an isentropic process, to the work input for the actual process. W turbine = η s,t (h 4 h 5,s ) (7) W pump = (h 8,s h 7) η s,p = υ 7 (P 8 P 7 ) η s,p (8) 120 III.2. Exergy analysis Exergy analysis is generally used to quantify the magnitudes of the irreversibilities in the thermal energy systems. Exergy balance, which is derived combining the energy and entropy balances, is applied to the components of the system to find the exergy flow rates at each state and the exergy destruction of each component. Exergy destruction can also be regarded the potential work lost due to irreversibilities. At the steady state conditions, the exergy destruction rate of a control volume can be found applying the exergy balance for a control volume, as follows: Ex d = (1 T 0 ) Q T j W cv + (n ex ) i (n ex ) o (9) j where n, T j, T 0, Q, Ex and Ex d are the molar flow rate, temperature of the boundary where heat transfer occurs, temperature of the environment, heat transfer rate between control volume and the environment, exergy flow rate, and rate of exergy destruction. The summation of the exergy destruction rate of the each component is called the total exergy destruction of the system. The contribution of the exergy destruction of each component in the total exergy destruction rate can be found by calculating the exergy destruction ratio as follows. y 1,i = Ex d,i Ex d,total (10) where Ex d,i and Ex d,total denote the exergy destruction rate of the component i and the total exergy destruction of the system. Alternatively, the exergy destruction rate of a component can be compared to the chemical exergy rate of the fuel (Colpan, 2005), which is taken as biomass for this study: y 2,i = Ex d,i Ex biomass (11) where Ex biomass and Ex d,i denote the exergy flow rate of the fuel and exergy destruction rate of the component i, respectively. The exergy includes the physical and chemical exergy, if the kinetic and potential exergies are neglected. ex = ch ex + ph ex (12) where ph ex is the specific physical exergy and ch ex is the specific chemical exergy at a given state. In general, if the chemical composition of a substance does not change at the inlets or exits of a control volume, the chemical exergy is not needed to be calculated to find the exergy destruction in that control volume. For the system studied, chemical exergy is included in the calculations only in the burner because

135 chemical reaction takes place in the combustion process. Physical flow exergy rate at a given state defined as: ph ex = (h h 0) T 0 (s s 0) (13) where h and s are specific enthalpy in molar basis and specific entropy in molar basis. s 0 and h 0 denotes specific enthalpy and specific entropy in molar basis for the dead state which defines conditions of the reference environment. In this study, The specific chemical exergy of an ideal gas mixture is defined as: ch ex = ex och + R T o ln x i (14) Here, x i is the mole fraction of species i and och ex is the standard specific chemical exergy (in molar basis) at the reference temperature and pressure. The chemical exergy of biomass can be defined as (Szargut, 2005): Ex biomass = β (n biomass (LHV biomass + w h fg ) (15) LHV biomass = biomass HHV h fg (21) According to Dulong s formula (Cho et al., 1995), the higher heating value of the biomass is a function of the dry-biomass weight percentages of carbon, oxygen, hydrogen, and sulphur, as shown in Eq. (22). HHV biomass = C (H O/8) (22) The exergy efficiency is defined as the ratio of the ratio of desired exergy outputs to exergy inputs expended to generate these outputs. As exergy efficiency is defined different each component due to having different working principles. For pump: η ex,pump = 8 ex 7 ex (23) w in For turbine: η ex,tur = w out 4 ex 5 ex For heat exchanger: (24) β = [ (H C ) (O C ) ( (H C ))] ( ( O C )) (16) η ex,hx = n ORC,fluid (ex 4 3) ex n bio,fluid (ex 1 2) ex (25) where n biomass is the molar flow rate of the biomass, w is the percentage of the moisture in the biomass, h fg is the molar specific enthalpy of vaporization of water and biomass LHV is the molar lower heating value of the biomass. C, H, O, and S denote the drybiomass weight percentages of carbon, oxygen, hydrogen, and sulphur. For regenerator: η ex,reg = (ex 3 8) ex (ex 5 6) ex (26) For the ORC and entire system exergy efficiencies can be found using Eqs. (27) and (28), respectively. III.3. Performance Assessment Parameters As a result of the energy analysis of the integrated system, the heat input to the burner and ORC, the net power output of the system, the heat transferred to ORC and the electrical efficiency of the ORC and the entire system can be found using Eqs. (17), (18), (19), (20) respectively. η ex,system = η ex,orc = W net Ex biomass W net n ORC,fluid (ex 1 2) ex IV. Results and discussion IV.1. Validation (27) (28) Q burner = n biomass LHV biomass (17) W net = W turbine W pump (18) Q ORC,in = n ORC,fluid (h 4 h 3) = n bio,fluid (h 1 h 2)(19) η el,orc = W net Q ORC,in (20) where n ORC,fluid and n bio,fluid denote working fluid circulating throughout the ORC and the fluid providing the heat transfer from biomass side to ORC. A computer code for the modeling equations presented in Section 3 is developed using the Engineering Equation Solver (EES) software. As a case study, the code developed is run considering the experimental data of a lab-scale ORC unit given in a study found in the literature (Gusev et al., 2014) as shown in Table 1. The results of the code are compared with those of the experiment for validating the model. The results of this comparison are given in Table 2. As can be seen from this table, the deviation between the numerical and experimental studies is less than 5%. The lower heating value of the biomass can be calculated knowing the higher heating value of the biomass, as shown in Eq. (21). 121

136 Table 1: Data taken from the experimental lab-scale ORC unit Parameter Value Type of working fluid in the ORC R245fa Type of heat transfer fluid Therminol 66 Mass flow rate of the working fluid in the ORC 0.3 kg/s (m ORC) Regenerator inlet pressure (P 5 ) Turbine inlet temperature (T 4 ) Heat exchanger pressure at the ORC side (P 3 ) Outlet temperature of the heat transfer fluid (T 2 ) Inlet temperature of the cooling water (T 9 ) Temperature difference between the inlet and outlet of cooling water stream (T 10 T 9 ) Heat transferred to the ORC side 171 kpa K 931 kpa 383 K K 3.7 K kw Table 2: Comparison between the experimental and numerical results Experimental Value Numerical Value Deviation Rate Turbine output 4.7 kw 4.9 kw %4 power (W T) Heat exchanger inlet K K %1 temperature at the ORC side (T 3 ) Regenerator outlet K K %0.4 temperature (T 6 ) IV.2. Parametric Studies After validation of the model, effects of some of the important input parameters on the performance of the system are investigated. These parameters include the turbine inlet temperature, excess air ratio, mass flow rate of the dry biomass, and the biomass types. In these studies, the baseline conditions used are given in Table 3. Using the data given in Table 3, the results that give all the thermodynamic properties of each state within the system are given in Table 4. Table 3: Baseline conditions used in the parametric studies Parameter Value Burner Type of fuel Wood Chemical composition of fuel CH 1.44 O 0.66 Mass flow rate of dry biomass kg/s Temperature of exhaust gases 400 K Excess air coefficient 0.2 Heat Transfer Fluid Type of heat transfer fluid Therminol VP-1 Mass flow rate of the heat transfer fluid 0.5 kg/s Pressure of the heat transfer fluid 780 kpa Temperature of the heat transfer fluid 673 K entering the heat exchanger ORC Type of working fluid in the ORC R134a Mass flow rate of the working fluid in the 0.3 kg/s ORC (m ORC) Pressure of working fluid entering the heat 3300 kpa exchanger (P 3 ) Condenser pressure (P 5 ) 700 kpa Turbine inlet temperature (T 4 ) 473 K Temperature difference between the inlet 10 K and outlet of cooling water stream (T 10 T 9 ) Isentropic efficiency of the turbine 0.68 Isentropic efficiency of the pump 0.8 Ambient temperature 298 K Ambient pressure 100 kpa 122 Table 4: Thermodynamic properties of each state of the integrated ORC system State Substance T (K) P h s n ex (kpa) (kj/kmol) (kj/ kmolk) (kmol/s) (kj/ kmol) 1 Therminol VP Therminol VP R 134a R 134a R 134a R 134a R 134a R 134a Water Water IV.2.1. Variation of Turbine Inlet Temperature Turbine inlet temperature (TIT) is one of the key operating parameters that affects the performance of the integrated system. The effect of the TIT on the electrical and exergy efficiencies are investigated and the results are shown in Fig. 2a. As shown in this figure, as the TIT increases, the net power output increases. As a result of this increase, the electrical and exergy efficiencies increase. Please note that as the heat transferred to the ORC is not a function of the TIT for this study as shown in Figure 2b, the trend of the electrical and exergetic efficiencies mainly depends on that of the net power output. To understand the reason of the trend of the change of exergy efficiency of the overall system shown in Fig. 2a, the exergy destrucion rates and effiencies of each component of the system are calculated. In this way, the components that have the highest exergy destruction rate (i.e. irreversibility rate) and lowest exergetic efficiency are found. Hence, the components that have more potential for improvement of the performance of the overall system are identified. Figures 3 and 4 show that the main reason of the increase in the exergy efficiency with the increase in TIT is the comperatively higher increase in the exergy efficiency of the heat exchanger or decrease in the exergy destruction rate. It can be seen from these figures that for this component, as the temperature increases from 377 to 470 K, exergy destruction rate decreases by 25%. This decrease can be atrributed to the increase in the change of the molar specific exergy change of the working fluid between the inlet and outlet the heat exchanger. Exergy destruction rate decrease in the condenser is the second significant reason for the increase in the exergetic performance of the overall system. The exergy destruction rate in this component mainly decreases because of the decrease in the heat transfer rate from the working fluid to the cooling water (i.e. enthalpy of the state 6 decreases while the enthalpy of state 7 does not change with turbine inlet temperature). On the other hand, these figures show that the exergy destruction rate of the regenerator and the turbine increases from kw to kw

137 and 2.97 kw to 3.5 kw, respectively, in this temperature range. Although there is an increase in the exergy destruction rate in these components, the exergy efficiency of the overall system increases as the total exergy destruction rate decrease is more than total exergy destruction rate increase. Exergy destruction rate and exergy efficiency of the pump and burner do not change because of the fact that inlet and outlet conditions of their control volumes do not change with respect to the turbine inlet temperature in this study; hence these components do not have an effect on the exergetic performance of the integrated system. Fig. 3: Change of exergy efficiency with the turbine inlet temperature Fig. 2: Change of (a) efficiency and (b) energy transfer with the turbine inlet temperature 123 Fig. 4: Change of exergy destruction rate of (a) heat exchanger and burner, and (b) pump, turbine, condenser, and regenerator with the turbine inlet temperature

138 IV.2.2. Variation of Excess Air Ratio A combustion process is complete if all the carbon, hydrogen, and sulfur (if any) in the fuel burns to CO 2, H 2 O, and SO 2, respectively. The minimum amount of air which allows the complete combustion of the fuel is called stoichiometric air. In this case, the products do not contain any oxygen. In practice, additional amount of air, which is called excess air, is fed to the burner. This excess air results in oxygen appearing in the products. Excess air also increases turbulence, which increases mixing in the combustion chamber. As there is more mixing of the air and fuel, these components have more chance to react. Hence, excess air helps to prevent the fuel from remaining unburned. On the other hand, supplying more than theoretical air provides safety. For ensuring complete safety, it is essential to control the levels of CO and check the amount of unburned hydrocarbon fuel. CO is a toxic gas that can be lethal in higher concentrations. Hydrocarbons which contains unburned fuel can cause explosions. The addition of excess air greatly lowers the formation of CO and unburned hydrocarbons by allowing them to react with O 2. High excess air ratio also reduces air pollution. As toxic compounds such as sulfur dioxide, carbon monoxide, nitrogen oxides can occur in high concentrations, smog, acid rain, and respiratory problems can occur (Basu et al., 2000). Excess air ratio is one of the important parameters affecting the performance of the system studied. In this study, the effects of this ratio on the energy and exergy efficiencies of the integrated system, and exergy efficiency and exergy destruction of each components are examined. Fig. 5a shows that the electrical and exergy efficiencies of the ORC increase with increase of the excess air ratio. This increase can be explained as follows. Heat transferred to ORC decreases in higher excess air ratios. As the excess air ratio increases, enthalpy of the state 2 increases whereas enthalpy of the state 1 does not change. When energy balance is applied to the control volume enclosing the heat exchanger, it is clearly seen that enthalpy of the state 3 (heat exchanger inlet of the ORC side) increases whereas enthalpy of the state 4 (turbine inlet) does not change. As the net power output does not change with the excess air ratio, both the electrical and exergy efficiencies of the ORC increase. The decreases in the exergy destructions in the heat exchanger, the regenerator and the condenser are responsible for the increase in the exergy efficiency of the ORC. The electrical and exergy efficiencies of the overall system also do not change as the heat occurred in burner does not change with excess air ratio. In addition, the increase in the exergy destruction rate of the burner is equal to the total decrease in the exergy destruction rates of other components; thus the exergy efficiency of the overall system does not change. On the other hand, excess air ratio has no effects on the exergy efficiency and exergy destruction rate of the turbine and pump. When the exergetic efficiencies of the components are observed, it can be seen that exergy efficiency of 124 the heat exchanger slightly increases as the decrease in the specific molar exergy difference between state 1 and 2 is greater than the decrease in the specific molar exergy difference between state 4 and 3. On the other hand, exergy efficiency of the regenerator increase significantly because the increase in the molar exergy difference between state 5 and 6 is less than the increase in the difference between state 3 and 8. Fig. 5: Change of efficiency (a) and (b) energy transfer with the turbine excess air ratio

139 Fig. 7: Change of exergy destruction rate of (a) burner, heat exchanger, and (b) pump, turbine, regenerator, condenser with the excess ratio IV.2.3. Variation of mass flow rate of dry biomass Fig. 6: Change of exergy efficiency of (a) burner, heat exchanger, and (b) pump, turbine, regenerator with the excess air ratio In this section, the effect of the mass flow rate of the dry biomass on the performance of the system is examined. An increase in the mass flow rate of dry biomass means increases in the heat gain from the burner and the heat transferred to ORC, which can also be interpreted from Eqs.17 and 19. These increases cause decreases of the electrical and exergy efficiencies because of the fact that the net output power of the cycle does not change with the change of mass flow rate of dry biomass. The trends of the changes of these efficiencies are shown in Fig. 8a. From the energy balance around a control volume enclosing the regenerator, it can be shown that an increase in the enthalpy of state 9 causes a decrease in the enthalpy of state 6. Enthalpy of the state 10 does not change with mass flow rate of dry biomass. Hence, heat transfer by the condenser increases with an increase in the mass flow rate of dry biomass. Exergy efficiency of the regenerator and heat exchanger denotes the ratio of exergy lost in the hot stream to the exergy gained in the cold stream as shown in Eq.26. The most considerable decrease of the exergy efficiency with an increase in the mass flow rate of biomass is in the regenerator. As the mass flow rate increases, exergy recovered from cold stream of the regenerator (stream that comes from pump) decreases more than its hot stream (stream that comes from turbine). Hence, the ratio becomes lower. Exergy efficiency of the heat exchanger slightly decreases with the increase of the mass flow rate because of the increase of the hot stream (stream that comes from burner) is more than the increase of the cold stream (stream that circulating in ORC). Exergy destruction rate of the regenerator increases with the increase of mass flow rate of dry biomass because of the fact that total molar specific exergy rates of the state 6 and 3 (outlet of the regenerator) increase with the increase of the mass flow rate of dry biomass while state 5 and 8 does not change. Exergy 125

140 destruction of heat exchanger also increases since the decrease of the specific molar exergy of the outlet conditions of the heat exchanger is more than decrease of the specific molar exergy of the intlet conditions of the heat exchanger. Hence exergy destruction rate of the heat exchanger decreases as shown in Eq. 9. Fig. 9: Change of exergy efficiency with the mass flow rate of dry biomass As shown in Fig. 10, decrease or increase of exergy destruction rate of the regenerator depends on mass flow rate. The increase in the specific flow exergy of turbine side outlet of the regenerator is less than the decrease in that of the pump side outlet of the regenerator until the mass flow rate value becomes 5 g/s. Hence exergy destruction increases. When the mass flow rate is higher than 5 g/s, the increase in the specific flow exergy of the turbine side outlet of the regenerator is more than the decrease in that of the pump side outlet of the regenerator. Thus exergy destruction increases. Fig. 8: Change of (a) efficiency and (b) energy transfer with the mass flow rate of dry biomass 126

141 burner. The more heat gained is, the more exergy destruction occured is (Fig. 14). Fig. 10: Change of exergy destruction rate of (a) heat exchanger, burner, and (b) pump, turbine, condenser, regenerator with the mass flow rate of dry biomass Fig. 12: Change of efficiency by the biomass type Variation of Biomass Fuels The lowest efficiency is gained when wheat straw is used as biomass fuel because of the fact that chemical structure of the wheat straw consists of the highest mass ratio of oxygen while paper s lowest. High oxygen mass ratio provides high higher heating value (HHV) and the high enthalpy of biomass. Hence the output heat increases as shown in Fig. 11. Fig. 13: Change of exergy efficiency by the biomass type Fig. 11: Change of energy transfer by the biomass type Power consumed and produced by the pump and the turbine does not change with the type of the biomass. As the net output power does not change, electrical efficiency of the ORC and system decrease when the biomass with high oxygen mass ratio is used as a fuel (Fig. 12). Accordingly, type of the biomass has no effects on the turbine and pump exergy efficiencies as shown in Fig. 13. Biomass with high oxygen mass ratio provides high amount of heat gained from Fig. 14: Change of exergy destruction rate by the biomass type V. Conclusions In this study effects of the change of various input parameters including turbine inlet temperature, mass flow rate of dry biomass, excess air ratio, and type of 127

142 biomass are examined to find out in what way and to what extent which parameter causes to change of the energy and exergy efficiency. Increase and decrease of the efficiencies are important as well as how this changes are occured. Because there are some important points that are related to the electrical and exergy efficiencies significantly. For instance, heat transfer to the ORC system, power output of the system, exergy destruction of the each components, etc. Hence this output parameters trends are examined with respect to change of input parameters. There are some significant conclusion of the study: High turbine inlet temperature provides high power output. Turbine inlet temperature has no effect on heat transferred to ORC or heat occurred in the burner. Hence electrical and exergy efficiencies of the ORC and entire system increase with the increasing turbine inlet inlet temperature. Variation of the excess air ratio does not affect the electrical and exergy efficiencies of the system because of the fact that change of the excess air ratio does not change the net output work and the lower heating value of the biomass. Higher mass flow rate of the dry biomass causes lower electrical and exergy efficiencies. The more biomass burns, the more heat occured. Increase of the heat generation causes the increase of the exergy destruction. Hence, both energy and exergy efficiency decreases in higher mass flow rate of the dry biomass. Type of the biomass has no effects to the power output of the turbine and the the power input of the pump. Because of the fact that turbine inlet temperature is taken input parameter. Hence exergy efficiencies of these components don t change with respect to the type of the biomass. Type of the biomass affects the amount of the heat generation and amount of the heat transfer to ORC system. Wheat straw provides high amount of heat gained from burner. In future studies, system can be modelled with more detailed conditions and thermal optimization can be applied. Furthermore, exergoeconomic analysis can be done. References Qiu G., Liu H., Riffat S., Expanders for micro-chp systems with organic Rankine cycle, Applied Thermal Engineering, 31, (2011). Quoilin S., Den Broek MV., Declaye S., Dewallef P., Lemort V., Techno-economic survey of Organic Rankine Cycle (ORC) systems, Renewable and Sustainable Energy Reviews, 22, (2013). Guo T., Wang HX., Zhang SJ., Selection of working fluids for a novel low-temperature geothermallypowered ORC based cogeneration system, Energy Conversation Management, 52, (2011). thermodynamic cycles and working fluids for the conversion of low-grade heat, Renewable and Sustainable Energy Reviews, 14, (2011). Maizza V., Maizza A., Working fluids in non-steady flows for waste energy recovery systems, Applied Thermal Engineering, 16, (1996). Marion M., Voicu I., Tiffonnet AL., Study and optimization of a solar subcritical organic Rankine cycle, Renewable Energy, 48, (2012). Saleh B., Koglbauer G., Wendland M., Fischer J., Working fluids for low-temperature organic Rankine cycles., Energy, 32, (2007). Lakew AA., Bolland O., Working fluids for lowtemperature heat source, Applied Thermal Engineering, 30, (2010). Al-Sulaiman FA., Exergy analysis of parabolic trough solar collectors integrated with combined steam and organic Rankine cycles, Energy Conversion Management, 77, (2014). Liu H., Shao Y., Li J., A biomass-fired micro-scale CHP system with organic Rankine cycle (ORC) - Thermodynamic modelling studies, Biomass and Bioenergy, 35, (2011). Qiu G., Shao Y., Li J., Liu H., Riffat SB., Experimental investigation of a biomass-fired ORC-based micro- CHP for domestic applications, Fuel, 96, (2012). Huang Y., Wang YD., Rezvani S., McIlveen-Wright DR., Anderson M, Mondol J., Zacharopolousa A., Hewitta NJ., A techno-economic assessment of biomass fuelled trigeneration system integrated with organic Rankine cycle, Applied Thermal Engineering, 53, (2013). Colpan CO., Exergy analysis of combined cycle cogeneration systems, Ms.C. Thesis, Middle East Technical University (2005). Szargut J., Exergy Method: Technical and Ecological Applications, WIT Press (2005). Cho KW., Park HS., Kim KH., Lee YK., Lee KH., Estimation of the heating value of oily mill sludges from steel plant, Fuel, 74, (1995). Gusev S., Ziviani D., Bell I., De Paepe M., Den Broek MV., Experimental comparison of working fluids for organic Rankine cycle with single-screw expander, 15 th International Refrigeration and Air Conditioning Conference, Purdue (2014). Basu, P., Cen, K.F., Jestin L., Boilers and burners, Springer, New York (2000) Chen H., Goswami DY., Stefanakos EK., A review of 128

143 Transient Analysis of an Absorption Solar Refrigerator with External and Internal Irreversibilities Yasmina Boukhchana 1*, Ali Fellah 2, Ammar Ben Brahim 3 1 Research Unit of Applied Thermodynamics, Department of Chemical and Processes Engineering, National School of Engineers of Gabes, University of Gabes, St Omar Ibn El-Khattab, 6029 Gabes, Tunisia Affiliation 2 Research Unit of Applied Thermodynamics, Technology Department, High Institute of Applied Sciences and Technology University of Gabes, 6029 Gabes, Tunisia 3 Research Unit of Applied Thermodynamics, Department of Chemical and Processes Engineering, National School of Engineers of Gabes, University of Gabes, St Omar Ibn El-Khattab, 6029 Gabes, Tunisia * Yasmina.Boukhchana@enig.rnu.tn Abstract The transient analysis of a solar absorption refrigeration cycle with external and internal irreversibilities is presented in this paper. The model consists of a flat plate solar collector, a refrigerator with three finite-size heat exchangers, namely, the evaporator between the refrigeration load and refrigerant, the condenser between the refrigerant and the ambient, and the generator between the solar collector and the refrigerant, and finally the refrigerated space. The total thermal conductance of the three heat exchangers is fixed.an empirical function is used to model the internal entropy generation of the cycle. The parameters of this function are estimated by fitting data obtained by simulation to the predictions of the THR model. The model is based on the first and second laws of thermodynamics, heat transfer equations at finite thermal source and sink capacities and entropy generation terms in order to consider the internal and external irreversibilities of the cycle. A thermodynamic analysing and optimization of the absorption cycle is then performed, reporting the operating conditions for minimum time to reach a prescribed cold-space temperature, thus maximum refrigeration rate, specifically, the optimal temperature of hot space and the optimal way of allocating the thermal conductance inventory. The results are presented in normalized charts for general applications. The collector temperature presents major influence on the conceptual and functional characteristics compared to the stagnation temperature influence. On the other hand the thermal load in the refrigerated space and the thermal conductance of the walls has analogous effects, therefore important to be considered in actual design. As a result, the model is expected to be a useful tool for simulation, design, and optimization of solar collector based energy systems. Keywords: Solar energy, Refrigeration, Absorption, irreversibilities, Optimization, Transient regime. I. Introduction Absorption refrigeration systems that could be used with solar energy or other sources of thermal energy such as waste heat are being developed for application in air-conditioning systems. The performance of absorption systems were studied expensively by detailed computer simulation (M. O. MC Linden and S. A. Klein, 1985; G. Grossman and al. 1987; K. Gommed and G. Grossman, 1990). The development of such computer codes require considerable effort and they also need as input the thermophysical properties of the working fluids. For preliminary design studies and for performance data representation it is useful to develop simplified models for absorption cooling systems. Such models can be used to represent performance characteristics of absorption machines when they form sub-components of a larger thermal system simulation programme. Several idealized models were developed recently using the three-heat-reservoir (THR) configuration of the absorption cycle (J. Chen and Z. Yan, 1989; 1989 bis; N. E. Wijeysundera, 1996). These models which take into account the external heat transfer irreversibilities of the cycle are able to provide realistic performance limits for the coefficient of performance (COP) and the cooling capacity of absorption refrigeration systems (N. E. Wijeysundera, 1996). However, if the THR models are to predict the performance of real absorption machines closely, the internal irrevesibilities of the cycle in addition to the external irrevesibilities have to be included in the analysis. Such models were used to obtain the optimum performance of commercial absorption chillers (J. M. Gordon and K. C. Ng, 1995; H. Tong Chua and al. 1996). Also by using a few fitting parameters, these models were able to reproduce performance data for absorption chillers (J. M. Gordon and K. C. Ng, 1995). Nevertheless, all those studies focus on the systems steadystate properties and ignore completely their dynamic behavior. Steady-state models are useful under many conditions although under strongly dynamic conditions that are often seen in real-life operation, these models can become unacceptably inaccurate 129

144 (Browne MW, Bansal PK. 2002). However, steady state models do not provide time dependent information on the thermal behavior of absorption refrigerators and are therefore not suitable for transient system simulations. In contrast, the model presented in this work allows the simulation of the dynamic absorption refrigerator behavior. It extends the range of applicable models for transient system simulations, where the time constants of the refrigerator significantly influence the system performance. The dynamic model of an irreversible absorption refrigerator allows the simulations of its transient behavior for changing input conditions or design parameters. This is important because absorption refrigerators usually have a high thermal mass, consisting of their internal heat exchangers, the absorbing solution and the externally supplied heat transfer media. The contribution of this work is the analysis of the transient irreversible three heat reservoir absorption heat transformers with Newton s heat transfer law. Thus, a transient mathematical model for a solar collector driven refrigeration plant is introduced. Finding an optimum heat transfer rate received from the solar collector to the generator and investigating the effect of time in solar collector stagnation temperature and collector temperature and heat rate are derived by minimizing the time required to reach a certain operation temperature in the refrigerated space. This issue becomes more important in large scale cooling applications in which the thermal inertia of the refrigerated space becomes very large. II. The transient model The main features of the absorption refrigerator-refrigerated space model are shown in Figure 1. The cycle has negligible work input. The cycle is driven by the heat transfer rate QH received from the source temperature TH, which is determined by the operation temperature of the generator. The refrigeration load QL is removed from the refrigerated space, at TL, and the heat transfer rate Q0 is rejected to the ambient, T0. The refrigerator shown in Figure 1 operates irreversibly due to the entropy-generation mechanisms that are present (for example, heat transfer, mixing, and throttling). The irreversible model takes into account the internal and external irreversibilities, which are fundamental features that will be present in the design of real absorption refrigerators. The instantaneous heat transfer interactions are given by Q UA T T (1) H H H HC Q UA T T (2) L L L LC Q UA T T (3) 0 0 0C 0 Q H Additionally, is proportional to the collector efficiency, where, without loss of generality, and negligible heat loss between the solar collector and the generator, as follows: Fig. 1: Problem sketch QH SCASCG (4) where AS.C represents the collector area, G is the irradiance at the collector surface. The efficiency of a flat plate collector can be calculated as: (Bejan and all., 1995; Sokolov and Hershgal, 1993) SC ab T H T 0 (5) where a and b are two constants that can be calculated, as discussed by Sokolov and Hershgal (Bejan and all., 1995; Sokolov and Hershgal, 1993bis). Eq. (5) can be rewritten by introducing the collector stagnation temperature Tst as follows: b T T (6) SC St H where Tst (for which SC 0 ) is given by: T T a b (7) St 0 The equation for heat input QH can be rewritten by combining Eqs. (4) and (6) as follows: Q A Gb T T (8) H SC st H The first and second law read: Q Q Q (9) H L Generator Condenser + Absorber Evaporator 0 0 ds Q0 Q Q dt T T T in H L 0C HC LC T HC T 0C T LC Q 0 Q L T 0 T L Q H T H Irreversible refrigerator (10) 130

145 We account for the transient cooling of the refrigerated space by writing the first law, dtl M C. UA T0 T Q Q W dt air v air L load L where W 0 L (11) UA T T accounts for the rate of heat gain from the ambient through the walls of the refrigerated space and Qload is the thermal load or rate of heat generated inside the refrigerated space. By writing the set of Eqs. (1) (10) for the absorption refrigerator and (11) for the refrigerated room, we take into account the fact that the thermal inertia of the refrigerated space is large enough such that the transient operation of the refrigerator can be neglected when compared to the time evolution of the temperature inside the refrigerated space. Generally, it is difficult to model all internal entropy generation sources in order to get an analytical variation law. We have chosen to consider the following approaches (Wijeysundera, 1997; Gordon and Ng, 2000). The entropy of the working fluid is represented by using linear variation law with temperature: ds dt in T T T T 1 HC 0C 2 0C LC, (12) where the parameters are to be estimated by fitting detailed simulation data to predictions. 1 2 To obtain the best estimates of the parameters 2 and from simulated performance data (Boukhchana and all., 2014, 2015) the following least-square procedure is used. According to the cycle model mentioned above, the rate of entropy generated by the cycle is described quantitatively by the second law as: ds Q0 Q Q dt T T T Tot H L 0 H L (13) The factors (UA)H, (UA)L, and (UA)0 represent the overall thermal conductances of the heat exchangers. The overall thermal conductance of the walls of the refrigerated room is given by (UA)W. The proposition here is to use the model to optimize the distribution of finite resources and the generator heat input QH, aiming to achieve maximum refrigeration rate, QL, in the transient regime. For that, since (UA)H, (UA)L, and (UA)0 are commodities in short supply, it makes sense to recognize the total external conductance inventory, UA (hardware), as a constraint: UA UA UA UA (14) H L In addition, we define the ratio w, which accounts for the size of the heat transfer area of the refrigerated room, as follows: UA w UA w The nondimensional version is (15) Q y (16) H H HC Q z (17) L L LC 1 1 Q y z (18) 0 0C Q B (19) H st H Q Q Q (20) H L 0 0 ds Q0 Q Q d in H L 0C HC LC dsin d ds Q Q Q0 d tot H L HC C C LC d L w 1 Q Q d H L L load L (21) (22) (23) (24) where we have appropriately defined the following nondimensional groups: TH T H, L T T L 0 0 T T T, 0 HC LC 0C T T T HC LC C QH QL Q0 Qload QH, QL, Q0, Qload UAT UAT UAT UAT tua A Gb S, B SC, S M C UA M C air v, air air v, air The conductance allocation ratios are y UA UA UA H, z UA L (25) (26) (27) (28) (29) We are interested in how the imperfect features (finite temperature differences) identified in the model influence the overall performance of the refrigeration plant.

146 III. Numerical method and results set point temperature (, L set = 0.97). The problem consists of integrating Eqs. (23) and (24) in time and solving the non-linear system (16) (22) at each step time. The objective is to minimize the time θset to reach a specified refrigerated space temperature, L, set, in transient operation. To generate the results shown in Fig. 2 8 some selected parameters were held constant and others were varied. The numerical method calculates the transient behavior of the system, starting from a set of initial conditions, then the solution is marched in time and checked for accuracy until a desired condition is achieved (temperature set points or steady state). The equations are integrated in time explicitly using an adaptive time step, 4th 5th order Runge Kutta method (Yang and all., 2005). Newton Raphson s method with appropriate initial guesses was employed for solving the above set of non-linear equations. During the integration of the ordinary differential equations, one time the set of fixed parameters H, st, B, y, z, w HC and Q load is defined Eqs (16) and (19) give. The system of Eqs (16) (22), at each time step of integration of Eqs (23) and (24), deliver and. Q, Q L, 0 0C LC To test the model and for conducing the analysis presented in this section, we assuming a small absorption refrigeration unit with a low total thermal conductance (UA = 400 W/K), we considered a total heat exchanger area A = 4m 2 and an average global heat transfer coefficient U = 0.1 kw/m 2 K in the heat exchangers and Uw = kw/m 2 K across the walls, which have a total surface area Aw = 54m 2, T0 = 25 C and Qload=0.8 kw. Considering a typical air conditioning application, the refrigerated space temperature to be achieved was established at TL,set = 16 C. Thus, the resulting dimensionless parameters that were kept fixed initially were: L, set =0.97. Q load =0.007, w=0.2, Fig. 2 shows that during the heat up period, the temperature of the evaporator starts to decrease linearly then it decreases very slowly. Here, the reaction of the evaporator is seen strongly affected by the generator behavior. His temperature starts rising linearly, then it becomes stable. As the temperature of the generator is higher causing more heat is absorbed in the evaporator. While, the temperature of the evaporator is decreasing very slowly the temperature of the generator still maintained quit constantly, indicating that the equilibrium state has reached (Abdullah and Hien, 2010). Also, there is an intermediate value of the collector size parameter B, between and 0.175, such that the temporal temperature gradient is maximum, minimizing the time to achieve prescribed L Fig. 2: The behavior refrigeration space temperature, in time ( H = 1.3, L st = 1.6) B=0.07 B=0.04 B=0.1 Fig. 3 and 4 show the behavior of θset versus B, while varying y and ΓH. The results stress the importance of identifying Bopt, mainly for lower values of ΓH. For ΓH = 1.3, there is a narrower range of values for B where the system operates in optimal conditions, outside of which the performance deteriorates dramatically. This effect is reduced as ΓH increases, as is demonstrated with the results for ΓH = 1.4. The existence of an optimum with respect to the thermal energy input is not due to the irreversible equations that model the system alone. However, an optimal thermal energy input results when the irreversible equations are constrained by the recognized total external conductance inventory, UA (hardware), Eq. (14), which is finite, and the operating temperature of the generator, ΓH. These constraints are the physical reasons for the existence of the optimum point. During the transient operation, to reach the desired ΓL,set = 0.97, there is an internal and a total entropy generated by the cycle, which is obtained by integrating Eqs. (21-23) in time. set y=0.2, z=0.3 y=0.3, z=0.2 y=z= B Fig. 3: Time to reach a refrigerated-space temperature setpoint for different thermal conductance allocations for H = 1.3 and st =

147 10 9 Bopt is simply the optimal collector size for which in the presence of a finite UA θset is minimum, which represents neither maximum efficiency nor minimum total entropy generated by the cycle. 8 set x y=0.3, z= 0.2 y=z=0.25 y=0.2, z= B Fig. 4: Time to reach a refrigerated-space temperature setpoint for different thermal conductance allocations for H = 1.4 and st = 1.6. Fig. 5 and 6 show the internal and total entropy generated by the cycle up to θset, versus B, while varying y and ΓH. We see that there are a minimum for internal and total entropy generated by the cycle for a certain dimensionless collector size parameter B. Note that Bopt, identified for minimum time to reach ΓL,set, does not coincide with the Bopt where minimum internal and total entropy occurs, although the values are close. (a ) S in S tot y=0.2, z=0.3 y=0.3, z=0.2 y=z= B (a) S in x 10-4 y=0.3, z= y=0.2, z=0.3 y=z= B B (b ) Fig. 6: Internal and total entropy generated during the time to reach a refrigerated-space temperature setpoint for different thermal conductance allocations and for H = 1.4, st = 1.6. According to our initial proposition, we seek the set of optimal values (Bopt, yopt) that minimize θ to reach ΓL,set, thus maximizing in the transient regimes. Figures 7 and 8 illustrate the behavior of θset,min and Bopt(θset,min) versus y, while varying ΓH, therefore identifying the set ( Bopt, yopt) which corresponds to our original set of fixed parameters, w, and L, set 0.02 y=0.2, z=0.3 y=0.3, z=0.2 y=z= Q L Q load. The results show that the thermal conductance should be divided equally between the generator and evaporator for maximum Q (y = 0.25). L S tot y=0.3, z=0.2 y=0.2, z=0.3 y=z= B (b) Fig. 5: Internal and total entropy generated during the time to reach a refrigerated-space temperature setpoint for different thermal conductance allocations and for H = 1.3, st =

148 Fig. 7: Minimum time to reach a refrigerated-space temperature setpoint for different coupling temperatures, with respect to the variation of the thermal conductance allocation. B opt set-min y H =1.3 H =1.4 H =1.3 H = y Fig. 8: Optimal collector size to reach a refrigerated-space temperature setpoint for different coupling temperatures, with respect to the variation of the thermal conductance allocation. V. Conclusions In this article, a transient irreversible model to study the absorption refrigeration cycle was presented and used to demonstrate the existence of an optimal way of allocating the thermal conductance inventory and an optimal collector size for maximum refrigeration rate. The model accounts for the internal and external irreversibilities. This means that these optima are fundamental features that will be present (and deserve to be identified and exploited) in the design of actual absorption refrigerators, no matter how complicated these designs may be. Appropriate dimensionless groups were identified and the generalized results reported in charts using dimensionless variables. The importance of the analysis of the absorption refrigeration system in the transient regime is this stressed. The most important conclusion is that 1. The maximum refrigeration rate, for minimum time to achieve a specified refrigeration load temperature, requires a narrow range of collector size parameter, mainly for low coupling temperatures. 2. The Optimal collector size and minimum time to reach a specified refrigerated-space temperature are influenced analogously by the thermal conductance of the walls. 3. In general, half of the total supply of thermal conductance has to be divided equally between the generator and evaporator, for maximum refrigeration rate. 4. Optimal size collector identified for minimum time to reach set point temperature in the refrigerated space does not coincide with Bopt where minimum total entropy occurs. Nomenclature A : Area, (m 2 ) a, b : Constant in Eq.(5) B : Dimensionless collector size parameter C : Specific heat, (kj/kg K) G : Irradiance on collector surface, (W/m 2 ) M : Mass of air in the refrigerated space, (kg) Q : Heat transfer rate, (W) S : Entropy generation rate, (kj/k) t : Time, (s) T : Temperature, (K) U : Global heat transfer coefficient, (W/m 2 K) W, y, z : Conductance fraction Greek letters Γ : Dimensionless temperature θ : Dimensionless time : Efficiency of a flat plate collector Superscripts 0 : Ambient air : Air H : Heat source L : Refrigerated space load : Cold space thermal load opt : Optimum SC : Solar collector Set : Setpoint St : Collector stagnation temperature References Abdullah M.O., Hien T.C., Comparative analysis of performance and technoeconomics for a H2O NH3 H2 absorption refrigerator driven by different energy sources. Applied Energy, 87, , (2010). Bejan A., Vargas J.V.C. and Sokolov M., Optimal Allocation of a Heat Exchanger Inventory in Heat Driven Refrigerators, International Journal of Heat and Mass Transfer, 38, , (1995). Boukhchana Y., Fellah A. and Ben Brahim A., Numerical Study of Entropy Generation in an Irreversible Solar-Powered Absorption Cooling Systems, 9 ème Congrès Francophone de Génie des Procédés, Agadir, Maroc, April 28-30, (2014). 134

149 Boukhchana Y., Fellah A. and Ben Brahim A., Transient modeling and simulation of an ammonia-water absorption solar refrigerator, International Journal of Mechanics and Energy, 3(1), 33-43, (2015). Browne MW, Bansal PK. Transient simulation of vapour-compression packaged liquid chillers. Int J Refrige, 25, , (2002). Engineering, 17, 12, , (1997). Yang W.Y., Cao W., Chung T.S., Morris J., Applied numerical methods using MATLAB, Wiley-Interscience, A John Wiley & Sons, Inc.; (2005). Chen J. and Yan Z., Equivalent combined systems for three-heat-source heat pumps. J. Chem. Phys. 90(9), (1989). Chen J. and Yan Z., An optimal endoreversible three-heat-source refrigerator. J. Appl. Phys. 65(l), l-4 (1989 bis). Gommed K. and Grossman G., Performance analysis of staged absorption heat pumps: water-lithium bromide systems. ASHRAR Trans., (1990). Gordon J. M. and Ng K. C., A general thermodynamic model for absorption chillers: theory and experiment. Heat Recovery CHP 15(l), (1995). Gordon J.M., Ng K.C., Cool Thermodynamics, Cambridge Int. Science Publishers, Cambridge, (2000). Grossman G., Gommed K.and Gadoth D., A computer model for simulation of absorption systems in flexible and modular form. ASHRAE Trans. 93(2), (1987). Linden M.O. MC and Klein S. A., Steady state modeling of absorption heat pumps with a comparison to experiments. ASHRAE Trans. 2(B), (1985). Sokolov M., Hershgal D., Optimal coupling and feasibility of a solar powered year-round ejector air conditioner, Solar Energy, 50, 6, , (1993). Sokolov M., Hershgal D., Solar-powered compression-enhanced ejector air conditioner, Solar Energy, 51, , (1993bis). Tong Chua H., Han Q., Choon Ng K., and Gordon J. M., Thermodynamic modeling and experimental evidence for the optimization and maximum-efficiency operation of absorption chillers. ECOS 96, Efkiency, Cost, Optimizarion,Simulation and Environmental Aspects of Energy Systems, June 25-27, Stockholm, , (1996). Wijeysundera N. E., Performance limits of absorption cycles with external heat-transfer irreversibilities. Appl. Thermal Engng 16(2) (1996). Wijeysundera NE., Performance of three-heat-reservoir absorption cycles with external and internal irreversibilities, Applied Thermal 135

150 A Study on Adsorption Characteristics of Activated Carbon-R134a and Activated Carbon-R404a Pairs Muhsin Kilic*, Ersan Gonul Uludağ University, Engineering Faculty, Department of Mechanical Engineering, Bursa, TR16059, Turkey * mkilic@uludag.edu.tr Abstract As one of environmentally friendly refrigeration methods, solid adsorption refrigeration, which can be powered by low-grade renewable and waste heat resources, has tracked much interest over the world. The physical adsorption process occurs mainly within the pores and surface of the adsorbent. It required the knowledge of adsorption characteristics when the temperatures and pressures are varying. The objective of this study is to evaluate adsorption characteristics of R134a and R404a on activated carbon experimentally by a constant volume variable pressure method at different adsorption temperatures ranging from 293 to 333 K and for pressures up to about 5 MPa. These data are useful for the design of adsorption cooling and refrigeration systems and are unavailable in the literature. Two samples of commercially available activated carbon with widely varying surface areas were chosen. The shapes of the isotherms obtained from the experimental data were similar in all cases and comparable to those reported in the literature. Adsorption parameters were evaluated from the isotherms using the Dubinin-Astakhov (DA) equation. The concentration dependence of the isosteric enthalpies of adsorption is extracted from the data. Further, the enthalpy of adsorption data were extracted, and correlations are provided for the two specimens investigated. Keywords: Adsorbent, activated carbon, adsorption system, refrigerant, R134a, R404a. I. Introduction Optimizing energy, protecting environment and sustainable development are all the main themes of the contemporary world in the 21 st century. As one kind of environmentally friendly refrigeration method, the research and developments on the adsorption refrigeration systems have attracted more attention in recent decades (Wang et al., 2009, Wang et al., 2010). The main heat sources for adsorption machines are waste heat and solar energy. Physical adsorption working pairs are usually preferred when solar energy is the heat source (Anyanwu and Ezekwe, 2003, Solmuş et al., 2010). It provides an alternative to conventional vapor compressor refrigeration, because the former can be driven by low grade heat sources such as solar energy and industrial waste heat. In addition, they have minimal moving parts. In contrast to vapor absorption cycles, adsorption cycles dispense with the heat exchangers (Wang et al., 2010). The properties of adsorbent/adsorbate pairs as well as the operating conditions have significant effects on the system performance (Solmuş et al., 2014, Wang et al., 2009). The isosteric heat of adsorption is a specific combined property of an adsorbent/adsorbate combination. The equilibrium adsorption properties at several adsorbent temperatures and adsorption chamber adsorbate pressures were studied for a wide range of pairs (Chan et al., 1984, Wang et al., 2009, Solmuş et al., 2011, Solmuş et al., 2011, Saham et al., 2008, Saha et al., 2007, Saha et al., 2009, Wang and Wang, 1999). Meanwhile, on the refrigerant field considerable impetus already exists to use natural and/or ozone friendly refrigerants. If the need is to use refrigerants that result in system pressures above atmospheric pressures, that are also non-toxic and ozone friendly, the choice narrows down to partly halogenated hydro fluorocarbon refrigerants such as R-134a (tetrafluoroetan CF3CH2F) and R-404a (CHF2CF3 / CH3CF3 / CF3CH2F) which is near a zeotropic blend of HFC-125/HFC-143a/HFC-134a. Thus, R-134a and R-404a based adsorption refrigeration cycles provide a perfect match for the current aspirations and expectations from adsorption cooling systems. The design of these refrigeration systems requires data on isotherms and the heats of adsorption for indenting heating inventories (Riffat et al., 1997, Solmuş et al., 2014, Anyanwu and Ezekwe, 2003, Banker et al., 2004). For example, in an adsorbtion cooling system, when the adsorbate gas is adsorbed by a solid adsorbent in a thermal compressor, the heat of adsorption has to be removed using a heat sink. Similarly, when the adsorbate gas is desorbed at a higher temperature and pressure there is a need to add the heat of adsorption. Therefore, the variation of the heat of adsorption as a function of loading, which in turn depends on the pressure and temperature at which adsorption/desorption occurs, has to be considered. It is known that the dependence of isosteric heat of adsorption on the loading is a measure of energetically homogeneous nature of the adsorbent surface. Detailed literature review on adsorption working pairs for refrigeration is given by Wang et al.(2009). Isosteric heat of adsorption is traditionally expressed 136

151 as a function of concentration due to its dependence on temperature is relatively weaker (Chakraborty et al., 2006, Chan et al., 1984, Parakash et al., 2000, Saha et al., 2007). For adsorption of fluids below their thermodynamic critical point, its magnitude is larger than the heat of vaporization of the adsorbate, which has a strong temperature dependence (Saha et al., 2009, Chakraborty et al., 2006). As a result, the difference between the two is a property of relevance in the design of adsorption refrigeration systems. It is a matter of regret that adsorption data are unavailable from the manufacturers of adsorbents. The characteristics of a new adsorbent like a kind of activated carbon may show differently than the known ones. In order to design adsorption based cooling cycle it is inevitable to evaluate adsorption isotherms of the assorted adsorbent/adsorbate pair as well as the isosteric heat of adsorption. In the view of the above mentioned perspectives, the present study reports an experimental study to obtain isotherm data for the adsorption of R134a and R404a refrigerants on the two different type of commercially available activated carbon (AC) specimens. Adsorption isotherms of R134a and R404a on the activated carbon specimens were measured over a temperature range of C and pressures up to about 5 bar using constant volume variable pressure (CVVP) method. Moreover, the isosteric heats of adsorption are evaulated from the present experimental data. II. Experimental Facility II.1. Setup tanks were heated by using hot water circulation at 60 C during 6 h, while the vacuum process is still running. At the end of regeneration process, the test system is purged with helium gas and evacuated further to achieve low vacuum conditions. The evacuation and helium purging are continued several times to ensure that there is no residual gas left in the system. Based on the measurements, there is no measurable interaction between the inert gas and the adsorbent. After evacuation, the charging cell is pressurized with the assorted refrigerant and left until it reaches an equilibrium state. During charging, it is necessary to keep the charging pressure lower than the saturation pressure of the refrigerant to ensure no condensation is occurred. At this state the initial pressure and temperature in the charging cell are measured before adsorption. Once equilibrium is achieved, the needle valve between the charging and adsorption tank is opened. The pressure and temperature in the adsorption tank are recorded to calculate the uptake of the assorted refrigerant by ensuring thermal equilibrium present. This process was repeated for the each charging step until the high pressure reached. By the use of a specimen, each isotherm was measured at a constant temperature over a range of pressure from 0 to about 5 bar. For each specimen with the known initial dry mass, experiments were performed at constant temperatures chosen at the range of 20 to 60 ºC for pressures up to about 5 bar. Experimental study was performed by the use of commercially available two different type of activated carbon (AC) specimens. Physical characteristics of the adsorbents used in the tests are presented in Table 1. The constant volume variable pressure (CVVP) experimental setup comprises (i) a charging tank with a volume of 3000 cc, (ii) an adsorption tank with a volume of 3000 cc, (iii) temperatures of the both the charging and adsorption tanks are controlled independently by the separate circulating water systems, (iv) a pressure transducer with an uncertainty of 0.15% of full scale and a pressure ranging from 0 to 1.6 MPa, (v) Pt 100 type thermometers with an uncertainty of 0.2% for temperature measurement, (vi) separate sensors with an uncertainty of 0.2% used with the adsorbent species for direct temperature measurement, (vii) a vacuum pump that achieves vacuum level of 0.5 mbar, and (viii) a computer used to control the test system and record the data. The volume of both charging and adsorption tanks are inclusive of the volumes of related piping and valves. II.2. Measurements Prior to adsorption process, the specimen of the adsorbent is placed in an oven for 24 h to desorb any residual gas. The oven temperature is kept constant at 120 C. Before starting adsorption test the system was evacuated to take out any gases and moisture from the bed using a vacuum pump to 5 mbar. The 137 Tab. 1: Physical characteristics of the adsorbents used in the tests. Activated Carbon Type I: ACG Type II: ACP Size (mm) D=4 Density (kg/m 3 ) Micro Pore Volume(cm 3 /g) Specific Surface Area (m 2 /g) Pore Diameter (Å) Shape Granulated Cylindrical Pellets II.3. Assessment of overall uncertainty There are some uncertainties associated with instrumentation, average adsorption cell temperature during adsorption and the void correction. Moreover, certain errors introduced due to the mathematical calculations. It is expected that the overall uncertainty will be within 3%. III. Mathematical approach The starting point for this analysis is the use of Dubinin Astakhov (D A) model of adsorption isotherm in the following form (Saha et al., 2009, El-Sharkawy et al., 2006, Akkimaradi et al., 2001):

152 W = W 0 exp { [ RT E (p s p )]n } (1) with W = Cυ a and W 0 = C 0 υ 0 (2) Here E is the characteristic energy of the assorted adsorbent/adsorbate pair which can be evaluated experimentally. The parameter n is an exponential constant which gives the best fitting of the experimental isotherms. The quantity C denotes the specific mass of adsorption (kg of adsorbate per unit mass of adsorbent), and v a is the specific volume of the adsorbed phase, which is given by υ a = υ b exp ( Ω(T T b )) (3) where Ω = ln(b/υ b )/ (T c T b ) (4) The quantity b denotes the van der Waals volume, vb is the saturated liquid specific volume at the normal boiling point (Saha et al., 2009, El-Sharkawy et al., 2006, Akkimaradi et al., 2001). T is the temperature with suffixes c and b referring to critical and normal boiling points, respectively. The parameter v0 can be obtained by using Eq. (3) at T = 0. Table 2 shows the properties and parameters of the adsorbates used in the present experimental study. Tab. 2: Properties and parameters of the adsorbates. R-134a R-404a Molecular Weight (MW) Boiling Point at 1 atm (Tb) C C Critical Temperature (Tc) C C Critical Pressure (Pc) 4059 kpa 3729 kpa Critical Density kg/m kg/m 3 b vb (m 3 /kg) v0 (m 3 /kg) Ω Eq. (1) can be rewritten as follows: ln p = ln p s E/(RT)[ln(c 0 ν 0 cν a )] 1/n (5) Differentiating Eq. (5) with respect to 1/T for the isosteric conditions (i.e. C is constant). Noting that va is also a function of temperature, one can get the following equation: ln p (1/T) = ln p s (1/T) (E R ) [ln(c 0ν 0 Cν a )] 1 n (ETΩ (nr))[ln(c 0 ν 0 Cν a )] ((1 n))/n (6) Isosteric heat of adsorption is defined by the Clausius Clapeyron relation at constant concentration as: Q ads CC = R( ln p)/ (1/T) (7) and for the heat of vaporization defined as h fg = R( ln p s )/ (1/T) (8) Substituting Eq. 7 and 8 into Eq.6, the following equation for the heat of adsorption can be derived (El-Sharkawy et al., 2007). Q ads = h fg + (E)[ln(C 0 ν 0 Cν a )] 1/n +(ETΩ n)[ln(c 0 ν 0 Cν a )] ((1 n))/n (9) The standard procedure for evaluation of isosteric heat of adsorption as described by Eq.7, is to plot the isosters on ln p versus 1/T plane. Normally, a constancy of slope is observed at temperatures well over the critical point of the adsorbate. As a result the classical treatment of isosteric heat of adsorption being shown as a function of relative uptake is a good approximation for adsorbent adsorbate combinations which broadly follow the Dubinin s isotherms (Saha et al., 2009, El-Sharkawy et al., 2006, Akkimaradi et al., 2001). Due to non-ideality of the gas phase, during an adsorbate molecule uptake to the assorted adsorbent is affected by the pressure and temperature changes (Saha et al., 2009, El-Sharkawy et al., 2006, Akkimaradi et al., 2001, Lin et al., 1999, Saham et al., 2008). In order to consider the effect of pressure and temperature changes, heat of adsorption can be calculated by using the Eq.9. IV. Results and discussions The experimental data was used to evaluate the adsorption parameters for the granulated activated carbon (ACG)-R134a, pellet activated carbon (ACP)-R134a, granulated activated carbon (ACG)-R404a and pellet activated carbon (ACP)-R404a pairs. Derived objection function is optimized by the use of a homemade code based on a genetic algorithm. Table 3 shows that computed values of the adsorption parameters (W 0, C 0, E and n) for the ACG-134a, ACP-R134a, ACG-R404a and ACP-R404a pairs. Tab. 3: Computed adsorption parameters of the different adsorbent-adsorbate pairs. Pairs W0 C0 E n (m 3 /kg) (kg/kg) (kj/kmol) ACG-R134a 0.380x ACP-R134a 0.211x ACG-R404a 0.338x ACP-R404a 0.192x Comparison of the experimental and the computed isotherms of the adsorbent-adsorbate pairs at 30 C are given in Fig.1. It can be seen from the Fig.1. that the measured results and the computed data obtained from D-A equation (Eq. 1) with the parameters given in Table 2 are in a very good agreement. The shapes of the isotherms obtained from the experimental data were similar in all cases and comparable to those reported in the literature for commercially available different adsorbents (Saha et al., 2009, El-Sharkawy et al., 2006, Akkimaradi et al., 2001, Lin et al., 1999, Saham et al., 2008). Then, Eq.5 and Eq.9 with the parameters provided in Table 3 were used to evaluate isosters and isosteric heat of adsorption of the adsorbent-adsorbate pairs. Fig.2 138

153 shows the isosters of adsorption of R134a and R404a on activated carbon specimens for C/C0 = 0.7. It can be seen from the Fig.2 that the variation of ln(p) with 1/T presents lineer variation for each adsorbent-adsorbate pairs. It is interesting to see that the isosters of the adsorbates of R134a and R404a with activated carbon is almost overlap each other, while the slopes of the lines are almost same for the all adsorbent-adsorbate pairs. molecules first penetrate into narrow pores of adsorbent, resulting in a stronger interaction between adsorbate and adsorbent. This implies a higher value of isosteric heat of adsorption at lower loading. After completely filling the smaller pores, adsorbate molecules are gradually accommodated in larger pores, in which the adsorption affinity becomes weaker. Therefore a monotonic decrease in isosteric heat of adsorption as a function of adsorbate uptake. Fig. 1: Comparison of the experimental and the computed isotherms of R134a and R404A on adsorbents at 30 C. Fig. 3: Variation of isosteric heat of adsorption of R134a on granulated activated carbon with temperature. Fig. 2: Obtained isosters of adsorption of R134a and R404a on granulated and pellet activated carbon for C/C0 = 0.7. Fig. 4: Variation of isosteric heat of adsorption of R404a on granulated activated carbon with temperature.. The maximum value of adsorption capacity decreases with the increase of adsorbent temperature. Comparing relative uptakes of the two refrigerants of R134a and R404a, the uptake magnitude of R134a on the same adsorbent is greater than the one of R404a. Meanwhile, the granulated activated carbon (ACG) have higher uptake comparing to the pellet activated carbon (ACP) at the same conditions considered in this study. A careful inspection on the Figs. 3 to 6, it can be seen that the isosteric heat of adsorption decreases with increasing adsorbate uptake for all the cases. The isosteric heat of adsorption varied with the temperature and the maximum value of isosteric heat obtained at the lowest temperature. In addition, the temperature have more effect on isosteric heat, where the maximum value of isosteric heat obtained with low temperature at 0 C. The adsorbate 139 Fig. 5: Variation of isosteric heat of adsorption of R134a on pellet activated carbon with temperature. Comparing the magnitude of the isosteric adsorption heat, ACP-R134a pair has the greatest one among the pairs tested in this study. It is also observed that replacing R134a with R404a do not significant effect on the magnitude of the isosteric adsorption heat on

154 the same adsorbent as seen Figs. 3 and 4 for ACG, and also Figs. 5 and 6 for ACP. Acknowledgements The authors gratefully acknowledge that this study is supported by Scientific and Technology Research Council of Turkey (TÜBİTAK), under project number: 112M163, and this study is also supported by Scientific Research Projects Funds of Uludağ University, under project number: KUAP(MH)-2015/60. Fig. 6: Variation of isosteric heat of adsorption of R404a on pellet activated carbon with temperature. In addition to all, the magnitude of the heat of adsorption is greater than that of the enthalpy of vaporization of R-134a and R-404a in the all range of the experimental tests performed as seen from the Figs. 3 to 6. V. Conclusions In this study, the adsorption properties of R-134a and R-404a on commercially available two different type specimens of activated carbon for adsorption process has been experimentally studied. The experiments were conducted over a temperature range from 20 C to 60 C and pressure up to about 5 bars. The measured experimental data were used to obtain the parameters in the Dubinin-Astakhov (D-A) equation for corresponding adsorption process. Adsorption characteristics such as isotherms were evaluated from the D-A equation with the obtained specific parameters for the each adsorbate-adsorbent pairs. Further, the isosteric heat of adsorption were obtained, and correlation parameters were provided for the each adsorbate-adsorbent pairs investigated. Comparison between the correlated results and the experimental data shows very good agreement. It is observed that the adsorption capacity per kg of adsorbent increases rapidly with increasing relative pressure at the beginning of the adsorption process. The maximum value of adsorption capacity decreases with the increase of adsorbent temperature. The isosteric heat of adsorption varied with the temperature and the maximum value of isosteric heat obtained at the lowest temperature. Comparing two refrigerants of R134a and R404a, the uptake magnitudes of R134a on the same adsorbent is greater than the one of R404a. Meanwhile, the granulated activated carbon (ACG) have higher uptake comparing to the pellet activated carbon (ACP) at the same conditions. In addition, it is also observed that the magnitude of the heat of adsorption is greater than that of the enthalpy of vaporization of R134a and R404a in the range of experimental conditions studied. This aspect is important in the design of thermal compressors in which the coolant requirements for removing the enthalpy of adsorption have to be assessed. 140 Nomenclature ACG : Activated carbon granulated ACP : Activated carbon pellet b : The van der Waals volume C : Adsorption uptake per kg of adsorbent (kg adsorbate/kg adsorbent) C0 : Maximum adsorption uptake per kg of adsorbent (kg adsorbate/kg adsorbent) E : Characteristic energy of adsorption pair (kj/kmole) MW : Moleculer weight (kg/kmole) p : Pressure (kpa or bar) R : Universal gas constant (8.314 kj/kgk) Qads : Isosteric adsorption heat (kj/kg) T : Temperature (K or ºC) va : Adsorbed phase specific volume (kg/m 3 ) vb : Saturated liquid specific volume at the normal boiling temperature (kg/m 3 ) n : Exponential constant W : Adsorbed volume per unit mass of adsorbent (m 3 adsorbate/kg adsorbent) W0 : Maximum adsorbed volume per unit mass of adsorbent (m 3 adsorbate/kg adsorbent) Subscripts a : adsorbed phase b : boiling point c : critical point CC : constant concentration ads : adsorption fg : vaporization enthalpy 0 : maximum s : saturation References Akkimaradi, B.S., Prasad, M., Dutta, P., Srinivasan, K., Adsorption of 1,1,1,2-tetrafluoroethane on activated charcoal, J. Chem. Eng. Data, 46, (2001). Banker, N.D, Srinivasan, K., Prasad, M. Performance analysis of activated carbon + HFC 134a adsorption coolers, Carbon, 42, (2004). Chakraborty, A., Saha, B.B., Koyama, S., Ng, K.C., On the thermodynamic modeling of the isosteric heat of adsorption and comparison with experiments, Appl. Phys. Lett., 89, , (2006). Chan, C.K., Tward, E., Boudale, K.I., Adsorption isotherms and heat of adsorption of hydrogen, helium, neon and nitrogen on activated charcoal, Cryogenics,

155 24, (1984). El-Sharkawy, I.I., Kuwahara, K., Saha, B.B., Koyama, S., Ng, K.C., Experimental investigation of activated carbon fibers/ethanol pairs for adsorption cooling system application, Appl. Therm. Eng., 26, (2006). Reviews, 14, (2010). Wang R.Z., Wang B.Q., Adsorption mechanism and improvements of the adsorption equation for adsorption refrigeration pairs, International Journal of Energy Research, 23, (1999). Lin, S.H., Lin, R.C., Prediction and experimental verification of HFC-134a adsorption by activated carbons. J. Environ. Sci. Health, 34 (1), (1999). Prakash, M., Mattern, A., Prasad, M., Sant, R., Subramanya, P., Srinivasan, K., Adsorption parameters of activated charcoal from desorption studies, Carbon, 38 (8), (2000). Riffat, S.B., Williams, M.D., Corr, S., Adsorption heat pump using HFC refrigerants, International Journal of Energy Research, 21, (1997). Saha B.B., Habib K., El-Sharkawy I.I., Koyama S., Adsorption characteristics and heat of adsorption measurements of R-134a on activated carbon, International Journal of Refrigeration, 32, (2009). Saham B,B,, Chakraborty A,, Koyama S,, Yoon S,H, Mochida I,, Kumja M, Yap C., Ng K.C., Isotherms and thermodynamics for the adsorption of nbutane on pitch based activated carbon, International Journal Heat and Mass Transfer, 51, (2008). Saha, B.B., Koyama, S., El-Sharkawy, I.I., Habib, K., Srinivasan, K., Dutta, P., Evaluation of adsorption parameters and heats of adsorption through desorption measurements. J. Chem. Eng. Data 52 (6), (2007). Solmuş I., Kaftanoğlu B., Yamalı C., Baker D., Experimental investigation of a natural zeolite water adsorption cooling unit, Applied Energy, 88, (2011). Solmuş I., Yamalı C., Kaftanoğlu B., Baker D., Çağlar A., Adsorption properties of a natural zeolite water pair for use in adsorption cooling cycles, Applied Energy, 87, (2010). Solmuş, I., Yıldırım, C., Theoretical analysis of the performance of an adsorption cooling system for various working pairs, J. of Thermal Science and Technology, 34(2), (2014). Wang, L.W., Wang, R.Z., Oliveira, R.G., A review on adsorption working pairs for refrigeration, Renewable and Sustainable Energy Reviews, 13, (2009). Wang, D.C., Li, Y.H., Li, D., Xia, Y.Z., Zhang, J.P., A review on adsorption refrigeration technology and adsorption deterioration in physical adsorption systems, Renewable and Sustainable Energy 141

156 Performance Investigation of a Geothermal Powered Organic Rankine Cycle for Natural Working Fluids Mustafa Alptekin 1,2*, Onder Kizilkan 1, Ahmet Kabul 1, Resat Selbas 1 1 Suleyman Demirel University, Faculty of Technology, Department of Energy Systems Engineering, Isparta, 32260, Turkey 2 Hakkari University, Faculty of Engineering, Department of Mechanical Engineering, Hakkari, 30000, Turkey. * mustafaalptekin@sdu.edu.tr Abstract Importance of geothermal energy, which is one of renewable energy resources, has been rapidly increasing in our country as well as all over the world. As power generation can be done using different resources, electric generation can be performed from a power plant with organic Rankine cycle, which operates low temperature, from inactive geothermal resources in our country. In this study, energy and exergy analyses of a geothermal powered Organic Rankine Cycle (ORC) were conducted, and net work output, total exergy destruction rate, thermal and exergy efficiencies of overall system were calculated using original data of a geothermal power plant in Denizli province. There are two organic Rankine cycles in the system, and the analyses were performed for organic working fluids n-pentane, R245fa and R600a. it was investigated effect of condenser pressure and turbine inlet temperature on system performance. As the condenser pressure increases, the thermal and exergy efficiencies decreases, and the turbine inlet temperature is directly proportional with system performance. Among the working fluids examined, R600a and R245fa demonstrate the best exergetic performance. For R245fa, the total exergy destruction rate and exergy efficiency are calculated to be 3675 kw and 76%, respectively. Keywords: Geothermal energy, energy and exergy analysis, Organic Rankine Cycle (ORC), natural fluids. I. Introduction In the recent years, there has been a significant increase in the usage of renewable and low-grade waste heat. The geothermal energy is considered as an alternative source instead of fossil fuels since it is reliable and one of least-expensive renewable energy source. The utilization of this energy may be by using organic Rankine cycle (ORC), which converts this energy to useful power. ORC has some advantages such as environment-friendly, safety, system components are available and high flexibility (El-Emam and Dincer, 2013; Long et al., 2014). Solar radiation, biomass combustion, geothermal energy and industrial waste heat Kaska, 2014] can be used as required heat source of an ORC (Kaska, 2014). There are many studies about ORC in the literature. Heberle and Brüggemann (2010) performed the exergy analysis of a combined heat and power generation system for low grade geothermal resources (below about 180 C). They investigated the first and second law efficiencies of the system for different operating conditions, and found that isopentane and R227ea can be preferred in series and parallel circuits, respectively. Kanoglu, (2002) carried out exergy analysis of a dual-level binary geothermal power plant, which had a power output of 12.4 MW, using actual plant data of the power plant. As a result of, he determined that the condenser had the highest exergy destruction rate, and the thermal and exergy efficiencies were found to be 5.8% and 29.1%, respectively. Unverdi and Cerci (2013) investigated the performance of Germencik geothermal power plant. The working fluid was water and geothermal water had a source temperature of 205 C. They compared this system with the other geothermal power plants in the world, and the exergy efficiency of the overall system was calculated as 35.4%. El-Emam and Dincer (2013) performed exergy and exergoeconomic analyses of geothermal regenerative organic Rankine cycle with optimization. They conducted the effect of operating parameters on the system energetic and exergetic efficiencies and economic parameters, and found the energy and exergy efficiency values to be 16.37% and 48.8%, respectively, for a net out power of 5 MW. Thermodynamic and economic analysis for the pre-feasibility of a binary geothermal power plant were performed by Budisulistyo and Krumdieck (2015). They conducted thermodynamic and economic analyses for key cycle design options and component selection parameters, and used n-pentane, R134a and R245fa as working fluid. The profitability analysis was done for the top three options by them. They stated that a standard Rankine cycle with a 2-stage turbine using n-pentane is the most thermo-economical design for the particular brine resource and re-injection conditions. Coskun et al. (2011) proposed a modified exergoeconomic model for geothermal power plants using exergy and cost accounting analyses in a case study for Tuzla geothermal power plant system (Tuzla GPPS), which has a total installed capacity of 7.5 MW. They conducted the analysis using actual system data to assess energy and exergy efficiencies, exergy losses 142

157 and loss cost rates. Besides, they determined that exergy efficiency values vary between 35% and 49%, and studied to provide a more comprehensive evaluation of the system six new exergetic cost parameters. Dagdas et al. (2005) performed thermodynamic optimization of a power plant using actual data, and obtained some important results. They found the optimum flashing pressure of 200 kpa, and determined isobutene as a working fluid. In addition, they calculated to be 8.80% and 38.58% the first and second law efficiencies of the power plant, respectively. Akpinar and Hepbasli (2007) evaluated, constructed and tested with a comparison of exergetic (second law analysis) analysis of two Ground Source Heat Pumps (GSHPs) in Turkey based on the actual operational data by using entropy and exergy balance equations. They determined exergy (second law) efficiency values for both systems and exergy destructions in each of the system components to assess the improvement potential, and indicated that this method presented here can be applied to other GSHP systems worldwide as a useful tool. Kecebas and Gokgedik (2015) carried out both conventional and advanced exergy analyses of an existing geothermal binary power system. Therefore, in-depth information was collected about the irreversibilities in the system and its parts. They used to simulate the Bereket Geothermal Power Plant, in Denizli, the Engineering Equation Solver (EES) and GateCycle software packages. They found that condensers have the highest improvement potential for both conventional and advanced exergy analysis, and the modified exergy efficiency and the total system efficiency are to be 18.26% and 9.60%, respectively, in the real conditions. Yamankaradeniz (2016) performed thermodynamic performance assessment of a geothermal district heating system (GDHS) by using advanced exergy analysis to identify the interactions among system components and the potential for improvement. He applied new exergetic parameters to the Bursa GDHS in Turkey. He concluded that the advanced exergetic analysis is a more meaningful and effective tool than the conventional one. In addition, he found the exergy efficiencies are 25.24% and 26.34% for the conventional and advanced ones, respectively. Tan and Kecebas (2014) assested thermodynamic and economic evaluation of a geothermal district heating system (GDHS) using advanced exergy-based methods. They splited into endogenous/exogenous and unavoidable/avoidable parts the exergy destruction and the total operating cost within each component of the system by the help of the advanced exergetic and exergoeconomic analyses. They found that the exergetic efficiency and the exergoeconomic factor of the overall system for the Sarayköy GDHS is 43.72% and 5.25%, respectively, according to the conventional tools while these values are 45.06% and 12.98%, respectively, according to the advanced tools. Yildirim and Ozgener (2012) performed a review study about thermodynamics and exergoeconomic analysis of two geothermal power plants in Turkey. They investigated the effects of thermal fluids used in 143 power plants on energy and exergy efficiencies. They presented improvement suggestions, and conducted exergoeconomic analyses while power plants investment costs and equipment maintenance costs are taken into consideration. Liu et al. (2015) carried out parametric optimization and performance analyses of geothermal organic Rankine cycles using R600a/R601a mixtures as working fluids. They optimized evaporator and condenser pressures and cooling water temperature rise, and analyzed ORC power output, parasitic power consumption, heat exchanger areas and turbine sizes using R600a/R601a. They found that a geothermal ORC using R600a/R601a generates 4 11% more power than with pure R600a, and determined that the evaporator and condenser area per unit power output using R600a/R601a are higher than that using pure R600a or R601a. In this study, thermodynamic analysis of a geothermal powered ORC system is investigated using actual data for different natural working fluids. The working fluids are selected to be R245fa, and R600a while the system uses n-pentane as working fluid in actual case. Fot these there fluids, the thermal efficiency, exergy efficiency and exergy destructions of the ORC based power plant are found. Besides, the effects of turbine inlet temperature and condenser pressure on system performance are also parametrically analyzed. II. System Description In Fig. 1, schematic view of the geothermal powered ORC power generation system is presented. There are two different cycles in the system, and condenser is a water-cooled condenser. The same fluid operates in both power cycles and as mentioned earlier, analyses are performed for different working fluids which are R245fa, R600a and n-pentane. Fig.1: Schematic view of ORC system Firstly, geothermal water which has have a temperature of 145 C transfers to some of amount of its heat to the organic working fluid in evaporator at first and second cycles, respectively. Secondly, it splits in half, and transfers to rest amount of its heat to the organic working fluid in preheaters at the first and second cycles. Finally, it is pumped back to underground as waste water after it exits from preheaters. The organic working fluid enters to turbine as saturated vapor in both two cycles after it

158 gets heat of the geothermal water in the preheaters and evaporators. After the working fluid leaves the turbine as superheated vapor at condenser pressure and leaves the condenser as saturated liquid. The working fluid is circulated by pump, and enters the preheater as compressed liquid. For the thermodynamic assestment of the system, the some assumptions are made as follows: 1) All the processes are assumed as steady state. 2) The pumps and turbines are adiabatic. 3) The heat transfer to/from ambient and pressure losses in the piping system and in the preheaters, evaporators, condensers of ORC system are neglected 4) The working fluid at the inlet of ORC pump is assumed as saturated liquid. 5) Potential and kinetic energy variations are neglected. 6) The dead state pressure P0 and temperature T0 are considered to be kpa and 27 C, respectively. III. Thermodynamic Evaluation Energy and exergy analysis generally involves applying the first and second laws of thermodynamics and the principles of conservation of mass, while energy analysis usually excludes considerations of the second law of thermodynamics. Neglecting kinetic and potential energies, the conservation of mass for steady-state processes can be expressed as follows (Dincer and Rosen, 2007) m in = m out (1) The first law of thermodynamics is an express of energy principle. It is expressed for steady state processes as follows (Cengel and Boles, 2007; Bejan, 1997); Q + (m h) in = W + (m h) out (2) Neglecting chemical, kinetic and potential exergies, exergy balance in a control volume in which a steady state process occurs can be written as (Dincer and Rosen, 2007; Akpinar and Hepbasli (2007)); E x Q E x W = (m ε) out (m ε) in + T 0 S gen (3) where; E x dest = T 0 S gen (4) where E x Q, E x W and ε represent exergy of heat, exergy of work and thermomechanical exergy (flow exergy), respectively. These expressions are shown as follows Kaska (2014); E x Q = Q ( T T 0 T ) (5) E x W = W (6) 144 ε = (h h 0 ) T 0 (s s 0 ) (7) S in + S gen = S out (8) where 0 subscript expressed reference conditions. The first and second law efficiencies of the all system are calculated as follows (Dincer and Rosen, 2007); η I = W Q (9) η II = E x W E x Q (10) The above equations were applied to the geothermal powered ORC system, and were obtained mass, energy and exergy balance equations for each component. Tab. 1: Input values to the system Parameters Values Pump isentropic efficiency 85 % Turbine isentropic efficiency 85 % Condenser temperature 30 C Turbine inlet temperature 100 C Geothermal water inlet temperature 145 C Inlet temperature to condensers of cooling water 20 C Outlet temperature from condensers of cooling water 27 C Pressure of geothermal water 600 kpa Mass flow rate of working fluid in the first ORC 58.6 kg/s Mass flow rate of working fluid in the second ORC 58.4 kg/s Inlet mass flow rate of geothermal water kg/s Preheater I Capacity 8770 kw Preheater II Capacity 8555 kw Ambient pressure kpa Ambient temperature 25 C IV. Results and discussions Analyses were made using Engineering Equation Solver (EES) program. The calculated thermodynamic data of the ORC were determined for n-pentane, R245fa and R600a working fluids. Table 2 shows at each location of the ORC system thermodynamic properties such as temperature, enthalpy and entropy for n-pentane. The analyses were performed using actual data of geothermal power plant in Denizli. IV.1. Effect of condenser pressure (Pcon) The geothermal powered ORC system was thermodynamically modelled and conducted the first and second law of thermodynamics analyses. The Figures 2 to 5 illustrate the effect of condenser pressure variation on net work output, total exergy destruction rate, thermal and exergy efficiencies. The turbine inlet temperature and isentropic efficiencies of pumps and turbines were kept constant as 100 C and 85 %, respectively. Figures 2 and 3 show the net work output and thermal efficiency of overall system as function of condenser pressure for different organic working fluids. When the condenser pressure increases from 150 kpa to 450 kpa, both the net work output and thermal efficiency decrease for all working fluids. While R600a has the highest thermal efficiency,

159 n-pentane has the lowest efficiency, respectively. The maximum net work output and thermal efficiency was obtained to be about 10 MW and 21%, respectively. Net work output - kw Tab. 2: Data of the ORC system for n-pentane n-pentane R245fa R600a P con, kpa Fig. 2: Net power output as a function of condenser pressure for different organic working fluids Thermal efficiency (h) m P Ref. Substance T ( C) (kg/s) (kpa) h (kj/kg) s ε (kj/kg.k) (kj/kg) 1 n-pentane n-pentane n-pentane n-pentane n-pentane n-pentane n-pentane n-pentane n-pentane n-pentane n-pentane Water Water Water Water Water Water Water Water Water Water Water Water Water Water Water Water ,24 0,2 0,16 0,12 0,08 0,04 n-pentane R245fa R600a P con, kpa Fig. 3: Thermal efficiency as a function of condenser pressure for different organic working fluids The effect of condenser pressure on the exergy destruction rate and exergy efficiency of ORC system. The effect of condenser pressure on the exergy destruction rate and exergy efficiency of ORC system is illustrated in Figs. 4 and 5. While the total exergy 145 destruction rate increases with increasing of condenser pressure, the exergy efficiency decreases as the condenser pressure increases for all working fluids. It is observed that the exergy destruction rate decreases from 7629 kw to kw when the condenser pressure increases for n-pentane. Moreover, this fluid have the highest destruction rate. R600a have the largest exergy efficiency and the lowest total exergy destruction rate. Although increment in destruction rate for R600a is larger than that of R245fa, R600a have larger than R245fa exergy efficiency at high pressures. Total exergy destruction rate - kw Exergy efficiency n-pentane R245fa R600a P con, kpa Fig. 4: Exergy destruction rate as a function of condenser pressure for different organic working fluids 1 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 n-pentane R245fa R600a P con, kpa Fig. 5: Exergy efficiency as a function of condenser pressure for different organic working fluids IV.2. Effect of turbine inlet temperature (Tturb,in) The effect of turbine inlet temperature variation on net work output, total exergy destruction rate, thermal and exergy efficiencies are presented in Figs The Figures 6 and 7 illustrate variation of the net work output and the thermal efficiency as the turbine inlet temperature increases 80 C to 110 C. Both the net work output and efficiency increase with increasing turbine inlet temperature. The highest thermal efficiency is obtained for n-pentane while the lowest one is for R245fa. Furthermore, the thermal efficiency of n-pentane and R245fa increase from about 10% to 14% and from about 10% to 14%, respectively, while the turbine inlet temperature increases.

160 The effect of turbine inlet temperature on the total exergy destruction rate and exergy efficiency of ORC system is presented for different working fluids in Figs. 8 and 9. While the total exergy destruction rate decreases for all working fluid with increasing turbine inlet temperature, the exergy efficiency of overall system increases with increasing turbine inlet temperature. It is observed that the lowest exergy destruction rate occurs for the case of R245fa. Therefore, this fluid have the highest exergy efficiency. n-pentane and R600a are compared to each other that for n-pentane destruction rate is higher than that of R600a at low turbine inlet temperatures. However, it is noticed that exergy destruction in n-pentane is lower than that of R600a at high temperatures. Net work output - kw n-pentane R245fa R600a T turb in, C Fig. 6: Net power output as a function of turbine inlet temperature for different organic working fluids Thermal efficiency (h) 0,16 0,14 0,12 0,1 n-pentane R245fa R600a 0, T turb in, C Fig. 7: Thermal efficiency as a function of turbine inlet temperature for different organic working fluids Total exergy destruction rate - kw n-pentane R245fa R600a T turb in, C Fig. 8: Total exergy destruction rate as a function of turbine inlet temperature for different organic working fluids 146 Exergy efficiency Fig. 9: Exergy efficiency as a function of turbine inlet temperature for different organic working fluids V. Conclusions Geothermal energy powered an Organic Rankine Cycle is evaluated in terms of first and second laws of thermodynamics. Energy and exergy analysis of ORC system are carried out using actual power plant data. The main results of the study can be summarized as follows: The thermal and exergy efficiencies are directly proportional with turbine inlet temperature whereas they decrease with increasing condenser pressure. R600a and R245fa have the best performance for effect of condenser pressure and turbine inlet temperature, respectively. For condensation pressure of 150 kpa and turbine inlet temperature of 100 C, the thermal and exergy efficiencies of R600a are 20.7% and 91% while those of n-pentane are 9% and 52.7%, respectively. It is found that turbine inlet temperature has significant effect on both system performance. Higher thermal and exergy efficiencies are obtained by increasing turbine inlet temperature. Used working fluid n-pentane in actual plant, has the lower performance than R600a and R245fa for two parameters. Therefore, it is proposed that R600a or R245fa should be used. Nomenclature E x h m P ORC Q s S T W : Exergy (kw) : Specific enthalpy (kj/kg) : Mass flow rate (kg/s) : Pressure (kpa) : Organic Rankine Cycle : Heat load (kw) : Specific entropy (kj/kg.k) : Work (kw/k) : Temperature (C) : Work (kw) Greek letters ε : Specific exergy (kj/kg) η I : Thermal efficiency : Exergy efficiency η II 0,78 0,77 0,76 0,75 0,74 0,73 n-pentane R245fa R600a 0, T turb in, C

161 Subscripts con : Condenser dest : Destruction gen : Generation in : Inlet out : Outlet 0 : Referance state References Akpinar E.K. and Hepbasli A., A comparative study on exergetic assessment of two ground source (geothermal) heat pump systems for residential applications, Building and Environment, 42, (2007). Bejan A., Advanced Engineering Thermodynamics, Wiley, New York, (1997). Budisulistyo D. and Krumdieck S., Thermodynamic and economic analysis for the pre-feasibility study of a binary geothermal power plant, Energy Conversion and Management, 103, (2015). Coskun C., Oktay Z. and Dincer İ., Modified exergoeconomic modeling of geothermal power plants, Energy, 36, (2011). Long R., Bao Y.J., Huang X.M. and Liu W., Exergy analysis and working fluid selection of organic Rankine cycle for low grade waste heat recovery, Energy, 73, (2014). Tan M. and Kecebas A., Thermodynamic and economic evaluations of a geothermal district heating system using advanced exergy-based methods, Energy Conversion and Management, 77, , (2014). Unverdi M. and Cerci Y., Performance analysis of germencik geothermal power plant, Energy, 52, (2013). Yamankaradeniz N., Thermodynamic performance assessments of a district heating system with geothermal by using advanced exergy analysis, Renewable Energy, 85, (2016). Yildirim D. and Ozgener L., Thermodynamics and exergoeconomic analysis of geothermal power plants, Renewable and Sustainable Energy Rewiews, 16, (2012). Cengel Y.A. and Boles M.A., An Engineering Approach Thermodynamics, Fifth Edition, 946, (2007). Dagdas A., Ozturk R. and Bekdemir S., Thermodynamic evaluation of denizli kizildere geothermal power plant and its performance improvement, Energy Conversion and Management, 46, (2004). Dincer I. and Rosen M.A., Environment and sustainable development, Elsevier Science, 472, (2007). El-Emam R.S. and Dincer I., Exergy and exergoeconomic analyses and optimization of geothermal organic rankine cycle, Applied Thermal Engineering 59, (2013). Kanoglu M., Exergy analysis of a dual-level binary geothermal power plant, Geothermics, 31, (2002). Kaska O., Energy and exergy analysis of an organic Rankine for power generation from waste heat recovery in steel industry, Energy Conversion and Management, 77, (2014). Kecebas A. and Gokgedik H., Thermodynamic evaluation of a geothermal power plant for advanced exergy analysis, Energy, 1-10 (2015). Liu Q., Shen A. and Duan Y., Parametric optimization and performance analyses of geothermal organic rankine cycles using r600a/r601a mixtures as working fluid, Applied Energy, 148, (2015). 147

162 An Experimental Investigation on Exergy Analysis of an Ejector Expansion Refrigeration System Nagihan Bilir Sag 1*, Halil Kursad Ersoy 2, Arif Hepbasli 3 1,2 Department of Mechanical Engineering, Faculty of Engineering, Selcuk University, Alaeddin Campus, Konya, Turkey 3 Department of Energy Systems Engineering, Faculty of Engineering, Yasar University, Bornova, Izmir, Turkey * nbilir@selcuk.edu.tr Abstract Usage of ejector as an expander for expansion work recovery in the conventional refrigeration cycle was experimentally investigated. In order to identify the magnitudes and locations of irreversibilities within the components of the ejector expansion refrigeration cycle, an exergy analysis was employed. The exergetic performance of the ejector refrigeration system was compared with that of a conventional vapor compression refrigeration system of the same cooling capacity under the same external conditions. It was found that the ejector expander system exhibited a lower total irreversibility and a higher exergy efficiency in comparison with the conventional system for all operating temperatures. When the ejector was used as the expander in the refrigeration system, the total irreversibility was lower than in the conventional system by %, while the exergy efficiency values were % higher than in the conventional system. Keywords: Refrigeration, ejector, throttling loss, exergy efficiency, irreversibility I. Introduction Energy consumption is increasing day by day in the world. So, the efficient use of energy is very important for the future of humanity. There are a lot of researches on efficiency improvement of systems that consume energy. One of the goals of these researches is to increase the energy efficiency of conventional vapor compression refrigeration systems. Recently conducted studies to improve the performance of vapor compression refrigeration systems have concentrated on reducing the throttling losses that increase irreversibility in the expansion valve. For this purpose, using a simple and low-cost ejector with no moving parts instead of expansion valve is a topic recently investigated in order to find out whether the performance of the system would increase. The idea of using an ejector instead of expansion valve was, for the first time, introduced by Gay in 1931 (Pottker, 2012). Kornhauser (1990) determined that using ejector in R12 conventional refrigeration system improves the cooling coefficient of performance (COP) by 21%, theoretically. However, an improvement of only % could be achieved experimentally (Menegay and Kornhauser, 1996). In a study by Bilir and Ersoy (2009), the performance of R134a ejector refrigeration system according to the conventional system could be theoretically improved up to of 22.3%. Additionally, in this study, it was calculated that even under off-design conditions, when an ejector was used in a refrigeration system, the COP of the system exhibited a higher value than a conventional cycle. In another study done by Ersoy and Bilir (2010), the exergy efficiency of the ejector system, irreversibility analysis, and the effects of ejector components on the system performance were theoretically investigated. In this study, it was found that while the efficiency of the ejector components increased, the system performance increased and the area ratio of the ejector decreased. It was determined that when ejector was used as an expander in the transcritic CO2 refrigeration cycle, the total irreversibility of the system reduced by 39.1% (Ersoy and Bilir, 2012). Tas et al. (2015) made a detailed review study on the ejector refrigeration cycle. They stated that the important factors affecting the system performance were operating conditions and refrigerant type. In a theoretical study, where R134a was used as a refrigerant in a bus air condition system that incorporated an ejector and double evaporator, Unal and Yilmaz (2015) reported that the system performance had a 15% higher COP than the conventional double-evaporator refrigeration system. There are very few experimental studies on the use of ejector instead of expansion valve in the conventional vapor refrigeration system with R134a as a refrigerant. A study by Harrell and Kornhauser published in 1995 stated that the COP improvement of an experimentally tested R134a refrigeration system varied between 3.9% and 7.6% (Pottker, 2012). The reason why this improvement remained low was associated with the ejector design based on singlephase flow knowledge and lack of two-phase ejector flow data. Experimental studies on the ejector refrigeration system were conducted under the leadership of Wongwises (Disawas ve Wongwises, 2004; Wongwises ve Disawas, 2005; Chaiwongsa ve Wongwises, 2007; Chaiwongsa ve Wongwises, 2008). 148

163 In these studies, the evaporator was wet-type and ejector partially recirculated the refrigerant on the low pressure side. However, when using the ejector, the amount of the improvement of the cooling performance was not clearly stated in these studies. An ejector-expander refrigeration cycle with noseparator and dual evaporator was experimentally compared with a conventional dual-evaporator refrigeration cycle by Lawrence and Elbel (2013). It was also found that the ejector refrigeration system with dual evaporator exhibited an exergy efficiency that was 8.5% higher than that of the conventional dual evaporator refrigeration system. Yılmaz (2015) experimentally found that the coefficient of performance of the dual-evaporator refrigeration cycle with the ejector increased by 8% compared to that of the conventional dual-evaporator refrigeration cycle. Bilir Sag et al. (2015) experimentally investigated the coefficient of performance of the ejector and the conventional refrigeration systems for fresh food refrigerator (at a condenser temperature of 40 ºC and evaporator temperature of 5 ºC) under the same external conditions. They found that the coefficients of performance were % higher than that of the conventional system while the exergy efficiency values were % higher than in the conventional system. In the literature, there have been a limited number of experimental investigations of the exergy efficiency of the R134a ejector-expander refrigeration systems. The main objective of this contribution is to experimentally carry out an exergetic comparison between two refrigeration cycles, a conventional vapor-compression refrigeration cycle (VCRC) and an ejector-expander refrigeration cycle (EERC), using R134a refrigerant and that the system should be suitable for an automobile air conditioner (condenser temperature at 60 ºC and evaporator temperature at 10 ºC). II. Experimental Setup A schematic view of an ejector expander refrigeration cycle, along with its P-h diagram, is given in Fig. 1. The way, in which this system operates, was presented in the previous studies of two of the authors of this article (Bilir and Ersoy, 2009). The experimental refrigeration setup shown in Fig. 2 was used for measuring the COP of both the VCRC and EERC under the same cooling capacity and external operating conditions. The same compressor, evaporator and condenser were used in both refrigeration systems. The differences of the EERC in comparison with the VCRC include the use of an ejector instead of an expansion valve as the main throttling element, the use of liquid-vapor separator at the diffuser exit of the ejector, and the use of a small expansion valve to ensure the pressure drop ( P 70 kpa) for a small throttling between the separator and the evaporator. Explanation and operation of the experimental set together with the details of the elements and properties of the measuring tools used in the experimental set are described in the study by Ersoy and Bilir Sag (2014). In investigating operating conditions, the software was written using the Engineering Equation Solver (EES) software package (Klein, 2011) in accordance with the mathematical model (Bilir and Ersoy, 2009; Ersoy and Bilir, 2013) of the ejector expander system. The important dimensions of ejector geometry (diameters of the motive nozzle throat and exit, the constant area mixing chamber diameter) were determined. Dimensions of the constant area ejector can be seen in Fig 3. 8 Qc 9 Condenser 5 P P cp Primary flow 2 Ejector 2 b 1 b 3 m Secondary flow Compressor Expansion Evaporator valve m b 2 b h 10 Q e 11 Fig. 1: Ejector expander refrigeration cycle with its P-h diagram. 149

164 P T T Inverter Power meter 4 P T 5 T P Compressor Oil separator Sight glass CO/EC P T Filter Dryer Condenser T Water heater 9 8 FM EO/CC CO/EC Sight glass City water T P T Accumulator Ejector P T FM T 6 Separator EO/CC 3 P 1 EO/CC 2 EO/CC P T T 10 FM EO/CC Expansion valve 7 CO/EC P T Evaporator T T 11 FM T P FM TCP Temperature sensor Pressure sensor Flow meter Temperature control panel Expansion valve TCP Electrical heater Brine tank Pump CO/EC CO/EC EO/CC On in conventional mode and Off in ejector mode On in ejector mode and Off in conventional mode Fig. 2: Schematic view of experimental setup for the VCRC and EERC Suction nozzle Constant-area mixing chamber Diffuser Flow from condenser (Primary flow) o 2 1.7º o Stream to separator (mixed flow) Primary nozzle Flow from evaporatory (Secondary flow) Fig. 3: Dimensions of the ejector and its schematic diagram (in mm) 150

165 III. Exergy Analysis The main purpose of using the ejector as an expander is to reduce the irreversibility of the conventional refrigeration system. Exergy analysis is important to determine the reduction of irreversibility for every element and the whole system of refrigeration cycle that uses an ejector instead of an expansion valve. Exergy analysis is also crucial in terms of determining whether there is any improvement in the exergy efficiency in comparison with the conventional system while it predicts the distribution, source and magnitude of irreversible losses in energy systems and hence, provides guidelines for efficient energy usage (Kotas, 1985). In the expansion valve for EERC: m m m 6 7 e (21) E x ( Ex E 7 ) (22) exp, EERC 6 x exp, Ex / E x (23) EERC 7 6 In the separator: m m cp m 3 e (24) m m 4 6 m cp m e (26) x Ex Ex E ) E sep ( 3 4 x6 (27) sep ( Ex 4 Ex 6 ) / E x3 (28) (25) The physical exergy of every point in a cycle is expressed as: E x m ( h h ) T ( s ) (1). 0 0 s0 The reference state values were taken to be ambient pressure of 100 kpa and the temperature of 27 ºC (Ersoy ve Bilir, 2010). The exergy destructions and efficiencies of each element of the cycle can be determined from the following equations (Kotas, 1985; Dincer and Rosen, 2007; Alsuhaibani et al.,2012). In the compressor: m m m 4 5 cp (2) E xcp ( Ex Ex ) W 4 5 cp (3) ( E x Ex 4 ) / (4) cp 5 W cp In the condenser: m 5 (5) m 1 m cp m 9 m w m 8 (6) E x c ( Ex 5 Ex 1) ( Ex 8 E x9) (7) Ex Ex ) /( Ex E ) (8) c ( x5 In the evaporator: m 2 m 7 m e (9) m 10 m 11 m br (10) E x e ( Ex 7 Ex 2) ( Ex 10 E x11) (11) Ex Ex ) /( Ex E ) (12) e ( x2 In the ejector: m m 1 cp (13) m m 2 e (14) m m cp m 3 e (15) x ( Ex Ex E 3 ) (16) E ej 1 2 x Ex /( Ex E 2 ) (17) ej 3 1 x In the expansion valve for VCRC: m m m 1 7 e (18) E x ( Ex E 7 ) (29) exp, VCRC 1 x (20) exp, VCRC Ex7 / Ex1 151 The total exergy destruction in the ejector expansion refrigeration system is thus calculated: E x Ex Ex Ex Ex Ex E, x (29) tot des cp c e When the total exergy destruction is calculated for a VCRC, and E are taken to be zero. E x ej xsep The exergy efficiency is defined as the ratio of total exergy output rate to the total exergy input rate (Aroraa and Kaushikb, 2008; Matawala, 2012). E x Ex 1 ( Ex Ex ) W, (30) sys o i tot des ej loss exp Here, the total exergy destruction rate cp sep E x tot, des is the amount lost because of irreversibility and the unused exergy. The exergy loss rate E is the exergy xloss rate dissipated from the system to the environment, which could be used by other systems (Matawala, 2012). The exergy loss from the system to the environment due to various causes (discharging of the condenser cooling water to the ambient and heat losses due to piping) is computed: E x 9 (31) loss ( Ex Ex 8) Ex heatloss IV. Results and Discussion In this study, the exergetic evaluations of the experimental results, that were obtained from the use of the ejector designed and are shown in Fig. 3, were made for a condenser temperature of 60 C and an evaporator temperature of 10 C (design condition). In order for the ejector system to operate at the design condition, the temperatures and volumetric flow rates of the condenser inlet water and brine fluid were kept constant. Thus, the system was set to the necessary external conditions. In order to carry out the experiments, the external operating conditions were set as follows: the condenser inlet water temperatures were 40.3, 42.9 and 46.2 ºC, the condenser water volumetric flow rate was 0.44 m 3 /h, the temperature and volumetric flow rate of brine fluid were 20 ºC and 0.58 m 3 /h,

166 respectively. First, the cooling capacity of the ejector expander refrigeration system was determined under the external conditions. Then, the experiments of the vapor compression refrigeration cycle were performed under the same external conditions and cooling capacity as the EERC. The same cooling capacity in the conventional system was obtained by the inverter of the compressor. Following this method, the experiments of the ejector expander and conventional system were performed under the same cooling capacity and the same external conditions. temperature, which leads to the increase in the condenser saturation temperature and saturation pressure. For the given operating conditions, the COP values for the ejector expander cycle are % higher than in the conventional cycle. The uncertainties associated with the COP due to the accuracy of the instrument measurements for both the conventional and ejector systems were calculated to range between ±2.73% and ±2.8%. The input and output properties for each component of the EERC and VCRC are given on Table 2 for the design conditions (condenser temperature ~ 60 ºC, evaporator temperature ~ 10 ºC). From Table 1, it is seen that the condenser and evaporator exit pressures are close to each other when the two cycles are compared. Fig. 5 shows the variations of the cooling capacity of the EERC and VCRC with the condenser water inlet temperature. While the water inlet temperature increases, it is observed that the cooling capacity of both the systems also increase. As the condenser water inlet temperature increases, the pressure of the condenser exit and the motive nozzle exit increases. This causes the flow rate of the refrigerant sucked from the evaporator to increase. Thus, it is also predicted to increase the cooling capacity of the system. The cooling capacity of the conventional system was adjusted according to the ejector system. The inverter of the compressor in the VCRC was used to provide it. Fig. 5 also shows the variations of the coefficients of cooling performance (COP) of the EERC and VCRC with the condenser water inlet temperature. It was found that for the condenser water inlet temperature of 46.2 C, the COP for the conventional system was 1.858, whereas it was for the ejector system. Under these conditions, the use of an ejector as an expander provided an improvement of 14.2% in the COP. The reasons of the obtained higher coefficient of performance from the ejector system are the work recovery in this system and the fact that the refrigerant entering the evaporator is almost as saturated liquid. According to Fig. 5, the decrease in the COP for both systems as the condenser inlet water temperature increases is an expected result because of the increased water Fig. 5: Variation of COP values with the condenser inlet water temperature Fig. 6 shows the irreversibility of every component and total irreversibility for the ejector and conventional systems under the design condition for the same external operating conditions and the cooling capacity (4.47 kw). It can be seen that amount of the irreversibility of every component of the ejector system is lower than that of the conventional system. According to Fig. 6, the use of an ejector as an expander in the system reduces irreversibility by 12.66% compared with the total irreversibility of the VCRC. The use of the ejector instead of the expansion valve reduces the total amount of the irreversibility of the refrigeration cycle, as expressed theoretically in a study by Ersoy and Bilir (2010). Fig. 6: Comparing the irreversibility of every component and total irreversibility of the ejector and conventional 152

167 refrigeration cycles Tab. 1: Inlet and outlet variables of every component in both the ejector-expander and conventional refrigeration systems for the same cooling capacities under the same external conditions and (design conditions). EERC VCRC m (kg.s-1) P (kpa) T ( C) h (kj.kg-1) s (kj.kg-1.k-1) m (kg.s-1) P (kpa) T ( C) h (kj.kg-1) s (kj.kg-1.k- 1) Condenser inlet Condenser outlet Condenser cooling water inlet Condenser cooling water outlet Compressor inlet/separator vapor outlet Compressor outlet Ejector inlet (primary flow) Ejector inlet (secondary flow) Ejector outlet (exit of diffuser)/separator inlet Evaporator inlet Evaporator outlet Brine inlet Brine outlet Expansion valve inlet Expansion valve outlet/separator liquid outlet Fig. 7 shows the variation of the total destruction for the ejector and conventional systems with the condenser water inlet temperature. As the condenser water inlet temperature increases, the amount of the total destruction of both systems also increases. This can be explained as follows: as the condenser water inlet temperature increases, the saturated temperature of the condenser increases. Accordingly, the compressor pressure ratio also increases. Hence, the total irreversibility in the systems components and the total exergy destruction increase. In Fig. 7, it is seen that the total exergy destruction amount decreases by % when ejector, instead of the expansion valve, is used in the system. This result shows us, the purpose of using an ejector has been reached experimentally. It is shown in Fig. 7 that as the condenser inlet water temperature increases from 40.3 ºC to 46.2 ºC, the total exergy destruction of the ejector system increases by 15.6%, while that of the conventional system increases by 22.7%. Fig. 8 shows a comparison of the exergy efficiencies for both the conventional and ejector cycles. According to Fig. 8, as the temperature of cooling water at the entrance to the condenser increases, the exergy efficiency for both refrigeration cycles decreases. On the other hand, it is determined that as the condenser cooling water inlet temperature increases the total exergy destruction for both cycles increases (Fig. 7). For this reason, it is an expected result that the exergy efficiency decreases as the condenser water inlet temperature increases. The consumed power of the compressor increases as the condenser cooling water inlet temperature increases for the constant brine temperature and flow rate. With regard to the exergy efficiency equation, it is clear that the exergy efficiency will decrease with the increase in the power consumption of the compressor. Fig. 7: Variation of total exergy destructions of the ejector and conventional cycles with the condenser water inlet temperature. 153 Fig. 8: Variation of exergy efficiency with condenser water inlet temperature for the EERC and VCRC From Fig. 8, it is observed that the exergy efficiency of the EERC is higher than the conventional system.

168 This is because, according to the ejector efficiency equation, when the exergy output of both systems is almost equal, the exergy input (the compressor consumed power) of the ejector system is less than that of the conventional system. It is clear from Fig. 8 that the exergy efficiency increases by % when ejector instead of the expansion valve is used in the system under the same external conditions. A similar result was reported by Lawrence and Elbel, (2013) for the ejector cycle with dual evaporators without separator. The uncertainty associated with the exergy efficiency due to the accuracy of the instrument measurements for both the conventional and ejector systems was found ±2.1%. Fig. 9 shows a Grassmann (exergy flow and loss) diagram of the exergy balance of the ejector cycle operating under the design conditions. It can be seen that the compressor has the highest irreversibility, and the expansion valve operating in small pressure range has the lowest irreversibility. According to the Grassmann diagram, the total destruction of the system is 64.49%. The exergy loss that spreads to the environment from the system due to various reasons (discharging of the condenser cooling water to the environment and heat loss to the environment from the pipeline) is 21.26%. Consequently, the exergy efficiency of the ejector system is 14.25%. In addition, it is seen in Fig. 9 that the irreversibility rate in the evaporator is W, while that in the evaporator of the conventional system is W under the same operating conditions for both systems (Fig. 6). Accordingly, the use of the ejector in the system decreases the irreversibility in the evaporator by 59.6% and increases the exergy efficiency of the evaporator by 51.53%. This is because, as mentioned previously, the refrigerant enters as almost saturated fluid to the evaporator of the ejector system. This improvement in the evaporator and expansion work recovery provided by the ejector creates a combined effect on the system performance. Fig. 9: Grassmann diagram for the EERC under the design conditions. V. Conclusions In this study, we have experimentally investigated usability of the ejector as an expander to reduce the irreversibility of the R134a conventional refrigeration system. We have performed some experiments on an experimental setup that can operate as a conventional or an ejector refrigeration system of the same cooling capacity under the same external operating conditions. b) The total irreversibility was lower than in the conventional system by % while the exergy efficiencies were % higher than in the conventional system. c) The coefficients of performance were determined to be % higher than those of the conventional system. d) Fort future works, exergoeconomics and exergoenvironmental analyses and assessments are recommended. We may summarize some concluding remarks obtained from the results of the present study as follows: a) As the condenser water inlet temperature increased, the amount of the total destruction of both systems also increased. 154 Acknowledgements The authors would like to thank the Scientific and Technical Research Council of Turkey (TUBITAK) for their financial support of the project with the Grant number 110M044. In addition, the present paper

169 constitutes part of the PhD thesis of Nagihan Bilir Sag. Nomenclature E x : Exergy destruction rate (kw) h : Specific enthalpy (kj.kg-1 ) : Mass flow rate (kg.s-1 ) P : Pressure (kpa) Q : Cooling capacity rate (kw) s : Specific entropy (kj.kg-1.k-1 ) T : Temperature (ºC or K) Greek letters : Exergy efficiency ( ) Subscripts b : Primary and secondary flow entrance state to the suction chamber br : Brine c : Condenser ci : Condenser inlet cp : Compressor dest : Destruction e : Evaporator EERC : Ejextor Expander Refrigeration Cycle ei : Eveporator inlet ej : Ejector exp : Expansion valve i : Input m : Constant area mixing chamber o : Output sep : Separator sys : System tot : Total w : Water VCRC : Vapour Compression Refrigeration Cycle 0 : Reference environment m References Alsuhaibani Z., Ersoy H.K., Hepbaşlı A., Exergetic and sustainability performance assessment of geothermal (ground source) ejector heat pumps, International Journal of Exergy, 11 (3), (2012). Aroraa A., Kaushikb S.C., Theoretical analysis of a vapour compression refrigeration system with R502, R404A and R507A, International Journal of Refrigeration, 31 (6), (2008). Bilir N., Ersoy H.K., Performance improvement of the vapour compression refrigeration cycle by a two phase constant area ejector. Int. J. Energy Res. 33 (5), (2009). Bilir Sag N., Ersoy H.K., Hepbasli A., Halkaci H.S., Energetic and exergetic comparison of basic and ejector expander refrigeration systems operating under the same external conditions and cooling capacities, Energy Conversion and Management, 90, (2015). (2007). Chaiwongsa A P., Wongwises S., Experimental study on R-134a refrigeration system using a two-phase ejector as an expansion device. Appl. Therm. Eng. 28, (2008). Dincer I., and M.A., Rosen., Exergy: Energy, Environment and Sustainable Development. Oxford, UK: Elsevier (2007). Disawas S., Wongwises S., Experimental Investigation on The Performance of The Refrigeration Cycle Using a Two-Phase Ejector as an Expansion Device, International Journal of Refrigeration, 27 (6), (2004). Ersoy H.K., Bilir N., The influence of ejector component efficiencies on performance of ejector expander refrigeration cycle and exergy analysis. Int. J. Exergy 30, (2010). Ersoy H.K., Bilir N., Performance characteristics of ejector expander transcritical CO2 refrigeration cycle. Proc. IMche Part A J. Power Energ. 226, (2012). Ersoy H.K., Bilir N., Ejector Design and Experimental Investigation of its Effects on Performance of a Compressorbased Refrigerator in Which the Ejector is Used as an Expander. The Scientific and Technological Research Council of Turkey (TUBITAK), MAG Project 110M044 (2013). Ersoy H.K., Bilir Sag N., Preliminary experimental results on the R134a refrigeration system using a twophase ejector as an expander. Int. J. Refrigeration, 43, (2014). Kornhauser A.A., The use of an ejector as a refrigerant expander. Proceedings of the 1990 USNC/IIR-Purdue refrigeration conference, Purdue University (1990). Klein, S.A EES (Engineering Equation Solver), Academic Professional Version D, F-Chart Software Madison, WI, USA. Kotas T.J., The exergy method of thermal plant analysis. London: Butterworths (1985). Lawrence N., Elbel S., Theoretical and practical comparison of two-phase ejector refrigeration cycles including First and Second Law analysis, International Journal of Refrigeration, 36, (2013). Matawala V.K., Exergoeconomic optimization of an industrial aqua ammonia vapour absorption refrigeration unit, Ph.D. thesis, The Maharaja Sayajirao University of Baroda, Vadodara (2012). Chaiwongsa P., Wongwises S., Effect of throat diameters of the ejector on the performance of the refrigeration cycle using a two-phase ejector as an expansion device. Int. J. Refrigeration, 30, Menegay P., Kornhauser A.A., Improvements to The Ejector Expansion Refrigeration Cycle, Proceedings of the 31th Intersociety Energy Conversion Engineering Conference, Washington DC,

170 (1996). Pottker G. Potentials for COP increase in vapour compression systems. PhD thesis, University of Illinois at Urbana-Champaign (2012). Tas H., Bilgin N., Senturk B., Gungor A., Soğutma Çevrimlerinde Ejektör Kullanımının Araştırılması, Tesisat Mühendisliği Dergisi, 149, (2015). Wongwises S., Disawas S., Performance of the twophase ejector expansion refrigeration cycle, International Journal of Heat and Mass Transfer, 48, (2005). Yilmaz T., Unal S., Thermodynamic analysis of the two-phase ejector air-conditioning system for buses, Applied Thermal Engineering, 79, (2015). 156

171 Thermodynamic Assessment of Ozone Friendly Cascade Refrigeration System Using Natural Refrigerants H. Cenk Bayrakci 1, Onder Kizilkan 2 *, Ahmet Kabul 3, Selin Cekin 4 1,2,3,4 Süleyman Demirel University, Faculty of Technology, Department of Energy Systems Engineering, 32260, Isparta, Turkey * onderkizilkan@sdu.edu.tr Abstract The purpose of this study is to make energy and exergy analysis of two-stage (cascade) refrigeration system which could be reached by spending less energy at lowers temperatures without causing extreme global warming and environment pollution by using natural refrigerants. In cascade system, R-744 and R-600a refrigerants were selected for low temperature circuit and for high temperature circuit respectively. Cooling load was determined as 10 kw at the system. The temperatures were taken 40 at condenser side and -30 at evaporator side. The effect of sub-cooling and superheating were investigated on system performance. Energy and exergy analysis were completed by using EES software. The results were presented by tables and graphically. Keywords: Natural refrigerant, Cascade system, Energy, Exergy I. Introduction Because of the global warming, usage of natural refrigerants is being common in large scale. CO2 refrigeration systems especially preferred at vehicle air conditioners and cascade refrigeration applications. Cascade systems are combined refrigerating systems used in industrial refrigerating plants at very lower temperature applications or at super refrigerating. If a system is to be operated at very lower temperatures, it means lower evaporation and lower condenser temperature, accordingly condensing temperature. In a system, if the heat is aimed quickly removed by condenser and if the refrigerant is aimed to condense totally at low temperatures, this could only be performed by the refrigerating of this system s condenser by other system. Cascade refrigerating systems are gained with the combined operation of two systems that is, with a system s condenser refrigerating other system s evaporator. In cascade refrigeration systems there are two different refrigeration systems as mentioned above. The biggest advantage of cascade system is to be able to provide different properties with different refrigerants. A different refrigerant is used at the system when obtaining low temperatures and as to the system s condenser refrigerating other system s evaporator, it is used a different refrigerant. The first system is called low temperature system and the other one is called high temperature system. In these systems refrigerating applications could be carried out between 70 C and 100 C (Dincer, 2003). In cascade systems there are many studies about the usage of different refrigerants in technical literature. 157 Kim and Kim (2002) have carried out an experimental study for cascade systems in which CO2 (R744), R744/134a and R744/290 have been used. Furthermore, they have compared these experimental data with a simulation suitable for system conditions. They have explained refrigerant flow rate, compressor power, cooling capacity; performance coefficient (COP) values changes with graphics (Kim and Kim, 2002). There are a number of studies about cascade refrigeration systems in the literature. Lee et al. (2006) were made thermodynamic analysis for the most suitable condensing temperature in the cascade condenser of a cascade cooling system which is operated with CO2/NH3 refrigerant couple in their study. In the study, they have taken the temperature difference in the system of evaporation temperature, condensing temperature and cascade condenser as design parameters. In the designed system they have written the exergy balances of each component separately and they have explained the temperature difference in evaporation temperature, condensing temperature and cascade condenser with graphics. With the results, they have presented equations concerning with the system s maximum performance coefficient depending on the temperature difference in evaporation temperature, condensing temperature and cascade condenser, by optimum condensing temperature of cascade condenser and again with the parameters. Gong et al. (2009) were measured cooling performance parameters by using dual (R170 + R23 and R170+R116) and triple (R170 + R23 + R116) azeotropic mixtures. Besides, R508B (R23+R116) refrigerant has also been used for comparison in similar conditions. R404A has been used in system s high temperature circuit. These four refrigerants COP, cooling capacity and condenser

172 temperature values have been determined in various condensing and evaporation temperatures. As a consequence of the indications, they have emphasized that R170+R116 dual mixture has % 10 higher performance coefficient value compared to R508B and mixtures have better potentials at low temperatures to -80 C. Bhattacharyya et al. (2005) were analyzed optimum cases of performance parameters (compressor outlet pressure, COPheating, COPcooling, COPsystem and second law efficiency) in the course of simultaneous heating and cooling. They were explained the results with graphics; they have emphasized that COP value has increased and also underlined that propane and carbon dioxide couple have been ideal in terms of thermal efficiency in these kind of systems. By using finite times method, Agnew and Ameli (2004) have showed the usage of this method at thermodynamics in a cascade circuit where R717 (ammoniac, known as eco-friendly) and R508b alternative refrigerant couple have been used. With this method, they have carried out optimization of a cascade circuit whose high temperature circuit is multi-stage and whose low temperature is singlestage. They have suggested several data to designers for these kind of systems operating with alternative refrigerants. Bhattacharyya et al. (2009) have analyzed simultaneous heating and cooling applications in a heating / cooling purpose cascade circuit where N2O and CO2 refrigerant couple have been used. In their study, they have presented COP (low temperature, high temperature and in general separately for the system) value and the change of Second Law efficiency according to the efficacy of the heat exchanger and inter stage temperature; and also they have given COP value and the change of Second Law efficiency as graphics according to the gas cooler outlet temperature. Therefore, they have introduced that the system s all performance is distinct from heat exchanger and they have stated that the gas cooler, evaporator and internal heat exchanger s design have affected all system s performance equally. Getu and Bansal (2008), were made thermodynamic analysis of an R744 R717 cascade refrigeration system. Their systems working range were between -50 and 40 C. They were showed COP variations with changing R- 717 evaporating temperature and condensing, evaporating and differential temperatures by graphics. They developed a mathematical model (multi linear regression analysis) for a guide of setting optimum thermodynamic design parameters. Duney et al. (2014), were used natural refrigerant propylene (R1270), which was been proposed for transcritical cascade refrigeration system and analyzed. In their study, Propylene was used in the low temperature (LT) cycle and carbon dioxide was used in the high temperature (HT) cycle of the cascade transcritical refrigeration system. They made also a thermodynamic analysis for a transcritical CO2/propylene (R744 R1270) cascade system for cooling and heating applications. They used EES software for making analysis. In the study variation of three important design parameters i.e. gas cooler outlet temperature TC, evaporating temperature TE 158 and overlap temperature in cascade heat exchanger is considered in order to determine system COP, optimum temperature in cascade heat exchanger and optimum mass flow ratio of LT and HT cycles. They developed regression equations for Topt, COPmax and optimum mass flow ratio to help thermal engineers to design an optimized transcritical cascade system. Yan et al (2015), were made thermodynamic analysis of an internal auto-cascade refrigeration cycle (IARC) with mixture refrigerant R290/R600a. R290/R600a mixture was a zeotropic mixture and used in domestic refrigerator-freezers. According to their study, performances of the IARC are evaluated by using a developed mathematical model, and then compared with that of the conventional refrigeration cycle (CRC). According to the simulation results, the IARC with R290/R600a has % improvement in coefficient of performance (COP), % improvement in volumetric refrigeration capacity and % reduction in pressure ratio of compressor compared with those of the CRC under the same given operating conditions. Therefore, many investigations of the cascade refrigeration system are attracting attention The purpose of this study is to make energy and exergy analysis of two-stage (cascade) refrigeration system which could be reached by spending less energy at lowers temperatures without causing extreme global warming and environment pollution by using natural refrigerants. II. System Description Natural refrigerants are naturally occurring substances, such as hydrocarbons (propane, isobutane), CO2, ammonia, water and air. These substances can be used as cooling agents (heat transfer medium) in refrigerators and air conditioners, don't harm the ozone layer and have no or negligible climate impact (Refrigerants Naturally, 2015). This kinds of fluids have been used as refrigerants for many years, however, they are now finding their way into applications where previously fluorocarbons were the preferred option (AGDE, 2015).They are now being used more extensively due to their low impact on the environment (Linde, 2015). The advantages of natural refrigerants are they do not damage the ozone layer and have a negligible impact on the greenhouse effect. From an economic perspective, these refrigerants are inexpensive, in some cases even cheaper than HFCs. Also, natural refrigerants are extremely energy-efficient, sometimes up to 40% more than HFCs (Shecco, 2015; Kizilkan, 2015). In this study, some different natural refrigerants and some chlorine based refrigerants were used in theoretical analysis. Analyses were made by using EES software (Klein, 2015). Schematic diagram of the system was shown in Figure 1. In Table 1, it could be seen that the Physical, safety and environmental properties of investigated the refrigerants.

173 III. Thermodynamic Modelling The performance characteristics of the vapor compression refrigeration cycle for the cold storage facility are assessed by applying first and second law analysis of thermodynamics. The balance equations are used to determine the work and heat interactions, energy and exergy efficiencies and exergy destruction rates for each system component. The general mass balance equation for a steady-state and steady-flow processes can be written as (Cengel and Boles, 2006) m in = m out (1) The energy balance equation is given below: E in = E out (2) Equation (2) can be written as: Q + m inh in = W + m outh out (3) Fig. 1: Schematic diagram of the cascade system where, m is the mass flow rate, E is the rate of net energy, Q is the rate of net heat, W is the rate of net work, and h is the specific the subscripts in and out stand for inlet and outlet respectively. ASHRAE number Tab. 1: Physical, safety and environmental properties of investigated refrigerants Critical Critical ODP GWP Molecular Safety Pressure Temperature (relative (relative formula group (kpa) ( C) to R11) to CO2) 159 Atmospheric life time (year) R1270 CH3CH=CH A R290 CH3CH2CH A R600 CH3CH2CH2CH A R600a CH(CH3) A R717 NH B R12 CCl2F A R22 CHClF A R134a CH2FCF A The second law of thermodynamics overcomes with concepts of entropy and exergy. Exergy analysis of systems allows determining irreversibility and available energy (exergy) in the system. These analyses reveal the efficiency of the systems in terms of first and second law of thermodynamics for a steady state operation, the general exergy balance equation can be defined as (Dincer and Rosen, 2007). Eẋ in = Eẋ out + Eẋ dest (4) In equation 4, the exergy balance equation can also be written as: Eẋ Q Eẋ W = m ine in m oute out + T 0 S gen (5) where, Eẋ Q and Eẋ W are the exergies of heat and work, respectively, e is the specific exergy, T0 is the dead state temperature and S gen is the entropy generation rate. In equation 5, the exergy of heat, the exergy work and entropy generation are given below (Kotas, 1985) Eẋ Q = Q ( T T 0 T ) (7) Eẋ W = W (8) Eẋ dest = T 0 S gen (6) The specific exergy is given below with relative to the environment conditions: e = (h h 0 ) T 0 (s s 0 ) (9) where s is entropy, P is the pressure and the subscript 0 indicates properties at the reference state. The performance of the cascade refrigeration system can be calculated using energy and exergy efficiency definitions: COP = Q E W C,low+W C,high (10)

174 η ex = Eẋ Q E Eẋ W C,low +Eẋ W C,high (11) where Q E represents evaporator refrigeration capacity and W C,low and W C,high represents compressor capacity of lower and higher cycles, respectively. The governing balance equations are given for all system components in Table 2 according to the reference points shown in Figure 1. Tab. 2: Energy and exergy balance equations for system components. Component Mass balance Energy balance Exergy balance Higher cycle compressor m 5 = m 6 = m high W C,high = m high (h 6 h 5 ) Ex 5 + W C,high = Ex 6 + Ex dest,wc,high Condenser m 6 = m 7 = m high Q con = m high (h 6 h 7 ) Higher cycle throttling valve Cascade heat exchanger Lower cycle compressor Lower cycle throttling valve m 7 = m 8 = m high h 7 = h 8 m 8 = m 5 = m high m 2 = m 3 = m low Q HEX = m low (h 2 h 3 ) Q HEX = m high (h 5 h 8 ) m 1 = m 2 = m low W C,low = m low (h 2 h 1 ) Ex Qcon = Q con [1 T 0 ] T con Ex 6 = Ex 7 + Ex Qcon + Ex dest,con Ex 7 = Ex 8 = Ex dest,hctv Ex 2 + Ex 8 = Ex 5 + Ex 3 + Ex dest,hex Ex 4 +Ex Qevap = Ex 1 + Ex dest,wc,low m 3 = m 4 = m low h 3 = h 4 Ex 2 + Ex 8 = Ex 3 + Ex 6 + Ex dest,lctw Evaporator m 4 = m 1 = m low Q evap = m low (h 1 h 4 ) Ex Qevap = [( T 0 T evap ) 1] Ex 4 + Ex Qevap = Ex 1 + +Ex dest,evap IV. Results and Discussion In order to simulate the cascade refrigeration cycle for natural refrigerants, the following assumptions were made: All operations are steady state and steady flow. Pressure losses through pipelines are neglected. Heat losses and heat gains from or to the system are neglected. The changes in potential and kinetic energies are neglected. The pump operations are adiabatic and isentropic. The directions of heat transfer to the system and work transfer from the system are taken positive. Using the balance equations, and under the assumptions given above, the analyses are performed for different natural refrigerants using EES software (Klein, 2015). The results of thermodynamic analyses of the cascade refrigeration system for the given cooling load are given in Table 3 for all refrigerants. The table is divided into two parts, natural refrigerants and chlorine based refrigerants. It can be seen from the table that the best COP value is obtained using R717 followed by, R600, R600a and R290 in natural refrigerants. The performances of these refrigerants are very similar to that of R12, R22 and R134a. Also the trend of exergy efficiency is the same as COP. For the exergy destruction rates, the highest destruction is occurred using R717 and R600 for the given refrigeration duty. Furthermore, the electricity consumption and pressure ratio of the compressor for all refrigerants are given in Table 3. Tab. 3: The results of thermodynamic analyses of the refrigeration system Eẋ Refrigerant COP η Dest, ex kw R R Natural R600a R R Chlorine R based R134a Also, some parametric studies were carried out to see the variation of performance indicators with different parameters. In Fig. 2, variation of condenser temperature with COP can be seen. If the COP values examined, the best COP value was obtained for R717 refrigerant. With the increasing condenser temperatures, COP values decreasing. R290 has the worst value for COP. In Fig. 3, variation of evaporator temperature with COP in can be seen. If the COP values examined, the best COP value was obtained for R717 refrigerant again like before. With the increasing evaporator temperatures, COP values also increasing. R290 has the worst value for COP again. 160

175 Fig. 2: Variation of condenser temperature with COP Fig. 5: Variation of evaporator temperature with exergy efficiency In Figure 6 and Figure 7, variation of condenser and evaporator temperatures with exergy destruction in the high pressure side and low pressure side could be seen respectively. With the increasing condenser temperature and evaporator temperature total exergy destruction values are decreasing. Fig. 3: Variation of evaporator temperature with COP In Fig. 4, variation of exergy efficiency with condenser temperature in the upper system could be seen. If the exergy efficiency values examined, the best value was obtained for R717 refrigerant again like before. With the increasing condenser temperature, exergy efficiency values are decreasing. R290 has the worst value for exergy efficiencies. Fig. 6: Variation of condenser temperature with exergy destruction Fig. 4: Variation of condenser temperature with exergy efficiency In Fig. 5, variation of evaporator temperature with exergy efficiency in the subcooling system could be seen. Same situation occurs here like Figure 3. R717 has best values for exergy efficiency. R290 has the worst values. 161 Fig. 7: Variation of evaporator temperature with exergy destruction V. Conclusion Thermodynamic assessment of cascade refrigaretion cycle was carried out for natural refrigerants while CO2 refrigerant was used for the lower cycle.

176 According to the graphics and results it could be seen clearly that the best values given by natural refrigerants (especially R717 and R600) in the cascade refrigeration systems. Because of low GWP, low exergy destruction, good COP and exergy efficiency values and ozone friendly properties, natural refrigerants could prefer by the refrigeration systems manufacturers. Because of the other refrigerants have higher GWP values and lower COP and exergy efficiency values, they are non-preferable for these systems. However, in the point of view of energy and exergy efficiency, the best alternative refrigerants for higher cycle are found to be R717 and R600. Such a comparison of energetic and exergetic performance of these refrigerants gives valuable and practical knowledge for refrigeration sector designers. References Agnew B., Ameli S.M., A finite time analysis of a cascade refrigeration system using alternative refrigerants. Applied Thermal Engineering, 24, , AGDE, (2015). Australian Government Department of the Environment. lications/natural-refrigerants-case-studies, accessed Bhattacharyya S., Mukhopadhyaya S., Kumar A., Khurana R.K., Sarkar J., Optimization of a CO2 C3H8 cascade system for refrigeration and heating, International Journal of Refrigeration, 28, , Bhattacharyya S., Garaia A., Sarkarb J., Thermodynamic analysis and optimization of a novel N2O CO2 cascade system for refrigeration and heating, International Journal of Refrigeration, 32, , Cengel Y.A., Boles M.A., Thermodynamics: an engineering approach, 5 th ed., McGraw-Hill, New York, USA, Dincer I., Refrigeration Systems and Applications, Wiley, England, Gong M., Sun Z., Wu J., Zhang Y., Meng C., Zhou Y., Performance of R170 mixtures as refrigerants for refrigeration at -80 0C temperature range, International Journal of Refrigeration, 32, , Kim S.G., Kim M.S., Experiment and simulation on the performance of an auto cascade refrigeration system using carbon dioxide as a refrigerant, International Journal of Refrigeration, 25, , Kizilkan O., A Comparative Investigation of Natural Refrigerants: A Case Study for Cold Storage Application, SDU International Journal of Technological Sciences, 7(3), 1-15, Klein, S.A., Engineering Equation Solver (EES), Version D, F-Chart Software, Kotas T.J., The exergy method of thermal plant analysis, Butter-Worths, London, UK, Lee, T.S., Liu, C., Chen, T., Thermodynamic analysis of optimal condensing temperature of cascadecondenser in CO2/NH3 cascade refrigeration systems, International Journal of Refrigeration, 29, , Linde, (2015). The Linde Group, Industrial gases. l_refrigerants/index.html, accessed Refrigerants Naturally, (2015). c/o HEAT International, accessed Shecco, (2015). Beyond HFCs accessed Yan G., Hu H., Yu J., Performance evaluation on an internal auto-cascade refrigeration cycle with mixture refrigerant R290/R600a, Applied Thermal Engineering, 75, , Dincer I., Rosen, M.A., Exergy: Energy, Environment and Sustainable Development, 1 st ed., Elsevier Science, Oxford, UK, Duney, A.M., Kumar, S., Agrawal, G.D., Thermodynamic analysis of a transcritical CO2/propylene (R744 R1270) cascade system for cooling and heating applications, Energy Conversion and Management 86, , Getu, H.M., Bansal, P.K., Thermodynamic analysis of an R744 R717 cascade refrigeration system, International Journal of Refrigeration, 31,45 54,

177 Thermodynamic Analysis of an Integrated System with A Concentrating Collector for Multi-Generation Purposes Yunus Emre Yuksel 1*, Murat Ozturk 2 1 Afyon Kocatepe University, Education Faculty, Department of Elementary Science Education, ANS Campus, Afyon, 03200, Turkey 2 Suleyman Demirel University, Faculty of Technology, Department of Mechatronics Engineering, Cunur, West Campus, Isparta, Turkey * yeyuksel@aku.edu.tr Abstract In this paper, thermodynamic analysis of an integrated system with concentrating collector is investigated for power, heating, cooling and domestic hot water production. The renewable energy based integrated system consists of four sub-systems; i-) a concentrating collector cycle, ii-) an energy storage process, iii-) a Rankine cycle, and iv-) a double effect absorption cooling system. The integrated system for multi-generation purposes is examined in two operating modes, i-) solar mode and ii-) storage system mode. Thermodynamic analysis based on the energy and exergy efficiency, and also exergy destruction rate for whole system and its components are presented for two operating modes. The overall energy and exergy efficiencies of the integrated system are calculated as 51.32% and 46.75%, respectively for the solar mode, whereas these efficiencies are found to be 47.44% and 45.43%, respectively for the storage system mode. In addition, the parametric studies including the variation of ambient temperature from 0 0 C to 30 0 C, and solar radiation flux from 500 W/m 2 to 1000 W/m 2 are presented for the integrated system and its components to investigate and compare the system efficiency. Keywords: Solar energy, concentrating collector, energy, exergy, multi-generation, efficiency. I. Introduction Nowadays, world energy production is based approximately 80% on fossil fuels, such as coal, oil and natural gas (Carvalho, et al., 2011). The problem about fossil fuels is not only shortening of them but damaging the environment as well. Gases such as CO2, SOx, and NOx emitting from burning of fossil fuels constitute greenhouse gases. Therefore, the effects of global warming and climate changes increase day by day, and also these effects can be easily seen by human being. It is time to change this current energy infrastructure with alternative energy sources which are abundant on earth and harmless to the environment. Solar energy is a reliable energy resource. Also, this renewable energy source is a well-known proven renewable energy system, because of its availability. According to the environmental view-point, solar energy systems do not have negative effects and harmful emission to the environment compared to the fossil energy sources, which continuously increase the earth s average ambient temperature and pollution (Al-Sulaiman, et al., 2011). There are a few solar thermal systems that can be used to produce electricity via thermal power plants, such as a solar tower system (STS), a parabolic dish collector (PDC) and a parabolic trough solar collector (PTSC). The PDC system is used to focus the concentrated solar energy on a working fluid to generate heat energy, and then change it to electricity in a conventional generator. The system uses one or more parabolic dishes called reflector to reflect the direct solar energy onto the receiver sub-system located the focus point of the concentrating collector. The thermal energy is collected in a heated working fluid. The high temperature working fluid is transferred to the steam generator through pipes to produce high pressure and super-heated steam. This super-heated steam is sent to a conventional high-efficiency steam turbine in order to generate electricity in this case. Also, the PDC system is the most advanced solar energy technology and has been utilized in large solar power plants for three decades. As a result, they are rather efficient at thermal energy absorption and power conversion processes (Al-Sulaiman, et al., 2011). Thus, in this paper, the PDC system is considered for increasing the temperature of the working fluid. In order to produce cooling, an absorption chiller system should be used integrated with the renewable energy resources, such as solar, geothermal and biomass. The other important advantages of the absorption cooling system should be given as i-) this system does not cause ozone layer depletion, ii-) this system use natural refrigerants possibly having less CO2 emissions, and iii-) this system is independent of the electric grid. The most common commercially suitable absorption refrigeration systems are single and double effect systems. In double effect absorption refrigeration systems, a secondary fluid (absorbent) is used to circulate and to absorb the primary fluid (refrigerant). The success of the absorption relies on the selection of an appropriate combination of absorbent and 163

178 refrigerant (Minciuc, et al., 2003). The most widespread absorbent and refrigerant combinations in absorption refrigeration systems are LiBr-H2O and ammonia-water. The LiBr-H2O pair is the most suitable one for air-conditioning and chilling applications. Gomri (2010) have also performed second law comparison of single effect and double effect vapor absorption refrigeration systems, and concluded that the exergy efficiency of double effect absorption system is higher than the single effect system. Also, in this paper, lithium bromide and water are chosen as working fluid mixture. Zhao et al. (2003) have investigated a new type double effect absorption cooling system based on the energy and exergy analysis. Also, balance equations of the mass, energy and exergy are given for the whole system and its components. Abo Elazm et al. (2011) have presented a study about comparison of single and double effect absorption coolers showing that coefficient of the performance of double effect absorption cooler is higher than single effect absorption system. In this paper, double effect absorption system is considered for cooling application. Increasing population and high energy demand need new alternatives for current energy infrastructure. In order to meet rising energy demand, energy sources should be used more efficiently. Integrated systems offer higher advantages than single output systems in terms of efficiency (Ahmadi, et al., 2013). Several studies have been conducted on multi-generation energy production systems. Buck and Fredmann (2007) have analyzed the efficiency of a tri-generation system based on a micro turbine assisted by a solar power tower. They have conducted an economic analysis on the use of the single and double effect absorption process. The authors have recommended that using the double effect process because it has showed better thermal performance and lower operating cost compared to the single effect absorption process. Khaliq et al. (2009) have presented a trigeneration system using waste heat. They have studied the energy, exergy efficiencies and electrical to thermal energy ratio with respect to both waste heat temperature and pressure of process heat. Kavvadias and Maroulis (2010) have analyzed the multi-objective optimization of a new tri-generation system for power, heating and cooling production. This optimization study has been carried out on technic, economic, energetic and environmental performance indicators in a multi-objective optimization framework. The results have indicated that tri-generation system should be more economically attractive, energy and exergy efficiently and environmental friendly than conventional system. Dincer and Zamfirescu (2012) have carried out energy and exergy based analyses for renewable energy based multigeneration, considering different options for generating such outputs as power, heat, hot water, heating and 164 cooling, hydrogen and fresh water. They have compared single and cogeneration systems in terms of payback time, it is found that cogeneration systems have 2.8 less payback time than single generation system. Ozturk and Dincer (2013) have presented a solar based multi-generation system with hydrogen, electricity, cooling and heating outputs. They have presented exergy destruction rate, energy and exergy efficiencies of each component and the whole system. Also, overall system performance depending on reference temperature has been analyzed. Several researchers have studied the utilizing of integrated systems in energy generation to improve the thermodynamic and environmental performance. Ozturk and Dincer (2013) have researched the integrated systems having rising interest in the last few decades so as to reduce energy consumption and accomplish more sustainable energy production. Al-Sulaiman (2013) has carried out an energy-based analysis of a concentrating solar collector integrated with steam and binary vapor cycles as a prime power for electricity generation. The author has applied an energy efficiency and power production analysis to find the best design parameters of the integrated system. Caliskan et al. (2013) have conceptually modelled hybrid renewable energy based hydrogen and electricity production and storage systems and analyzed them in detail with energy, exergy and sustainable approaches. As a case study they have designed a hybrid wind-solar renewable system. Results show that maximum energy efficiencies of wind turbine, solar PV panel, electrolyzer, and PEMFC are 26.15%, 9.06%, 53.55%, and 33.06% respectively. Maximum exergy efficiencies of the same subsystems are 71.70%, 9.74%, 53.60%, and 33.02% respectively. Ghosh and Dincer (2014) have proposed a novel multi-generation system, which combines three different renewable energy resources, such as solar, wind and geothermal energy, for production multi-outputs as power, heating and cooling, drying and fresh water. The mathematical expressions of the mass, energy, entropy and exergy balance are also given. Also, the meteorological parameters that affect the renewable energy systems for multi-generation are considered. Padilla et al. (2014) have conducted an exergy analysis to parabolic trough collectors (PTC) in order to investigate the effects of operational and environmental parameters on performance of PTC. The main parameters considered for the analysis are: inlet temperature and mass flow rate of heat transfer fluid, wind speed, pressure or vacuum in annulus and solar radiance. According to results, inlet temperature of heat transfer fluid, solar irradiance and vacuum in annulus have significant effect on the thermal and exergetic performance, however the effect of wind speed and mass flow rate of heat transfer fluid is negligible. Al-Ali and Dincer (2014) have presented a new multigenerational integrated geothermal-solar system to produce electrical power, cooling, space

179 heating, hot water and heat for industrial use. They have applied energy and exergy analyses and compared the results of single generation, cogeneration, trigeneration and multigeneration systems. To investigate the effects of operating conditions and environmental parameters, a parametric study is exercised. As a comparison, energy efficiencies of single generation and multigeneration are found as 16.4% and 78% respectively, 26.2% and 36.6% in exergy efficiency respectively. Mamaghani et al. (2015) have modelled a molten carbonate fuel cell-gas turbine (MCFC-GT) hybrid plant in view of energetic, exergetic, economic and environmental analyses. They have optimized the system by using a multi-objective optimization. Target of multi-objective optimization has been 51.7% exergy efficiency and the total cost of million USD per year. They have concluded that operating pressure has the most significant effect on the exergetic efficiency of the plant according to sensivity analysis on variations of system parameters. Chitsaz et al. (2015) have analysed a novel trigeneration system driven by a solid oxide fuel cell (SOFC) in terms of exergy efficiency, exergy destruction rate and greenhouse gas emissions. In the study, four operation cases have been investigated: electrical power generation, electrical power and cooling cogeneration, electrical power and heating cogeneration, and trigeneration. A maximum improvement in the exergy efficiency has been found as 46% in trigeneration system equipped with solid oxide fuel cell as a prime mover compared to the case when the SOFC is used as a standalone unit. Khalid et al. (2015) have analyzed three new developed HVAC systems for heating and cooling applications. To compare these three systems, energy and exergy analyses have been applied for each case and the effects of parameters on energy and exergy efficiencies have been evaluated. The maximum overall efficiency has been found in natural gas operated system with vapour absorption system chiller at 27.5% while minimum energy efficiency has occurred in photovoltaic and solar thermal operated system with vapour compression chiller at 19.9%. A multigeneration energy system based on sawdust biomass fuel with five useful outputs has been analysed in terms of energy and exergy analyses by Soltani et al. (2015). Energy and exergy efficiencies of the multigeneration system are found as 60% and 25%, respectively, while corresponding energy and exergy efficiencies of a biomass system with only electricity generation are 11% and 13%, respectively. Of the several heat recovery options from exhaust gases, electricity generation and wood drying result in the highest exergy efficiency while district heating and drying lead to the highest energy efficiency. Hassoun and Dincer (2015) have developed a new organic Rankine cycle based multigenerational system to meet the demands of a net zero energy building and assessed such a system for an application to a net zero energy house in Lebanon. Energy and exergy analyses have been applied and a parametric study 165 has been conducted. In addition, exergoeconomic analysis and an optimization study for optimizing the total system cost to the overall system efficiency have been carried out. Bade and Bandyopadhyay (2015) have used a pinch analysis based methodology to integrate gas turbine and regenerator with a process plant to minimize fuel consumption. Thermodynamic analysis of this combined heat and power plant has been applied on gas turbine pressure ratio versus power to heat ratio diagram. By using this novel diagram, it is expected to optimize the integration of gas turbine with a process plant. The specific objectives of this study are to investigate a thermodynamic analysis of the multi-generation system supported by the PDC system with an energy storage sub-system, a Rankine cycle and a double effect absorption cooling sub-system, and to reduce the environmental impacts and system cost. The other main sub-objectives of this paper should be detailed as listed below; To develop an advanced Engineering Equation Solver (EES) software code for analyzing a novel integrated system using the PDC as a prime mover. To determine the exergy contents for each stream of the integrated system. To calculate the exergy efficiencies and destructions of the system components and whole system for two operating modes. To conduct a complete parametric study to analyze the impacts of the varying some significant parameters on the integrated system performance. II. System design Fig. 1 shows the schematic representation of the integrated system with the PDC sub-system, the Rankine cycle, the energy storage sub-system and the double effect absorption cooling system. The PDC system collects the solar radiation and then concentrates in order to boil the working fluid for obtaining thermal energy. The working fluid leaving a collector pump at point 6 enters the PDC system to be heated up to 600 C. This outlet temperature is assumed the maximum suitable temperature for the selected working fluid in the PDC system. The mass flow rate of the PDC system without looping is kg/s. At point 1, heated working fluid leaves the concentrating collector and goes through the heat exchanger-i (HEX-I) and HEX-V for solar energy storage and electricity generation, respectively. At point 15, water enters a hot storage tank and then water is pumped into HEX-III at night time. Heat transfer coefficient of the heat water storage tank integrated with multi-generation system is taken as W/m 2 K. The working fluid heated by the PDC

180 system is used to heat the working fluid in the Rankine cycle sub-system for power generation. The Rankine cycle considered in this system has one turbine and feed water unit with a condenser, a pump and a heat exchanger. At point 16, water vapor leaving heat exchanger with approximately 420 C goes through turbine in order to be expanded and produce electricity. After this process, temperature and pressure of water vapor decrease at point 17 where steam enters the condenser-i. Steam at this point is generally high quality saturated liquid-vapor mixture. Steam is condensed at constant pressure in the condenser-i, and leaves the condenser-i as saturated liquid. Water entering pump-iii at point 18 is compressed isentropically to the operating pressure of the HEX-III and V for the night time and solar time, respectively. Figure 1. Schematic diagram of the integrated system with concentrating collector The Rankine cycle of the integrated system produces electricity, and a part of this producing electricity is used for operations of the system devices. The waste heat from the HEX-II is used to produce cooling using by the double effect absorption cooling sub-system. In order to use waste heat of system, the double effect lithium-bromide-water absorption system is chosen instead of a conventional refrigeration system. In this paper, required energy for the double effect absorption system is supplied from the Rankine cycle and the energy storage sub-system. As it can be seen from Fig. 1, the most important components of the double effect absorption system are the high and low pressure generator, the high and low temperature heat exchanger, the solution and refrigerant pump, an absorber, a condenser and an evaporator. It should be noted that, integrated system is modeled according to the optimum operating parameters for the double effect 166 absorption sub-system. Since the solar energy inputs to the system changes with time, the solar-based integrated system has a dynamic process characteristic. Solar radiation increases from zero at the sunrise to its maximum at solar noon time, after that decreases until it equals to zero at the sunset. To reach a continuously processing solar based system, an additional prime mover or an energy storage sub-system should be integrated with the solar system. In this paper, a thermal energy storage system is combined with the solar-based integrated system. This energy storage sub-system stores the excess parts of the solar thermal energy during the solar time, and provides operating the integrated system at night time. Hence, design parameters of the solar thermal energy storage sub-system are very important for a continuously running solar thermal system. In this paper, it is assumed that, 65% of the solar thermal energy during the solar-time is stored in the energy

181 storage sub-system with the purpose of meeting the heat loss from the heat storage tank and to provide heat in the heat exchanger for continuously energy production. Also, a thermocline or single tank system is selected as the hot-storage tank. III. Assumptions In order to make thermodynamic modeling of the efficiency and structure of the integrated system, some assumptions should be accepted, and numerical analysis also should be made for the enthalpy, entropy, temperature, pressure, mass flow rate and exergy of the inlet and outlet flows. The following assumptions are used for this paper; All the system components operate in the steady state conditions. References temperature and pressure are assumed as 25 C and 1 bar, respectively. All flow steams are ideal gases. The changes in the kinetic and potential terms in the energy and exergy balance equations are negligible. No pressure and heat losses are considered in the flow channels. The turbine and pumps in the integrated system are adiabatic. There is no chemical reaction in the system components. IV. Thermodynamic Analysis Thermodynamic analysis is used to evaluate the performance of the system and its components in terms of the first and second laws of thermodynamic. In this section, thermodynamic analysis consisting of the mass, energy, and exergy analysis is employed in order to evaluate performance and improvement potential of the integrated system with concentrating collector system. In the most general principle, a balance equation for a quantity in the given process should be written as follows; Input + Generation Output Consumption = Accumulation (1) This equation gives the quantity balance for a process. The difference between input to the system with generated quantity in the system boundary and output quantity from the system with consumed quantity in the system is equal to the accumulated quantity. In the steady state condition, the accumulation terms in the Eq. (1) are equal to zero, because all properties in the process are unchanging with time (Dincer & Rosen, 2013). IV.1. Mass balance analysis Mass balance is the first analysis in evaluating any system thermodynamically. In the steady state condition, mass balance equation for any system 167 can be written as follows; m i = m e (2) where m is the mass flow rate and subscripts i and e indicates the inlet and exit flow of the matter, respectively. IV.2. Energy balance analysis Energy balance is a key analysis for any system that is investigated. According to the first law of thermodynamic, energy is neither created nor destroyed in a system. So the sum of energy content in a process is always constant, therefore energy is conserved. By neglecting the kinetic and potential energy changes, energy balance analysis should be written as follows showing that the sum of input energy is equal to the sum of output energy; i E i + Q i = e E e + W (3) where Q and W are the heat and work transfer rate, respectively. Neglecting potential and kinetic energy, the above equation should be written as follows; Q + m i h i = W + m eh e (4) where h is the specific enthalpy. The energy analysis and balance equation is performed for each main components of the integrated system as given below. IV.2.1. Concentrating collector The PDC system has a parabolic shape with covered reflecting materials, and a receiver is placed to the focal point of the concentrating collector. Reflected solar energy from the reflecting surface to the collector receiver should be given as follows (Kalogirou, 2009) (Duffie & Beckman, 2006); Q R = ρ r,c α r,c I ds A C (5) where ρ r,c and α r,c are the reflectivity and absorptivity of the concentrating collector reflecting surface material, I ds is the direct solar radiation and A C is the area of the reflector. The produced useful power from the PDC system should be calculated as follows; Q u = F R [C o (ρ r,r α r,r )Q R U L (T c T o )A R εσ(t R 4 T o 4 )A R ] (6) where F R is the heat removal factor, C o is the collector concentrating ratio, ρ r,r and α r,r are the reflectivity and absorptivity of the receiver, U L is the overall heat loss coefficient of the concentrating collector, T c and T R are the collector and receiver temperature, respectively, ε is the receiver emissivity and σ is the Stefan-Boltzmann constant.

182 To analyze the concentrating collector performance on the basis of thermodynamic assessment, design parameters of the collector are given in Table 1. Tab. 1: Concentrating collector parameters Parameter Values ρ r,c 0.9 ρ r,r 0.85 α r,c 0.9 α r,r 0.85 A C 100 m 2 A R 0.4 m 2 F R 0.9 C o 250 U L 25 W/m 2 K ε 0.2 IV.2.2 Rankine cycle As seen from Fig. 1, the high temperature working fluid goes to the Rankine cycle turbine at point 16 and after expansion leaves from here at point 17. To analyze the inlet and outlet enthalpies and turbine power output, the energy balance equation of the turbine should be written as follows; m 16h 16 = m 17h 17 + W T (7) Energy balance equation for the condenser-i in the Rankine cycle is given by; m 17h 17 = m 18h 18 + Q cond I (8) The Rankine cycle pump work should be written using by the energy balance equation; W pump IIII = m 16(h 18 h 19 ) (9) In order to exchange heat energy more from the PDC to the Rankine cycle, HEX-V is used. Inlet and outlet enthalpies of the HEX-V should be calculated by simply applied to the energy balance equation; m 7(h 7 h 8 ) = m 16(h 16 h 20 ) + Q loss,hex V (10) IV.2.3 Double-effect absorption cooling sub-system To analyze the inlet and outlet conditions of the generator-i (or high temperature generator), energy balance equation should be written as follows; m 22h 22 + m 30h 30 + Q gen I = m 23h 23 + m 31h 31 + m 34h 34 (11) An energy balance equation of the generator-ii (or low temperature generator) is given as follows; m 33h 33 + m 34h 34 + Q gen II = m 35h 35 + m 37h 37 + m 38h 38 (12) An energy balance equation of the condenser-ii in the cooling sub-system should be expressed as follows; m 36h 36 + m 37h 37 = m 24h 24 + Q con II (13) The following energy balance equation should be used to calculate the heat absorbed from the evaporator; m 25h 25 + Q eva = m 26h 26 (14) To obtain the heat rejected from the absorber, the following energy balance equation should be used; m 26h 26 + m 40h 40 = m 27h 27 + Q abs (15) IV.3. Exergy balance analysis Exergy is defined as the maximum useful work that could be obtained from the system at a given state in a specified environment. Exergy analysis provides more meaningful information about a system or a process than energy analysis. Because energy analysis deals with conserved quantities, it provides one sided view of a process or a system. Unlike energy analysis, exergy analysis shows inefficiencies and wastes occurring in the process. According to the second law of thermodynamic, different kinds of energies exist, all energies are not equal, and quality decreases in any system. Exergy analysis based on the second law of thermodynamic can be used to investigate a system for more economical and effective use of energy sources (Yuksel & Ozturk, 2014). Thus, integrated systems offering more effective use of energy sources should be evaluated with both energy and exergy analyses. By applying exergy analysis, where and how inefficiencies occur can be found, and so efficiency of the system can be increased. Based on the first and second law of thermodynamics, an exergy balance equation should be written as follows (Dincer & Rosen, 2013); i m inex in + E x Q = e m outex out + E x W + E x D (16) where E x Q and E x W are the heat and work exergy flow rates through the boundary at temperature Tj at location j, respectively. E x D is the exergy destruction rate. Heat exergy flow rate is given as; E x Q = (1 T o T i ) Q i (17) where T i is the temperature in the i th given state. E x W = W (18) Exergy is generally comprised of four parts which are physical exergy (exph), chemical exergy (exch), kinetic exergy (exke), and potential exergy (expe). The 168

183 specific exergy should be given as follows; ex = ex ke + ex pe + ex ph + ex ch (19) In this study, we have neglected the kinetic, potential and chemical exergy, as elevation difference is low and speeds in the process are negligible, and there is no chemical reaction. Physical exergy can be defined as maximum effective work as a process interacts with the environment (Bejan, et al., 1996). The physical exergy rate of the i th flow is written as follows; ex ph = (h i h o ) T o (s i s o ) (20) The exergy rate of the material flow should be calculated as follows; E x i = m ex i (21) The exergy destruction rate for each components of the solar-based integrated system are written based on the given above procedure, and shown in Table 2. Tab. 2: Exergy destruction rate equations System Exergy destruction rate equations components Parabolic Q E x D,PDC = E x 6 E x 1 + E x Solar collector HEX- I E x D,HEX I = E x 2 + E x 14 E x 3 Ex 15 Collector E x D,CP = E x 5 E x 6 + W CP pump Pump-I E x D,P I = E x 13 E x 14 + W P I Hot storage Ex D,HST = E x 15 E x 9 tank Cold storage E x D,CST = E x 12 E x 13 tank Turbine E x D,tur = E x 16 E x 17 W T Condenser-I Q E x D,Con I = E x 17 E x 18 E x Con I Generator-I E x D,Gen I = E x 22 + E x 30 E x 23 E x 31 E x 34 Expansion E x D,EV I = E x 32 E x 33 Valve-I Generator-II E x D,Gen II = E x 33 E x 34 + E x 35 + E x 37 + E x 38 Condenser-II Q E x D,Con II = E x 36 E x 37 E x 24 E x Con II Evaporator Q E x D,Eva = E x 25 E x 26 + E x Eva Absorber Q E x D,Ab = E x 26 + E x 40 E x 27 E x Ab IV.4. Energy and exergy efficiencies of the integrated system The definition of the energy efficiency of the system is the ratio of useful energy outputs by the system to the total energy inputs. In this paper, solar based integrated system is considered in two modes. Energy efficiency equations of the main sub-system should be written as follows for the solar mode (SM); Ƞ Rankine SM = W Rankine SM W p III Q HEX V+W p III Ƞ storage SM = Q storage tank Q HEX I+W p I Ƞ absorption SM = Q cooling SM+Q heating SM Q HEX II+W p IV (22) (23) (24) 169 Ƞ system SM = (W Rankine SM W p I W p III W p IV W CF) + Q PDC Q cooling SM+Q heating SM+Q storage SM Q PDC (25) The energy efficiency equations for the storage system mode (SSM) are given as follows; Ƞ Rankine SSM = Ƞ storage SSM = W Rankine SSM Q HEX III+W p III Q HEX III+Q HEX IV Q storage tank+w p II Ƞ absorption SSM = Q cooling SSM+Q heating SSM Q HEX IV+W p IV (26) (27) (28) Ƞ system SSM= (W Rankine SSM W p II W p III W p IV) Q storage + Q cooling SSM+Q heating SSM Q storage (29) The evaporator provides the cooling applications and the energetic coefficient of performance (COP en ) of the double effect absorption system should be written as follows; COP en = Q cooling Q gen I+W p IV (30) Exergy efficiency equations of the integrated system for the solar system mode should be written as follows; Ψ collector SM = E Q x PDC Q E x solar +W CF Ψ Rankine SM = W Rankine SM W p III Q E x HEX V +W p III Q Ψ storage SM = E x storage tank Q E x HEX I +W p I Q Ψ absorption SM = E x cooling SM Q +E x heating SM Q E x HEX II +W p IV Ψ system SM = (W Rankine SM W p I W p III W p IV W CF) Q E x cooling SM (31) (32) (33) (34) Q + E x PDC Q Q +E x heating SM +E x storage SM Q (35) E x PDC Exergy efficiency equations of the integrated system for the storage system mode should be given as; Ψ Rankine SSM = Ψ storage SSM = E x HEX III W Rankine SSM Q E x HEX III +W p III Q Q +E x HEX IV Q E x storage tank +W p II Ψ absorption SSM = E Q Q x cooling SSM +E x heating SSM Q E x HEX IV +W p IV Ψ system SSM = (W Rankine SSM W p II W p III W p IV) Q Q E x cooling SSM +E x heating SSM E x storage Q + E x storage (36) (37) (38) Q (39) Exergetic coefficient of performance (COP ex ) of the double effect absorption cooling system should be written as follows;

184 COP ex = E Q x cooling Q E x gen I +W p III (40) The exergy efficiency equations for each component of the solar based integrated system are given in Table 3. IV. Results and discussion In this chapter, the outcomes of thermodynamic assessment of the integrated system using renewable energy are exhibited and discussed. In this analysis, the mass flow rates of the working fluids are kept constant. The thermodynamic data at each state number, such as mass flow rate (kg/s), pressure (kpa), temperature ( C), enthalpy (kj/kg), entropy (kj/kgk), energy rate (kw) specific exergy (kj/kg) and exergy rate (kw) for both solar and storage sub-system modes of the multi-generation system are calculated by using the EES software program (Klein, 2007). The performance of each component of the integrated system is analyzed through different thermodynamic parameters, i.e., exergy destruction rate (kw), exergy destruction ratio (%), exergy efficiency (%) and power or heat transfer rate (kw). These thermodynamic parameters are calculated by the equations, and analysis results are presented in Table 4. Tab. 3: Exergy efficiency equations for the solar based integrated system System components Exergy efficiency equations Parabolic dish collector Ψ PDC = E x 1 E x 6 Q E x Solar HEX- I Ψ HEX I = E x 15 E x 14 E x 2 E x 3 Collector pump Ψ CP = E x 6 E x 5 W CP Pump-I Ψ P I = E x 14 E x 13 W P I Hot storage tank Ψ HST = E x 9 E x 15 Cold storage tank Ψ CST = E x 13 Turbine Ψ Tur = E x 12 W T E x 16 E x 17 Q Condenser-I Ψ Con I = E x Con I E x 17 E x 18 Generator-I Ψ Gen I = E x 31+E x 34 E x 30 E x 22 E x 23 Expansion valve-i Ψ EV I = E x 33 E x 32 Generator-II Ψ Gen II = E x 34 E x 35 E x 37 +E x 38 E x 33 Expansion valve-ii Ψ EV II = E x 36 E x 35 Condenser-II Ψ Con II = Evaporator Ψ Eva = E x Col Absorber Ψ Ab = Q E x Con II E x 36 +E x 37 E x 24 Q E x 26 E x 25 Q E x Ab E x 26 +E x 40 E x 27 Moreover, the overall exergy destruction rate is analyzed to ensure better understanding of the magnitudes of the useful work losses in the all sub-systems. In order to investigate the system performance, the exergy efficiency analysis can assist to identify the ineffectiveness within the integrated system. The system performance should 170 also be improved by reducing the heat losses in the integrated system by the re-design or modification studies. Table 6 shows that the highest exergy destruction rate occurs in the PDC system with 2042 kw, and its exergy efficiency is 80.73% for the solar mode. The component having the highest exergy efficiency is HEX-I with 98.69% for the solar mode. According to the thermodynamic assessment results, it is necessary to improve the development aims on this concentrating collector model for the more efficiency solar based integrated system design. As seen in Table 7, for storage sub-system mode, the highest exergy destruction rate belongs to the heat storage tank with kw, and also exergy efficiency of this component is 12.69%. Two components have very high exergy efficiencies; these are expansion valve-iv and expansion valve-i with 99.67% and 99.27%, respectively. In addition, exergy destruction rate of these components for the storage system mode are 0.31 kw and 0.65 kw, respectively. Tab. 4: Thermodynamic analysis results for solar mode of the renewable energy based integrated system devices Devices Exergy destruction rate (kw) Exergy destruction ratio (%) Exergy efficiency (%) Power or heat transfer rate (kw) Parabolic dish collector HEX-I Hot storage tank HEX-V HEX-II Collector pump Pump-I Pump-III Condenser-I Turbine Generator-I Generator-II HEX-VI HEX-VII Pump-IV Condenser-II Expansion valve-i Expansion valve-ii Expansion valve-iii Expansion valve-iv Evaporator Absorber Energy and exergy efficiencies of each sub-system of the renewable energy based integrated system for the solar mode and storage system mode are illustrated in Fig. 2 and 3, respectively. As it is observed, energy and exergy efficiency of the whole system for both modes are higher than all other sub-systems, because integrated systems have higher efficiency values than single output systems. According to the results shown in Fig. 2 and 3, the

185 Rankine sub-system has the highest energy and exergy efficiency among all sub-systems for each mode, mainly due to the work producing by working fluid passes successively through the Rankine turbine, and the concentrating collector has the lowest energy and exergy efficiency for the solar mode principally for irreversibility associated with the high temperature difference between the collector and ambient air. The absorption cooling sub-system has the lowest energy and exergy efficiency for the storage system mode, mainly due to the temperature differences of working fluid streams, and also due to the pressure drops in the cooling system components. It is recommended that, it should likely be important to focus development studies on the concentrating collector and absorption cooling sub-system. Otherwise, the thermodynamic analysis results illustrated that the cooling sub-system does not show important exergy destruction rate, principally this sub-system does not directly use the fuel energy but instead utilizes waste heat produced by the concentrating collector sub-system. The energetic COP of the cooling sub-system is calculated as 0.98 for two modes. This is much greater than the exergetic COP which are for solar mode and for storage system mode, respectively. parameters have important roles in the integrated system outputs. The effects of the varying ambient temperature from 0 C to 30 C on the exergy destruction rate and exergy efficiency of the integrated system for the solar mode are presented in Fig. 4 through the Fig. 8. Exergy destruction rate and exergy efficiency of the PDC system for the solar mode with respect to reference temperature is shown in Fig. 4. As seen from this figure, when the ambient temperature increases, the exergy destruction rate of the PDC system decreases from 1190 kw to 1135 kw, and exergy efficiency of the collector increases from 14.42% to 15.36%, respectively. This originates from the decrease in the temperature difference between ambient and concentrating collector. Ex D,collector (kw) Ex D,collector ycollector 0,154 0,152 0,15 0,148 0,146 ycollector , T 0 ( o C) Fig. 4: Exergy destruction rate and exergy efficiency of the parabolic dish collector system for solar mode depending on the reference temperature changes Fig. 2: Energy and exergy efficiencies for sub-systems of the integrated system for solar mode Fig. 5 illustrates the variation in the ambient temperature and their corresponding impacts on the exergy destruction rate and exergy efficiency of the Rankine sub-system for the solar mode. While the ambient temperature increases from 0 C to 30 C, exergy destruction rate of the Rankine sub-system decreases from 1700 kw to 850 kw, and also the exergetic efficiency increases from 37.21% to 41.05%. The main reason for this outcome, output exergy rate from the Rankine sub-system increases by the increasing ambient temperature, and consequently exergy efficiency of the system increases simultaneously Fig. 3: Energy and exergy efficiencies for sub-systems of the integrated system for storage system mode To analyze the efficiency variation of the renewable energy based integrated system in terms of exergy destruction rate and exergy efficiency, main design parameters such as ambient temperature and solar radiation intensity are investigated. These design 171 Ex D,Rankine-SM (kw) Ex D,Rankine-SM y Rankine-SM T 0 ( 0 C) Fig. 5: Exergy destruction rate and exergy efficiency of the Rankine system for solar mode depending on the reference temperature changes yrankine-sm

186 Similarly, in Fig. 6, by increasing the ambient temperature of the storage tank sub-system for continuously system operation, the exergy destruction rate decreases from 952 kw to 412 kw, and the exergy efficiency increases exponentially from 9.1% and 21.84%, respectively. This is due to fact that the increase in the ambient temperature for the heat storage tank sub-system requires less heat energy. In contrast, as seen in Fig. 7, exergy destruction rate of the absorption cooling system for solar mode increases by the increasing ambient temperature and exergetic efficiency decreases. This is due to increasing the temperature difference between ambient air and cooling sub-system requires more cooling load. Ex D,HST-SM (kw) T 0 ( o C) Fig. 6: Exergy destruction rate and exergy efficiency of the hot storage tank for solar mode depending on the reference temperature changes Ex D,absorption-SM (kw) Ex D,HST-SM y HST-SM Ex D,absorption-SM yabsorption-sm Fig. 7: Exergy destruction rate and exergy efficiency of the absorption cooling system for solar mode The impacts of varying ambient temperature on the whole system exergy destruction rate and exergy efficiency is shown in Fig. 8. It can be observed that, increasing ambient temperature decreases the exergy destruction rate, and increases the exergy efficiency for the solar mode. Because the solar radiation intensity varies during the solar daylight, the variations of the integrated system efficiency are analyzed. Fig. 9 demonstrates the variations of exergy destruction rate and exergy efficiency of the concentrating collector sub-system for different values of solar radiation intensity. As seen in this figure, exergy destruction rate and exergy efficiency of the PDC system for the solar , T 0 ( o C) 0.1 0,184 0,18 0,176 0,172 0,168 yhst-sm yabsorption-sm 172 mode increase with increasing solar radiation intensity. This is because increasing the solar radiation intensity increases the outlet temperature of the working fluid for the PDC sub-system. Ex D,system-SM (kw) , T 0 ( o C) Fig. 8: Exergy destruction rate and exergy efficiency of the whole system for solar mode depending on the reference temperature changes Ex D,collector-SM (kw) Fig. 9: Exergy destruction rate and exergy efficiency of the parabolic dish collector system for solar mode depending on the solar radiation intensity changes The effects of the increasing solar radiation intensity on the exergy destruction rate and exergy efficiency of the whole system for the solar mode are given in Fig. 10. Similarly, these two thermodynamic properties increase with increasing solar radiation intensity. Ex D,system-SM (kw) Ex D,system-SM ysystem-sm 0,47 0,46 0,45 0,44 0, I b (W/m 2 ) Ex D,collector-SM y collector-sm Ex D,system-SM y system-sm I b (W/m 2 ) Fig. 10: Exergy destruction rate and exergy efficiency of the whole system for solar mode depending on the solar radiation intensity changes ysystem-sm ycollector-sm ysystem-sm

187 Similar results between the exergy destruction rate and exergy efficiency based on the variable ambient temperature for the integrated-system components for the storage system mode are obtained, and the parametric studies results are shown in Figs Ex D,Rankine-SSM (kw) T 0 ( o C) Fig. 11: Exergy destruction rate and exergy efficiency of the Rankine system for storage system mode depending on the reference temperature changes Ex D,HST-SSM (kw) Ex D,Rankine-SSM y Rankine-SSM Ex D,HST-SSM y HST-SSM T 0 ( o C) Fig. 12: Exergy destruction rate and exergy efficiency of the hot storage tank for storage system mode depending on the reference temperature changes For storage mode, as seen in Figs. 11, 12 and 14, increase of the ambient temperature decreases the exergy destruction rate of the Rankine system, storage tank system and whole system, respectively, and increases the exergy efficiency of these systems. In contrast, as seen in Fig. 13, the exergy destruction rate of the absorption cooling sub-system for the storage system mode increases, and exergy efficiency decreases with increasing ambient temperature. Therefore, this situation improves the exergetic COPs of the integrated systems. Maximum exergy efficiency of whole system for the solar mode and storage system mode which are 46.8% and 46.4%, respectively, occurs at 30 C ambient temperature. At this temperature condition, exergy destruction rates of whole system are about 3500 kw and 2000 kw, respectively, for the both solar and storage system mode yrankine-ssm yhst-ssm Ex D,absorption-SSM (kw) , T 0 ( o C) Fig. 13: Exergy destruction rate and exergy efficiency of the absorption system for storage system mode depending on the reference temperature changes Ex D,system-SSM (kw) Fig. 14: Exergy destruction rate and exergy efficiency of the whole system for storage system mode depending on the reference temperature changes V. Conclusions Ex D,absorption-SSM yabsorption-ssm Ex D,system-SSM ysystem-ssm 0,158 0,156 0,154 0,152 0,148 In this paper, thermodynamic analysis results of an integrated system powered by the PDC system are presented. Because energy analysis cannot provide adequate information about the energy losses, exergy analysis is performed in order to see real efficiencies and destructions of whole system and its components. In addition, parametric studies are conducted in order to understand how environment temperature affects the exergy efficiencies of each systems and whole system. Also, solar radiation intensity impacts on the exergy destruction rate and exergy efficiency of the integrated system considered for the solar mode are investigated for a more efficient system design. According to the thermodynamic analysis results based on the first and second laws, integrated systems offer higher efficiency than single output systems. Finally, these concluding points can be drawn from the analyses: Exergy efficiencies of the Rankine cycle, PDC system, hot storage tank, absorption cooling sub-systems and whole system for the solar mode are found as 43.05%, 15.36%, 17.43%, 16.46% and 46.75%, respectively. 0,16 0,15 0,47 0,46 0,45 0,44 0, , T 0 ( o C) yabsorption-ssm ysystem-ssm 173

188 Exergy efficiencies of the Rankine cycle, storage tank, absorption cooling sub-systems and whole system for the storage system mode are calculated as 38.48%, 18.27%, 14.73% and 45.62%, respectively. Increasing ambient temperature causes an increase in the exergy efficiency of the all sub-systems except for the absorption cooling sub-system and whole system for both solar and storage system modes. Increasing solar radiation intensity causes an increase in the exergy efficiency of the PDC sub-system and whole system for the solar mode. With regard to exergy analysis results, the highest exergy destruction ratio occurs in the PDC system and hot storage tank for the solar and storage system mode with 51.30% and 36.61%, respectively. The expansion valve III in the absorption cooling sub-system has the lowest exergy destruction rates in the multi-generation system for both solar and storage system modes, because the absorption cooling components create relatively low exergy destruction in the proposed system. Nomenclature A C : Collector area (m 2 ) C : Concentrating ratio I ds : Direct solar radiation (W/m 2 ) E : Energy rate (kw) COP en :Energetic coefficient of performance COP ex :Exergetic coefficient of performance ex : Specific exergy (kj/kg) ex ch : Chemical exergy (kj/kg) ex ph : Physical exergy (kj/kg) E x : Exergy rate (kw) E x D : Exergy destruction rate (kw) E x Q : Exergy transfer associated with heat transfer (kw) E x W : Exergy transfer associated with work (kw) F R : Heat removal factor h : Specific enthalpy (kj/kg) HEX : Heat exchanger U L : Heat loss coefficient (W/m 2 K) m : Mass flow rate (kg/s) T : Temperature (K) PDC : Parabolic dish collector SM : Solar mode SSM : Storage system mode Q : Heat rate (kw) W : Work rate (kw) Greek Letters : Energy efficiency : Exergy efficiency ρ r,c : Collector reflectivity ρ r,r : Receiver reflectivity : Collector absorptivity α r,c 174 α r,r ε σ Subscript ch e i ke pe ph R u o References : Receiver absorptivity : Receiver emissivity : Stefan-Boltzmann constant : Chemical exergy : Exit : Inlet : Kinetic exergy : Potential exergy : Physical exergy : Receiver : Useful : References state Ahmadi, P., Dincer, I. & Rosen, M., Development and assessment of an integrated biomass-based multi generation energy system. Energy, Issue 56, pp Al-Ali, M. & Dincer, I., Energetic and exergetic studies of a multigenerational solar-geothermal system. Applied Thermal Engineering, Issue 71, pp Al-Sulaiman, F., Energy and sizing analyses of parabolic trough solar collector integrated with steam and binary vapor cycles. Energy, Issue 58, pp Al-Sulaiman, F., Dincer, I. & Hamdullahpur, F., Exergy modeling of a new solar driven trigeneration system. Solar Energy, 85(9), pp Al-Sulaiman, F., Hamdullahpur, F. & Dincer, I., Trigeneration: A comprehensive review based on prime movers. International Journal of Energy Research, 35(3), pp Bade, M. & Bandyopadhyay, S., Analysis of gas turbine integrated cogeneration plant: Process integration approach. Applied Thermal Engineering, Issue 78, pp Bejan, A., Tsatsaronis, G. & Moran, M., Thermal Design and Optimization. s.l.:john Wiley & Sons Inc. Buck, R. & Friedmann, S., Solar-assisted small solar tower trigeneration systems. Journal of Solar Energy Engineering, 129(4), pp Caliskan, H., Dincer, I. & Hepbasli, A., Energy, exergy and sustainability analyses of hybrid renewable energy based hydrogen and electricity production and storage systems: Modeling and case study. Applied Thermal Engineering, Issue 61, pp Carvalho, M., Serra, L. & Lozano, M., Optimal synthesis of trigeneration systems subject to environmental constraints. Energy, 36(6), pp

189 Chitsaz, A., Mahmoudi, S. & Rosen, M., Greenhouse gas emission and exergy analyses of an integrated trigeneration system driven by a solid oxide fuel cell. Applied Thermal Engineering, Issue 86, pp Dincer, I. & Rosen, M., Exergy: energy, environment and sustainable development. 2 dü. s.l.:elsevier, Oxford, UK. Dincer, I. & Zamfirescu, C., Renewable-energy-based multigeneration systems. International Journal of Energy Research, 46(15), pp Duffie, J. & Beckman, W., Solar engineering of thermal processes. s.l.:john Wiley & Sons, Inc.. Elazm, M. A., Shahata, A., Elsafty, A. & Aboelnasr, M., Parametric thermodynamic analysis for single and double effect absorption systems. Istanbul, Turkey, 10th International Conference on Sustainable Energy Technologies. Ghosh, S. & Dincer, I., Development and analysis of a new integrated solar-wind-geothermal energy system. Solar Energy, Issue 107, pp Gomri, R., Thermal seawater desalination: possibilities of using single effect and double effect absorption heat transformer systems. Desalination, 253(1-3), pp Hassoun, A. & Dincer, I., Analysis and performance assessment of a multigenerational system powered by Organic Rankine Cycle for a net zero energy house. Applied Thermal Engineering, Issue 76, pp Kalogirou, S., Solar energy engineering: processes and systems. s.l.:elsevier. Kavvadias, K. & Maroulis, Z., Multi-objective optimization of a trigeneration plant. Energy Policy, 38(2), pp system. Applied Thermal Engineering, Issue 77, pp Minciuc, E. et al., Thermodynamic analysis of trigeneration with absorption chilling machine. Applied Thermal Engineering, 23(5), pp Ozturk, M. & Dincer, I., Thermodynamic analysis of a solar-based multi-generation system with hydrogen production. Applied Thermal Engineering, Issue 51, pp Ozturk, M. & Dincer, I., Thermodynamic Assessment of an Integrated Solar Power Tower and Coal Gasification System for Multi-generation Purposes. Energy Conversion and Management, Issue 76, pp Padilla, R. et al., Exergy analysis of parabolic trough solar receiver. Applied Thermal Engineering, Issue 67, pp Soltani, R., Dincer, I. & Rosen, M., Thermodynamic analysis of a novel multigeneration energy system based on heat recovery from a biomass CHP cycle. Applied Thermal Engineering, Issue 89, pp Yuksel, Y. & Ozturk, M., Thermodynamic modelling of an integrated energy system for polygeneration design. Istanbul, Turkey, Clean Energy Conference. Zhao, Z., Zhou, F., Zhang, X. & Li, S., The thermodynamic performance of a new solution cycle in double absorption heat transformer using water/lithium bromide as the working fluids. International Journal of Refrigeration, Issue 26, pp Khalid, F., Dincer, I. & Rosen, M., Development and analysis of sustainable energy systems for building HVAC applications. Applied Thermal Engineering, Issue 87, pp Khaliq, A., Kumar, R. & Dincer, I., Performance analysis of an industrial waste heat-based trigeneration system. International Journal of Energy Research, 33(8), pp Klein, S., Engineering Equation Solver (EES), Academic Commercial, F-Chart Software. Mamaghani, A., Najafi, B., Shirazi, A. & Rinaldi, F., Exergetic, economic, and environmental evaluations and multi-objective optimization of a combined molten carbonate fuel cell-gas turbine 175

190 Heat Recovery Analysis of a Rotary Kiln in Cement Industry Ahmet Yakup Cumbul 1, Mehmet Akif Ezan 1*, Ismail Hakki Tavman 1, Arif Hepbasli 2 and M. Ziya Sogut 3 1 Dokuz Eylul University, Department of Mechanical Engineering, Izmir, Turkey 2 Yasar University, Department of Energy Systems Engineering, Izmir, Turkey 3 Bursa Orhangazi University, Department of Mechanical Engineering, Bursa, Turkey * mehmet.ezan@deu.edu.tr Abstract Energy efficiency of a cement production process is quite small due to large amounts of heat loss from the systems. Rotary kilns have been widely used in the cement industry to produce clinker. The surface temperature of the rotary kiln reaches up to 300 C. Considering the higher temperature difference and higher surface area of the kiln (Diameter: 4.8 m, Length: 70 m), heat losses to the environment become significant while those from the furnace determine the overall efficiency of the cement production. In the current work, a heat recovery unit was proposed to recovery the heat losses from the rotary kiln first. A mathematical model was then developed using the Engineering Equation Solver (EES) software package to resolve the coupled balance equations. Finally, parametric results were obtained from varying the ambient conditions and working parameters of the system. It was determined that the useful heat rate recovered from the rotary kiln could reach up to 350 kw. Keywords: Rotary Kiln, heat recovery, mathematical model I. Introduction Portland cement is one of the most widely used construction material for buildings. Cement is a fine powder material and it is the core ingredient of concrete. In an industrial cement production process, the raw materials are heated inside large rotating furnaces, which is also known as rotary kiln, to produce cement clinker. Four essential elements, silicon, aluminum, iron, and calcium are used in the production of cement. Most cement production plants are settled nearby the mineral deposits of limestone. Other ingredients that are necessary for cement production are smaller amounts of clay, sand, mill scale, shale, bauxite and fly ash. The raw materials undergo several processes before entering the rotary kilns. Materials are broken into small pieces inside a crusher and then blended to prepare a mixture in a proportion of ingredient. Materials are then ground into powder and sent through the preheater and increase the temperature of the mixture before entering the furnace. Rotary kilns are slightly inclined cylindrical vessels in which the counter-current hot gases are passed through the materials and increase the temperature of the materials. The final product, which is known as cement clinker, gradually move through the rotary kiln to the bottom output and reaches to the cooling unit. The cement production process is illustrated in Figure 1. Inside the rotary kiln, the highest temperature of the raw materials reaches nearly 1500 C while the supplied flame temperature is up to 1800 C, which is roughly one-third of the surface temperature of the sun. The rotary kilns are commonly made of mild steel, which becomes weaken above 480 C. In cement plants, rotary kilns are monitored by remote temperature sensors to avoid such overheating problems and thermal deteriorations. The outer surface of the rotary kilns is exposed to the ambient to provide a natural heat rejection and provide a continuous production process. When the surface temperature of rotary kiln tends to increase up to a pre-defined upper limit, an additional cooling unit, generally an air blower unit, becomes activate to reject excessive amount of heat. Figure 1. Schematic representation of production process (Liu et al. 2015) In a recent report of The European Cement Association (Cembureau), it is revealed that the cement production sharply increases since 2001 (CEMBUREAU, 2015). Figure 2 shows the evolution of World cement production by regions. It is clear that the growth in production maintains in South America, Africa, and Asia in These areas produced the 4.4%, 4.8% and 80.4% of world cement production, respectively. According to the same report. China holds the first place in cement production with 2,438.0 million tons per year, which corresponds the 56.5% of the world cement production. European 176

191 Union countries and Turkey, on the other hand, hold positions at 3 rd and 6 th places, respectively with the production capacities of and 71.2 million tons per year, respectively. Figure 2. World cement production by regions (CEMBUREAU, 2015) Cement production is one of the most energy-consuming industries in the world. To produce one kilogram of cement clinker approximately kj of energy is required without considering the heat losses (Luo, 2015).In a real process, on the other hand, due to the significant amount of heat losses, the energy requirement increases up to kj/kg (Luo, 2015). Waste heat recovery units can be utilized to reduce the energy consumptions and improve the system efficiency. Engin and Ari (2005) assessed the possible heat recovery approaches for a dry type cement rotary kiln system. It is stated that the total energy loss of a cement plant is approximately 40% of the total input. Detailed thermodynamic analysis revealed that 19.15% is through hot flue gas, 5.61% is from cooler stack and 15.11% is due to the heat loss (convective and radiative losses) from the kiln surface. It is claimed that with a proper waste heat recovery design, 15.6% of energy could be recovered. Söğüt et al. (2010) developed a mathematical model to examine the performance of a heat recovery unit for the rotary kiln. It is found that 73% of waste heat could be recovered with the proposed heat exchanger geometry and transferred to the working fluid. Karamarković et al. (2013) stated that in a magnesium production company, the heat losses from the rotary kiln and exhaust gases are 26.35% and 18.95% of the input energy, respectively. To reduce the heat loss, they have proposed an annular duct heat exchanger. The annular heat exchanger can reduce the fuel consumption of the kiln by 12% and increases the energy and exergy efficiency of the system by 7.53% and 3.81%, respectively. Liu et al. (2014) established a combined analytical model that consist of raw material preheating & decomposition (I), clinker calcination (II) and clinker cooling processes (III). The impact of each process on the overall energy efficiency of the system is evaluated. It is found that process (I) has the highest impact on the 177 plant efficiency. It is followed by the process (III) and process (II). Caputo et al. (2011), on the other hand, analyzed a heat recovery unit to capture waste heat from the external surface of the rotary kiln. Parametric heat transfer and economic analyses have been conducted to find the optimum design and working conditions. Results confirm that the proposed heat exchanger model is appropriate both technically (efficient) and feasible (low-cost).wang et al. (2013) stated that in a cement production plant, 85% of total energy is consumed in a rotary kiln to produce the clinker. Due to convection and radiation effects on the external surface of the rotary kiln, the heat loss can reach up to 15%. They have conducted an experimental study for a prototype rotary kiln and examined the influence of working conditions on the performance of the heat recovery unit. It is found that increasing the temperature of the working fluid tends to reduce heat losses through the ambient. Recently, Luo et al. (2014) proposed a thermoelectric waste heat recovery unit to produce electricity directly from the temperature difference between the surface of the rotary kiln and the ambient. They have developed a mathematical model to predict the possible power generation and the savings by utilizing a waste heat recovery unit on a rotary kiln with dimensions of 4.8 m in diameter and 72 m long. The proposed unit produces approximately 221 kw electricity which corresponds nearly 32% of the heat loss through the kiln surface. In this study, a mathematical model is developed to predict the heat recovery from a rotary kiln under steady-state conditions. Comparative results are obtained by varying the surface temperature of the rotary kiln, mass flow rate of the working fluid, and also ambient conditions. II. Material and method III.1. Definition of the problem In the current work, the annual surface temperature variation of a rotary kiln is obtained from a cement production plant in Turkey. The outer diameter and the total length of the furnace are 4.8 m and 70 m, respectively. Considering the construction limitations on the plant site the length of the heat exchanger is decided to be 6 m. It is also known that the surface temperature of the kiln varies along the furnace. That is, parametric analyses have been conducted by varying the thermal boundary condition on the kiln so that the influence of the surface temperature on the useful (recovered) heat is obtained. The geometry of the proposed heat exchanger is given in Figure 3. The heat exchanger is designed as two half cylinders that are positioned around the kiln. The openings on the each side of the heat exchanger are intentionally designed to prevent overheating of the furnace. The surface temperature of the surface can be monitored remotely from the side opening and if the temperature reaches the upper limit, the blowers nearby the furnace are turn on to reduce the temperature. There is a total of 158 pipes

192 (carbon-steel) with an inner diameter of D in = 55.4 mm. It is proposed that there are two collectors to supply and collect the heat transfer fluid (HTF) to the pipes. The total mass flow rate of the HTF is varied in the range of 1 kg/s to 4 kg/s depending on the surface temperature of the kiln. The air gap between the furnace wall and the pipes is assumed to be Δr gap =15 cm. The outer face of the tubes is insulated (Δr ins =15 cm) to avoid heat loss through the ambient. calculated regarding the mass flow rate and the enthalpy variation of the HTF, q m h h (3) useful HTF out in HTF The heat lost from the outer surface of the insulation is composed of radiation and convection, q, A, T, T 4 4 lost sur ins ins outer ins outer sky h, total Ains, outer T ins, outer T (4) where sky temperature is defined according to the modified Swinbank equation (Hendricks & Sark, 2011) T T 0.32T (5) sky 1.5 (a) General view The total heat transfer coefficient (h,total) covers both the forced (by the wind) and natural (by buoyancy) components. The heat lost inside the air gap is expressed as q m c T T (6) lost, gap air air out in air To solve the Eqs. (1) to (6) useful heat rate and heat lost inside the air gap are defined regarding overall heat transfer coefficient. Useful heat rate is given as follows, (b) Close view Figure 3. Geometry of the proposed heat recovery unit III.2. Mathematical model & solution method Under steady-state working conditions, heat balance for the proposed system can be written as follows, q q q q (1) rad, kiln useful lost, gap lost, sur where q rad,kiln represents the net radiative heat transfer rate between the outer surfaces of furnace and pipes. q useful is the total rate of useful heat transferred by HTF. The last the terms on the right-hand side, on the other hand, represent the heat losses from the system. q lost, the gap is heat lost from kiln to the ambient air inside the gap and q lost,sur is the lost from the outer surface of the insulation. The rate of radiative heat exchange between two concentric cylinders is defined as (Cengel & Ghajar, 2011), q rad, kiln 4 4 Tkiln Tpipe, out Akiln 1 1 pipe Dkiln kiln pipe Dhx (2) Useful heat rate, on the other hand, can be 178 q useful UA T, (7) lm HTF where UA is calculated from the overall thermal resistance which is defined between the working fluid and the outer surface of the pipe. ΔT lm,htf is a log-mean temperature difference. Similarly, heat loss inside the air gap is defined in terms of the overall heat transfer coefficient as below, q UAT (8) lost, gap lm, air The numerical code is developed in Engineering Equation Solver (EES) software to solve the heat balance equations that are given in Eqs. (1) to (8). The efficiency of the heat recovery unit is defined regarding to the rate of useful heat to the rate of radiative heat transfer from the kiln surface as q useful (9) q rad, ki ln III. Results and discussions The chemical reactions inside the rotary kiln significantly affect the surface temperature of the furnace. Experimental measurements indicate that there are five different temperature zones along the length of the furnace: T kiln = 473 K, 523 K, 573 K, 623 K and 653 K. The ambient (T ) and sky (T sky)

193 temperatures vary throughout the year. That is, according to the monthly average temperature data of Izmir, Turkey (MGM, 2015), three different ambient temperatures are considered (T = 273 K, 283 K, 293 K and 303 K). Figure 4 represents the influence of ambient temperature on the performance of the heat recovery unit for T kiln = 573 K. In Figure 4(a), the variations of the outlet temperature of the HTF are given regarding the mass flow rate and also ambient temperature. It is clear that the outlet temperature decreases as the flow rate and outdoor temperature increase. Lowering the ambient from 303 K to 273 K, cause nearly 2 K reduction in outlet temperature of the HTF. On the other hand, the mass flow rate has a significant effect on the outlet temperature of the HTF. Increasing the flow rate from 2.7 to 3.5 kg/h the outlet temperature reduces nearly by 3 K. From Figure 4(b), one can see that the efficiency decreases from 61% to 54% by decreasing ambient temperature from 303 K to 273 K. On the other efficiency does not significantly influenced by varying mass flow rate for the selected parameters. Figure 4(c), indicates that increasing the flow rate from 2.7 to 3.5 kg/h the useful heat rate increases approximately by 3%. Besides, decreasing the ambient from 303 K to 273 K reduces the useful heat rate by 12%. (c) Useful heat rate Figure 4. Effect of ambient conditions on the performance of heat recovery unit Figure 5 shows the effect of surface temperature of the kiln on the performance of the heat recovery unit for T = 293 K. In Figure 5(a), the variations of the outlet temperature of the HTF are given as a function of mass flow rate and also furnace temperature. The outlet temperature increases as the flow rate decreases or the furnace temperature increases. The influence of kiln surface temperature becomes significant for lower values. At a flow rate of 1.5 kg/s, when the kiln surface rises from 473 K to 523 K the outlet temperature increases by 9 K. The outlet temperature reaches up to 400 K for T kiln = 653 K at the lowest flow rates. (a) HTF outlet temperature (a) HTF outlet temperature (b) Efficiency of the heat exchanger (b) Efficiency of the heat exchanger 179

194 (c) Useful heat Figure 5. Effect of kiln surface temperature on the performance of heat recovery unit Regarding the outlet temperature of the HTF, the difference between the highest and lowest furnace temperature cases is nearly 75 K at 1.5 kg/s. Increasing the flow rate lessen the temperature gap. At a flow rate of 3 kg/s the difference between the highest and lowest furnace temperature cases is less than 15 K. Furthermore, it is also evident that the effect of mass flow rate becomes significant at higher furnace temperatures. Increasing the flow rate from 1.5 kg/s to 5 kg/s reduces the outlet temperature by 35 K. Figure 5(b), shows that the efficiency slightly changes by varying mass flow rate but strongly influenced by the kiln temperature. At a flow rate of 5 kg/s, the efficiency of the heat recovery unit is obtained as 29%, 48%, 60%, 68% and 72% at T kiln = 473 K, 523 K, 573 K, 623 K and 653 K, respectively. Figure 5(c), on the other hand, depicts that the useful heat rate hardly depends on the furnace wall temperature. Increasing the furnace temperature from 473 K to 653 K, the useful heat rate improves more than 10 times. IV. Conclusions We have proposed a numerical model to predict the steady-state heat transfer of a heat recovery unit for a rotary kiln in a cement production plant in this study. We have considered an insulation thickness of 15 cm around the tubes while we have not included the heat bridges in the model. We have drawn the following concluding remarks from the results of the present study: a) The ambient temperature slightly affects the performance of the recovery unit. At lower ambient temperature heat lost through the environment increases. b) The useful heat rate recovered from the rotary kiln can reach up to 350 kw. c) The useful heat is recovered around the kiln could be directly used in the preheater of the furnace or could be integrated with secondary systems for providing heating or hot water to the cement plant. 180 d) In a cement production facility, the thermal management of the control system is crucial to ensure the continuity of the production line. e) The proposed recovery unit may also be integrated with an absorption cooling system to reject excessive heat from the control cabinet. f) The accuracy of the current model may be increased using geometry-specific Nusselt correlations. As a further study, the authors suggest performing an in-depth CFD (computational fluid dynamics) analysis to predict the effects of convection inside the air gap. Moreover, the heat transfer inside the tube can be calculated by dividing the tube into small segments, so that the variation of tube wall temperature along with the flow direction could be taken into account. References Caputo, A. C., Pelagagge, P. M., & Salini, P. (2011). Performance modeling of radiant heat recovery exchangers for rotary kilns. Applied Thermal Engineering, 31(14), CEMBUREAU, Activity Report 2014, The European Cement Association, May 2015 <retrieved from: Cengel, Y. A., & Ghajar, A. J. (2011). Heat and Mass Transfer: A Practical Approach, McGraw-Hill Education. Engin, T., & Ari, V. (2005). Energy auditing and recovery for dry type cement rotary kiln systems A case study. Energy conversion and management, 46(4), Hendricks, J. H. C., & Sark, W. G. J. H. M. (2013). Annual performance enhancement of building integrated photovoltaic modules by applying phase change materials. Progress in Photovoltaics: Research and Applications, 21(4), Karamarković, V., Marašević, M., Karamarković, R., & Karamarković, M. (2013). Recuperator for waste heat recovery from rotary kilns. Applied Thermal Engineering, 54(2), Liu, Z., Wang, Z., Yuan, M. Z., & Yu, H. B. (2015). Thermal efficiency modelling of the cement clinker manufacturing process. Journal of the Energy Institute, 88(1), Luo, Q., Li, P., Cai, L., Zhou, P., Tang, D., Zhai, P., & Zhang, Q. (2015). A Thermoelectric Waste-Heat-Recovery System for Portland Cement Rotary Kilns. Journal of Electronic Materials, 44(6), MGM, 2015, <retrieved from: Söğüt, Z., Oktay, Z., & Karakoç, H. (2010).

195 Mathematical modeling of heat recovery from a rotary kiln. Applied Thermal Engineering, 30(8), Wang, K., Du, W. J., & Cheng, L. (2013). Experimental Investigation on a Heat Recovery Device Installed on Cement Rotary Kiln. Applied Mechanics and Materials, 345,

196 SOLAR ENERGY 182

197 Experimental Analysis of Latent Thermal Energy Storage for Solar Heating Applications: Preliminary Results Onder Kizilkan 1*, Ahmet Kabul 2, Sefika Yildirim 3, Gamze Yildirim 4 1,2,3,4 Süleyman Demirel University, Faculty of Technology, Department of Energy Systems Engineering, 32260, Isparta, Turkey * onderkizilkan@sdu.edu.tr Abstract In this study, the preliminary results of an experimental study is given for latent thermal energy storage. For this aim, an experimental setup is built for solar assisted thermal energy storage using phase change materials. The system consists of two solar air collectors and a heating coil. The phase change material is selected to be Sodium Acetate Trihydrate and it is located in a small tank inside of the heating coil. During the experiments, the air is heated up by air collectors by the help of solar energy. Then it enters to the heating coil where it gives some amount of its thermal energy to the phase change material and the rest of the air is blew to the room. The results are showed that after sunset time, the room is heated for two hours using the latent heat of PCM. Keywords: Thermal energy storage, solar energy, phase change material, latent heat I. Introduction Technological developments and increase in world population rising energy consumption rapidly. Nowadays, energy production and sustainability are important issues for humanity. In the world, energy is mostly provided from fossil fuels. However the burning of fossil fuels brought the largest environmental issue ever, which is climate change caused by CO2 emission. On this occasion, scientists had begun to research in renewable energy technologies in order to turn the tide of climate change and achieve a sustainable development for human beings (Gok, 2010). Renewable energy sources are inexhaustible resource has the potential to be different and to renewable and conventional energy sources. Production of energy from this source due to the clean and environmentally friendly sources of renewable energy sources in our country and around the world is rapidly evolving. The most important energy source in renewable forms of energy is the sun. Solar energy is abundant, renewable and free energy source. But constantly to have wavy and intermittent power characteristics of solar energy in terms of electricity supply constitute some problems. Overcoming these problems and energy storage applications in order to increase the usage rate of renewable energy is on the agenda. Especially wind and solar origin of different energy storage methods are used for the purpose of performance enhancement in renewable energy sources. Solar energy potential of renewable energy sources, taking into consideration the heat without harming the environment with solar energy produced is used for heating and storing this energy phasechange material (PCM) in the night environment. 183 Energy storage plays important roles in conserving available energy and improving its utilization, since many energy sources are discontinuous in nature. Short term storage of only a few hours is essential in most applications, however, long term storage of a few months may be required in some applications. Solar energy applications require an efficient thermal storage. Thus, the successful application of solar energy depends, to a large extent, on the method of energy storage used (Khudhair et al, 2004). Thermal energy storage (TES) is one of the key technologies for energy conservation, and therefore, it is of great practical importance. One of its main advantages is that it is best suited for heating and cooling thermal applications. TES can contribute significantly to meeting society s needs for more efficient, environmentally benign energy use. TES is a key component of many successful thermal systems, and a good TES should allow little thermal losses, leading to energy savings, while permitting the highest reasonable extraction efficiency of the stored thermal energy (Dincer et al, 2002). Amongst above thermal heat storage techniques, latent heat thermal energy storage is particularly attractive due to its ability to provide high energy storage density and its characteristics to store heat at constant temperature corresponding to the phase transition temperature of phase change material (PCM). Phase change may be in the following form: solid solid, solid liquid, solid gas, Liquid gas. Phase Change Materials (PCM) is latent heat storage material. As the source temperature rises, the chemical bonds within the PCM break up as the material changes phase from solid to liquid (as is the case for solid-liquid PCMs which are of particular interest here). The phase change is a heat-seeking

198 (endothermic) process and therefore, the PCM absorbs heat. Upon storing heat in the storage material, the material begins to melt when the phase change temperature is reached. The temperature then stays constant until the melting process is finished (Sharma et al., 2004) A literature survey about energy storage using PCM show that there is an increasing interest on TES applications with PCMs. Bhargava (1983) examined a water heater utilizing a material which changes phase for storage of solar energy. Serale et al (2014) analyzed some of the thermo-physical and rheological properties and material behavior that interest flatplate solar thermal collectors with slurry PCM as the heat carrier fluid. Concepts of solar thermal systems filled with a slurry phase change material were proposed and a prototypal system as presented. Possible advantages and drawbacks of this technology was also discussed. Arjun and Hayavadana (2014) worked on thermal energy storage materials, development of PCMs, classification, working principle of PCMs and working of PCMs in clothing. The study also summarizes the evaluation of textiles containing PCMs and different applications. Guichard et al. (2015) studied a new configuration of a complex roof using PCM installed on a dedicated test cell. For the first time and in field conditions, an experimental device using phase change material was set up at Reunion Island, location having a tropical and humid climate. They concluded that in the tested configurations and for a non-ventilated air layer, the measured temperatures were on either side of the PCM's melting point. D'Avignon and Kummert (2016) conducted an experimental study carried out to assess the performance of PCM storage tank in various operating conditions in a dynamic test bench. Soares et al. (2016) evaluated the heat transfer through small thermal energy storage (TES) units filled with different phase change materials (PCMs): free-form and microencapsulated PCMs. They reported that the experimental results were very useful for benchmarking and validation of numerical models to be used in the design and optimization of new TES systems for buildings. Li et al. (2016) studied on the influences of thermal conductivity enhancers on heat transfer performance inside the PCM during the melting and solidification processes for solar chimney application. Heinz and Streicher (2006) investigated different ways of the integration of PCMs into a thermal energy storage. Different PCM materials, with and without enhancement of the thermal conductivity, were used, and their performance concerning the resulting charge/discharge power of a storage tank was tested experimentally. In this study, it is aimed to utilize solar based latent thermal energy storage technique for the heating of a laboratory which is located in Suleyman Demirel University, Isparta. For this goal, two air solar collectors are used for heating up the air and then it is used to transfer its heat energy to PCM which is 184 placed inside of a fan coil unit, inside the laboratory. Sodium Acetate Trihydrate is selected to be the phase change material for its good properties. The results given here are the preliminary outcomes of the experimental measurements. II. Phase Change Materials The aim to use PCMs for the purpose of storing thermal energy is to make use of the latent heat of a phase change, usually between the solid and the liquid state. Since a phase change involves a large amount of latent energy at small temperature changes, PCMs are used for temperature stabilization and for storing heat with large energy densities in combination with rather small temperature changes. The successful usage of PCMs is on one hand a question of a high energy storage density, but on the other hand it is very important to be able to charge and discharge the energy storage with a thermal power, that is suitable for the desired application. One major disadvantage of latent thermal energy storage is the low thermal conductivity of the materials used as PCMs, which limits the power that can be extracted from the thermal energy storage (Heinz and Streicher, 2006) Some of the important properties required for PCMs are; High latent heat of fusion per unit mass, so that a lesser amount of material stores a given amount of energy. High specific heat that provides additional sensible heat storage effect and also avoid sub cooling. High thermal conductivity so that the temperature gradient required for charging the storage material is small. High density, so that a smaller container volume holds the material. A melting point in the desired operating temperature range. The phase change material should be nonpoisonous, non-flammable and nonexplosive. No corrosiveness to construction material. PCM should exhibit little or no super cooling during freezing (Ravikumar and Srinivasan, 2008) III. System Description The solar assisted thermal energy storage system is designed for heating of a laboratory located at Suleyman Demirel University, in Isparta. The laboratory is 46.9 m 2 in area (width: 7 m, length: 6.7 m, height: 3 m) and it is aimed to be heated by the heat energy stored in the phase change material by solar energy. The schematic drawing of the test facility is given in Fig. 1.

199 For the dimensioning of the experimental system, the calculations are made first which are given in section IV. According to the calculations, the air speed of fan is found to be 7 m/s in the solar collector system so that it is kept constant during the experiments. Each channel diameter 100 mm, and the length of the fancoil unit is 1300 mm. In order to ensure sufficient heat, two solar air collectors are mounted (Fig. 2). Fig. 1: The schematic drawing of the test facility Fig. 2: The installation of experimental setup For the phase change material, Sodium Acetate Trihydrate (CH3COONa.3H2O) is selected for latent energy storage because of its good heat transfer properties. The PCM is located in a separated cylinder inside the fan-coil unit so that the heated air flows over the PCM and transfers its heat energy to it (Fig. 3). The properties of the Sodium Acetate Trihydrate is given in Tab. 1. Tab. 1: Properties of sodium acetate trihydrate Phase Solid Color colorless Density ~ 1.42 gr/cm 3 (20 C) Solubility ~ 613 g/lt (20 C,H 2O) Flashpoint temperature > 250 C (anhydrous) Melting point ~ 58 C ph ~ (50 g/lt,h 2O,25 C) 185

200 Re = m D A μ (9) where μ is the dynamic viscosity. In equation 9, A and D can be found from: A = s W (10) D = 2 s (11) Fig. 3: Sodium acetate trihydrate located inside the fan-coil IV. Thermodynamic Calculations For the thermodynamic calculations of the solar air collector, the mathematical formulation given in reference Kalogirou (2009) is used. Also, similar design calculations can be found in some other literature (Duffie and Beckman 2013; Tiwari 2003). The useful energy absorbed by the solar collector is defined as (Kalogirou 2009): Q u = A c F R [S U L (T i T a )] (2) where FR is the heat removal factor, S is the absorbed solar radiation, UL is the heat loss coefficient, Ti is the inlet air temperature, Ta is the ambient temperature. F R = m c p A c U L {1 exp [ U L F A c m c p ]} (3) where m is the mass flow rate of air, cp is the specific heat of air, F is the collector efficiency factor. where s is the depth of air channel, W is the collector width. h[1] = σ (T p+t b ) (T p 2 +T b 2 ) ( 1 ɛp )+( 1 ɛ b ) 1 (12) where, ɛ p is the emissivity of absorber plate, ɛ b is the emissivity of back plate, Tp is the temperature of absorber plate, Tb is the temperature of back plate. Fan selection is made and the air flow should be done in the laboratory for the design of the device must be known. Therefore intended to be heated the quantity of air required for laboratory; Q L = V m H d (13) where, Q L is the air flow required for laboratory, Vm is the volume of laboratory, Hd is the number of hour weather cycle For the pressure drop of the whole system, the formula given below is used: ΔP t = ΔP s + ΔP d = ( l R + Z) + P E (14) T 0 = T i + ( 1 U L ) [S U L (T i T a )][1 exp ( A c U L F m c p )] (4) where T0 is the exit air temperature, Ti is the inlet air temperature and Ta is the ambient temperature. S = G t (τ α) (5) Here Gt is the total insolation, (τ α) is the effective coefficient. In equation 4, F is given below: h F = (6) h+u L where h is the convection heat transfer coefficient. h = h[2] + ( 1 ( 1 h[1] )+( 1 h[2] )) (7) where h[1] is the radiation heat transfer coefficient from the absorber to the back plate, h[2] is the convection heat transfer coefficient. h[2] = ( k D ) (Re)0.8 (8) Here k is the heat transfer coefficient, D is the hydraulic diameter, Re is the Reynolds number. 186 Where, l is channel length (m), R is unit pressure drop (Pa/m), Z is pressure drop of fittings elements (Pa), PE is the total pressure drops (Pa) due to devices such as filters, heaters, measurement devices, etc. Additionally, ΔP s and ΔP d are the static and the dynamic pressure drops, respectively. V. Result and Discussion The experiment were made in June during a sunny day. The measurements are made for solar radiation intensity, outer temperature, indoor blowing air temperature at the exit of fan-coil and indoor temperature. In Fig. 1, the variation of solar radiation intensity is given with variation of time for Isparta region. As can be seen from the figure, the radiation intensity is about 650 W/m 2 at 11 am while it reaches to a maximum value of about 1100 W/m 2 between 1 pm to 3 pm. These values show that Isparta city has a great potential of solar energy. In Fig. 2, the variation of blowing air temperature at the exit of solar collectors is given with the variation of time. The first measurements have shown that the air temperature after the collectors was lower than the melting temperature of PCM. As seen from Fig. 2, maximum air temperature after the collectors was

201 measured about 44 C. This is obviously lower than the melting point of sodium acetate trihydrate (58 C). It was expected according to the calculations that, the air temperature at the exit of the collectors to be about 60 C. This unexpected result is due to improperly designed channels inside the collectors and also the number of collectors can be increased to 3. Another option to solve this problem in order to store the solar energy in PCM is to utilize a different PCM which has got a melting temperature around C. Fig. 3: Variation of outdoor temperature with time Fig. 1: Variation of solar radiation intensity with time Fig. 2: Variation of blowing air temperature with time In Fig. 3 and 4, the variation of outdoor and indoor temperatures are given with the variation of time, respectively. It is obvious from the figures that, the temperature difference after the sun rise is getting bigger when compared to the day time. From this situation, it can be said that in spite of not reaching to the melting temperature of PCM at the exit of air collectors, a small amount of sensible heat has been transferred to the ambient. Fig. 4: Variation of indoor temperature with time To ensure the effect of solar energy storage system after the sunrise, the measurements were continued till 9 pm in the evening. Figs. 5-7 show the variation of indoor, outdoor and blowing air temperatures between 7 pm 9 pm. As can be seen from the figures that after sun rise, the indoor temperature decreases from 28 C to 26 C while the outdoor temperature decreases from 23 C to 21 C. As mentioned earlier, this temperature difference is due to sensible heat storage. Fig. 5: Variation of outdoor temperature with time after sunset 187

202 References Arjun D., Hayavadana J., Thermal Energy Storage Materials (PCMs) for Textile Applications, Journal of Textile and Apparel Technology and Management, 8(4), 1-11, Bhargava A.K., A solar water heater based on phasechanging material, Applied Energy, 14(3), , Fig. 6: Variation of indoor temperature with time after sunset Duffie J.A., Beckman W.A., Solar Engineering of Thermal Processes, John Wiley & Sons Inc., New Jersey, USA, Gok O., Increasing Energy Efficiency In Dishwashers By Using Thermal Energy Storage In Phase Change Materials, PhD Thesis, Cukurova University, Turkey, D'Avignon K., Kummert M., Experimental assessment of a phase change material storage tank, Applied Thermal Engineering, 99, , Dincer I., Rosen M.A., Thermal energy storage, systems and applications. John Wiley and Sons, England, Fig. 7: Variation of blowing air temperature with time after sunset VI. Conclusions Utilization of thermal energy storage applications are being increased for sustainable use of energy. Latent heat thermal energy storage with PCMs is particularly attractive due to its high energy storage density amongst TES techniques. In this study, a solar assisted latent thermal energy storage application was established experimentally for heating of a laboratory at Suleyman Demirel University, Isparta. Sodium acetate trihydrate was used as PCM. The preliminary results of the experimental measurements showed that, the blowing air has been heated up to a maximum 44 C which is not adequate for melting of the PCM which has got a melting temperature of 58 C. There were two options discussed to solve the problem. The first one was to use 3 solar air collectors instead of 2 and the other one is to use a different PCM which has got a melting point range between C. The next step of this study will be mounting automatic control system to TES application and also a different PCM will be used. Guichard S., Miranville F., Bigot D., Damour B.M., Boyer H., Experimental investigation on a complex roof incorporating phase-change material, Energy and Buildings, 108(1), 36-43, Heinz A., Streicher W., Application of Phase Change Materials and PCM-Slurries For Thermal Energy Storage, Proceedings of the Ecostock Conference, USA, Kalogirou S.A., Solar Energy Engineering: Processes and Systems, Academic Press, Oxford, UK, Khudhair A.M., Farid M.M., A review on energy conservation in building applications with thermal storage by latent heat using phase change materials, Energy Conversion and Management, 45, , Li Y., Liu S., Shukla A., Experimental analysis on use of thermal conductivity enhancers (TCEs) for solar chimney applications with energy storage layer, Energy and Buildings, 116, 35-44, Ravikumar M., Dr. Pss. Srinivasan, Phase change material as a thermal energy storage material for cooling of building, Journal of Theoretical And Applied Information Technology, , Seralea G., Casconea Y., Capozzolia A., Fabriziob E., Perinoa M., Potentialities of a Low Temperature Solar 188

203 Heating System Based on Slurry Phase Change Materials (PCS), Energy Procedia, 62, , Sharma S.D., Kitano H., Sagara K., Phase Change Materials for Low Temperature Solar Thermal Applications, Res. Rep. Fac. Eng. Mie Univ., 29, 31-64, Soares N., Gaspar A.R., Santos P., Costa J.J., Experimental evaluation of the heat transfer through small PCM-based thermal energy storage units for building applications, Energy and Buildings, 116, 18-34, Tiwari G.N., Solar Energy: Fundamentals, Design, Modelling and Applications, Alpha Science International Ltd., Pangbourne, UK,

204 A review of Solar Energy Status in Iraq and Current Status Ahmed Emad*, Ahmet Kabul, Onder Kizilkan Süleyman Demirel University, Faculty of Tchnology, Department of Energy Systems Engineering, 32260, Isparta, Turkey * ahmed.emad846@gmail.com Abstract There is no secret how renewable energy is important and useful by increasing concern into improve its efficiency and utilize it as an alternative energy clean and sustain comparing with fossil energy and other energy recourses. the paper release how to utilize solar energy in Iraq depending on studies of sun radiation and lights incidence area, and annual hot weather temperature per a year, in addition study of changing climate,so we could use these previous studies to analyze and compare of using solar energy types, depends of results duration life time usage sustain and costs, this paper show different ways to use renewable solar energy in Iraq,and looking for finding a way to use a better kind of renewable energy with high efficiency and long life duration usage of power, whereas Iraq suffers from short age power usage and the effects of war led to collect all efforts to find a way to improve the power energy production. Keywords: Renewable energy, solar thermal energy, Concentrated solar power I. Introduction Renewable energy sources: are energy resources that are inexhaustible within the time horizon of humanity. Renewable types of energy can be subdivided into three areas: solar energy, planetary energy and geothermal energy. Even if the use of fossil fuels can be reduced significantly, and accepting that nuclear power is no long-term alternative, the question remains as to how the future supply of energy can be secured. The first step is to significantly increase the efficiency of energy usage, i.e. useful energy must be produced from a much smaller amount of primary energy, thus reducing carbon dioxide emissions. Renewable energies will be the key to this development, because they are the only option that can cover the energy demand of Earth in a climatically sustainable way (Quaschning, 2005). Now, Renewable energy become an abundant, wellestablished technology and the main ingredient is free. It is a well-known fact that eight countries have 81% of all word crude oil reserves, six countries have 70% of all natural gas reserves, and eight countries have 89% of all coal reserves. More than half of Asia, Africa and Latin America import over half of all their commercial energy. Most of these countries export crops that fetch low prices, but import energy at high prices, which leads to a drain on foreign exchange earnings. This problem is worsened by the fact that power generation is continuously increasing in these countries (Table 1) additionally, the world population keeps increasing at 1.3 2% per year, so that we are doubling our population every 60 years. Therefore in the year 2060, we expect our population will be in excess of 12 billion (Sari, 2004). Tab. 1: World total final consumption (Mtoe) (Source: (IEA, 2016) a Coal Oil Gas Electricity Heat Renewables Total final consumption a Average annual growth rate, in percent. The sun is an excellent source of radiant energy, and is the world s most abundant source of energy. It emits electromagnetic radiation with an average irradiance of 1353 W/m2 on the earth s surface. To put this into perspective, if the energy produced by 25 acres of the surface of the sun were harvested, there would be enough energy to supply the current energy demand of the world (Chaichan and Abaas, 2012). Recently, there is an increase needed for energy, especially electrical energy. Not only are oil prices increasing but pollution continues to rise due to the burning of fossil fuels, and the probability of oil supply depletion remains. All of these Issues encourage the investigation of using solar, wind and other Renewable energies for the generation of electrical power (Kazem and Chaichan, 2012), Renewable resources gained more attention in the last two decades due to persisting energy demand coupled with decrease in fossil fuel resources and its environmental effect to the earth. This paper reviewed and discussed the status of renewable energy used and developments in Arab countries especially in Iraq, how to apply and improve these systems as an essential element for the 190

205 sustainable economic development of these countries, despite their wealth in oil and gas. The paper presents a review of the renewable energy resources in Arab countries and sheds light on some achieved and/or ongoing renewable energy projects in the (table.2). Also show how Iraq suffers from electricity reduction, and many challenges will have to face the future increases in electrical prodution.as well as the impact of war, Years of underinvestment in the power sector. In Iraq, the electric power generated is not enough to meet the power demand of domestic and industrial sectors. Tab. 2: Solar-energy resources (kwh/m 2 /day) Country Solar Country Solar energy energy Algeria 5 7 Oman 5 6 Bahrain 5 8 Palestine 4 6 Egypt 5 9 Qatar 5 6 Iraq 5 6 Saudi Arabia 6 8 Djibouti 4 6 Sudan 5 8 Jordan 5 7 Somalia 6 9 Kuwait 5 8 Syria 5 6 Lebanon 4 6 Tunisia 5 7 Libya 5 7 UAE 5 6 Mauritania 6 Yemen 4 6 Morocco 5 7 II. Literatures overview By concerning of sustainable renewable energy in Arab countries especially in Iraq this paper refer to how this region could utilize available solar energy in real, associating with other reviews and studies of other researches. Quaschning (2005), described the most important technical systems for using renewable energy sources, and introduces important calculation and simulation methods for these. The main focus was on technologies with high development potentials such as solar thermal systems, photovoltaics and wind power. Sari, (2004), discussed the growing need of energy in both developed and developing countries, and the acute population growth. In the results, it was observed that renewable energy penetration into the energy market was much faster than was expected in recent years and by Chaichan and Abaas (2012), focused on the feasibility of improving concentrating solar power (CSP) planet efficiency, by manufacturing a diminished prototype. They investigated three states, coloring the central target with selective black color, enclose the target by a glass box and coloring the glass encloser by selective black color. The tests were conducted in Baghdad- Iraq summertimes weathers The results showed improvement in thermal storage when using the glass encloser. This improvement varied form month to month. Kazem and Chaichan, (2012), reviewed and discussed the status and future of renewable energy in Iraq. The uses of renewable energy sources, such as solar, wind and biomass have been reviewed. Concluding with recommendations for the utilization of these energy resources, investigated found that solar, wind and biomass energy were not being utilized sufficiently at 191 present, but according to them, these energies could play an important role in the future of Iraq s renewable energy. The authers showed aims to review and discuss the status and future of renewable energy in Iraq. The uses of renewable energy sources, such as solar, wind and biomass, have been reviewed. Chedid and Chaaban (2003), reviewed and discussed status of renewable-energy (RE) developments in Arab countries (AC) as an essential element for the sustainable economic development of these countries, despite their wealth in oil and gas. In conclusion, they have given a detailed status of RE developments in AC. Also RE resources were introduced, and success stories in selected countries were extracted and discussed. Valenzuela (2010), studied and simulated the viability of introducing a 50 MW solar power plant in the locality of Barletta, Italy. The study was divided in two main parts: The first theoretical one which was about solar energy and explained the two main processes with which profit could be taken from the sun: photovoltaic energy and thermal energy. In the second part, the 50 MW plant was studied and simulated to arrive to the final design. Dickes et al., (2014) have desigened a lab-s solar power plant and for experimental purposes and dynamic analysis. The test rig included an Organic Rankine Cycle (ORC) unit, a field of parabolic trough collectors and a thermal energy storage system presents the results of an experimental campaign conducted on the ORC module alone. This power unit, designed for a 2.8 kw net electrical output. Al-Karaghouli and Kazmerski (2010) addresses the need for electricity of rural areas in southern Iraq and proposes a photovoltaic (PV) solar system to power a health clinic in that region, analysis shows that the optimal system s initial cost, net present cost, and electricity cost, respectively. These values for the PV system are compared with those of a generator alone used to supply the load. Using the HOMER software computer model, determined that the most economic system for a remote health clinic in southern Iraq having a daily load. Ismael Mohammed Saeed et al (2016) worked at study the current and future energy issue such as the energy policy revolution, the power sector expansion strategy and the corresponding environmental impact in Iraq. It aims to introduce the capabilities of renewable and nuclear energy and deals with the environmental impact of renewable and nuclear energy, especially in greenhouse-gas mitigation. Also review the capabilities of renewable and nuclear energy in Iraq. Long-range Energy Alternative and Planning (LEAP) System was used to analyse the electricity generation including the future expansion plan and simulates the environmental impact for every technology used in electricity generation. III. Solar Energy Potential of Iraq Solar energy is a highly renowned alternative energy type. The intensity of the Sun s irradiation that reaches 92 billion the tons of globe petroleum. A calculation from 2002 states that the energy received from the sun in one hour was greater than the world

206 used in one year. Due to the latter, the investigation of this source has lasted for years and continues to be a matter of great importance and relevance today There are two main ways of taking profit of the energy, sun s which are photovoltaic plants and solar collector plants (Valenzuela, 2010). Iraq is well-known for long hours of sunshine. Studies have shown that Iraq receives more than 3000 hours of solar radiance per year in Baghdad alone. The hourly solar intensity varied between 416 W/m 2 in January to 833 W/m 2 in June. Even the hours of sunshine in Spain cannot compete with the levels observed in Iraq (Kazem and Chaichan, 2012). Iraq has excellent solar energy potential, ranging from 1800 to 2390 kwh/m2/yr of direct normal irradiation, and much of the flat Iraqi landscape is appropriate as shown in (Figure 1.) Also shown are the locations where the Ministry of Electricity plans to issue concessions for the 3,500 MW in total capacity via fossil-fuel power plants. Fig. 1: Iraq s solar irradiation (Source: The German Aerospace Center (DLR), Iraq Ministry of Electricity) There are two basic principles for converting solar energy. Photovoltaic systems convert solar energy directly into electrical energy. Tracking systems are used in concentrated photovoltaic (CPV) in particular. These use photovoltaic in power plants for the central energy supply in order to generate higher efficiency. Solar thermal systems have reflectors that concentrate the sun s rays in an absorber to heat a fluid that generates water vapor using a heat exchanger. The water vapor is used to drive turbines and generators as in conventional thermal power plants. Solar thermal Power plants include parabolic trough, Fresnel and solar tower power plants. Dish Stirling power plants operate differently here concentrated sunlight directly drives a Stirling engine that in turn drives a generator. One thing all these solar applications have in common is a single-axis or Double-axis tracking system that continuously aligns the reflectors with the course of the sun (Schaeffler, 2013). Iraq is one of the hottest countries in the world (with summer temperatures up to C), and summer temperatures are steadily increasing. About 50% of overall electricity demand is due to air conditioning, People in Baghdad, especially, are desperate to buy, and hopefully have enough electricity to use, air conditioners, as noted frequently in media. Lack of electricity during the critical summer months affects national productivity and makes it difficult to work in the stifling heat. As a result of the electricity shortages and demand for air conditioning, 90% of Iraqi households rely on some sort of diesel power generation operated by private independent operators (Figure2). 192

207 Fig. 2: Ad-hoc power distribution grids established by entrepreneurs who run diesel generators to feed demand when the grid is down (Undp, 2015) III.1. Photovoltaic (PV) systems Due to uniform distribution of solar radiation throughout Iraq, solar PV technology is suitable for producing electricity through-out Iraq. Solar PV technology is also suitable for off-grid electricity generation in power plants in rural desert areas. The efficiency of PV cells is influenced by high air temperature and dust contamination. Due to the dusty weather in Iraq, it is important to investigate the type of dust, density of dust, rate of accumulation of dust, and the effect of dust on the PV performance. Iraqi experiments using photovoltaic (PV) cells were unsuccessful. In Iraq, photovoltaic cells were used in community street lights but were unsuccessful because the cells had a low efficiency factor and Iraqi weather is characterized by dusty days (Figure 3 and 4). In addition, power from PV systems is currently more expensive comparing with large-scale concentrating solar power These factors acted and reduced the range of use of PV cells, though the cells did find limited application in individual home rooftop systems, community water pumping stations, and areas where the terrain makes it difficult to access the power grid (Kazem and Chaichan, 2012). Fig. 3: Solar street lights installed in Iraq Also Iraqi government started building dams in early 1950, to reduce the flood impact and developing the agricultural sector in the country. The company that constructed the first dam proposed to the Iraqi government to exploit the dam also for producing electricity, and the government accepted. This was a good step in electricity generation in Iraq. The first hydroelectric power station to supply the national grid network was the Dokan station in Initially, it applied 84 MW installed power and later increased it to 400 MW. Iraq has five main hydroelectric plants, with about total initial capacity of 2500 MW. The electricity produced by hydroelectric plants is affected by many factors, thus a continuous steady production is impossible. Among the factors is the height of stored water in the dam, which is decided by the rate of rainfall. Dams in Iraq are also used for irrigation; this means faster depletion of the stored water in drought years, which happened in Iraq frequently. Tables 3 and 4 show the currently operational dams and those under con-struction. Iraq aims to increase her installed capacity from hydroelectricity to 5500 MW after completion of these projects(saeed.et al,2016) Tab. 3: Dams with hydroelectric plants being constructed in Iraq and their maximum capacities (Movr, 2016) Name Location Capacity/MW Status Mosul Ninawa 1010 Active Haditha Anbar 660 Active Dokan Sulaimaniyah 400 Active Derbandikhan Hexagonal 240 Active Samara Salahuddin 75 Active Hamrin Dyala 50 Active Hindiya Karbala 15 Active Kufa Karbala 6 Active Total 2456 Fig. 3: Electrcity generation from PV sysmtems in Iraq 193

208 Tab. 4: Under-construction dams with hydroelectric plants in Iraq with expected capacity (Movr, 2016) Name Location Capacity/ MW Status Bekhme Erbil 1500 Under Construction Mandawa Erbil 620 Under Construction Taq Taq Erbil 300 Under Construction Al-Baghdadi Anbaar 300 Under Construction Badush Ninawa 171 Under Construction Bakerman Ninawa 24 Under Construction Total 2615 III.2. Solar thermal systems Among the different technologies being developed to this end, Concentrated Solar Power systems is a promising renewable technology. A standard CSP technology uses solar collectors and a tracking system in order to concentrate solar rays during sunshine hours. This concentrated beam is then absorbed and used as the heat source for a thermodynamic cycle. (Figure 4) illustrates the working principle of such system using parabolic trough collectors as sol receivers (Dickes et al., 2014). Fig. 5: Concentrating solar power distribution in the world (Trec, 2016) Fig. 6: Global hours of bright sunlight (Wikipedia, 2016) Fig. 4: Parabolic trough collectors as sol receivers Concentrated solar power (CSP) is expected to be very well suited to the long days of sunshine and the high temperatures found in Iraq (Figure 5-6). Investigations are underway in Iraq to improve the use of CSP during high temperature weather conditions. The use of solar water heater systems by domestic loads has increased. PV and/or CSP system implementations have shown that their efficiency and reliability depend on many factors, including orientation (longitude and latitude), environment (solar intensity, temperature, humidity, wind, dust, rain, pollution, etc.) And the PV technology used. Thus, before committing to a large-scale (in megawatts) PV or CSP project, a thorough investigation of the above factors is essential (Kazem and Chaichan, 2012). III.3. Hybrid Power Systems (HPS) Solar energy is copious, but capturing it is not cheap, which is the primary reason that solar power contributes only a tiny fraction of global energy production. Solar remains many times more expensive than power derived from fossil fuels, even as oil and natural gas prices rise. The intermittency issue could be solved by combining a solar thermal plant with a natural gas plant, enabling the solar power shortage during Iraq s winter months to be covered by gas-generated power. Maximum solar output during Iraq s intensely sunny summers, meanwhile, would coincide with peak demand and lessen the need for additional peaking capacity from gas. Colocating with a gas plant would also answer the solar plant s need for accessible transmission lines, a critical siting criterion for many renewable projects. Most importantly, combining a solar parabolic trough plant with a gas-powered plant would reduce costs because both can utilize the same steam turbine, generator, and associated equipment (Doyle and Jaafar, 2010). Hybrid power systems (HPS) are any autonomous electricity generating systems combining renewable energy sources and conventional generators. Winddiesel systems, which combine wind turbines and diesel generators are a subclass of HPS. The purpose of these systems is to produce as much energy as 194

209 possible from the renewable sources while maintaining an acceptable power quality and reliable supply. Furthermore, the fuel savings and lower generation costs obtained with the HPS should at least balance the high investment costs (Figure7) for renewable energy generators, controllers, dump loads, storage units, converters, etc. (Pereira, 2000). Financial support for studies that investigate renewable energy in Iraq and its applications is required. Introducing solar thermal collectors in public buildings to produce hot tap water can be considered a first step towards reducing dependence on fossil fuel resources. Focusing and working on using hybrid systems by combining renewable energy source with conventional generator its suitable efficiency comparing with climate change in this area and standalone renewable energy systems. Referemces Al-Karaghouli A., Kazmerski L., Optimization and lifecycle cost of health clinic PV system for a rural area in southern Iraq using HOMER software, National Renewable Energy Laboratory, 1617 Cole Blvd., Golden, CO 80401, USA, Fig. 7: current and potential future costs for largescale solar power systems in high-resource locations vs. fossil power Most of the rural areas in southern Iraq are still undeveloped and in a chaotic state after the invasion, and there is a need to provide these areas with electricity. Small standalone photovoltaic (PV) electrification systems can play a strategic role in the region s development. The region enjoys a huge amount of solar radiation during the entire year. Although capable of providing plentiful and reliable electricity, HOMER software developed by the National Renewable Energy Laboratory (NREL) to assist the design of micropower systems. HOMER is a computer model that simplifies the task of evaluating design options for both off-grid and grid-connected power systems for remote, stand-alone, and distributed-generation (DG) applications. HOMER s optimization and sensitivity analysis algorithms allowone to evaluate the economic and technical feasibility of a large number of technology options and to account for variation in technology costs and energy resource availability. HOMER models both conventional and renewable-energy technologies (Al- Karaghouli and Kazmerski, 2010). IV. Conclusions Chaichan M.T., Abaas K.I., Practical investigation for measurement of concentrating solar power prototype for several target cases at Iraqi summertime weathers, Anbar Journal for Engineering Sciences, 5(1), 76-87, 2012 Chedid R., Chaaban F., Renewable-energy developments in Arab countries: a regional perspective, Applied Energy 74, , Dickes R., Dumont O., Declaye S., Quoilin S., Bell I., Lemort V., Experimental Investigation of an ORC System for a Micro-Solar Power Plant, Proceedings of the 2014 Purdue Conferences, Purdue University, USA, Doyle P., Jaafar K., Iraq Has an Opportunity to Become a Solar Leader, Iraqi Solar, Developmentsarticle.pdf, IEA, International Energy Agency, Kazem H.A.,Chaichan M.T., Status and future prospects of renewable energy in Iraq, Renewable and Sustainable Energy Reviews, 16(8), , By all reviews and studies above it is concluded that that Iraq is suitable area for applying sustainable renewable energy so that, by study these reviews we can illustrate the factors that could be play an important rules to improve sustainable renewable energy in Iraq: The solar energy density in Iraq is among the highest in the world. Additionally, there is significant wind energy potential in several areas in Iraq. Government support is required for implementing small, renewable energy pilot projects, especially those that serve people in rural areas. 195 Movr, Ministry of Water Resources Iraq, Pereira A.L., Modular supervisory controller for hybrid power systems, Risø National Laboratory, Roskilde, _Modular_supervisory_controller_for_hybrid_power_ systems, Quaschning V., Understanding Renewable Energy Systems, Earthscan, United Kingdom, Saeed I.M., Ramli A.T., Saleh M.A., Assessment of sustainability in energy of Iraq, and achievable

210 opportunities in the long run, Renewable and Sustainable Energy Reviews, 58, , Sari A, Renewable energy for a clean and sustainable future, Energy Sources, 26, , Schaeffler Technologies AG & Co. KG Issued: 2013, April. Trec, Undp, United Nations Development Programme, NDPIQ_catalysing_solarenergy_ProDoc.docx/_jcr_c ontent/renditions/page.html, Valenzuela J, Performance of a 50 MW concentrating Solar Power Plant, Politecnico Di Bari University, Wikipedia, media/file:solargis-solar-map-world-map-en.png,

211 Effect of Solar - Geothermal Heat Exchanger Design and Fluid Type on the Thermodynamic Performance of a Power Plant Anil Erdogan 1*, Duygu Melek Cakici 1, Can Ozgur Colpan 2 1 Dokuz Eylul University, Graduate School of Natural and Applied Sciences, Mechanical Engineering Department, Izmir, Turkey 2 Dokuz Eylul University, Faculty of Engineering, Mechanical Engineering Department, Izmir, Turkey * anilerdogan1992@hotmail.com Abstract A design problem for shell and tube heat exchanger that combines a Parabolic Trough Solar Collector (PTSC) and Organic Rankine Cycle (ORC) was formed for the given fluids using their temperatures, pressures and mass flow rates. The design problem was modeled and solved using Engineering Equation Solver (EES). Using a thermal model for the PTSC, first, glasscover temperature of the PTSC was calculated, and then useful energy gain and the temperature of the fluid leaving the solar collector were found. The design problem of heat exchanger formed was solved using Logarithmic Mean Temperature Difference (LMTD) method; and heat transfer surface area, overall heat transfer coefficient, pipe side and shell side heat transfer coefficients, and the pressure drop along the heat exchanger were found. A parametric study was performed to study the effect of some of the important parameters (e.g. pipe diameter, pipe length, and baffle spacing) on the output parameters (e.g. overall heat transfer coefficient and pressure drop). The results show that R600 as the tube side fluid and Therminol VP1 as the shell side fluid give better performance. When the solar irradiation intensity changes between 800 and 2700 kwh/m 2, overall heat transfer coefficient decreases from 1579 W/m 2 -K to 1491 W/m 2 -K, heat transfer surface area increases from 7 m 2 to m 2, and pumping power increases from kw to kw. Keywords: Parabolic trough solar collector (PTSC), shell and tube heat exchanger, logarithmic mean temperature difference (LMTD), engineering equation solver (EES) I. Introduction Solar collectors absorb the incoming solar radiation and transfer the heat absorbed into a fluid (e.g. air, water, or thermal oil) circulating through the tubes of the collector (Mills, 2004). This fluid is then directly used or transfers its heat to another fluid for the desired purpose (e.g. production of hot water or steam). In order to reach high temperatures with high efficiency, Parabolic Trough Solar Collectors (PTSC) are generally preferred. These collectors are light structures and have low cost and used for process heat applications between 50 C and 400 C (Duffie and Beckman, 2013; Fernández-García et al., 2010). PTSCs are made of a reflective material sheet which has a form of parabolic shape. The schematic of a PTSC is given in Figure 1. As can be seen in this figure, the receiver consists of a metal tube (colored black) surrounded by a glass cover which is used to reduce the undesirable heat losses. PTSCs provide higher solar concentration levels according to flat plate collectors. Collector performance, which depends on design and material, is significantly affected by factors such as reflectivity, receiver, absorptivity, heat transfer fluid, and its flow rate tracking mechanism, among others (Kalogirou, 2004; Kalogirou, 2009). Tracking mechanism must be safe and able to follow the sun with certain degree of accuracy. It returns the collector to its original position at the end of the day. Also it should work during periods of sparse cloud cover. In addition, tracking mechanisms are used for protection of collectors. When the collectors turn, mechanisms protect from hazardous environmental and working conditions (e.g. storm, overheating, and the failure of working fluid) (Nuwayhid et al., 2001). Tracking mechanisms are divided into two categories, namely mechanical and electronic controlled systems. Electronic controlled tracking mechanisms are generally used for PTSCs. These system improved accuracy and reliability. Additionally, electronic controlled tracking mechanisms together with the sensor and computer controlled motors measure the solar radiance on the receiver (Islam et al., 2015). Fig. 1: Schematic of a PTSC (Modified from Islam et al., 2015) There are some studies on the PTSC design and modeling in the literature. Kalogirou et al. (1996) conducted a performance test of a PTSC according to 197

212 ASHRAE standards. The collector efficiency and incidence angle were measured and as a result of these measurements the collector s acceptance angle was obtained in the range of ±0.5. In addition, when the maximum error of the tracking mechanisms was ±0.2, the system worked continuously at the maximum possible efficiency. In the study by Odeh et al. (1998), efficiency of PTSCs was determined for its operation with Syltherm 800 oil and water as the working fluids. Absorber emissivity effects and internal working fluid convection effects were evaluated. An efficiency equation was developed and used in a simulation model to evaluate the performance of the system for different radiation conditions and different absorber tube sizes. Herrmann et al. (2004) proposed a concept, where a less expensive liquid medium such as molten salt is utilized as storage medium rather than the heat transfer fluid itself. Detailed performance and cost analyses were performed. The study concluded that specific cost for two tank molten salt storage is in the range of US $30-40/ kwh depending on storage size. Kalogirou et al. (2004) discuss the different types of solar collectors (flat plate, compound parabolic, evacuated, parabolic trough, Fresnel lens, heliostat field collectors) and their applications. Optical, thermal and thermodynamic analyses of the collectors, and a description of the methods used to evaluate their performance were given. Brooks et al. (2005) characterized the performance of PTSC. Low temperature testing was performed using water as the working fluid. Evacuated glass-shield and unshielded receiver were tested and it was found that the thermal efficiencies are 53.8% and 55.2%, respectively for these receivers. Brooks et al. (2005) developed a PTSC in a similar size to the small-scale commercial modules. In this study, the working collector length is 5 m, aperture width is 1.5 m, and rim angle is In addition, optical error analysis was conducted to estimate the intercept factor. In the study by Qu et al. (2006), a performance model of a solar collector based on a linear and tracking parabolic trough reflector was programmed using the software EES. The model included fundamental radiative and convective heat transfer, and mass and energy balance relations. Typical performance showed that when the hot water at 165 o C flows through a 6 m by 2.3 m PTSC with 900 W/m 2 solar insulation and 0 incident angle, the collector efficiency is estimated as 35%. Patnode et al. (2006) developed a model for a solar collector driven Rankine cycle using TRNSYS simulation program. Both the PTSC and Rankine cycle models were validated with the measured temperature and flow rate data of the SEGS VI plant for the years between 1998 and These models were used to evaluate the effects of solar field collector degradation, flow rate control strategies, and alternative condenser design on the plant performance. A heat exchanger is a device that is used to transfer heat between two or more fluids. Liquids flow in separate plates or tube surfaces and they do not leak. Heat exchangers are used in many different 198 engineering applications such as air conditioning, space heating, waste heat recovery, and geothermal power plant systems. Heat exchangers are typically classified according to the flow arrangement (e.g. parallel flow, counter-flow and cross-flow), number of fluids (one fluid or two fluids), and construction type (e.g. shell and tube, plate, and fin) (Kakaç et al., 2002; Shah and Sekulic, 2002). Shell and tube heat exchangers are preferred for space heating, power production, and chemical processing applications. The main advantages of this heat exchanger type over other types can be listed as follows: There is substantial flexibility regarding their materials to accommodate corrosion and other concerns. There is substantial flexibility regarding their materials to accommodate corrosion and other concerns; they can be used in systems with higher operating temperatures and pressures; and tube leaks are easily located and plugged since pressure test is comparatively easy (Perry et al., 1997; Subramanian, 2010a; Subramanian, 2010b). However, this heat exchanger requires more space and cleaning and maintenance is difficult since a tube requires enough clearance at one end to remove the tube nest. Shell and tube heat exchangers are classified and built according to the widely used Tubular Exchange Manufactured Association (TEMA) standards (Harrison, 2007). This type of heat exchangers differ according to the number of tube passes. The simplest form is one tube pass as shown in Figure 2a. Two tube passes and three tube passes configurations are shown in Figure 2b and Figure 2c, respectively. Baffles are often installed to increase heat transfer coefficient of the shell side fluid. Additionally, baffles are fixed to the tubes to reduce tube vibration. In TEMA standards, these heat exchangers are classified according to front end, shell types, and rear end head types. Some common types of front ends are Type A (head channel and removable cover) and Type B (integral cover), shell types are Type E (one shell pass) and Type F (two shell passes), rear end head types are Type U (U tube bundle) and Type M fixed tubesheet like B stationary head (Harrison, 2007; Perry et al., 1997). Fig. 2: Types of shell and tube heat exchangers (a) one tube pass, (b) two tube passes, and (c) three tube passes The thermal design of shell and tube heat exchangers is done according to the principles of thermodynamics, heat transfer, and fluid dynamics. As a result of this

213 design, shell types, flow arrangement, geometry of the heat exchanger, and tube and shell materials are determined for the specified heat transfer. The mass flow rate of shell and tube side fluids, inlet temperatures, and outlet temperature of one of the fluids are generally used as input parameters in this type of design problem. There are some studies on the design, modeling, and optimization of the shell and tube heat exchangers in the literature. For example, Kara et. al. (2004) created 240 alternative exchanger configurations and the computer program that they developed selects the optimum configuration among the all possible exchanger configurations. The shell diameter, baffle spacing, number of passes are the parameters that can be changed in this program. The program then determines the overall dimensions of the shell and the optimum heat transfer surface area required to meet the specified heat transfer duty by calculating minimum or allowable shell-side pressure drop. The results of their study showed that triangular tube pitch layout with one or two tube passes yields the best performance. Reppich et al. (1995) developed a computer based design model to determine the optimum dimensions of segmentally baffled shell and tube heat exchangers by calculating optimum shell side and tube side pressure drops from the equations provided in his work. The six optimized dimensional parameters are number of tubes, tube length, shell diameter, number of baffles, baffle spacing, and baffle cut. The proposed model also includes a cost analysis. Selbas et al. (2006) created a mathematical model of the heat exchanger, which was coded and solved in MATLAB environment. Dittus-Boetner equation and Bell and Delaware method were used to calculate the tube side heat transfer coefficient and shell side heat transfer coefficient, respectively. LMTD method was used to analyze the heat exchanger. They also performed a cost analysis to study the effect of six different variables (outer tube diameter, tube layout, number of tube passes, shell diameter, baffle spacing, and baffle cut) on the heat transfer surface area and shell side fluid velocity. In the paper by Walraven et al. (2014), the system optimization of different configurations of ORCs with both plate heat exchangers and shell-andtube heat exchangers were compared. They also created a mathematical model for integrated ORC systems, and compared the performance of the systems when different fluids are used. Their study showed that ORCs with all plate heat exchangers perform mostly better than ORCs with all shell-andtube heat exchangers. Different tube configurations were investigated in the paper by Walraven et al. (2014). Their study concluded that the 30 tube configurations should be used for the single phase heat exchangers and the 60 tube configurations for the two phase heat exchangers. Fettaka et al. (2013) developed a mathematical model of the shell and tube heat exchangers, and nine different optimization works were carried out by using different geometric parameters (tube layout, number of the tube passes, baffle spacing, baffle cut, tube to baffle diametrical clearance, tube length, tube outside diameter, tube wall thickness, and shell to baffle diameter clearance). 199 Taguchi method is a commonly used experimental design method based on statistical approaches to determine the optimum conditions for a given problem. In recent years, the use of Taguchi method in the world has increased rapidly, especially in the industry. This method was developed and pioneered by the Japanese engineer, Dr. Genichi Taguchi (Ghani et al., 2004). This technique suggests a simple and systematic way to optimize design for performance quality and cost. Two important criteria used in Taguchi design are signal to noise (S/N), which measures quality; and orthogonal arrays, which establish many design factors simultaneously. Although this method has been generally used for experimental studies, it has been shown that it could be effectively applied to mathematical modeling studies (e.g. Ghani and Choudhury, 2004; Sasmito et al., 2015; Yang and Tarng, 1998). The literature survey discussed above shows that studies that include the shell and tube heat exchanger design for PTSC systems are limited; and there is no performance or optimization study on shell and tube heat exchangers that combine a geothermal brine fueled ORC and a PTSC. In this study, a shell and tube heat exchanger that combines a PTSC and ORC systems was designed. For this purpose thermal models for the PSTC and heat exchanger were first developed. Then, a parametric study was conducted to study the effect of important design parameters (e.g. outer diameter, tube length, baffle spacing, number of passes, and tube and shell side fluids) on the output parameters (e.g. the overall heat transfer coefficient and the total pressure drop). In addition, an optimization study was carried out using Taguchi method to find the optimum design parameters that maximizes the performance and minimizes the cost. In this method, six key parameters were selected such as tube diameter, tube length, baffle spacing, number of passes, shell side fluid, and tube side fluid. Three values for each parameters were taken for the first stage of Taguchi method. Then according to the results found, a further refinement of the optimum parameters was done using second stage Taguchi analysis. In addition, the effect of solar irradiation intensity on the optimum parameters was assessed. II. Mathematical Modeling As the objective of this study is to design and optimize a shell and tube heat exchanger that combines a PTSC and an ORC system under different solar irradiation intensity, mathematical models for both PTSC and heat exchanger are developed. The schematic of the integrated system is shown in Figure 3. The modeling approach and equations for both of these components are given in detail in the following subsections.

214 Fig. 4: Geometric properties of the PTSC Fig. 3: The schematic of the integrated system consisting of a PTSC, an ORC, and a shell and tube heat exchanger. II.1. Mathematical Modelling of PTSC In this section, the modeling approach and equations for the PTSC are given. The aim of the PTSC model developed is to find the exit temperature of the PTSC system for a given set of input parameters consisting of the design and operating parameters of PTSC as well as the meteorological data. Modeling of PTSC discussed in this section is based on the approach and equations given in the Refs. (Duffie and Beckman, 2013; Kalogirou, 2009). The rate of useful energy delivered by a single collector can be found using the Eq. 1. Q u = F R (SA a A r U L (T i T a )) (1) Where F R is the heat removal factor, S is the heat absorbed by the receiver, Aa is the aperture area, Ar is the receiver area, UL is the solar collector overall heat loss coefficient, Ti is the entering fluid temperature, and Ta is the ambient temperature. Then, the heat absorbed by the receiver is defined as: S = G b η r (2) Where G b is the direct irradiation intensity and η r is the receiver efficiency, which can be found using Eq. (3). η r = ργταk (3) Where ρ is the reflectance of the mirror, γ is the intercept factor, τ is the transmittance of the glass cover, α is the absorptance of the receiver, and K is the incidence angle modifier. Geometric properties of the PTSC are shown in Fig. 4. Where m is the mass flow rate of the collector fluid and c p is the heat capacity of the collector fluid. F is the collector efficiency factor defined as: F = 1 U L 1 + D o +( D o U L h fi D i 2k +lnd o ) D i (5) Where D i and D o are the inner and outer diameters of the receiver, respectively. These diameters and the diameter of the glass cover (Dg) are shown in Figure 4. In this equation, k is thermal conductivity of the receiver tube. h fi is the heat transfer coefficient inside the receiver tube, which can be calculated using Eq. (6). h fi = Nu k fi D i (6) Where Nu is the Nusselt number of the fluid flowing through the receiver. If the flow inside the receiver tube is turbulent (Re>2300), Nusselt number can be evaluated using the following correlation (Duffie and Beckman, 2013): Nu = 0.023(Re 0.8 )(Pr 0.4 ) (7) Where Re is Reynolds number of the flow inside the receiver, Pr is the Prandlt number of the collector fluid. If the flow is laminar, (Re<2300) Nusselt number is taken as constant as The solar collector heat loss coefficient is defined as: U L = ( 1 A r ) (h c,g a +h r,g a )A g (8) Where A r is receiver area, A g is glass cover area. A r and A g can be found as: A r = πd o L PTSC (9) A g = πd g L PTSC (10) The heat removal factor is given by: F R = m h c p (1 exp ( U L F A r )) (4) A r U L c p m 200 Fig. 5: Schematics of receiver tube and its diameters.

215 Aperture area, length of solar collector and width of the PTSC are shown in Fig. 6. E b = εσt s 4 (17) Where ε is a radiative property of the surface termed the emissivity. The net rate of radiation heat transfer from the surface is: " q rad = εσ(t 4 s T 4 sur ) (18) Where T sur (T sur = T a ) is surrounding temperature. Then, the net radiation heat transfer is in following the form: " q rad = h r (T s T sur ) (19) Then, from Eqs. (20) and (21), the radiation heat transfer coefficient, h r,g a, for the glass cover to the ambient is calculated as follows. Fig. 6: Design parameters of the PTSC The convection heat transfer coefficient, h c,g a, for the glass cover to the ambient can be calculated as: h c,g a = Nu k D g (11) Where Nu is the Nusselt number of air and k is the thermal conductivity of air. Before calculating the Nusselt number, temperature of glass cover, Tg, is assumed to be close to the ambient temperature. Then, the average temperature ( T ave ) is found between the ambient temperature and glass cover temperature. T ave = T a+t g 2 (12) Then, Reynold number (Re) is calculated according to the average temperature. εσ(t g 4 T a 4 ) = h r (T g T a ) (20) h r,g a = ε g σ(t g + T a )(T g 2 + T a 2 ) (21) Where ε g is the glass cover emittance and σ is the Stephan-Boltzman constant. T g is the glass cover temperature and T a is the ambient temperature. These temperatures are shown in Fig. 7. The net radiation exchange between two black surfaces is given by Q 12 = A 1 F 12 σ(t 1 4 T 2 4 ) A 2 F 21 (T 1 4 T 2 4 ) (22) Where T 1 and T 2 are the temperatures of surfaces associated with the surfaces A 1 and A 2. The term 1 = 1 A 1 F 12 A 2 F 21 represents the resistance due to the geometric configuration of the two surfaces. If surface is a real surface, ε = α, and ρ = 1 α = 1 ε. Re = ρ V D g μ (13) Where ρ is the density of air, V is the wind velocity, D g is glass cover outer diameter, and μ is the kinematic viscosity of the air. Then, Nusselt number is calculated as follows. Nu = (Re) 0.52 (0.1 < Re < 1000 (14) Nu = 0.3(Re) 0.6 (1000 < Re < 50000) (15) The maximum emissive power at a given temperature is the blackbody emissive power (E b ). Emissive power is prescribed by the Stefan-Boltzmann law. E b = σt s 4 (16) Where T s (T s = T g ) is the surface temperature and σ is the Stefan-Boltzmann constant. A black body is an ideal emitter. But the energy emitted by a real surface is less than a black body at the same temperature and is defined as follows. 201 Fig. 7: Different temperature types in the receiver tube The general form of radiation heat transfer coefficient, h r, between the receiver tube and the glass cover is defined as: h r = σ(t 1+T 2 )(T 2 1 +T2 2 ) 1 ε1 +A 1 A2 ( 1 (23) ε2 1) Where T 1 is the receiver temperature, T 2 is the glass

216 cover temperature, ε 1 is the receiver emittance, ε 2 is the glass cover emittance, A 1 is the aperture area, and A 2 is the glass cover area. For our case, the subscript 1 and 2 are replaced by g, and r. h r,r g can be calculated as follows. h r,r g = σ(t g+t r )(T g 2 +Tr 2 ) 1 εr +A r Ag ( 1 εg 1) (24) Finally, since U L is based on the assumed T g value, we need to check if the assumption made was correct. Therefore, T g can be obtained from energy balance: T g = A rh r,r g T r + A g (h r,g a +h c,g a )T a A r h r,r g +A g (h r,g a +h c,g a ) (25) To find the exit temperature of the PTSC (or the inlet temperature of the heat exchanger for the shell side), an energy balance around the PTSC should be applied, as shown in Eq. (26). T h,i = T i,oil + Q u (26) m PTSC c p,oil II.2. Mathematical Modeling of a Shell and Tube Heat Exchanger The modeling of the heat exchanger is done using the LMTD method. Using this model, a design problem is formed to calculate the overall heat transfer coefficient, the heat transfer surface, the pressure drop across the heat exchanger and the pumping power. The following assumptions are made in this model: The heat exchanger runs under steady-state conditions. Heat losses to the surroundings are negligible. The temperature of each fluid at the inlet and exit is uniform across the cross-sectional area of the shell and tube. Changes in the kinetic and potential energies of the flowing streams from inlet to exit can be neglected. The temperature change of the fluid between the inlet to exit can be considered negligible. The temperature of the ORC side fluid can be found applying an energy balance around a control volume enclosing the heat exchanger as shown in Eq. (27). material conductivity, and R" fi and R" fo are the tube side and shell side fouling factors, respectively. Fouling factor is found according to the fluid types shown in Table 1: Table 1. Fouling Factors for different fluid types (Harrison, 2007) Fluid R" f (m 2 K/W) Sea water and treated boiler feedwater (below 50 C) Sea water and treated boiler feedwater (above 50 C) River water (below 50 C) Fuel oil Refrigerant liquids Steam (non-oil bearing) Firstly, velocity of the fluid in the inner tube is found for calculating h i (heat transfer coefficient of the tube side). V i = 4 m N π d i 2 ρ T npass (29) m is the mass flow rate of tube side fluid, N T is the number of tubes and n pass is the number of passes. Then the Reynolds number is calculated using the following equation. Re = ρ V i D i μ (30) According to Eq. (30), flow regime is identified as either laminar or turbulent. According to the identified regime, Nusselt number is calculated using the equations given in Table 2. Table 2: Nusselt number for laminar and turbulent flow in a circular tube (Kakaç, et al., 2002) Regime Equation Condition Laminar Nu = 4.36 Re 2300 f (Re 1000)Pr Nu = 8 Turbulent ( f 1 8 ) < Re (Pr 3 1) < 5x10 4 Where f = (0.79ln (Re) 1.64) 2 After finding the velocity and the Nusselt number for the inner tube as discussed above, heat transfer coefficient h i can be calculated using Eq. (31). h c,o = h c,i + Q m c (27) h i = Nu k d i (31) To perform the heat transfer analysis of an heat exchanger, the major parameters are the heat transfer rate q, heat transfer surface area A, overall heat transfer coefficient U, and cold and hot fluid inlet and outlet temperatures. Overall heat transfer coefficient is calculated as follows: 1 = D o 1 + ln(d o D i ) U o D i h i 2πkL + 1 h o + D o D i R" fi + R" fo (28) Where D i and D o are the inner and outer diameters of the tube, respectively; h i and h o are the tube side and shell side heat transfer coefficients, k is the tube 202 For calculating the shell side heat transfer coefficient (h o ), equivalent diameter should be identified firstly. If the pipe alignment is triangular (Kakaç et al., 2002): D e = 1.27 d o (P t d o 2 ) (32) If pipe alignment is square (Kakaç et al., 2002): D e = 1.10 d o (P t d o 2 ) (33)

217 Where P t is the tube pitch and this parameter changes according to the tube diameter (Harrison, 2007). Cross sectional area of the shell perpendicular to the flow direction A s, can be calculated as follows. A s = (P t d o ) e D s P t (34) Where D s is the shell diameter, e is the baffle spacing. Shell diameter is calculated using Eq. (35) D s = CL 1 d o 2 CTP [A(PR)2 ] L (35) CL and CTP depend on the geometry of heat exchanger as shown in Table 3. Table 3: CL and CTP Values (Kakaç et al., 2002) Tube Layout Angle CL For 45 o and 90 o 1 For 30 o and 60 o 0.85 Number of Passes CTP One Pass 0.93 Two Passes 0.9 Three Passes 0.85 Velocity of the shell side fluid is can be found as follows m V o = (36) ρ A s Reynolds number of the shell side is: Re = ρ V o D e μ (37) To calculate the heat transfer coefficient of the shell side, ho, Nusselt number should be first calculated using the equation developed by McAdams, which is shown in Eq. (38) (Kakaç et al., 2002) Nu = h od e k = 0.36 Re 0.55 Pr 1 3 ( μ b μ w ) 0,14 (38) Where μ b and μ w are the kinematic viscosities of shell fluid at bulk temperature and wall temperature, respectively. Wall temperature is defined as: T wall = T b,h+t b,c 2 (39) Where T b,h, and T b,c are the average temperatures of hot and cold fluids between the inlet and exit. In this study, for modeling and analysis of shell and tube heat exchanger, Logarithmic Mean Temperature Difference Method (LMTD) is used. For a counter flow type heat exchanger, the logarithmic mean temperature difference is calculated as follows. T lm = (T h,i T c,o ) (T h,o T c,i ) ln (T h,i T c,o) (T h,o T c,i ) (40) Where T h,i and T h,o are the inlet and outlet temperatures of hot and cold fluid; T c,i and T c,o are inlet and outlet temperatures of cold fluid respectively. The effective mean temperature difference: T m = F T lm (41) Where F is the correction factor for the heat exchanger and the figures to find this factor can be found in Ref. (Kakac et al., 2002). In this figures, F is a function of P and R, which can be calculated using Eqs. (42) and (43). P = (T c,o T c,i ) (T h,i T c,i ) R = (T h,i T h,o ) (T c,o T c,i ) (42) (43) Finally, heat transfer rate is calculated using Eq. (44). q = UA T lm (44) Where A is the heat transfer surface area. This term can be calculated as follows. A = πd o LN T (45) Where L is the tube length and N T is the number of tubes. Pressure drop across the heat exchanger and pumping power are important parameters in the design of a heat exchanger. The pressure drop in the shell side of a heat exchanger can be calculated as follows (Shah and Sekulic, 2002). P shell = f 2 fric G shell (N b +1) D shell (46) 2 ρ D e φ shell Where G shell is the mass velocity of shell side fluid, N b is the number of baffles, f fric is the friction factor and φ shell is the viscosity ratio between bulk and wall temperatures. These parameters are calculated as follows. f fric = exp ( ln(re)) (47) G shell = m A s (48) φ s = ( μ b μ wall ) 0.14 (49) N b = ( L + 1) (50) e To calculate the pressure drop across the tubes of the PTSC, head loss due to frictional losses in the pipe is first calculated using Eq. (51). h l = f L d V 2 2g (51) 203

218 f is the friction factor, which can be found from the Moody chart (White, 2009). The pressure change across the PTSC tubes for a single row can be found as follows. P PTSC = P o P i = h l ρg (52) The total pressure including both the PTSC and the shell side of the heat exchanger can be found as shown below. Finally, the pumping power can be calculated using Eq. (53). W pump = m vδp total η pump (53) Where η pump is the pump efficiency. ΔP total can be calculated using Eq. (54): ΔP total = P PTSC + P shell (54) III. Results and Discussion In this section, the results and discussion on the parametric studies are presented. The parametric studies include the effect of outer diameter of the heat exchanger tube, tube length, baffle spacing, number of passes, and tube layout angle on the performance parameters including number of tubes required, overall heat transfer coefficient, and total pressure drop. It was considered that in the solar farm, there are 5 rows, and there are 8 modules of PTSC in each row. The type of the collector selected is SkyTrough collector (SkyFuel, 2009). The geometric data of PTSC is given in Table 4. Table 4: Geometric data of SkyTrough PTSC (SkyFuel, 2009) Single collectors width 6 m Single collectors length 14 m Receiver inner diameter m Receiver outer diameter m Glass cover diameter m Transmissivity of the receiver 0.94 Absorptivity of the receiver 0.97 Reflectivity of the aperture surface 0.96 Intercept angle 1º Receiver emittance 0.92 Glass cover emittance 0.87 Intercept factor 0.93 In this study, some assumptions were made as follows: The receiver temperature is taken T r = 300 C Entering temperature of heat transfer fluid is taken T i = 290 C The dead state properties are taken as T a = 25 C and P a = kpa As the baseline condition for the simulations, one shell and one tube pass type heat exchanger is used in the system shown in Fig. 2. The input parameters of the heat exchanger model for the baseline conditions are 204 given in Table 5. The parametric studies were conducted considering the minimum and maximum values for outer diameter of the tube, tube length, and baffle spacing considering the standards given in TEMA standards (Harrison, 2007). These values are shown in Table 6. Number of passes is taken as 1, 2, or 3; whereas tube layout angle is taken as 30, 45, 60, or 90. Table 5: Input parameters of the heat exchanger model for the baseline conditions Name of the parameter Value PTSC (Shell) side fluid Therminol VP1 ORC (Tube) side fluid R134a Number of passes 1 Tube length 12 m Inner diameter of the tube m Wall thickness m Tube pitch m Tube layout angle 90 Baffle spacing 1.5 m Thermal conductivity of tube 63.9 W/m K Mass flow rate of cold fluid kg/s Mass flow rate of hot fluid 27.5 kg/s Hot fluid pressure kpa Cold fluid pressure 4370 kpa Outlet temperature of the hot fluid 563 K Inlet temperature of the cold fluid K Table 6: The values of the design parameters used in the parametric study Outer Diameter of the Tube (m) Tube Length (m) Baffle Spacing (m) Minimum Maximum III.1 The effect of outer diameter of the tube on the performance of the system The effect of outer diameter of the tube on the performance of the system was assessed using the data given in Table 5. The different values for this diameter were taken from TEMA standards (Harrison 2007). In TEMA standards, tube diameters have 10 values, which are m, m, m, m, m, m, m, m, m, and m. Using the code developed in EES, a parametric study was conducted for these values for the outer diameter of the tube. As a result of this study, the change of overall heat transfer coefficient, and pressure drop with respect to this diameter was found. The parametric study was repeated for different number of passes (1, 2 and 3) and layout angles (30, 45, 60, and 90 ), and the results for these studies are shown in Figures 7 and 8, respectively. Fig. 8a and 8b show the change of overall heat transfer coefficient, and total pressure drop for different values of outer diameter of the tube and the number of passes, respectively. The results show that increasing the outer diameter of the tube, the overall heat transfer coefficient fluctuate. It can be seen that taking this parameter as small as possible ( m), the overall heat transfer coefficient required get its maximum value. The reason of this trend can be attributed to change the flow regime. On the other

219 hand, as it can be seen from Figure 8b, the effect of this parameters on the total pressure drop is more significant when the outer diameter of the tube is less than m. When this diameter is m, the total pressure drop is at its maximum values. This finding can be explained as follows. As the outer diameter increases, the equivalent diameter (Eq. 32) increases and thus pressure drop decreases. These figures also show that the number of passes does not have a significant effect on the results but it only changes the number of tubes required. Taking the number of passes as 1 yield slightly lower overall heat transfer coefficient compared to a heat exchanger with 2 or 3 tube passes. (b) Fig. 9: The effect outer diameter of tube on the (a) overall heat transfer coefficient, (b) total pressure drop for different tube layout angles III.2 The effect of tube length on the performance of the system (a) Ten different values of tube lengths can be found in TEMA standards. These values are m, m, m, m, m, 7.32 m, 8.53 m, 9.75 m, 10.7 m, and m (Harrison, 2007). Fig. 10a and 10b show the change of the overall heat transfer coefficient and the total pressure drop for different values of tube length respectively. The results show that increasing the tube length, the overall heat transfer coefficient increases. This trend may be due to the change in the flow from laminar to turbulent when the tube length increases; and thus the tube side heat transfer coefficient increases. When the tube length is taken as high as possible (11.58 m), the overall heat transfer coefficient gets it maximum value. (b) Fig. 8: The effect outer diameter of tube on the (a) overall heat transfer coefficient, (b) total pressure drop for different number of passes. Fig. 9a and 9b show the change of overall heat transfer coefficient, total pressure drop for different values of outer diameter of the tube and the tube layout. The results show that increasing the outer diameter overall heat transfer coefficient, and total pressure drop fluctuate. If tube layout angle is selected 30 or 60 instead of 45 or 90, both overall heat transfer coefficient and total pressure drop increase. (a) (b) Fig. 10. The effect of the tube length on the (a) overall heat transfer coefficient, (b) total pressure drop for different number of passes. (a) Figs. 11a and 11b show the overall heat transfer coefficient and total pressure drop for different values 205

220 of tube length and layout angle, respectively. These figures present that when the tube layout angle is 30 or 60, the overall heat transfer coefficient, the total pressure drop are higher. The tube layout angle depends on the tube layout constant. If tube layout angle is 30 or 60, the value of CL and shell diameter are lower than the values of those when the tube layout angle is 45 or 90. Thus, shell side fluid velocity, overall heat transfer coefficient, total pressure drop slightly increase. where the baffle spacing is 0.32 m and then this parameters almost remain constant with a further increase in the baffle spacing. This result is mainly due to the change in the flow regime. When the baffle spacing changes between 0.08 m to 1.5 m, the Reynolds number decreases from to 72189; hence the flow regime gets closer to laminar flow when we increase the baffle spacing. (a) (a) (b) Fig. 11. The effect of the tube length on the (a) overall heat transfer coefficient, (b) total pressure drop for different tube layout angles. III.3 The effect of baffle spacing on the performance of the system Minimum and maximum baffle spacing values are shown in Table 6. Minimum baffle spacing is found using the TEMA standards (Harrison, 2007); whereas the maximum baffle spacing is taken as 29.5d o 0.75 where d o is in meters (Fettaka et al., 2013). (b) Fig. 12. The effect of the baffle spacing on the (a) overall heat transfer coefficient, and (b) pressure drop for different number of passes. Figs. 13a and 13b show the change of overall heat transfer coefficient, and total pressure drop for different values of baffle spacing and tube layout angle, respectively. If tube layout angle is 45 and 90, overall heat transfer coefficient decreases. It can also be seen from the Figs. 13b that the effect of tube layout angle on the pressure drop is not significant. Figs. 12a and 12b show the change of overall heat transfer coefficient, and total pressure drop for different values of baffle spacing and number of passes, respectively. The results show that increasing the baffle spacing decreases the overall heat transfer coefficient. This trend may be due to the fact that as the baffle spacing decreases, the effect of the turbulence and thus the shell side heat transfer coefficient decrease. When the baffle spacing is 0.08 m, overall heat transfer coefficient is at its maximum value (1040 W/m 2 K). On the other hand, if the baffle spacing is 1.5 m, overall heat transfer coefficient takes its minimum value (115 W/ m 2 K). These results show that the baffle spacing should be selected as low as possible (0.08 m). The effect of number of passes on the results seems to be negligible. Fig. 12a and 12b shows that increasing the baffle spacing first sharply decreases the total pressure drop up to the point (a) (b) Fig. 13. The effect of the baffle spacing on the (a) heat transfer surface area (b) pressure drop for different tube layout angles. IV. Conclusions In this study, the design of a shell and tube heat 206

221 exchanger, which combines a PTSC and an ORC, was done by applying the principles of thermal sciences. For this purpose, thermal models of the PTSC and heat exchanger were first developed and then solved using Engineering Equation Solver for a case study. Parametric studies were conducted to find the effect of some of the key design parameters on the output parameters of the model such as overall heat transfer coefficient and the total pressure drop. The main conclusions derived from the parametric studies are listed below. Outer tube diameter does not have significant effect on the pressure drop. However, increasing the diameter decreases the overall heat transfer coefficient. When the tube length increases, the overall heat transfer coefficient increases; but the pressure drop increases. If baffle spacing decreases, Reynolds number of shell side increases and the flow regime might become turbulent. Thus, the overall heat transfer coefficient increases as the spacing decreases. If baffle spacing is taken between 0.08 m and 1.5 m, the overall heat transfer coefficient decreases from 1040 W/m 2 K to 115 W/m 2 K. The parametric studies showed that if tube layout angle is chosen 30 or 60 instead of 45 or 90 both the overall heat transfer coefficient and total pressure drop increase. The number of passes does not have a significant effect on the overall heat transfer coefficient and the pressure drop. Acknowledgements The authors would like to thank Mehmet Akif Ezan and Gokhan Fidan for their guidance in some parts of the modelling of the heat exchanger. Nomenclature As cross sectional area of the shell perpendicular to the flow direction (m 2 ) A area (m 2 ), heat transfer surface area (m 2 ) b collector width (m) cp specific heat capacity (kj/kg K) CL tube layout constant CTP tube count constant D diameter (m) Ds shell diameter (m) e baffle spacing (m) F correction factor F collector efficiency factor ffric friction factor Fr heat removal factor g gravitational acceleration (m/s 2 ) Gb direct irradiation intensity (kw h/m 2 ) Gshell mass velocity of shell side fluid (kg m/s) h heat transfer coefficient (W/m 2 K), enthalpy (kj/kg) hc,g-a convective heat transfer coefficient for the glass cover to the ambient (W/m 2 K) hfi heat transfer coefficient of inside receiver tube (W/m 2 K) 207 hi tube side heat transfer coefficient (W/m 2 K) hloss head loss (m) ho shell side heat transfer coefficient (W/m 2 K) hr,g-a radiative heat transfer coefficient for the glass cover to the ambient (W/m 2 K) K incidence angle modifier k thermal conductivity (W/m K) L single collector length (m), length of a heat exchanger (m) m mass flow rate (kg/s) Nb number of baffles npass number of passes Nt Number of tubes Pr Prandtl number Pt tube pitch (m) Q heat transfer rate (W) q reflectivity Rf " fouling factor (m 2 K/W) Re Reynolds number S heat absorbed by the receiver (W/m 2 ) T temperature (K) U overall heat transfer coefficient (W/m 2 K) UL heat loss coefficient for the solar collector (W/m 2 K) Tlm logarithmic mean temperature difference (K) Greek Symbol α absorptivity γ specific weight (N/m 3 ) ε emissivity ηpump isentropic efficiency of the pump μ viscosity φ viscosity ratio between bulk and wall temperatures ρ reflectance of the mirror, density (kg/m 3 ) σ Stefan Boltzmann constant (W/m 2 K4 ) ϑ specific volume (m 3 /kg) transmittance of the glass cover τ o Subscript a ambient, aperture b bulk c cold e equivalent fric friction g glass cover h hot i inner, inlet o outer, outlet r receiver s surface sur surroundings w wall References Brooks, M., Mills, I., Harms T., Design, Construction and Testing of a Parabolic Trough Solar Collector for a Developing-Country Application, In Proceedings of the ISES Solar World Congress, Orlando, FL, (2005). Brooks, M. J., Mills, I., Harms, T.M., Performance of a Parabolic Trough Solar Collector. Journal of Energy in Southern Africa 17(3): 71 80, (2005).

222 Duffie, J., Beckman, W., Solar Engineering Of Thermal Processes, Wiley, (2013). Fernández-García, E., Valenzuela, Z.L., Pérez, M., Parabolic-Trough Solar Collectors and Their Applications. Renewable and Sustainable Energy Reviews 14(7): , (2010). Fettaka, S., Thibault, J., Gupta Y., Design of Shelland-Tube Heat Exchangers Using Multiobjective Optimization, International Journal of Heat and Mass Transfer 60: , (2013). Ghani, J.A, Choudhury, I.A., Hassan H.H., Application of Taguchi Method in the Optimization of End Milling Parameters. Journal of Materials Processing Technology 145(1): 84 92, (2004). Harrison, J., Standards of the Tubular Exchangers Manufacturers Association. Eight Ed., New York: Tubular Exchangers Manufacturers Association, (2007). Herrmann, U., Kelly, B., Price H., Two-Tank Molten Salt Storage for Parabolic Trough Solar Power Plants, Energy 29(5-6): , (2004). Islam, M.K., Hasanuzzaman, M., Rahim, N.A., Modelling and Analysis of the Effect of Different Parameters on a Parabolic-Trough Concentrating Solar System, RSC Adv. 5(46): (2015). Kakaç, Sadik, Liu, H. Pramuanjaroenkij, A. Heat Exchangers: Selection, Rating, and Thermal Design. Second Ed., CRC Press, (2002). Kalogirou, S., Parabolic Trough Collector System for Low Temperature Steam Generation: Design and Performance Characteristics. Applied Energy 55(1): 1 19, (1996). Kalogirou S., Solar Thermal Collectors and Applications, Progress in Energy and Combustion Science, 30(3), , (2004). Kalogirou S., Solar Energy Engineering Processes and Systems. Elsevier Inc., (2009). Kara Y. A., Ozbilen G., A Computer Program for Designing of Shell-and-Tube Heat Exchangers. Applied Thermal Engineering 24(13): (2004). Collectors. Solar Energy 62(6): , (1998). Patnode A., Simulation and Performance Evaluation of Parabolic Trough Solar Power Plants. Ph.D Thesis, University of Wisconsin-Madison, (2006) Perry, R.H., Green D.W., Maloney J.O., Chemical Engineers Handbook, Seventh Ed., (1997). Ming Q., Archer David H., Sophie V. M.,. A Linear Parabolic Trough Solar Collector Performance Model. Renewable Energy Resources and a Greener Future 8(3), (2006). Reppich M., Zagermann S., A New Design Method for Segmentally Baffled Heat Exchangers. Computers & Chemical Engineering 19(95): , (1995). Sasmito, A. P., Jundika C. K., Shamim T., Arun S. M., Optimization of an Open-Cathode Polymer Electrolyte Fuel Cells Stack Utilizing Taguchi Method, Applied Energy. (2015). Selbaş, R., Kizilkan O., Reppich M., A New Design Approach for Shell-and-Tube Heat Exchangers Using Genetic Algorithms from Economic Point of View. Chemical Engineering and Processing: Process Intensification 45(4): (2006). Shah, R.K., Sekulic D.P., Fundamentals of Heat Exchanger Design, Wiley (2002). SkyTrough Next-Generation Solar Parabolic Trough Technology. 4. Link: ghbrochure.pdf. (Last Access: ) Subramanian, R S., How to Design a Shell-and-Tube Heat Exchanger.(2010a) Link: ch302/notes/designshelltube.pdf (Last Access: ). Subramanian, R. S., Shell-and-Tube Heat Exchangers. (2010b) Link: ch302/notes/shelltube.pdf (Last Access: ). Walraven, D., Laenen, B., William D., Comparison of Shell-and-Tube with Plate Heat Exchangers for the Use in Low-Temperature Organic Rankine Cycles. Energy Conversion and Management 87: (2014). Mills D. Advances in Solar Thermal Electricity Technology, Solar Energy 76(1-3): (2004). Nuwayhid R. Mrad Y.F., Abu-Said R. The Realization of a Simple Solar Tracking Concentrator for University Research Applications, Renewable Energy 24: (2001) Odeh S.D., Morrison G.L., Behnia, M.. Modelling of Parabolic Trough Direct Steam Generation Solar 208 Walraven, D., Laenen, B., William, D., Optimum Configuration of Shell-and-Tube Heat Exchangers for the Use in Low-Temperature Organic Rankine Cycles. Energy Conversion and Management 83: , (2014). White, F. M., Fluid Mechanics, Seventh Ed., Mc Graw Hill, (2009). Yang, W.H., Tarng, Y.S., Design Optimization of

223 Cutting Parameters for Turning Operations Based on the Taguchi Method, Journal of Materials Processing Technology 84(1-3): (1998). 209

224 Thermal Regulation Enhancement of Concentrated Photovoltaic Systems Using Phase- Change Materials Mohamed Emam 1*, Mahmoud Ahmed 1, Shinichi Ookawara 1, 2 1 Egypt-Japan University of Science and Technology (E-JUST), Department of Energy Recourses Engineering, New Borg-El- Arab city, Alexandria, 21934, Egypt. 2 Tokyo Institute of Technology, Tokyo, Japan. * mohamed.emam@ejust.edu.eg Abstract Concentrated photovoltaic system (CPV) is one of the most promising applications of solar energy. However, due to high concentration ratios (CR), a significant increase of its temperature occurs which reduces the conversion efficiency and increases the potential to damage the CPV system. To avoid such problems, thermal regulation of CPV system is of great importance. It can be achieved by integrating the phase change materials (PCMs) with phase transition temperature close to the CPV optimal operating temperature. One of the main obstructions for such application is how to enhance the low thermal conductivity of the PCMs in order to achieve a fast thermal dissipation response. In the present work, the insertion of metal fins inside the PCMs for improving heat transfer is proposed. To investigate the thermal performance of the proposed CPV-PCM system, a comprehensive 2-D model for CPV layers integrated with PCM is developed. This model couples a thermal model for CPV layers and thermo-fluid model that takes into account the phase-change phenomenon using enthalpy method. The model is numerically simulated at different internal fin arrangements at CR = 20. The numerical results are validated using the available experimental and numerical results. It is found that the use of fins increases the heat transfer inside the PCM and achieves a significant reduction of solar cell temperature compared with that of the system without using fins. Keywords: Concentrated photovoltaic, thermal regulation, PCM, solar cell efficiency I. Introduction Concentrated photovoltaic (CPV) systems are widely recognized as the most efficient form of Photovoltaic (PV) power generation due to its high solar energy gain with small capital cost, viz., for getting more PV power output by using less solar cell material than other conventional non-concentrated PV systems. In CPV systems, relatively inexpensive materials such as plastic lenses or mirrors are used to capture the incident solar radiation on a fairly large area and concentrate that energy onto small solar cell (Du et al. 2012). However, due to high concentration ratios (CR), a significant increase of solar cell temperature occurs which reduces the conversion efficiency and increases the potential to damage the CPV system (Ma et al. 2015). Therefore, thermal regulation of CPV systems is of great importance. Many researchers are seeking to develop an effective cooling system to mitigate the impact of excessive temperature rise in the CPV conversion efficiency by removing heat from CPV module surfaces to keep a good performance as much as possible. The effective cooling method would achieve high efficiency, long lifetime, and enhance the possibility of using concentrators. Numerous previous investigations had been carried out to incorporate phase change materials (PCM) within PV systems for thermal regulation. The PV- PCM system absorbs a considerable amount of energy as a latent heat during the phase transition from solid to liquid over a very narrow range of transition temperature which increases the electric conversion efficiency by preventing the overheating of the system during the day and releasing it during the night. In addition, PV-PCM system stands out from the conventional thermal control systems with its compactness, lightness, and high efficiency. However, thermal management of CPV systems using PCM is relatively rare. Therefore, the incorporation of PCMs with phase transition temperature near to the PV normal operating temperature of 25 C for thermal regulation of CPV systems under relatively high concentration ratios is a new contribution for thermal regulation of CPV systems with relatively high CRs. One of the main obstructions for such application is how to enhance the low thermal conductivity of the PCMs to achieve a quick thermal dissipation response, especially at relatively high CRs. The literature demonstrate that thermal performance improvements achieved using metal fins were significant since the temperature distribution in the PCM container became more uniform than systems without fins, and the PV module temperature rises were clearly restrained. A review of experimental and computational studies to improve the low thermal conductivity of PCMs that were conducted over many decades had been carried by (Fan & Khodadadi 2011). (Huang et al. 2004; Huang et al. 2006b; Huang et al. 2011) investigated the effect of fin spacing, width, and fin type on the PV-PCM system performance. It was 210

225 noticed that the insertion of fins improved the effective thermal conductivity of PCMs and enhanced the thermal performance of PV-PCM system. As the fin spacing reduced, the maximum temperature decreased, and the temperature uniformity of PV cells was achieved. (Malvi et al. 2011) used conductive fins, mesh or encapsulation inside the PCM layers to enhance PCM thermal conductivity. It was found that an increase of PCM conductivity by 10% can improve PV output by 3%. In the present work, the insertion of metal fins inside the PCMs for improving its thermal conductivity is proposed. To investigate the thermal performance of the proposed CPV-PCM system, a comprehensive 2- D model for CPV layers integrated with PCM is developed. This model couples a thermal model for CPV layers and thermo-fluid model that takes into account the phase-change phenomenon using enthalpy method, the conversion of solar incident radiations and real-time heat loss boundary conditions. The model is numerically simulated at different internal fin arrangements. It was reported (Huang et al. 2006a) that the 2-D model can accurately reproduce the prediction of the 3-D model for simple line-axis system with simple boundary conditions. II. Physical model The schematic diagram of the proposed 2-D CPV- PCM system with dimensions and boundary conditions is presented in Fig. 1. This system was introduced to study the effect of fins on the CPV-PCM system temperature control under high values of solar incident radiations. As shown in the figure, the PCM was placed between two aluminum flat plates and then attached to the rear side of the CPV module. The aluminum front/back walls of 3 mm thickness were included to achieve uniform temperature distribution over the front surface of the system. Moreover, they protected the PCM and provided a high rate of heat transfer so that the PCM absorbed heat easily from the CPV module. This was enhanced further by a series of aluminum fins with 3 mm thickness extended into the PCM from the front wall. The interior dimensions of the container were 125mm height by 200 mm depth. The upper and lower ends of the CPV- PCM system were assumed to be adiabatic, so the heat flows were assumed to be symmetrical and occurred through the perpendicular direction of the CPV-PCM cell only (Huang et al. 2004; Huang et al. 2006b; Park et al. 2014). As the solar radiation incident on the surface of the PV module, part of the solar incident radiation G(t) was absorbed at the front face of the CPV-PCM system and converted to electric power by the PV cells and another part was lost by convection and radiation to the surrounding. The rest of radiation energy was conducted through the PV cells and its mounting plate to PCM causing the system temperature rise while a small fraction was dissipated from the rear surface of the system. 211 Fig. 1: Schematic diagram of CPV-PCM system with internal fins. The main criteria for selection of a suitable PCM for a particular application are its phase transition temperature which should be near to the PV characterizing temperature of 25 C. In addition, other important parameters including high values of thermal conductivity and latent heat. Additionally, stability to cycling heat process must be taken into account for making an appropriate decision (Dhaidan et al. 2013). In the present work, the selected PCM is salt hydrate CaCl2.6H2O, and the thermo-physical properties of the PCM and aluminum are shown in Table 1. CPV- PCM systems without fins and with a different number of fins were investigated at CR = 20. Also, the effect of the fin length on the CPV thermal regulation was discussed. Table 1: Thermo-physical properties of selected PCM (Hasan et al. 2015) and aluminum. Thermo-physical properties CaCl 2-6H 2O (PCM) Aluminum Melting point, ( C) 29.8 N/A Heat of fusion, (kj/kg) 191 N/A Thermal conductivity Solid, (W/m C) Liquid, (W/m C) Density Solid, (kg/m 3 ) Liquid, (kg/m 3 ) Specific heat capacity Solid, (kj/ kg K) Liquid, (kj/ kg K) N/A 2675 N/A N/A Thermal expansion coefficient, (k -1 ) N/A Thermal cyclic stability Yes (Tyagi & Buddhi 2008) - Chemical classification Salt hydrate - N/A: Not Applicable. III. Mathematical model III.1. Governing equations In the current study, a comprehensive 2-D model for CPV layers integrated with PCM is developed to predict the transient temperature distribution within the CPV-PCM system with and without fins. This model comprises the energy equations for CPV layers and the thermo-fluid model for the transient analysis of PCM. The computational domain is shown in Fig. 2

226 which consists of the aluminum front plate, aluminum fins, PCM and the aluminum back plate. conversion efficiency and temperature coefficient at a reference temperature, T ref =25 C respectively. The reference solar radiation G(t) is equal to 1000W/m 2. These values are provided by the manufacturer data sheet and are available for most PV cells (Tiwari & Swapnil 2010). The front loss, E f of thermal energy by the effect of the wind speed and radiation can be determined as follow (Zelin Xu & Kleinstreuer 2014): E U f f U ( T Ta ) (7) f g K g sc 1 1 (8) h f Fig. 2: Computational domain of CPV-PCM system. The front wall heat flux q"w can be calculated from the energy balance equations of the CPV module where the total energy absorbed by the CPV cell can be written as follow: E gg (t) (1) sc sc sc The total energy absorbed by the tedlar can be written as: 1 ( t) E gg (2) T sc sc Then the total energy absorbed by both CPV cell and its tedlar can be written by: 1 G( t) G( t) E (3) sc T g sc According to the CPV cell efficiency, part of the total absorbed energy (E) at the front surface of the CPV- PCM system is converted into electricity E el, another part (E f) is lost from the front surface of CPV cell to the surrounding by means of convection and radiation. The remaining part is conducted through the CPV cells and its mounting plate to PCM causing the system temperature rise as shown in Fig. 2. The amount of thermal energy passes through the CPV cell to PCM can be estimated according to the following relation: q E E E (4) w el f The electric power produced by solar cell can be written as follows: E gg(t) (5) el sc sc sc g The front surface heat transfer coefficient for the glass surface of the CPV to the ambient can be calculated by (Agrawal & Tiwari 2011). h (9) f V w In order to deal with phase change problem, the enthalpy-porosity technique is used. In this technique, the solid-liquid interface is not tracked explicitly. The presence of the solid or liquid phase is instead monitored using a quantity known as a liquid fraction (λ). The PCM is assumed to be Newtonian, incompressible, and unsteady. Also, the PCM density variation in the buoyancy term is modeled by the Boussinesq approximation for involving thermal buoyancy. Thus, the governing equations include the continuity equation, momentum equations, and the energy equation for the 2-D transient laminar flow could be written as follows (Hosseini et al. 2012; Taylor 2007): u v 0 x y u u u P u u u v t x y x 2 2 x y v v v u v t x y S x P v v g y 2 2 x y (10) (11) T T ref Sy (12) With μ = μ s =, ρ = ρ s in solid regions of PCM, and μ = μ l, ρ = ρl in liquid region of PCM. S In Eqs 11 and 12, is the Darcy's law damping terms (as source term) that are added to the momentum equation due to phase change effect on convection. It is defined as: Where, η sc is the solar cell efficiency which is a function of solar cell temperature as reported in (Zelin Xu & Kleinstreuer 2014) and can be written as follows: 1 S A mush V (13) 1 ( T T )) (6) sc ref ( ref sc ref Where: η ref and β ref are the CPV cell electrical 212 Where ε is a small number, typically around 10-3 introduced to avoid the singularity and A mush is the mushy zone constant which describes how steeply

227 the velocity is reduced to zero when the material solidifies and its value depends on the morphology of the medium. The value of A mush used here for the computations is the standard one, i.e kg.m -3 s -1. Energy equation for melt: H H H u v t x y Energy equation for solid: H t H H l l (14) x x y y H H s s (15) x x y y The enthalpy of the material is computed as the sum of the sensible enthalpy, h, and the latent heat, ΔH: H h H Where, T (16) h href cpdt (17) Tref The latent heat content can be written in terms of the latent heat of the material, L: L (18) Where ΔH may vary from zero (solid) to L (liquid). Therefore, the liquid fraction, λ, can be defined as: H 0 T Tsolid L H T Tsolid Tsolid T Tliquid (19) L Tliquid Tsolid H 1 T Tliquid L III.2. Boundary conditions The computational domain is a 2-D rectangular cavity with dimensions (L H) in the x-y plane, and δ is aluminum plate thickness, so the no-slip boundary conditions are: u = v = 0, at x = δ, (L- δ) and y = 0, H for t 0 P = 0, at x = δ and y = 0 for t 0 The initial values of u, v and P, are set to be zero. At the exterior front boundary at x = 0, where the computational domain is exposed to heat flux, the boundary condition is: T q w k x x0 Thermally coupled boundary condition is applied at the interface between the aluminum front plate and 213 PCM at x =δ, T T k k x x Al, x PCM, x At the exterior back boundary at x = L, the boundary condition is: T k x x l h b T x l T a Where: h b is the heat transfer coefficient from back exterior wall to the surrounding; h b 2.8 3V (20) w For adiabatic boundary condition is applied on the upper and lower ends, the boundary conditions are as follow: T y T y 0, H 0 0, x, ytini The phase change occurs at a set temperature. In the case of constant specific heat capacities for each phase, the temperature field can be defined as: E c T T m T E L m cl 0 E L, E L, T T m T T m T T m ( Solid phase) s (21) ( melt zone) ( liquid phase) The latent heat value E of the PCM in the melt zone is modeled as high sensible heat value in each time step and accumulated with time. At any time, when the accumulated heat is larger than the specified latent heat L of PCM, the PCM is changed to the liquid phase. The values of CPV cell characteristics and design parameters used in the CPV-PCM are indicated in Table 2. The temperature of the aluminum front wall will be estimated from the numerical simulation. Then the temperature of CPV cell can be calculated by the following equation (Tiwari & Swapnil 2010). T sc T w qw k sc sc T k T (22) Table 1: Characteristics and design parameters used in the CPV-PCM. Parameter Value Parameter Value H sc 125 mm α sc τ g α T δ g 0.32 mm 1 G(t) 1000W/m 2 k g 1 2, 3 η ref δ sc 0.2 mm 1 k T k sc δ T 0.3 mm 1 β sc β ref , (Zhou et al. 2015); 2, (Hedayatizadeh et al. 2013); 3, (Zelin Xu & Kleinstreuer 2014); 4, (Z. Xu & Kleinstreuer 2014); 5, (Dubey & Tay 2013).

228 III.3.Computational procedures and Model validation. Firstly, the upper wall heat flux q"w can be calculated from the energy equation of the CPV module by guessing initial value for the electric conversion efficiency. The governing equations subjected to the boundary and initial conditions for the computational domain are solved by using the commercial software ANSYS 17.0 using the finite volume technique. The SIMPLE algorithm has been used to solve the pressure velocity coupling. The first-order upwind scheme was used for solving the momentum and energy equations, whereas the PRESTO (PREssure STaggering Option) scheme was adopted for the pressure correction equation. By solving the governing equations at each time step, liquid mass fraction has been updated using Eq. 19. First-order implicit time integration scheme has been employed. The time step in the calculations was selected as small as 0.5 s and the number of iterations for each time step was 400. The number of triangular 2-D elements is discretized the physical domain. This number of grids and time step are considered after careful examination of the results to achieve the grid independency and accommodate both the required solution accuracy and convergence at a relatively low run time. The convergence was checked at each time step, with the convergence criterion of 10 6 for all variables. The computational results are validated with the experimental data of Huang et al. (Huang et al. 2011) by comparing the average predicted and measured temperature on the front surface of the system with time as shown in Fig 3. The incident solar radiation and ambient temperature used were 750 W/m 2 and 19 C, respectively. The top, bottom, and back surfaces were adiabatic. Reasonable agreement is obtained between the current computational results and experiments of Huang et al. (Huang et al. 2011). Furthermore, comparisons of the predicted front surface temperature with the numerical results reported by Huang et al. (Huang et al. 2004) is conducted as shown in Fig 4 at an ambient temperature of 20 C, incident solar radiation of 1000 W/m 2 and an initial system temperature of 20 C. Good agreement is found between both results. Fig. 3: Comparison between the predicted average temperature evaluations on the front surface with the corresponding experimental results of (Huang et al. 2011) with the ambient temperature set at 19 C and incident radiations of 750 W.m -2. Fig.4: Comparison between the predicted average temperature evaluations on the front surface of the numerical results of (Huang et al. 2004) with the ambient temperature set at 20 C with incident radiations of 1000 W.m -2 and an initial system temperature of 20 C. 214 IV. Results and discussion IV.1. Solar cell temperature without PCM In order to investigate the performance of CPV-PCM system, a thermal model for the PV reference cell without PCM is developed for comparison. This model includes a complete energy balance on PV layers such as glass cover, polycrystalline silicon solar cell, and PV-tedlar as documented in Eqs 1-9. The computed temperature is compared with the nominal operating temperature of the polycrystalline silicon

229 (MSX- 60) solar cell given by the manufacturing data sheet at solar radiation of 800 W.m -2, wind speed of 1m/s, and ambient temperature of 20 o C is compared with the computed temperature at the same conditions. The comparison shows a very good agreement where the predicted temperature is about o C, and the given nominal temperature is 47 o C. The deviation from the nominal value is estimated to be about 1.2% (Hedayatizadeh et al. 2013). Figure 5 indicates the variation of steady state solar cell temperature at different values of CR ranged from 1 to 20 without using PCM as a cooling medium. It is clear that increasing the CR leads to an increase in the solar cell temperature where it increases from 50 to about 510 o C. This occurs due to the increasing of the incident solar energy as CR increases. Fig. 5: variation of solar cell temperature without PCM versus CR, at an ambient temperature of 25 C and 1 m.s -1 wind velocity. occurred. Therefore, the liquid-solid interface changes from parallel lines near to the front wall to curved one as the time elapses as shown in Fig. 6. Along the height at the center of the non-finned CPV- PCM system at points A, B, and C, heat transfer, is initially dominated by conduction with a linear temperature increase with time. After one hour, the melting interface reaches point A; then heat transfer is dominated by convection. Due to convection, the temperature increased sharply towards the solar cell temperature while the temperatures in the solid phase locations (B, C) maintained their slow conductiondominated increase. The same behavior is observed at point B and C after two hours and three hours respectively. In addition, two stages of temperature variation of the solar cell are observed during the phase change process of the cooling material.firstly, a steep increase in the temperature with time followed by a gradual increase until reaching the peak. This variation is most likely due to sensible heating of PCM by conduction heat transfer through the aluminum plate, followed by the phase transition of the PCM adjacent to the aluminum front plate which is causing a thin melt layer in the PCM, absorbing the CPV thermal energy as latent heat. During this period, the PCM acts as an insulation material for the CPV cell while the heat transfer is dominated by conduction rising the CPV front surface temperature. Lastly, further increase of time results in a decrease in cell temperature followed by an increase until reaching the complete melting point. This stage indicates the start of the heat transfer by convection which balances heat transfer by conduction in the PCM. With increasing the time, heat is removed by convection and conduction until reaching a fully liquid phase. IV.2. Thermal performance of the CPV-PCM system without fins. Thermal regulation of the CPV-PCM system depends on the thermal behavior of the PCM during melting. The transient variation of the average solar cell temperature of the non-finned CPV-PCM system with 200 mm PCM thickness and CR=20 is presented in Fig. 6. The same figure presents the temperature variation with elapsed time along the height at the center of the non-finned CPV-PCM system ( points A, B, and C locations, see Fig. 6). As indicated in Fig. 5, it was found that the temperature of the solar cell without PCM could reach 510 o C at CR=20. Using PCM salt hydrate CaCl2.6H2O with a melting point 29.8 o C with similar conditions, the CPV-PCM without fins could maintain the solar cell at an average temperature of 64 o C for 2.0 hours while the temperature at the complete melting point of PCM is around 119 o C. With continued energy input, the melting of PCM occurred from top to bottom with clear temperature stratification because of the natural convection 215 Fig. 6: Average predicted solar cell temperature and the temperature variation with time in the center vertical line of CPV-PCM system with the time evolution of the solid-liquid interface of CPV-PCM system at an ambient temperature of 25 C and 1 m.s -1 wind velocity.

230 IV.3. Thermal performance comparison for CPV- PCM systems with a different number of fins. In the present work, a detailed analysis of the effect of the insertion of a different number of aluminum fins on the thermal regulation of the CPV-PCM system is proposed. Figure 7 presents the transient variation of the average solar cell temperature of the CPV-PCM system with a different number of fins ( each is 75 mm length) and no fins. From this figure, it is noticed that the CPV-PCM system with two fins could maintain the solar cell at an average temperature of 57 o C for 2.0 hours while the temperature at the complete melting point of PCM is around 111 o C. For the CPV-PCM system with four fins, the temperature of the solar cell was maintained at 54 o C for 2.0 hours while the temperature at the complete melting point of PCM is around 109 o C. It is clear that the use of the fins increases the heat transfer inside the PCM and achieves a significant reduction of solar cell temperature compared with the system without fins. PCM temperature begins to rise quickly. This flow pattern is maintained until the PCM is fully molten. Once the PCM in the uppermost section is fully molten after two hours, the temperature of the CPV/PCM system increases rapidly. t = 0.5 hr t = 1.0 hr t = 1.75 hr t = 2.0 hrs Fig. 7: Average predicted solar cell temperature of CPV - PCM system with different number of fins and no fins at CR= 20, an ambient temperature of 25 C and 1 m.s -1 wind velocity The predicted temperature distributions during the PCM melt process within the CPV-PCM system with four fins during the PCM melt process is presented in Fig. 8. From the figure, it is noticed that when aluminum fins are added to the system, the formation of a deep cavity in the upper part of the CPV-PCM system is reduced and divided into several small shallow cavities between the fins which reduced the thermal stratification within the system. After one hour, a natural convection flow of hot molten PCM passes through the gap at the end of the fins into the upper part of the system then turns to flow downward through the gap and near to the liquid-solid interface into the lower section. After 1.75 hours, the molten PCM reaches the aluminum rear plate, and its temperature rises causing an increase in the heat transfer rate from this side to the PCM adjacent to it. Thereby, the melting velocity increases and the CPV- 216 t = 3 hrs Fig. 8: Predicted isotherms of CPV-PCM system with four fins at different times In order to investigate the effect of using fins on the temperature uniformity for the CPV-PCM system.

231 Figure 9 presents the local solar cell temperature for the CPV-PCM with a different number of fins at a time of two hours. From this figure, it is noticed that the temperature difference between the top and base of the solar cell (125 mm height) for the non-finned CPV- PCM system equals to 27 o C. The use of aluminum fins provides improved thermal control while the temperature difference is reduced to 17 o C for the CPV-PCM system with two fins. Increasing the number of fins to four can decrease the temperature difference to 15 o C and also reduce the solar cell temperature. Fig. 10: Average predicted solar cell temperature of CPV-PCM system with four fins at a different fin length. V. Conclusions Fig. 9: Variations of local solar cell temperatures of CPV - PCM system without fins, with two and four fins after two hours. IV.4. Effect of fin length on the thermal control for CPV-PCM system. The average solar cell temperature of the CPV-PCM with four fins at different fin length of 75, 125, 200 mm is shown in Fig. 10. The variation of fin length is shown to be a marginal factor in improving the thermal performance of the CPV-PCM system. For the CPV- PCM system with 75 mm fin length, the temperature of the solar cell was maintained at 54 o C for 2.0 hours. Increasing the fin length to 200 mm could maintain the solar cell at the same value of temperature but for 2.2 hours. Furthermore, the temperature at the complete melting point of PCM is decreased from 109 o C to 98 o C by increasing the length of fins from 75 mm to 200 mm. A hybrid system including CPV and phase change material (PCM) is suggested as a single module to attain a significant reduction in solar cell temperature, particularly in a high concentration ratio. One of the main obstructions for such application is how to enhance the typical low thermal conductivity of the PCMs to achieve a quick thermal dissipation response. In the present work, the insertion of a different number of aluminum fins with the various lengths inside the PCMs for improving heat transfer is proposed. To investigate the thermal performance of the proposed CPV-PCM system, a comprehensive 2D model for CPV layers integrated with PCM is developed. This model couples a thermal model for CPV layers and thermo-fluid model that takes into account the phase-change phenomenon using enthalpy method. It was found that fins increased the heat transfer inside the PCM and achieved a significant reduction of solar cell temperature compared with the system without fins. For CPV-PCM system with two fins, the temperature of the solar cell was maintained at 57 o C for 2.0 hours while the temperature at the complete melting point of PCM is around 111 o C. Increasing the number of fins to four could reduce the temperature of the solar cell to 54 o C for 2.0 hours while the temperature at the complete melting point of PCM is around 109 o C. In addition using fins could enhance the temperature uniformity of the CPV-PCM system where the temperature difference between the top and base of the solar cell was reduced from 27 o C for the non-finned CPV-PCM system to 15 o C for the CPV-PCM system with four fins. The obtained results can be used to optimize the design of CPV-PCM systems to achieve a quick thermal dissipation response with longer of thermal regulation in CPVs. 217

232 Acknowledgements The work described in the current paper is financially supported by the Egyptian government especially Ministry of Higher Education (MoHE). As well as the authors would like to acknowledge the Egypt-Japan University of Science and Technology (E-JUST) for offering the facilities and tools. Nomenclature Amush : Mush zone constant [Kg.m -3.s -1 ] C : Specific heat [J.kg -1.K -1 ] E : Rate of energy per unit cell area [W.m -2 ] g : Gravity acceleration [m.s -2 ] G(t) : Incident solar radiation [W.m -2 ] h : Convection heat transfer coefficiecnt [W.m -2.K -1 ] H : Solar cell height [m] and total enthalpy K : Thermal conductivity [W.m -1.K -1 ] L : Latent heat [J] P : Pressure [N.m -2 ] : Momentum source term S t : PCM thickness [mm], Time [hours] T : Temperature [ o C] u,v : Velocity in x and y-direction respectively [m.s -1 ]. Greek letters µ : PCM viscosity [Pa.s] ρ : PCM density [kg.m -3 ] α τ λ β : Absorptivity and thermal diffusivity [m 2.s -1 ] : Transmissivity : Liquid fraction : Backing factor, thermal expansion coefficient [K -1 ] and cell temperature coefficient [K -1 ]. η : Solar cell efficiency δ : Thickness [m] Subscripts Sc : Solar cell a : Ambient l : liquid ref : Reference condition,g=1000 w.m -2, and T=25 o C s : Solid el : Electrical w : Wall and wind g : Glass b : Back m : melting T : Tedlar ini : initial References Agrawal, S. & Tiwari, A., Experimental validation of glazed hybrid micro-channel solar cell thermal tile. Solar Energy, 85(11), pp (2011). Dhaidan, N.S. et al., Experimental and numerical investigation of melting of phase change material/nanoparticle suspensions in a square 218 container subjected to a constant heat flux. International Journal of Heat and Mass Transfer, 66, pp (2013). Du, B., Hu, E. & Kolhe, M., Performance analysis of water cooled concentrated photovoltaic (CPV) system. Renewable and Sustainable Energy Reviews, 16(9), pp (2012). Dubey, S. & Tay, A. a O., Testing of two different types of photovoltaic-thermal (PVT) modules with heat flow pattern under tropical climatic conditions. Energy for Sustainable Development, 17(1), pp.1 12 (2013). Fan, L. & Khodadadi, J.M., Thermal conductivity enhancement of phase change materials for thermal energy storage: A review. Renewable and Sustainable Energy Reviews, 15(1), pp (2011). Hasan, a. et al., Increased photovoltaic performance through temperature regulation by phase change materials: Materials comparison in different climates. Solar Energy, 115, pp (2015). Hedayatizadeh, M. et al., Thermal and Electrical Assessment of an Integrated Solar Photovoltaic Thermal (PV/T) Water Collector Equipped with a Compound Parabolic Concentrator (CPC). International Journal of Green Energy, 10(August 2014), pp (2013). Hosseini, M.J.J. et al., A combined experimental and computational study on the melting behavior of a medium temperature phase change storage material inside shell and tube heat exchanger. International Communications in Heat and Mass Transfer, 39(9), pp (2012). Huang, M.J. et al., Natural convection in an internally finned phase change material heat sink for the thermal management of photovoltaics. Solar Energy Materials and Solar Cells, 95(7), pp (2011). Huang, M.J., Eames, P.C. & Norton, B., Comparison of a small-scale 3D PCM thermal control model with a validated 2-D PCM thermal control model. Solar Energy Materials and Solar Cells, 90, pp (2006a). Huang, M.J., Eames, P.C. & Norton, B., Phase change materials for limiting temperature rise in building integrated photovoltaics. Solar Energy, 80(9), pp (2006b). Huang, M.J., Eames, P.C. & Norton, B., Thermal regulation of building-integrated photovoltaics using phase change materials. International Journal of Heat and Mass Transfer, 47(12-13), pp (2004). Ma, T. et al., Using phase change materials in photovoltaic systems for thermal regulation and electrical efficiency improvement: A review and

233 outlook. Renewable and Sustainable Energy Reviews, 43, pp (2015). Malvi, C.S., Dixon-Hardy, D.W. & Crook, R., Energy balance model of combined photovoltaic solarthermal system incorporating phase change material. Solar Energy, 85(7), pp (2011). Park, J., Kim, T. & Leigh, S.B., Application of a phasechange material to improve the electrical performance of vertical-building-added photovoltaics considering the annual weather conditions. Solar Energy, 105, pp (2014). Taylor, P., Numerical Heat Transfer : An International Journal Of Computation And Methodology Enthalpy- Porosity Technique For Modeling Convection- Diffusion Phase Change : Application To The Melting Of A Pure Metal., (June 2013), pp (2007). Tiwari, G.N. & Swapnil, D., Fundamentals of Photovoltaic Modules and their Applications, Cambridge: Royal Society of Chemistry (2010). Tyagi, V. V. & Buddhi, D., Thermal cycle testing of calcium chloride hexahydrate as a possible PCM for latent heat storage. Solar Energy Materials and Solar Cells, 92(8), pp (2008). Xu, Z. & Kleinstreuer, C., Computational Analysis of Nanofluid Cooling of High Concentration Photovoltaic Cells. Journal of Thermal Science and Engineering Applications, 6(3), p (2014). Xu, Z. & Kleinstreuer, C., Concentration photovoltaicthermal energy co-generation system using nanofluids for cooling and heating. Energy Conversion and Management, 87, pp (2014). Zhou, J. et al., Temperature distribution of photovoltaic module based on finite element simulation. Solar Energy, 111, pp (2015). 219

234 Solar Radiation Exergy and Enviroeconomic Analysis for the West Black Sea Region in Turkey Yusuf Kurtgoz 1*, Emrah Deniz 2 1 Karabuk University, Graduate School of Natural & Applied Sciences, Department of Mechanical Engineering, 100. Yil. Mah., Karabuk, 78050, Turkey 2 Karabuk University, Faculty of Engineering, Department of Mechanical Engineering, 100. Yil. Mah., Karabuk, 78050, Turkey * ykurtgoz@karabuk.edu.tr Abstract The determination of useful solar energy amount is important to design solar energy systems. For this purpose, exergy values of solar radiation are calculated for the cities and districts of the West Black Sea region in Turkey. The monthly mean global solar radiation and sunshine hour data are obtained using solar energy potential atlas developed by the General Directorate of Renewable Energy. The long term monthly mean temperature values for the cities and districts are taken from the General Directorate of Meteorology. The approaches of Petela, Spanner and Jeter are used to calculate exergy-to-energy ratios for determining the maximum utilizable solar radiation energy. Exergy-to-energy ratios, solar radiation exergy values, locations having the highest solar exergy potential and solar radiation exergy maps are presented for the West Black Sea region in Turkey. Obtained results can be used for design of solar energy systems in this region. Finally, enviroeconomic analysis is performed and its results are shared in this study. Keywords: Solar radiation, exergy, solar energy, solar radiation exergy map I. Introduction In future, renewable energy sources are important for providing a sustainable environment. Solar energy as a renewable energy source is a very abundant resource in some regions in the world. For this reason, many studies about solar energy were performed. Studies mainly focused on following five solar energy research areas; a) PV/T system or solar thermal collector efficiencies b) single or hybrid solar-based electricity generation systems, c) generating hydrogen by using solar-powered systems, d) solar energy applications about sustainable energy buildings or zero energy, e) feasibility of solar energy applications. Knowledge of solar radiation distribution is important for designing and analysing solar energy systems (Coskun et al., 2011). The usefulness of energy is briefly known as exergy. In literature, as there are plenty of studies about energy and exergy, studies about exergy of solar energy are also performed by the researchers (Svirezhev and Steinborn, 2001; Petela, 2003; Svirezhev et al., 2003; Kabelac, 2005; Chu and Liu, 2009; Zamfirescu and Dincer, 2009; Jiménez-Muñoz et al., 2012). Enviroeconomic analysis is an effective tool providing price comparison in terms of CO2 emissions. There are several studies using enviroeconomic analysis tools. Enviroeconomic analysis of experimental systems used photovoltaic modules were presented by (Agrawal and Tiwari, 2013; Rajoria et al., 2013, 2015; Tiwari et al., 2015). Enviroeconomic analysis of a novel air cooler was performed by (Caliskan et al., 2012). The solar, biomass, and electrical energy based building heating systems were compared using enviroeconomic analysis by (Caliskan, 2015). The aim of this study is to determine the solar radiation exergy distribution and locations having the highest solar exergy potential and to demonstrate solar radiation exergy maps for West Black Sea region in Turkey. Enviroeconomic analysis is also performed in terms of exergy values. II. Materials and Methods The monthly mean global solar radiation and sunshine hours data are obtained using solar energy potential atlas developed by the General Directorate of Renewable Energy for the cities of Bartin, Bolu, Duzce, Karabuk, Kastamonu, Sinop and Zonguldak with their districts in the West Black Sea region. The meteorological values measured between 1985 and 2006 are modelled by the General Directorate of Renewable Energy to prepare this atlas. The long term monthly mean temperature values for the same cities and districts in the region are taken from the General Directorate of Meteorology. The monthly mean global solar radiation (W/m 2 ) for the cities and districts are calculated by using the monthly mean global solar radiation (kwh/m 2 ) and sunshine hours (h) data obtained from General Directorate of Renewable Energy. 220

235 II.1. Exergy of Solar Radiation CO2 is CO2 mitigation per year. The exergy concept shows beneficial amount of energy. The energy concept alone does not represent its quality aspect (Hepbasli and Alsuhaibani, 2014). Therefore, it is important to determine the maximum amount of work output that can be obtained from a solar energy system (Zamfirescu and Dincer, 2009). The relative potential of the maximum energy available from radiation () was expressed separately by Petela (Eq. 1), Spanner (Eq. 2) and Jeter (Eq. 3) to determine the exergy efficiency: 4 1 T0 4 T0 p 1 3 T 3 T (1) 4 T0 S 1 3 T (2) T0 J 1 T (3) where, T0 is the ambient temperature and T is the solar temperature (Petela, 2003; Hepbasli and Alsuhaibani, 2014). The international carbon price is between13 $/tco2 and 16 $/tco2. Environmental cost is calculated by Eq. 6 (Caliskan et al., 2012): Z CO z 2 CO CO where (6) z CO2 2 2 is the average international carbon price per tco2 (14.5 $/tco2) and Z CO2 is the enviroeconomic (environmental cost) parameter (CO2 mitigation price per year) ($/year). III. Results and Discussion According to cities, solar radiation energy comparisons by three different approaches are shown separately in Figure 1. Average exergy-to-energy ratio is for Petela and Spanner s approaches, therefore very similar results are obtained from these approaches. Average exergy-to-energy ratio is obtained as from Jeter s approach for the region. The ratio of the exergy Exrad to the energy Erad called as the maximum conversion efficiency is equal to (Eq. 4). Ex E Rad rad (4) Exergy of solar radiation can be computed by using exergy efficiency equations (Eq.1, Eq. 2 and Eq. 3) and Eq. 4. Firstly, monthly exergy efficiencies of the cities and districts are calculated separately using Petela, Spanner and Jeter s approaches. Then using Eq. 4, monthly solar radiation exergy values are obtained for three approaches. (a: Petela) II.2. Enviroeconomic Analysis Enviroeconomic analysis is the most effective tool to encourage the use of renewable energy technologies as it does not emit carbon (Tiwari et al., 2015). People and countries evaluate this payment obtained from enviroeconomic analysis for carbon emission and are motivated to decrease the carbon emissions (Caliskan, 2015). The enviroeconomic analysis is based on the price of CO2 emission into the environment. CO2 mitigation per year from the overall solar radiation exergy is calculated by Eq. 5 (Tiwari et al., 2015): CO 2 where CO CO 2 Ex 2 Rad (5) 1000 is the average CO2 equivalent amount for generating electricity from coal (2.0 kgco2/kwh), Ex is the annual overall solar radiation exergy and Rad 221 (b: Spanner) (c: Jeter) Fig. 1: Comparison of the monthly mean solar radiation exergy for the cities

236 The Monthly mean solar radiation values are listed in Table1 for the cities in the region. Because exergy-toenergy ratio values are close for three approaches, results obtained using Petela s approach are presented in this study. The Monthly mean solar radiation exergy values are also listed in Table 2 for the cities according to Petela s approach. Table 3 shows annual and seasonal monthly means of solar radiation exergy values for the cities according to Petela s approach. For Petela s approach solar radiation exergy maps of the West Black Sea region are shown month by month in Figure 2. Seasonal solar radiation exergy maps of the region are shown in Figure 3. Annual average solar radiation exergy map of the region is shown in Figure 4. Tab. 1: Monthly mean solar radiation values (W/m 2 ) for the cities Month City Bartin Bolu Duzce Karabuk Kastamonu Sinop Zonguldak Tab. 2: Monthly mean solar radiation exergy values (W/m 2 ) for the cities according to Petela s approach Month City Bartin Bolu Duzce Karabuk Kastamonu Sinop Zonguldak (3: March) (1: January) (4: April) (2: February) (5: May) 222

237 (6: June) (12: December) Fig.2: Monthly solar radiation exergy maps of West Black Sea region (7: July) (1: Spring) (8: August) (2: Summer) (9: September) (3: Autumn) (10: October) (11: November) 223 (4: Winter) Fig.3: Seasonal solar radiation exergy maps of West Black Sea region Spring, summer, autumn and winter monthly average solar radiation exergy values are W/m 2, W/m 2, W/m 2 and W/m 2, respectively for the region. Annual average solar radiation exergy value is W/m 2 for the overall region.

238 amount, the results in this study can be used to design solar energy systems in the region. Following conclusions have been listed: Fig.4: Annual average solar radiation exergy map of West Black Sea region Obtained results showed that the Bolu city has the highest (536.9 W/m 2 ) and the Bartin city has the lowest (490.7 W/m 2 ) annual monthly average solar radiation exergy potential in the West Black Sea region. Tab. 3: Annual and seasonal monthly means of solar radiation exergy values (W/m 2 ) according to Petela s approach Yearly Seasonal Means City Mean Spring Summer Autumn Winter Bartin Bolu Duzce Karabuk Kastamonu Sinop Zonguldak The results showed that the monthly average solar radiation exergy values are highest for April (615.5 W/m 2 ) and May (673.9 W/m 2 ) months and lowest for November (371.2 W/m 2 ) and December (374.3 W/m 2 ) months in the West Black Sea region, according to data of cities and districts. According to the average values of districts, annual CO2 mitigations and environmental costs of overall exergy per m 2 are listed in Table 4 for the cities in the West Black Sea region. Tab. 4: Annual CO2 mitigation and environmental cost of overall exergy City CO 2 mitigation per year Environmental cost per year CO (tco 2 2/year) Z CO ($/year) 2 Bartin Bolu Duzce Karabuk Kastamonu Sinop Zonguldak The values of CO2 mitigation per year are between tco2/year and tco2/year. Environmental cost values of overall exergy are between $/year and $/year in the West Black Sea region. IV. Conclusions In this study, solar radiation exergy potential is presented for the West Black Sea region in Turkey. Because it is important to know useful solar energy 224 Average exergy-to-energy ratios of Petela and Spanner s approaches are and average exergy-to-energy ratio is obtained as from Jeter s approach in the region. The results showed that the Bolu has the highest and the Bartin has the lowest solar radiation exergy potential in the West Black Sea region. Results showed that the highest monthly average solar radiation exergy value is for May and the lowest value is for November in the West Black Sea region, according to data of cities and districts. According to the seasons, the highest monthly average solar radiation exergy value which is W/m 2 occurred in spring in the region. Annual monthly average solar radiation exergy value is W/m 2 for the overall region. According to the enviroeconomic analysis results, the highest annual CO2 mitigation and environmental cost has been occurred in the Bolu city. The annual CO2 mitigation and environmental cost values per m 2 for the Bolu are 2.67 tco2/year and $/year respectively in terms of overall solar radiation exergy. Enviroeconomic analysis is an effective tool providing price comparison in terms of CO2 emissions. Thereby, motivation for decreasing carbon emissions can be increased. Furthermore, similar studies can be performed for different regions or overall Turkey in future. Nomenclature :Exergy efficiency (exergy-to-energy ratio) T0 : Ambient temperature (K) T : Solar temperature (K) Exrad : Radiation exergy values (W.m-2) Erad : Monthly mean solar radiation values (W.m- 2) CO 2 : Average CO2 equivalent amount for generating electricity from coal (2.0 kgco2.kw-1.h-1) : CO2 mitigation per year (tco2.year-1) CO2 z CO 2 Z CO 2 References : Average international carbon price per tco2 ($.tco2-1) : Enviroeconomic (environmental cost) parameter ($.year-1) Agrawal, S., Tiwari, G. N., Enviroeconomic analysis and energy matrices of glazed hybrid photovoltaic thermal module air collector, Solar Energy, 92, , (2013). Caliskan, H., Thermodynamic and environmental analyses of biomass, solar and electrical energy options based building heating applications, Renewable and Sustainable Energy Reviews, 43,

239 , (2015). Caliskan, H., Dincer, I., Hepbasli, A., Exergoeconomic, enviroeconomic and sustainability analyses of a novel air cooler, Energy and Buildings. (Cool Roofs, Cool Pavements, Cool Cities, and Cool World), 55, , (2012). Chu, S. X., Liu, L. H., Analysis of terrestrial solar radiation exergy, Solar Energy, 83, , (2009). Coskun, C., Oktay, Z., Dincer, I., Estimation of monthly solar radiation distribution for solar energy system analysis, Energy, 36, , (2011). Hepbasli, A., Alsuhaibani, Z., Estimating and Comparing the Exergetic Solar Radiation Values of Various Climate Regions for Solar Energy Utilization, Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 36, , (2014). Jiménez-Muñoz, J. C., Sobrino, J. A., Mattar, C., Recent trends in solar exergy and net radiation at global scale, Ecological Modelling, 228, 59 65, (2012). Kabelac, S., Exergy of solar radiation, International journal of energy technology and policy, 3, , (2005). Petela, R., Exergy of undiluted thermal radiation, Solar Energy, 74, , (2003). Rajoria, C. S., Agrawal, S., Tiwari, G. N., Exergetic and enviroeconomic analysis of novel hybrid PVT array, Solar Energy, 88, , (2013). Rajoria, C. S., Agrawal, S., Tiwari, G. N., Chaursia, G. S., Exergetic and enviroeconomic analysis of semitransparent PVT array based on optimum air flow configuration and its comparative study, Solar Energy, 122, , (2015). Svirezhev, Y. M., Steinborn, W. H., Exergy of solar radiation: information approach, Ecological Modelling, 145, , (2001). Svirezhev, Y. M., Steinborn, W. H., Pomaz, V. L., Exergy of solar radiation: global scale, Ecological Modelling, 169, , (2003). Tiwari, G. N., Yadav, J. K., Singh, D. B., Al-Helal, I. M., Abdel-Ghany, A. M., Exergoeconomic and enviroeconomic analyses of partially covered photovoltaic flat plate collector active solar distillation system, Desalination, 367, , (2015). Zamfirescu, C., Dincer, I., How much exergy one can obtain from incident solar radiation?, Journal of Applied Physics, 105, /1-5, (2009). 225

240 Comparison of Regression Analysis, ANN and ANFIS Methods in the Prediction of Monthly Mean Global Solar Radiation: A Case Study Yusuf Kurtgoz 1*, Emrah Deniz 2 1 Karabuk University, Graduate School of Natural & Applied Sciences, Department of Mechanical Engineering, 100. Yil Mah., Karabuk, 78050, Turkey 2 Karabuk University, Faculty of Engineering, Department of Mechanical Engineering, 100. Yil Mah., Karabuk, 78050, Turkey * ykurtgoz@karabuk.edu.tr Abstract In this study, the monthly mean daily global solar radiation in Adana province of Turkey is modelled by regression analysis, artificial neural network and adaptive network-based fuzzy inference system (ANFIS) methods. For this aim, three different models using these methods are evaluated and compared each other. The monthly means of daily sunshine duration, air temperature, relative humidity, wind speed, soil temperature and global solar radiation data of Adana province between 2000 and 2010 are obtained from General Directorate of Meteorology and used to design models. Seven input parameters are determined using these meteorological data and some geographical equations. Root mean square error (RMSE), mean absolute percentage error (MAPE) and correlation coefficient (R) are used to evaluate the performance of the models. For these performance indicators, the best values are obtained with ANN model. Keywords: Solar radiation, estimation, regression analysis, artificial neural network, ANFIS I. Introduction Knowledge of solar radiation is important in solar energy research. However, most of the meteorological stations do not have equipment for solar radiation measurement. (Yadav and Chandel, 2014; Li et al., 2015). Empirical models have been designed for global solar radiation using some climatic variables such as sunshine hours, extraterrestrial radiation, mean temperature, maximum temperature, relative humidity, altitude, latitude, cloudiness and evaporation (Bakirci, 2009). Artificial neural network (ANN), linear and nonlinear, and fuzzy logic models have been used in the prediction of solar radiation. Root mean square error, mean absolute percentage error and mean bias error are used as statistical error indicators in evaluation of designed models (Khatib et al., 2012). There are various studies using regression analysis. Ibrahim et al. developed a linear regression model using temperature and solar radiation data to predict solar radiation in Perlis, Northern Malaysia for The correlation coefficient value was (Ibrahim et al., 2012). In a study (Pu and Lin, 2000), step-by-step linear regression procedure was used to predict monthly total solar radiation on horizontal surfaces at Kunming. Multi linear regression and artificial neural network were compared for predicting solar radiation in Turkey. The ANN model results were better than multi linear regression (Sahin et al., 2013). Togrul and Onat (1999) developed multiple linear regression models for six geographical and meteorological parameters to estimate the monthly mean daily global solar radiation. In a study (Linares-Rodriguez et al., 2013), ANN ensemble model was developed and clear-sky estimates and satellite images were selected as input variables for the model. Accuracy of solar radiation estimates were relatively good. RMSE and correlation coefficient values were 6.74% and 99%, respectively. Azadeh et al. (2009) developed an integrated ANN model for prediction of solar radiation using climatological and meteorological parameters. Accuracy result of the model was 94%. Estimated values were compared with angstrom s model and 7.28% enhancement was provided. In another study (Amrouche and Le Pivert, 2014), the authors combined spatial modeling and ANN techniques in a methodology to predict solar radiation and used daily weather estimations in the method. Obtained results were at sufficient level. In literature there are also several studies using ANFIS technique to predict solar radiation. Latitude, longitude, and altitude inputs were used to estimate the mean monthly clearness index and daily solar radiation data in isolated areas with ANFIS method (Mellit et al., 2008). Multi-layer perceptron, local linear regression, Elman neural network, neural network auto-regressive model with exogenous inputs and ANFIS models were developed for solar radiation forecast and compared by (Moghaddamnia et al., 2009). To model solar radiation, fuzzy genetic method was compared with ANN and ANFIS techniques using latitude, longitude, altitude and month of the year data as inputs for seven cities in Turkey (Kisi, 2014). In another study (Piri and Kisi, 2015), ANFIS and neural network auto-regressive model with exogenous inputs methods were 226

241 compared with Angstrom and Hargreaves Samani empirical models. In estimation of solar radiation, artificial intelligence models results were better than Angstrom and Hargreaves Samani empirical models. Regression analysis, ANN and ANFIS methods are often used to model solar radiation. Therefore, the aim of this study is to evaluate and compare these methods performance in solar radiation forecast. For this aim, a case study is performed for Adana province of Turkey. II. Material and Methods The monthly means of daily sunshine duration (n), relative humidity (Hum), wind speed (WS), air temperature (T), soil temperature (Ts) and global solar radiation data ( ) of Adana province are obtained from General Directorate of Meteorology. Using the equations (Eq. 1, 2 and 3) and obtained data, Extraterrestrial radiation (H0), the ratio of sunshine duration to day length (n/n), Sine of declination angle (Sin), Hum, WS, T and Ts parameters and H parameter are determined as inputs and output, respectively. H0 is calculated by Eq. 1 and day length (N) for any day is calculated by Eq. 2 (Turk Togrul and Onat, 1999). H 0 = G SC ( cos 360n π 360 ) (cos cosδsin(w ss ) + πw ss 180 sin sinδ) (1) 2 N cos 1 tan tan 15 (2) where, Gsc is the solar constant, δ is the solar declination angle, is the latitude of site, wss is the mean sunset hour angle for the given month and nꞌ is the number of days of the year starting from first of January. The solar declination angle can be calculated by the following equation (Turk Togrul and Onat, 1999); n δ = 23.45sin ( ) (3) The monthly mean values of Adana province obtained from General Directorate of Meteorology between 2000 and 2009 are arranged as shown in Table 1 for seven inputs and one output. Similarly, the mean values belong to 2010 to be used for the test are also shown in Table 2. Using the data in Table 1, global solar radiation prediction models are developed with ANN, Regression Analysis and ANFIS methods. These prediction models are tested and compared using 2010 data in Table 2. Tab.1: Adana province monthly mean values between 2000 and 2009 Months H H 0 n/n Sin Hum WS T T s January February March April May June July August September October November December Tab. 2: Adana province monthly mean values by 2010 Months H H 0 n/n Sin Hum WS T T s January February March April May June July August September October November December To compare measured and estimated values of the monthly mean global solar radiation, the root mean square error (RMSE), the mean absolute percentage error (MAPE) and correlation coefficient (R) are used as statistical indicators. The equations for these indicators are: RMSE 1 MAPE n i1 1 n n H ic H im n i1 i1 H ic H H im ic ic im i im im (4) (5) (H 1 ic - Hic)(Him - Him) i R 2 2 (6) n n (H - H ) (H - H ) n where n is the number of data pairs, H ic and H im are the average estimated and measured values of the monthly mean global solar radiation, respectively. II.1. Artificial Neural Network (ANN) In this study, global solar radiation is predicted using multilayer perceptrons (MLP) feedforward neural networks trained by back-propagation (BP) algorithm. Proposed MLP model is shown in Fig. 1. A network structure which has 7 neurons in input layer, 8 neurons in hidden layer and 1 neuron in output layer is built for prediction. Sigmoid function is used as activation function in hidden layer and output layer. 227

242 The measured solar radiation values for the year 2010 are compared with the solar radiation values calculated with the developed formula. II.3. Adaptive Network-Based Fuzzy Inference System (ANFIS) The Adaptive Network-Based Fuzzy Inference System (ANFIS) was developed by Jang (Jang, 1993). ANFIS method aims to determine the behavior of imprecisely complex dynamical systems (Moghaddamnia et al., 2009). Fig. 1: Proposed MLP model As a result of different trials in the network, the value 0.8 is used for the learning rate and the value 0.7 is used for the momentum. MLP network training continues until the training error is smaller than , or the program has performed training iterations (epoch). The data belong to in Table 1 and belong to 2010 in Table 2 are used as training and test data respectively. Training and test data are scaled with following equation. Thus, data which is not in range of 0 to 1 is reduced to same scale. x xmin x' (7) x x max min Scaled outputs obtained from training and testing of MLP network must be transformed to original form again. For this purpose, the following formulation is obtained by reversing the above equation. x = x (x max x min ) + x min (8) II.2. Regression Analysis The regression analysis methods are used to determine relationship between a dependent variable and independent variables. The linear regression analysis demonstrates the variation of a y variable depending on a variable and the multi linear regression analysis (MLRA) demonstrates the variation of a y variable depending on multiple x variables (x1, x2, x3...). Expression of the multi linear regression analysis is shown in Eq. 9. ANFIS includes an adaptive ANN and a fuzzy inference system (Oguz and Guney, 2010) and uses a hybrid-learning rule merging gradient-descent, back-propagation, and a least-squares algorithm to determine a set of parameters (Mellit et al., 2008). A typical five layered ANFIS structure is shown in Figure 2 (Moghaddamnia et al., 2009; Kulaksiz, 2013): Layer 1: In this layer, every node is an adaptive node with a node function such as generalized bell membership function (Eq. 11), Gaussian membership function (Eq. 12). 1 A ( x) 2 x c i 1 a i i b i (11) 2 x c i A ( x) exp (12) i ai where, ai, bi and ci are parameters. Also x is the input to node i and Ai is the linguistic label (hot, cold, etc.) associated with this node function. In this layer, parameters are referred to as premise parameters. Premise parameters change the shape of the membership function. y = a+bx1+cx2+dx3+... (9) According to the values, the following equation including seven variables (Ho, n/n, Sinδ, Hum, WS, T, and Ts) with a correlation coefficient of is obtained by using MLRA to calculate the global solar radiation in Adana. H = H n/N Sinδ Hum WS T Ts (10) Fig. 2: A typical ANFIS structure Layer 2: In this layer, every node labeled multiplies the incoming signals and sends the product out (Eq. 13). Each node output shows the firing strength of a rule. wi ( x) ( x), i 1, 2 (13) A i B i 228

243 Layer 3: In this layer, every circle node labeled N represents the normalized firing strength of each rule. The ith node calculates the ratio of the ith rule s firing strength to the sum of two rules firing strengths by using Eq. 14. wi wi, i 1, 2 (14) w w 1 2 Layer 4: In this layer, every node is an adaptive node with a node function (Eq. 15), indicating the contribution of the ith rule towards the overall output. w z w p x q y r ) (15) i i i ( i i i where, w i is the output of layer 3 and pi, qi and ri are parameters. Parameters in this layer will be referred to as consequent parameters. Layer 5: In this layer, there is a single node labeled that calculates the overall output as the summation of all incoming signals (Eq. 16). i wi zi Z wi zi (16) w i i i ANFIS structure in this study with seven input parameters is shown in Figure 3 for global solar radiation. Estimated Values (MJ/m 2 ) Actual Values (MJ/m 2 ) Fig. 4: Distribution graph of ANN forecast results versus actual values for the year 2010 The distribution graph is shown in Figure 5 which belongs to real values versus forecast results that are calculated with Eq. 10 obtained for the year Estimated Values (MJ/m 2 ) R² = R² = Actual Values (MJ/m 2 ) Fig. 5: Distribution graph of MLRA forecast results versus actual values for the year 2010 The distribution graph which belongs to the ANFIS forecast results versus actual values for the year 2010 is shown in Figure 6. Fig. 3: ANFIS structure for seven input parameters III. Results and discussion The distribution graph which belongs to the ANN forecast results versus real values for the year 2010 is shown in Figure 4. Estimated Values (MJ/m 2 ) R² = Actual Values (MJ/m 2 ) Fig. 6: Distribution graph of ANFIS forecast results versus actual values for the year 2010 According to developed ANN, MLRA and ANFIS models in this study, measured and estimated global solar radiation values, RMSE, MAPE and R values for 2010 are shown in Table 3. RMSE, MAPE and R values are resulted between , and , respectively. 229

244 Comparison of three models with measured values is also shown in Figure 7. According to Table 3 and Figure 7, better estimation results are obtained with ANN model. Tab. 3: Measured and estimated values for the year 2010 Month Measured ANN MLRA ANFIS January February March April May June July August September October November December Solar Radiation (MJ/m 2 ) RMSE MAPE R Fig. 7: Comparison of measured and estimated values for the year 2010 IV. Conclusions Measured ANN MLRA ANFIS Month Furthermore, similar comparative study can be made also for different locations and input combinations in future. Nomenclature : Latitude ANFIS : Adaptive network-based fuzzy inference system Gsc : Solar constant (1367 W.m-2) H : Monthly mean daily global solar radiation (MJ.m-2) H0 : Monthly mean daily extraterrestrial radiation incident on horizontal surface, (MJ.m-2) Hic : Estimated value of monthly mean global solar radiation (MJ.m-2) Him : Measured value of monthly mean global solar radiation (MJ.m-2) Hum : Monthly mean humidity MAPE : Mean absolute percentage error MLP : Multilayer perceptrons MLRA : Multi linear regression analysis n : Monthly mean daily sunshine duration (h) N : Monthly mean day length (h) nꞌ : Day sequence number in year, starting from 1st January R : Correlation coefficient RMSE : Root mean square error T : Monthly mean air temperature ( C) Ts : Monthly mean soil temperature ( C) WS : Wind speed (m/s) wss : Sunset hour angle δ : Declination angle References Amrouche, B., Le Pivert, X., Artificial neural network based daily local forecasting for global solar radiation, Applied Energy, 130, , (2014). Seven meteorological and geographical parameters are used separately in ANN, MLRA and ANFIS methods to estimate global solar radiation in Adana province. The prediction results of these methods are compared with measured data obtained from General Directorate of Meteorology data center by using RMSE, MAPE, and R indicators. Following conclusions can be given for this study: Results of each three methods are quite good. When results of methods compared to each other, best RMSE and MAPE values are obtained with ANN model. According to RMSE indicator, ANN results are 44% and 28% better than MLRA and ANFIS results, respectively. For MAPE indicator, ANN results are 61% and 55% better than MLRA and ANFIS results, respectively. Obtained results showed that using the ANN model would be more appropriate for predicting global solar radiation and this method provides quite convergence to the measured values to evaluate the solar potential of a location. 230 Azadeh, A., Maghsoudi, A., Sohrabkhani, S., An integrated artificial neural networks approach for predicting global radiation, Energy Conversion and Management, 50, , (2009). Bakirci, K., Models of solar radiation with hours of bright sunshine: A review, Renewable and Sustainable Energy Reviews, 13, , (2009). Ibrahim, S., Daut, I., Irwan, Y. M., Irwanto, M., Gomesh, N., Farhana, Z., Linear Regression Model in Estimating Solar Radiation in Perlis, Energy Procedia. (Terragreen 2012: Clean Energy Solutions for Sustainable Environment (CESSE)), 18, , (2012). Jang, J.-S. R., ANFIS: adaptive-network-based fuzzy inference system, IEEE Transactions on Systems, Man and Cybernetics, 23, , (1993). Khatib, T., Mohamed, A., Sopian, K., A review of solar energy modeling techniques, Renewable and Sustainable Energy Reviews, 16, ,

245 (2012). Kisi, O., Modeling solar radiation of Mediterranean region in Turkey by using fuzzy genetic approach, Energy, 64, , (2014). Yadav, A. K., Chandel, S. S., Solar radiation prediction using Artificial Neural Network techniques: A review, Renewable and Sustainable Energy Reviews, 33, , (2014). Kulaksiz, A. A., ANFIS-based estimation of PV module equivalent parameters: application to a stand-alone PV system with MPPT controller, Turkish Journal of Electrical Engineering and Computer Science, 21, , (2013). Li, H., Cao, F., Bu, X., Zhao, L., Models for calculating daily global solar radiation from air temperature in humid regions A case study, Environmental Progress & Sustainable Energy, 34, , (2015). Linares-Rodriguez, A., Ruiz-Arias, J. A., Pozo-Vazquez, D., Tovar-Pescador, J., An artificial neural network ensemble model for estimating global solar radiation from Meteosat satellite images, Energy, 61, , (2013). Mellit, A., Kalogirou, S. A., Shaari, S., Salhi, H., Hadj Arab, A., Methodology for predicting sequences of mean monthly clearness index and daily solar radiation data in remote areas: Application for sizing a stand-alone PV system, Renewable Energy, 33, , (2008). Moghaddamnia, A., Remesan, R., Kashani, M. H., Mohammadi, M., Han, D., Piri, J., Comparison of LLR, MLP, Elman, NNARX and ANFIS Models with a case study in solar radiation estimation, Journal of Atmospheric and Solar-Terrestrial Physics, 71, , (2009). Oguz, Y., Guney, I., Adaptive neuro-fuzzy inference system to improve the power quality of variable-speed wind power generation system, Turkish Journal of Electrical Engineering and Computer Science, 18, , (2010). Piri, J., Kisi, O., Modelling solar radiation reached to the Earth using ANFIS, NN-ARX, and empirical models (Case studies: Zahedan and Bojnurd stations), Journal of Atmospheric and Solar-Terrestrial Physics, 123, 39 47, (2015). Pu, S., Lin, W., Correlations to estimate monthly total solar radiation on horizontal surfaces at Kunming, China, Energy Conversion and Management, 41, , (2000). Sahin, M., Kaya, Y., Uyar, M., Comparison of ANN and MLR models for estimating solar radiation in Turkey using NOAA/AVHRR data, Advances in Space Research, 51, , (2013). Turk Togrul, I., Onat, E., A study for estimating solar radiation in Elazig using geographical and meteorological data, Energy Conversion and Management, 40, , (1999). 231

246 Reduction of Entropy Production by Using of Solar Cooling Systems Based on SOLITERM Parabolic Trough Collectors Combined with Double Effect Absorption Chillers Ahmet Lokurlu* SOLITERM GROUP, Süsterfeldstr. 83, Aachen, Germany * info@solitermgroup.com Abstract The worldwide rapidly growing energy demand, e.g. for cooling and process steam generation, and the boundary conditions of the existing fossil energy supply structures are leading to an increasing lack of fossil energy carriers and growing greenhouse gas emissions, which pollute the ecosystem and increase the Entropy production rapidly. With the efficient utilisation of solar energy by Parabolic Trough Collectors for heating, cooling and process steam generation as well as desalination the demand of resources and the emission of climate depending harmful substances can be reduced and through this the increasing costs of these shrinking resources. On more twenty locations in ten countries SOLITERM Parabolic Trough Collector systems have been installed. Some of them deliver heat for cooling with double effect absorption chillers as well as steam for laundry of the hotel and other systems offer steam for industrial application as well as desalination of see water and electricity production. Each SOLITERM-System has at least two operation modes. These modes are designed to replace high economic fossil fuel or electricity costs with available free solar power. The operation modes are also appending on main energy consumers like laundry or air conditioning in hotels. So one operation mode replaces electricity consumption of compression chillers during high tariff times in summer with solar driven double effect absorption chillers, or it replaces expensive LPG for heating of existing boiler with solar generated steam. Another operation mode uses Parabolic Trough Collectors during winter for heating of buildings or warm water preparation. SOLITERM is the only supplier of efficient solar high temperature technology for buildings and factories worldwide. The SOLITERM System uses special developed parabolic trough collectors to generate temperatures of up to 450 C. With these collectors, SOLITERM opens the market for high efficient solar cooling with double effect absorption chillers. This is one of the most important technical solutions of the 21st century, which is particularly applicable for the needs of hotels, hospitals and factories. A major reason for the choice of the parabolic trough system is their effectiveness of high efficiency and as a result of it minimisation of Entropy production. For its development, SOLITERM is recipient of the internationally renowned more than 30 Awards, including Planet Hero of the Environment from Time Magazine and 3 times Energy Globe Award, several Innovation and sustainability Awards etc. Keywords: Entropy, parabolic trough collector, solar cooling, double effect absorption chiller I. Introduction The worldwide rapidly growing energy demand, for cooling and process steam generation, and the boundary conditions of the existing fossil energy supply structures are leading to an increasing lack of fossil energy carriers and growing greenhouse gas emissions, which pollute the ecosystem and increase the entropy production promptly. With the efficient utilisation of solar energy through Parabolic Trough Collectors for heating, cooling and process steam generation as well as desalination, the demand of resources and the emission of climate depending harmful substances can be reduced therefore also entropy. At more than twenty locations in ten countries, the SOLITERM Parabolic Trough Collector systems have been installed worldwide. Some of them deliver (steam) heat for cooling with double effect absorption chillers as well as steam for laundry of the hotel and other systems provide steam for industrial applications as well as desalination of sea water and electricity production. Each SOLITERM-System has at least two operation modes. These modes are designed to replace economically the demanding fossil fuels and electricity costs by freely available solar power in an economic way. The operation modes are also depending on main energy consumers like laundry or air conditioning in hotels. So one operation mode replaces electricity consumption of compression chillers during high tariff periods in summer with solar driven double effect absorption chillers or it replaces 232

247 expensive LPG for heating of an existing boiler with solar generated steam. Another operation mode uses Parabolic Trough Collectors during winter for heating of objects or for warm water preparation. throughout the day with a peak system output during the noon hours of the day when the cooling demand is the highest. SOLITERM is the only supplier of efficient solar high temperature technology for buildings and factories worldwide. The SOLITERM Systems use specially developed parabolic trough collectors to generate temperatures of up to 450 C. With these collectors, SOLITERM opens the market for highly efficient solar cooling with double effect absorption chillers. This is one of the most important technical solutions of the 21st century, which especially meets the needs of the hotels, hospitals and factories. Parabolic Reflector Absorber Tube PTC (Parabolic Trough Collector) Water / Thermal Oil Glass Tube Solar Solar Beam Beam Hot Water / Thermal Oil II. General explanations II.1. Conventional Solar Systems Fig. 1: The working principle of a parabolic trough collector (Source: Soliterm) The biggest problem of flat plate and evacuated tube collectors is that their low outlet temperature prevents them from being more efficient, especially in use for air conditioning due to a COP (Coefficient of Performance) of approx. 0.4 to 0.5 by small sized applications. In order to satisfy a certain demand, a larger collector area is needed. Many locations like hotels or roof tops of the buildings always have some limitations with respect to availability of space. As a result, the application range of conventional solar systems is also limited. II.2. The SOLITERM PTCs (Parabolic Through Collectors) To overcome these problems and to widen the application range for solar energy, SOLITERM developed several different sized Parabolic Trough Collectors (PTC) namely The SOLITERM PTC 1800 for large applications as well as the small sized SOLITERM PTC 1100 for residential buildings and PTC 3000, PTC 4000 as well as PTC 5000 filed applications for large scale systems. Both the collectors have a supply temperature of 200 C to 250 C and are capable for roof mounting. Parabolic trough systems use curved mirrors to focus the suns energy onto a receiver tube that runs down the focal line of a trough. In the receiver tube, a high-temperature heat transfer fluid (such as a synthetic oil) absorbs the sun energy, reaching temperatures of 450 C or even higher, and passes through an absorption chiller for cooling or through a heat exchanger to produce hot water and steam. The steam can be used to drive a conventional steam turbine power system to generate electricity. For instance the SOLITEM PTC-1800 collector concentrates the solar beam 40 times onto the absorber and reaches temperatures between 150 C-250 C. Single axis trackers are equipped with motors and advanced sensors allow the system to track the sun. Therefore high operating temperatures are achieved 233 The high temperatures achieved by the PTCs, can be used to power a double stage absorption chiller. Double stage absorption chillers have average COPs about of 1,25 to 1,5. This means that for every kw of solar energy 1.25 kw to 1.5 kw of cooling power are produced. The higher efficiency of PTC systems offsets the high initial investment costs and makes it the most competitive system for solar cooling, heating and process steam applications while considering the lifecycle costs. The high temperatures achieved by the PTCs, can be used to power a double stage absorption chiller. Double stage absorption chillers have average COPs about of 1,25 to 1,5. This means that for every kw of solar energy 1.25 kw to 1.5 kw of cooling power are produced. The higher efficiency of PTC systems offsets the high initial investment costs and makes it the most competitive system for solar cooling, heating and process steam applications while considering the lifecycle costs The solar combined heat, cooling and power is a crucial economical, ecological and highly efficient application. In the first step of the application, solar energy is used for electricity generation and in the second step, the heat supplied by the power process is used for generating chilled water and for supplying heat consumers. This example shows a clear advantage of parabolic trough systems in comparison to solar electricity generation with photovoltaic systems. III. A model Application for an Integrated Energy Supply An example for an integrated supply system developed by SOLITERM is the parking area installation at a Hotel in Antalya, Turkey, which is equipped with an air conditioning system that is operated with a SOLITERM system application on the parking place of the hotel.

248 Fig. 2: Application of PTC s on the top of parking place of a Hotel (Source: Soliterm) As one can see in the schematic diagram below, the main principle of the SOLITERM system is that the collector field that delivers high temperature heat at the level of 200 C to 250 C in the first cycle. In a steam generator that is placed between the first and the second cycle of the process, steam is produced at 4 bars, which can be used for the air conditioning in the double effect absorption chiller as well as for the steam supply of the laundry of the hotel. For different seasons, e.g. for summer and winter applications, an improved system configuration has also been developed for this location to maximize the economic use of solar energy. An important aspect of the use of the SOLITERM system is that this system is not limited to only supplying buildings with air conditioning and/or laundries and kitchens with steam during summer time: it can also be used for the heating of buildings or swimming pools in transition periods so that profitability increases significantly. Fig. 3: Schematic diagram of the SOLITERM System, (Source: Soliterm) The absorption system installed at the hotel is supplied by solar generated steam with parameters of up to 4 bar saturated steam with approx. 144 C. Furthermore, a small storage tank is needed to regulate the energy demand over the day. In case the outlet temperature of the collectors is not high enough to drive the steam generator (and thus the absorption chiller), auxiliary heating with the fossil fuel boiler is necessary. In the case of this particular installation, the SOLITERM system is connected to the existing steam system of the hotel, so that the remaining power can be used from the steam power supply of the hotel. 234 The aim was the optimal design of the components of the plant, so that the capacities of the storage tank, the boiler and especially the collector area, to cover the cold and heat demand of the building. The configuration of the plant is economically viable, if a high level of fuel costs savings with a low capital investment can be attained. In operation, the system is regulated in a way that the greatest possible part of cold and heat load is covered by solar energy supplied by the installed collector area. In Turkey, electricity costs are different depending on the time of day. Between 5 and 10 p.m., electricity costs are higher than during the normal tariff zone. In

249 this high tariff zone, it is important to reduce the share of the cooling supply of the conventional cooling system, increasing the solar contribution of the SOLITERM System to achieve greater efficiency. IV. Energy System Description The SOLITERM PTC 1800 roof mounted Parabolic Trough Collector series for temperature levels of up to 250 C has been developed. The collector system is tracked by a one axis tracking system and consists of modules with an aperture width of 1.8 m and a length of 5 m each. Solar radiation is focussed on an absorber tube which has a diameter of 38 mm. The focussing element is an aluminium reflector with a concentration ratio of up to 40. IV.1. SOLITERM PTC 1800 Collectors The figure 4 shows a Parabolic Trough Collector with reflector surfaces. The thermal losses at the high operating temperature level are reduced by a glass envelope tube. Due to higher costs involved, the glass envelope tube is not evacuated. The torsional stiffness is enhanced by an additional metal tube (torsion tube) behind the collector modules. A major part of the collector components are manufactured in Ankara, where a subsidiary of SOLITERM has been established. The rest of the components are produced in Germany and Switzerland. The tracking system of the of the system shown above, operates with wire ropes and allows the tracking of six collector rows each containing eight modules. A controller for two-axis tracked mirrors was modified for control requirements for the tracking of the Parabolic Trough Collectors. A calculation of the position of the sun gives the input for an approximate positioning of the collectors while sun sensors installed in different places of the collectors provide data for a precise positioning (fine tuning) of the collectors. IV.2. System Operation Modes The SOLITERM system has at least two operation modes. These modes are helpful to replace high fossil fuel and electricity costs with the available solar power, which depends on general boundary conditions like price level for electricity and fuel etc. But the operation modes are also dependent on the behaviour of energy consumers, time of day and energy consumption. SOLITERM s solar powered energy systems are designed to provide maximum efficiency based on the optimized and precisely produced collectors and a computerized tracking system. Operational efficiency is controlled through SOLITERM s online monitoring and overall management system. The data transfer as well as service and maintenance can be realized easily. By using this online monitoring system, our engineers cannot only control the system itself, but this achieves an optimization of energy consumption within certain premises. Via online monitoring system we can see how much energy is demanded and how much energy is actually being used and based on that we can adjust consumption in either way, which is bringing it to the most efficient point possible. Heat losses, over-consumption and all sorts of problems can be solved with our intelligent online monitoring system, resulting in a significant reduction in entropy for the environment and particularly for the citizens. SUN Collector Field Small Storage Tank Hot Water 180 C Existing Fossil System Steam Steam Generator Steam 2- stage AC Steam (4 bar) Cold Water 6 C Hotel Cooling Laundry The collector field of the pilot plant consists of five parallel rows with four modules each and a total aperture area of 1440 m². Pressurised water is used as the heat transfer medium. The inlet temperature of the collector field is about 155 C. The measured outlet temperature of the field is 180 C). With a downstream steam generator, saturated steam with 4 bars and 144 C is generated. Fig. 4: Parabolic Trough Collector with reflector surfaces (Source: Soliterm) 235 Heat Exchanger Warm Water C Heating, Swimming Pool Fig. 5: Operation modes of the PTC System for cooling and Steam production (Source: Soliterm) IV.3. Operation modes of the SOLITERM System Winter Mode An intelligent coupling of these operation modes has the advantage of replacing the electricity costs of the compression air conditioning system during high tariff times and the replacing of expensive fossil fuels during the rest of the day. In a third operation mode, buildings and/or swimming pools can be heated during winter time. In case of no cooling demand, for instance in winter, the collector field is streamed with low temperature and the heat is transferred to the warm water grid of the hotel by an additionally installed heat exchanger cycle. The absorption chiller can be operated exclusively by solar generated steam during peak load of the collector field. Because of the two-stage design, the chiller reaches 1250 kw cooling capacity at the fully loaded level with a guaranteed Coefficient of Performance (COP) of 1.3 and with a

250 cooling water inlet temperature of 29 C. The conversion process of solar heat in chilled water for air conditioning by the absorption chiller can be fundamentally improved if it is operated on part load. In the part load operation area, the COP value can reach up to 1.5. installing and maintaining the system. We have a wide range of high-tech products that are of a very high quality and we can use them for different purposes. With our PTC system we can do desalination, steam generation, cooling, heating as well as electricity generation. During the initial phase of operation, adaptation problems occurred, which have been eliminated by the optimisation of the installation under consideration of the existing structure. Since the optimisation of the installation has been finished, the next step is to follow this year. The dimension of the system will be extended significantly by a doubling of the collector field area. With the magnification of a solar energy deposit, a thermal capacity of about 850 kw will be made available. With this energy amount, cooling demands of up to 1250 kw and steam demands of up to 900 kw can be covered at the same time. V. Conclusion As explained above, SOLITERM is a company that that has a long history. Established in 1999, SOLITERM GmbH has more than 15 years experience in the field of renewable energies and has proven to be a strong brand name. We have realized more than 20 different projects around the world in many different industries, meeting the demands and requirements while providing a high level of service to our customers. The fact that we have this worldwide first unique production line enables us to work automated and efficient, giving us a strong position in the market by creating competitive advantage over other firms. We have the capability within our premises to produce 240 MW annually, which corresponds to twenty four 10 MW projects within a short time period for instance. So far SOLITERM has been present in many countries via our CEO, Dr. Ahmet Lokurlu, creating awareness by joining certain important events, participating press conferences and discussion rounds with important business people, politicians and committed celebrities. SOLITERM has achieved to become a brand with a strong brand image within the industry of renewable energies and now we have to use our capabilities to take this even further and develop what we have so far. The potential we have is far beyond than what other companies could possibly achieve and therefore we need to take further actions. Another important and key aspect is to use our strong German reputation, as the world always appreciates the quality, ethic and commitment that Germans have towards their work and this is why German brands are so popular currently. The way our engineers work, the way we manage things and the way we present ourselves are all important factors of our success. In technical terms we are able to do so many different things and the fact that we not only produce the system, but also provide our customers with all kinds of services from proposing the project to delivering, 236 A key aspect is that we are able to combine our system with already existing conventional systems, making it much easier for the whole installation process. Our storages and back-up systems provide us the opportunity to store as much energy as possible, which means that we are able to produce energy not only when the sun is shining, but throughout the day, basically 24 hours. The integration process has come so far that it is referred to as automated process, such as for the production itself. Our lightweight collectors can be installed on the ground, on rooftops as well as on facades of buildings, making our system very flexible towards different boundary conditions and customer requirements. Our new idea of installing floating systems is a solution for coastal countries with space problems. This is a highly unique and innovative idea, which can change the dimension of the renewable energy industry. Countries such as Malta, Cyprus and Singapore are only a few examples of target countries. We currently have some project proposals concerning this new technology, which we are about to finalize very soon. The world has to see this technology and the attention of the media will be on our product. We keep doing what we are best at and this is most certainly realizing commercial projects as well as doing R&D projects in order to find out about new possibilities we might have and how we can develop to become even stronger than we are now. SOLITERM offers investors the opportunity to become part of a strong movement and invest their capital wisely into something that is about to replace fossil energy usage in many parts of the world, providing an incredible foundation for multiplying their investments over short time periods. With all this provided information it can be said that this is a unique chance to make some changes in the world. The overall impact of our system is the reduction of economic dependence on fossil fuels, the enhancement of environment friendly technology usage and creating the awareness society needs, leading to the ultimate result of entropy reduction in the long-term. References Lokurlu, A., Krüger, D., Hennecke, K. Solare Kälteund Dampfversorgung mit Parabolrinnenkollektoren im Sarigerme Park Hotel an der Türkischen Mittelmeerküste, in erneuerbare Energie, AEE.

251 Lokurlu, A., Richarts, F., Krüger, D.: High efficient utilisation of solar energy with newly developed parabolic trough collectors (SOLITEM PTC) for chilling and steam production in a hotel at the Mediterranean coast of Turkey in International Journal of Energy Technology and Policy (IJETP), Vol. 3, No. 1/2, Inderscience Publishers. 237

252 Optimal Off-Design Conditions for Solar-Driven Organic Rankine Cycles Caglan Sevinc 1, Eray Uzgoren 2* 1 Sustainable Environment and Energy Systems, Middle East Technical University, Northern Cyprus Campus, Kalkanli, Guzelyurt, TRNC, Mersin 10, Turkey 2 Mechanical Engineering, Middle East Technical University, Northern Cyprus Campus, Kalkanli, Guzelyurt, TRNC, Mersin 10, Turkey * uzgoren@metu.edu.tr Abstract Solar-driven organic Rankine cycles (ORCs) often operate at off-design conditions, which can significantly affect its operating thermal efficiency. Control strategies need to address target pressure levels and refrigerant mass flow rate to maximize ORC s power output for a given set of available thermal energy and ambient temperature levels, characterized by the flow rate and the inlet temperature of heat transfer fluids (HTFs) for both heat source and heat sink. This study develops a numerical algorithm to determine the off-design states of a solar-driven ORC using the following inputs: (i) inlet temperature and the flow rate of the hot HTF, (ii) inlet temperature and the flow rate of the cold HTF, and (iii) flow rate of the working fluid. The novelty of the algorithm is related its capability to find steady-state condenser, evaporator pressure levels, and all internal states of the cycle so that rate of thermal energy input, rate of power output can be identified. A series of simulations revealed that a given ORC with the same heat input can theoretically operate at multiple off-design conditions. Using parate front optimization method, the best operating point for each conditions are determined for an ORC system utilizing R245fa as the working fluid. It is found that optimal conditions typically occur when the degree of superheat at the exit of the expander is minimum for dry fluids. Keywords: Organic Rankine cycle, off-design, solar energy, R245fa, Pareto front optimization I. Introduction Organic Rankine cycles (ORCs) can produce power from low-grade heat resources by utilizing positive displacement expanders and organic fluids suitable for application s temperature limits. Advances in ORC technology have matured enough and enabled manufacturers build ORCs for the market to be used with various heat resources. Among several ORC manufacturers, most target applications for geothermal and waste heat recovery while some extend their application areas to include biomass and solar thermal as reviewed by Velez et al. (2012). Intermittent nature of solar energy bring uncertainty in design and control strategies in concentrating solar power (CSP) applications. In such cases, off-design performance characteristics of ORCs need to be studied to address optimal output of CSP plants. Fu et al. (2014) investigated the effect of off-design heat source temperature on the system performance and showed that cycle efficiency and net power output can vary between -11.5% to 17.4% and -13.6% to 22.6% respectively. Song et al. (2012) analyzed the offdesign characteristics of an ORC utilizing R123 as working fluid (WF) and found out that the net power generated increases as the inlet temperature of heat source increases. They also revealed that a decrease in inlet temperature of cooling water assures an increase in the power generation and thermal efficiency. Hu et al. (2015) investigated off-design performance analysis of an ORC which is connected to a geothermal source and concluded that both work output and cycle efficiency increases as the condensation pressure decreases. In addition, Yamamoto et al. (2001) showed in their study that for dry fluids, best thermal efficiency is achieved when the refrigerant is saturated vapor at the inlet of the expander; which is validated, verified and expanded for isentropic fluids by Dai et al. (2009). Maraver et al. (2014) stated that such conclusion cannot be withdrawn for wet fluids, where re-heating and superheating causes an increase in the efficiency. Yang et al. (2014) pointed out that the efficiency increases as the amount of superheating is increased, for wet fluids. Abovementioned studies findings are founded on different constraints, hinting that there is no forthright generalized conclusions for optimal conditions. Yet it is possible to visualize a characteristic map for an ORC through a thermal energy input versus efficiency plot. In fact, this study demonstrate how such maps can be populated by varying mass flow rates and temperatures of hot and cold streams and how it would be useful during power plant design as well as during devising control strategies for ORC manufacturers. Present study utilizes a novel numerical model for finding steady state operating conditions of an ORC for a given input of inlet temperature and the flow rate of the hot heat transfer fluid (HTF), inlet temperature and the flow rate of the cold HTF and flow rate of the working fluid. A 25 kw ORC is used to demonstrate its off-design performance characteristics by changing mass flow rates and HTF temperatures by 238

253 20% up and down. In the model, it is shown that heat input is a function of mass flowrates of hot HTF, cold HTF and working fluid; therefore, it is possible to achieve same heat input when at least two of these parameters are changed simultaneously. Likewise, change in only one of these parameters should always require a different thermal energy input for its steady state operation. It is of great importance that it is possible to have different cycle efficiencies, even if the heat inputs are the same, because of having different mass flow rates. This phenomenon is shown by a single plot of efficiency vs. heat input in the latter sections of the current report. Methodology of the current study and mathematical model developed is explained in detail in the following section, which is followed by presenting results found and conclusions drawn. II. Numerical method A numerical method is used to quantify performance of a fixed ORC system, which consists of pump, evaporator, expander and pump. Input parameters to the model are as follows: (i) Inlet temperature and the flow rate of the hot HTF (ii) Inlet temperature and the flow rate of the cold HTF (iii) Flow rate of the working fluid First, condenser pressure, p L, is arbitrarily set to an initial value and temperature at pump s inlet, T 1, is taken as saturated liquid temperature of p L. Using the pump s characteristic curve (f p1 ), evaporator pressure, p H, is calculated at the specified refrigerant flow rate, m r, as follows: p= m r/ρ f@pl (1) p H p L = f p1 ( p) (2) η p = f p2 ( p) (3) W p = p(p H p L ) η p (4) T 2 T sat@pl (5) After calculating evaporator temperature, T 2, evaporator model is utilized in order to be able to find its exit state using effectiveness-ntu method for a counter-flow heat exchanger: Q ev = ε ev C p,h (T h1 T 2 ) = m r(h 3 h 2 ) (6) ε ev = f ev (m h, T h1, m r, T 2 ) (7) h 3 = h 2 + Q ev m r (8) T 3 = φ r (h 3, p H ) (9) Expander s exit is found following a similar approach used in pump model, as follows: p H p L = f e1 (m r) (10) 239 η e = f e2 (m r) (11) W e = p(p H p L )η e (12) h 4 = h 3 W e m r (13) T 4 = φ r (h 4, p L ) (14) Finally, condenser outlet is found using a similar methodology of evaporator model as it depends on mass flow rates and temperature differences at the inlets. Q co = ε co C p,c (T c1 T 4 ) = m r(h 4 h 1 ) (15) ε co = f co (m c, T c1, m r, T 4 ) (16) η cycle = (W e W p ) Q ev = F(p L, m r, m h, T h1 ) (17) However, Equation 14 does not guarantee that enthalpy at the condenser s exit to be equal to the enthalpy at the pump s inlet, yielding a non-zero energy balance equation for the overall cycle. Condenser pressure can be modified to influence both heat exchangers through evaporator and condenser temperatures in Eqns. (1)-(17) until energy balance equation is satisfied. This reveals that there is only one single value of condenser pressure that will enable steady state conditions for this cycle. In summary, cycle s thermal efficiency, thermal energy addition and removal becomes a function of mass flow rates and temperatures as provided as inputs to the model: η cycle = F 1 (m r, m h, T h1, m c, T c1 ) (18) Q in = F 2 (m r, m h, T h1, m c, T c1 ) (19) Q out = F 3 (m r, m h, T h1, m c, T c1 ) (20) W net = F 4 (m r, m h, T h1, m c, T c1 ) (21) As it can be seen from above relations, Q in and Q out are a function of mass flowrates of cold HTF, hot HTF and WF. Therefore, changing any of these parameters would always yield in a different heat transfer rate at steady state. Likewise, if it is desired to attain same level of Q in or Q out, at least two of the mass flowrates have to be changed simultaneously. This suggests that it is possible to have the same cycle in two separate cases for an ORC for which Qin is same but conversion efficiencies are different. IV. Results and discussions The described methodology is applied for a typical ORC unit to produce its characteristic map. The selected ORC s rated power is 25 kwe at its design conditions defined in Smith et al. (2007) as summarized in Table 1.

254 Table 1. Performance characteristics of ORC at design conditions Parameter Value mr [kg/s] 1.79 Tc [ 0 C] 21 mc [kg/s] 7.95 Th [ 0 C] 100 mh [kg/s] 2.65 Wgen [kw] 25.3 Wnet [kw] 24.1 Qin [kw] 447 Qout [kw] 422 Pl [kpa] 204 PH [kpa] 695 η [%] 5.41 In addition to this, increasing mass flowrate of cold fluid would result in an efficiency increase. However, such straightforward conclusion cannot be withdrawn for m h. Although m r always increases efficiency for lower efficiencies, it starts to have a parabolic behavior after around 0.7 m r. Even more, if the mr is significantly high (~ >2mr), efficiency tends to drop for every increment of m h. In addition to these, the effect of m h increase as mc increases and the magnitude is inversely proportional with m r. So, in order to adjust ORC to obtain maximum efficiency for varying conditions, priority should be given first to m r, then either m c or m h depending on m r. As can be seen from Table 1, useful work generated is 25.3 kw, which is in agreement with the reference. Related T-s diagram can be found in Figure 1. Fig. 2 Possible efficiencies for the current ORC Fig. 1 T-s Diagram of ORC at Design Conditions All five of the parameters are varied by 20% up and down of their original values in order to be able to work on off-design conditions. Increments are adjusted to 10% to have a reasonable simulation time while working on ordinary PCs. Figure 2 shows efficiency versus heat input graph in which 5 5 points exist. Following any constant x-line of Figure 2, there is no unique efficiency for the same heat input; instead, there exists different possible efficiencies for a given heat input. Keeping in mind that every point in Figure 2 represents a different input vector, it is possible to manipulate these inputs in a way that efficiency is increased as much as possible for the current ORC configuration. Providing different power inputs to pumps can yield in higher efficiency even if the heat input is not adjusted to be constant. The drawback of Figure 2 is that the input vector of each point cannot be gathered. Therefore, in order to have a better understanding of the phenomena, Figure 3 is plotted by grouping the same refrigerant flow rates for constant Tc and Th. It is important to state that m r and m c values increase from left to right, and bottom to top respectively. From Figure 3, it is easily seen that decreasing mass flowrate of refrigerant causes an increase in efficiency. Fig. 3 Performance map of ORC ORC approaches to maximum efficiency if the dry/isentropic refrigerant is at saturated vapor state at the inlet of expander. Such conclusion is only partially viable as efficiency is a function of Qin and not all of the maximum efficiencies corresponding to a particular Qin range obey to that assumption. In order to explain this issue clearly, Pareto frontier is drawn and shown in Figure 4. Within a variance of 20% of input parameters, Pareto frontier of ORC is shown in Figure 4; so the plot shows the highest efficiencies, with respect to Qin, that can be achieved with the current ORC. Sharp increase and decreases in the plot can be overcome if the change in variables, which is 10% as stated before, is reduced. 240

255 Fig. 4 Pareto Frontier for ORC Furthermore, the most efficient point corresponding to Qin range of, for instance kw, is shown on Figure 4 as well. Temperature entropy of that case is shown in Figure 5. Fig. 6 T-s diagram of the most efficient state for a heat input between 675 and 700 kw Fig. 5 T-s diagram of the most efficient state for a heat input between 150 and 175 kw As it can be seen from Figure 5, the refrigerant is still at a saturated mixture state at the expander inlet even though the highest possible efficiency is achieved for a Qin range of kw for the current, already designed, ORC. So, it is a must to have saturated vapour at turbine inlet to achieve the highest efficiency for dry / isentropic conclusion made in literature is not de-facto. In addition to these, if T-s diagram of the most efficient case for 675 kw < Qin < 700 kw is investigated, it is seen that the refrigerant is superheated at the turbine inlet, as can be seen in Figure 6. Even though all the related T-s plots are not presented here for brevity, as heat input increases, state at inlet expander corresponding to the most efficient case for that specific heat input- shifts from left to right, i.e. quality increases if the refrigerant is saturated mixture or degree of superheating increases if it is superheated. Fig. 7 Importance of x for the same Qin Finally, even if same quality is achieved for a higher Qin, the efficiency decreases as shown in Figure 8. Furthermore, lower saturated mixture quality results in lower efficiency as the heat input increases as can be seen in Figure 8. Also, as Qin increases, effect of x value increases as well, resulting in the fact that a drastic increase in quality would cause significant efficiency increase for high Qin. In addition to these, for the same heat input, efficiency peaks as the mass flowrates are adjusted in a way that quality is achieved as high as possible. This phenomenon is shown on Figure

256 s C x m W Q : Entropy (W/K) : Heat capacity rate (W/K) : Quality : Mass flowrate (kg/s) : Volumetric flowrate (m3/s) : Power (W) : Heat transfer rate (W) Fig. 8 Importance of x for different Qin V. Conclusions In this paper, optimal off-design conditions of a solardriven 25 kw ORC was investigated in order to find out how mass flowrates influence system parameters. It is illustrated that heat input is a function of m r, m c and m h when HTF inlet temperatures are fixed. It is possible to have two different ORC cases having same Q in but with a different efficiency. It is recommended that ORC manufacturers supply a performance map to assure that the system that it is integrated to operates around optimal off-design conditions. Moreover, it is shown that a generalized perfect state does not exist, even for dry/isentropic fluids that provide the most efficiency as this perfect state is a function of Qin. In addition, refrigerant flow rate is identified as dominant influencer on the cycle efficiency. Finally, it was found out that drastic changes in mass flowrates of hot and cold streams would cause a drastic change in efficiency only for high Qin. This work can further be improved by working with different refrigerants and power outputs. Acknowledgements This research was supported by a Marie Curie International Reintegration Grant within the 7th European Community Framework Programme. Nomenclature T p h : Temperature (0C) : Pressure (kpa) : Enthalpy (J/kgK) 242 Greek letters : Viscosity (Pa.s) η : Efficiency ԑ : Effectiveness ρ : Density (kg/m 3 ) Subscripts h : Hot H : High L : Low c : Cold r : Refrigerant in : Into the ORC out : From the ORC max : Maximum References Dai, Y., Wang, J., and Gao, L., Parametric optimization and comparative study of organic Rankine cycle (ORC) for low grade waste heat recovery, Energy Convers. Manag., 50(3), pp Fu, B. et al., Effect of off-design heat source temperature on heat transfer characteristics and system performance of a 250-kW organic Rankine cycle system. Applied Thermal Engineering, 70, pp Hu, D. et al Preliminary design and off-design performance analysis of an Organic Rankine Cycle for geothermal sources. Energy Conv. Manag., 96, pp Maraver, et al., Systematic optimization of subcritical and transcritical organic Rankine cycles (ORCs) constrained by technical parameters in multiple applications, Appl. Energy, 117, pp Smith, I.K. et al., Cost effective small scale ORC systems for power recovery from low enthalpy geothermal resources. Geothermal Resources Council Transactions, 31. Song, J., Gu, C., Ren, X. Parametric design and offdesign analysis of organic Rankine cycle (ORC) system. Energy Convers. Manag., 112, pp Vélez, F. et al., A technical, economical and market review of organic Rankine cycles for the conversion of low-grade heat for power generation. Renewable and Sustainable Energy Reviews, 16(6), pp Yamamoto, T. et al., Design and testing of the

257 Organic Rankine Cycle. Energy, 26(3), pp , Yang, K. et al., Effects of Degree of Superheat on the Running Performance of an Organic Rankine Cycle (ORC) Waste Heat Recovery System for Diesel Engines under Various Operating Conditions, Energies, 7(4), pp

258 One-Dimensional Transient Thermal Model for Parabolic Trough Collectors Using Closed-Form Solution of Fluid Flow Eray Uzgoren 1* 1 Mechanical Engineering, Middle East Technical University, Northern Cyprus Campus, Kalkanli, Guzelyurt, TRNC, Mersin 10, Turkey * uzgoren@metu.edu.tr Abstract The present paper develops a one-dimensional transient thermal model for parabolic trough collectors. The novelty of the model is the use of a closed-form solution of a linearized one-dimensional transient energy equation for the fluid flow. Any time-dependent variations of the parameters is captured by updating them in a time advancing numerical scheme. The analytical solution can be used alone to predict HTF s temperature profile provided that the overall heat loss coefficient from the fluid to the surroundings is already known a priori. In case it is unavailable, the fundamental model can be extended to include the thermal network for the absorber tube and the glass cover. The model is shown to capture both short- and long-term response of parabolic trough collectors at an accuracy level similar to existing one-dimensional models available in the literature while it is by far superior in terms of computation time as it eliminates spatial discretization. Keywords: Parabolic trough collector; transient convective heat transfer; closed-form I. Introduction Parabolic trough collectors (PTCs) direct beam solar radiation to a line of focus using a reflective parabolic shaped surface. A typical performance measure for PTCs accounting for optical and thermal losses is its efficiency, which is the ratio of energy gained by the heat transfer fluid (HTF) to the available direct solar radiation. Tube alignment/straightness, collector s reflectivity, geometric concentration ratio as well as irregularities affect the optical losses while thermal losses can be reduced simply using a glass cover. Annulus, the space between the absorber tube and the glass cover, is kept under vacuum pressure for high temperature collectors while low cost collectors contain air at atmospheric pressure due to manufacturing and mechanical challenges. There are a large number of studies developing or using numerical models for predicting PTC s thermal behavior in terms of thermal efficiency, thermal losses and HTF s outlet temperature. These models differ in the way that they select the thermal network, and simplifications made in governing equations. Fig. 1 illustrates a detailed and a simplified thermal resistance network examples. Fig. 1. Cross-section of a parabolic trough collector depicting absorbed heat with thermal losses and representative thermal resistance network. 244

259 Many studies focus on thermal losses by investigating steady behaviour while time adaptive systems need to rely on unsteady models. For example, Odeh et al. (1998) represented the experimental data of Dudley et al. (1994) through curve-fitting to predict thermal losses as a function of absorber and ambient/sky temperatures. Odeh and Morrison (2006) applied this representation in a lumped transient model to simulate and optimize a solar water heating system. Forristal (2003) considered a detailed thermal network with the thermal losses at the end brackets and solved zero- and one-dimensional steady energy equations using Engineering Equation Solver (EES) and concluded that one-dimensional equations including axial directions are only necessary for long collectors. System level models rigorously attempt to simplify the governing equations of the solar field to reduce the computation time. For example, Patnode (2006), while studying a full SEGS VI power plant simulation using TRNSYS, considered 1D steady state model for the solar field due to the computational overhead while modeling other components, i.e. thermal storage, as time dependent. They claimed that their results are mostly reasonable except during the morning warm up conditions. Similarly, Padilla (2011) used a steady model to predict heat transfer fluid s temperature at the PTC s exit for driving a dynamic steam generator of a regenerative Rankine cycle. Recent examples at system level simulations (Quoilin et al. 2011; He et al. 2012), which use TRNSYS to simulate PTC coupled with organic Rankine cycles, and many others reveal the need for a simple, fast and yet sufficiently accurate model for the solar field to estimate solar field s outlet temperature. Garcia-Valledares and Velazquez (2009) developed a transient one-dimensional numerical model that adopts finite volume method for axial discretization. Their dynamic simulations are used to obtain steady state results which remained within 6% error margin against the data presented by Dudley et al. (1994). Zaversky et al. (2013) used a similar thermal network and hence a similar set of equations to develop a dynamic model using Modelica libraries. They focused on the trade-off between accuracy and computational time based on number of control volumes in axial direction. Their 6 hour PTC simulation using 48 nodes with 2 m spacing takes about 6 minutes using a standard desktop computer. They concluded that an annual simulation at this accuracy level can be computed in one week while using a single control volume would reduce it to 2 hours with loss of accuracy. All existing one-dimensional time dependent PTC models use spatial discretization techniques, i.e. finite volume or finite difference methods. The accuracy of these methods increase with the number nodes employed in the axial direction but bring computational burden. In addition, discretization brings further difficulties, such time step selection as discussed in Zaversky et al. (2013) and numerical 245 diffusion of discontinuities as identified by the present study. In this paper, a closed-form solution to an unsteady, one-dimensional for the heat transfer fluid is developed. The closed form solutions of the fluid flow so far have only been used in either steady state onedimensional, e.g. by Xu & Wiesner (2015) or unsteady but zero-dimensional models, e.g. by Odeh & Morrison (2006). The present model obtains a closedform solution for both unsteady and one-dimensional energy equation to track HTF temperature profile as a function of both time and axial position. The model is validated against two independent experimental data those are available in the literature. Both models presented in this study provides a simple, fast and yet accurate engineering tool for full dynamic annual simulations of PTCs at sub-hour intervals. II. Materials and methods The flow is considered to be incompressible, single phase, uniform, and one-dimensional. In addition, it is assumed that dimensionless velocity, Peclét number, is high enough (i.e. Pe > 100) so that the axial conduction is neglected. Temperature variation of the fluid in axial direction and in time, T f (x, t) can be obtained through the energy equation of the fluid flow, which is given as follows: πd ρc 2 T (x, t) + m c T p 4 t p (x, t) = η x optwi b (t) + Uπd(T T(x, t)) (1) where ρ is the heat transfer fluid s (HTF s) density, c p is the HTF specific heat, d is the inner diameter of the absorber pipe, and m is the mass flow rate, η opt is the optical efficiency of PTC, W is the collectors aperture, I b is the direct solar irradiance, U is the overall heat transfer coefficient and T is the ambient temperature. All material properties are assumed to be invariant of time and calculated at their respective film temperatures. Eqn. (1) is solved analytically considering that all other outer parameters are invariant in time, i.e. solar irradiance intensity, absorber wall temperature distribution is invariant in time. T t (x, t) + a T x (x, t) + bt(x, t) = c (2) where coefficients, a, b, and c, are defined as follows: a = 4m 4U b = ρπd2 ρc p d c = 4η opt W I b + 4U πρc p d 2 (3) ρc p d T (4) Solution can be obtained by considering steady state solution and transient solution separately. Steady state solution can be found by dropping the time dependent term, / t, which makes the solution independent of the initial condition but only a function

260 of the inlet temperature, T i, of the HTF flow. The solution is given as follows: T(x, t ) = c b + (T i c b ) exp ( b a x) (5) For the transient solution, Lagrangian approach, which focuses on a fluid element moving with the flow, is considered in the form of the following equation: dt (t) + bt(t) = c (6) dt where x is embedded in a material derivative. is summarized below: There is no limitation on time stepping in terms of numerical stability. In addition, time steps can be non-uniform in time. The present algorithm becomes similar to a steady state solver when large time steps are selected, i.e. Δt > τ. The algorithm reaches steady state in a few steps, which is necessary to handle linearized radiative heat transfer related terms. In the limit of infinitesimal time steps, i.e. Δt 0, exact solution is recovered. The solution of Eqn. (6) is considered for an initial condition with arbitrary temperature distribution, T o (x). For constant mass flow rate, the temperature of a particle reaching location, x, is originated from x o = x at, which transforms the initial condition as T o (x at) for Eqn. (6). Note that initial HTF temperature distribution, T o (x), is only defined for positive values of x o. When time, t, reaches a value yielding a zero or negative value for x o, the inlet condition, T i, is used instead of the previous condition, T o (x), which produces the steady state solution as given in Eqn. (5). The overall solution can be represented as follows: c b T(x, t) = { + (T o(x o ) c ) exp ( bt) b t τ c + (T b i c ) exp ( b x) t τ (7) b a Fig. 2. HTF inlet and exit temperatures comparing the developed model against the experimental data. III. Results and discussion τ = ρπd 2 ai x (8) 4m Eqn. (7) represents a discontinuous function. It only becomes continuous provided that T o (x = 0) = T i while its derivatives remain discontinuous. Difference operators in discretization schemes for a discontinuous solution bring additional numerical errors, i.e. numerical diffusion unless special care such as front tracking is employed. This is illustrated in Fig. 2, which shows the difference between the analytical solution using Eqn. (7) and 1D finite volume discretization. Fig. 2 represents a water flow with a mass flow rate of kg/s in a 7 cm diameter absorber pipe with a loss heat transfer coefficient of 600 W/m 2 K. The inlet temperature of the flow is 29.5 o C. Two cases are considered; initial temperature of 29.5 o C and 39.5 o C. The finite volume solution is obtained using 10 and 200 control volumes and compared against Eqn. (7) at three different locations. While both cases illustrates the numerical diffusion for discretization schemes, it is more prominent when initial and inlet HTF temperatures are different. Time and temperature dependent variation in solar beam irradiance, overall heat transfer coefficient, mass flow rates and ambient temperatures is incroperated using a time advancing scheme. Numerical consequences of the closed form solution 246 Many previously reported numerical models have utilized the SEGS LS-2 collectors data given in Dudley et al. (1994) for validation purposes. Similarly, results of the present study using the extended model is compared to their steady state results in terms of HTF exit temperature. In addition, time dependent experimental data for SOLTERM facility (Rome, Italy) as given in Zaversky et al. (2013) are presented to characterize the accuracy and the computational time of the developed model during a representative day. The specifications of the collectors considered in each study are presented in Table 1. Table 1. Collector specifications used in benchmark studies Dudley et al. (1994) Zaversky et al. (2013) HTF Water/Syltherm-800 Molten Salt Length/Width 7.8 m x 5 m 48 m x 5.76 m Optical 72.36% 79.3% Efficiency Absorber Tube Steel / Luz Cermet Coating Outer diameter 0.07 m 0.07 Thickness 3 mm 3 mm Absorptivity Emissivity 0.06 at 100 o C; 0.15 at 400 o C Steel / Cermet Coating 0.1 at 400 o C; 0.14 at 550 o C; 0.15 at 580 o C Outer diameter 0.07 m 0.07 Thickness 3 mm 3 mm Absorptivity Transmissivity Emissivity

261 III.1.Case 1: SEGS LS-2 collectors The developed model is compared with the time constant test of Dudley et al. (1994), in which cold water was used as the HTF and the absorber tube has Luz coating. The test was performed by letting the system operate at zero-flux condition to obtain stable conditions and suddenly placing collectors in-focus. Hata! Başvuru kaynağı bulunamadı. shows the comparison of the time-history of HTF s heat gain and efficiency between the developed model and the benchmark data. The model considers a constant solar beam radiance at a level 870 W/m 2 at 170 sec after 9:45 AM. The initial slope for both efficiency and heat gain for both the model and the experiment show a sharp increase while the changes with time slow down considerably around 9:49 AM. Considering that the developed model is a one-dimensional simplified model, Hata! Başvuru kaynağı bulunamadı. shows a reasonable agreement between the developed model and the experimental data. conditions during digitalization of plots, the currently developed model performs similarly with the numerical model. Their simulation at 30 seconds time intervals was performed on a standard desktop PC and took 343 seconds while the present model on standard desktop PC completed the whole 6 hour simulation in only The reduction in computational time by a factor of 303 is considerably advantageous. In addition, present model s accuracy is not affected by the size of the field whereas discretization techniques needs to increase the number of nodes for longer PTC assembly s to preserve the level accuracy. IV. Conclusions This work presents a one-dimensional dynamic thermal model for parabolic trough collectors (PTCs) which is based on the analytical solutions of onedimensional energy equation for the fluid flow. The validation of the developed model is carried out against independent experimental studies of Dudley et al. (1994) and Zaversky et al. (2013) to illustrate that the present model s accuracy is comparable to the other numerical thermal models currently available in the literature while it reduces the computational time significantly. It is also possible to integrate this algorithm within a more detailed thermal resistance network. This would result in an intrinsic estimation of the overall heat transfer coefficient. Fig. 3. Comparison of the model with the time constant experiment. III.2.Case 2: SOLTERM facility The developed model is also compared against the experimental data of Zaversky et al. (2013) in which the dynamic behavior of a collector field is examined. They collected a six hours long data at 30 seconds intervals in SOLTERM facility in Rome, Italy, for air temperature, normal solar irradiance, HTF mass flow rate, HTF s inlet and exit temperature. Because of the fluctuations in solar irradiance, their data is well suitable to verify the dynamic thermal behavior of the collectors. The experimental data for all outside conditions was extracted from Zaversky et al. (2013) using plot digitizer software. The data for mass flow rate, air temperature and solar beam radiance can be found in Zaversky et al. (2013) and not given to avoid repetition. Fig. 4 shows the comparison between the developed model and the data. The root mean square (RMS) error as defined in the benchmark study is found to be 3.2 o C where the experimental data has an uncertainty of ±2 o C. Numerical model developed by Zaversky et al. (Zaversky et al. 2013) using spatial discretization with 48 control volumes (CV) obtained an RMS error 2.32 o C. Considering the additional error in the input 247 The present models strength over already existing 1D models can be summarized as follows: Transient, 1D analytical solution for HTF is utilized to eliminate spatial discretization and all issues it brings, i.e. trade-off between accuracy and computation time, stability, etc. Computation time is reduced significantly enabling annual simulations at sub-hour intervals, Discontinuities are maintained to handle sharp changes at the inlet HTF temperature, It is concluded that the developed model can easily be used in system level simulations focusing on longer periods using shorter time steps. Annual simulations of a PTC field at 30 second intervals would be completed in minutes using a standard desktop computer and the computation time is independent of PTCs length. It is also possible to reduce this time by increase the time stepping based on the time constant of the collector. Acknowledgements This research was supported by a Marie Curie International Reintegration Grant within the 7th European Community Framework Programme. References Dudley, V., Kolb, G. & Kearney, D., SEGS LS2 Solar Collector - Test Results, Report of Sandia National Laboratories.

262 Forristall, R.E., Heat transfer analysis and modeling of a parabolic trough solar receiver implemented in engineering equation solver, National Renewable Energy Laboratory. Available at: troughnet/solarfield/docs/34169.pdf [Accessed September 1, 2014]. García-Valladares, O. & Velázquez, N., Numerical simulation of parabolic trough solar collector: Improvement using counter flow concentric circular heat exchangers. International Journal of Heat and Mass Transfer, 52(3 4), pp He, Y.-L. et al., Simulation of the parabolic trough solar energy generation system with Organic Rankine Cycle. Applied Energy, 97, pp Odeh, S.D. & Morrison, G.L., Optimization of parabolic trough solar collector system. International Journal of Energy Research, 30(4), pp Odeh, S.D., Morrison, G.L. & Behnia, M., Modelling of parabolic trough direct steam generation solar collectors. Solar Energy, 62(6), pp Padilla, R.V., Simplified methodology for designing parabolic trough solar power plants. University of South Florida. Available at: [Accessed February 11, 2014]. Patnode, A.M., Simulation and Performance Evaluation of Parabolic Trough Solar Power Plants. M.S. University of Wisconsin-Madison. Available at: Quoilin, S. et al., Performance and design optimization of a low-cost solar organic Rankine cycle for remote power generation. Solar Energy, 85(5), pp Xu, R. & Wiesner, T.F., Closed-form modeling of direct steam generation in a parabolic trough solar receiver. Energy, 79, pp Zaversky, F. et al., Object-oriented modeling for the transient performance simulation of parabolic trough collectors using molten salt as heat transfer fluid. Solar Energy, 95, pp

263 Design and Performance Analysis of Linear Fresnel Reflector Melik Ziya Yakut 1 *, Arif Karabuga 2, Ahmet Kabul 3, Resat Selbas 3 1 Suleyman Demirel University, Technology Faculty, Mechatronic Engineering, Isparta, 32260, Turkey 2 Suleyman Demirel University, Keciborlu Vocational School, Departmen of Electrical Energy Generation, Isparta, 32260, Turkey 3 Suleyman Demirel University, Technology Faculty, Energy Systems Engineering, Isparta, 32260, Turkey * ziyayakut@sdu.edu.tr Abstract Lighting, heating, transport, industrial output: without energy we would have none of these essential day-today services without which we and our businesses cannot function. Renewable energy refers to natural energy resources (solar, wind, bio gas, tidal, geothermal etc.) alternative to fossil and nuclear fuels. It s shortly re-existing energy resource in natural evaluation. Solar energy is most important part of the renewable energy resources such that is an essential building block for our future global sustainable energy system. Solar energy technologies have to two groups. The first group is photovoltaic power system and second group is thermal power system. Linear Fresnel Reflector (LFR) is one of the solar thermal energy applications, which focuses sunlight to heat the heat transfer fluid (HTF) circulating through the receiver. Linear Fresnel reflector is thermal power system. This paper, we researched design and performance analysis of linear Fresnel reflector. Keywords Energy, Renewable Energy, Solar Energy, Linear Fresnel Reflector I. Introduction The use of renewable energies for power generation is main issue it today s society in order to avoid energy dependence and reduce the impact of greenhouse emissions by Sait et al. (2015). Medium and high temperature heat can be produced by using concentrating solar technologies. Based on the reflector configurations, the solar concentrating technologies may be classified as: linear Fresnel reflector, parabolic trough, parabolic dish and power tower. Among these, the linear Fresnel reflector (LFR) system is simple in design and cost effective system for medium temperature (100 o C-400 o C) applications. The performance of LFR system significantly depends on the receiver design. The earlier research work on LFR collector has been reported in the following order. The design and performance investigation of LFR has been reported first and followed by heat loss analyses from the cavity receiver for LFR collector. (Reddy and Kumar, 2014). Linear Fresnel reflector (LFR) technology is regarded as a prospective method to concentrate solar power (CSP), solar industrial heating and solar cooling due to its simplicity in structural design and low manufacture costs. Generally, an LFR solar collector (Fig. 1) consists of three main components: mirror field, receiver and tracking system. The direct solar radiation can be reflected by an array of parallel mirrors to a fixed focal line at which the receiver is mounted. (Lin et al., 2013). The sun is the only star of our solar system located at its center. The earth and other planets orbit the sun. The sun has a surface temperature of approximately 5500 K, giving it a white color, which, because of atmospheric scattering, appears yellow. The sun generates its energy by nuclear fusion of hydrogen nuclei to helium. Sunlight is source of energy to the surface of the earth that can be harnessed via a variety of natural and synthetic processes. Basically all the forms of energy in the world as we know it are solar in origin. Oil, coal, natural gas and wood were originally produced by photosynthetic processes, followed by complex chemical reactions in which decaying vegetation was subjected to very high temperatures and pressure over a long period of time. Even the energy of the wind and tide has a solar origin, since they are caused by differences in temperature in various regions of the earth. Many alternative energy sources can be used instead of fossil fuels. The decision as to what type of energy source should be utilized in each case must be made on the basis of economic, environmental and safety considerations. Because of the desirable environmental and safety aspects it is widely believed that solar energy should be utilized instead of other alternative energy forms because it can be provided sustainably without harming the environment. Renewable energy resources is maybe most importantly, that wide usage area of solar energy. Solar energy is a basic need of living plants and human being on the earth. It is intermittent in nature, eco-friendly and nonpolluting energy. Solar energy can be used for direct conversion into electricity (by photovoltaic conversion) and into thermal energy. Further, thermal energy conversion can be classified into three categories namely; 1- Low temperature range (<10 C) 2- Medium temperature range (< C) 3- High temperature range (>150 C). The greatest advantage of solar energy as compared with other forms of energy is that it is clean and can 249

264 be supplied without environmental pollution. Over the past century, fossil fuels provided most of our energy, because these were much cheaper and more convenient than energy from alternative energy sources and until recently, environmental pollution has been of little concern. The sun s energy has been used by both nature and humankind throughout time in thousands of ways, from growing food to drying clothes; it has also been deliberately harnessed to perform a number of other jobs. Solar energy is used to heat and cool buildings (both actively and passively) heat water for domestic and industrial uses, heat swimming pools, power refrigerators, operate engines and pumps, desalinate water for drinking purposes, generate electricity for chemistry applications and many more operations. Solar collectors may be classified as non-tracking and tracking type. Non-tracking collectors are eternal fixed in position and do not track the sun. Three main types of collectors fall into this category; Flat-Plate Collectors (FPCs), Stationary Compound Parabolic Collectors (CPCs) and Evacuated Tube Collectors (ETCs). Also solar energy is used for a variety of heating systems such as domestic hot water systems and industrial applications. The use of solar energy for domestic hot water and steam generation in industry is economical and environmentally friendly. The most common way of using solar energy is through hot water by solar water heaters. Hot water is required for domestic and industrial uses such as houses, hotels, hospitals, and mass-production and service industries. Along with the development of utility-size solar power plants in the past few years, related performance acceptance testing standards are building up progressively. On one hand, the American Society for Testing and Materials (ASTM) has codified a Standard Test Method E2848 for photovoltaic (PV) power plants; on the other hand, procedures for testing fluid heating solar collectors are defined by ISO for solar thermal systems. In the frame of Solar Paces Task I regarding Concentrating Solar Power (CSP) plants, the testing guidelines provided by National Renewable Energy Laboratory (NREL) concerning only parabolic trough and power tower solar systems are widely applied, since the Performance Test Code referenced as PTC 52 is still under revision by a committee formed through the American Society of Mechanical Engineers (ASME). In this study, we present design and performance analysis of a linear Fresnel reflector. Such that linear fresnel collector system is investigated theoretical and because of this used EES (Engineering Equation Solver) software. Solar energy technologies have to two groups. The first group is photovoltaic power system and second group is thermal power system, which is used collector. Solar energy collectors are special kinds of heat exchangers that transform solar radiation energy to internal energy of the transport medium. The major component of any solar system is the solar collector. This is a device that absorbs the incoming solar radiation, converts it into heat and transfers the heat to a fluid (usually air, water or thermal oil) flowing through the collector. 250 Figure 1. Construction principle of linear fresnel collector Linear Fresnel collectors, as shown figure 1 use long, segment of mirrors, located at the focal axis of the different rows, to focus sunlight onto a fixed absorber located at a common focal point of the reflectors. The fixed absorber includes one or several absorber tubes. A secondary concentrator is used to reflect the rays within accepting angel. This concentrated energy is transferred through to absorber into thermic fluid. Through a heat exchanger energy is extracted to use to generated power or other commercial applications. Linear Fresnel Reflectors (LFR) shows a number of advantages when compared to other CSP technologies when it comes to industrial applications, including the following: Direct Steam Generation capability (capability possible in other CSP technologies but rarely used), eliminating the need for heat exchangers that increase plant costs and reduce thermal energy production efficiency; Highly variable temperature and pressure ranges that can be adapted depending on the industrial application; Lowest land occupancy, knowing that a large chunk of steam-consuming facilities are located in industrial zones where land availability is scarce; High modularity, ranging from a few hundred kilowatts to several megawatts; Low environmental impact due to the limited raw material use and deletion of synthetic fluids (capability possible in other CSP technologies but rarely used); Lowest levelized cost of energy (LCOE) due to its modularity, simple and efficient design, DSG and low O&M requirements. Also fresnel reflectors can be very useful for solar energy, especially in large installations. The potential integration of LFR technology has been studied for a wide spectrum of industries that use steam for thermal applications. These applications address direct human needs such as water and food,

265 assistance of other energy / energy-consuming industries such as Oil-Gas, petrochemicals and mining, and temperature regulation needs. II. Design and Performance of Linear Fresnel Collector Skew of each mirror element is very important in optical design. Position of the sun, relative to the axis of rotation of the linear Fresnel reflector elements, is determined from the solar profile angle. In this linear Fresnel reflector, skew of each mirror element is chosen such that a ray normal to the plane of the aperture of the collector. The distance between two adjacent mirror radiation reflected from any mirror. Tubular absorber of appropriate size placed in the focal plane of linear Fresnel reflector solar concentrator would intercept all the solar radiations from the constituent mirror elements. The design has been used in 10 reflective mirrors. Each mirror is chosen 610 mm and 2000 mm in width. II.1. Design Parameters The location at the first mirror, Q1 = R + f tan (ξ0) (1) Figure 2. Geometry of linear Fresnel reflector Here; Q is the distance of the first mirror, R is the radius of the absorber, f is the focal distance from the aperture plane, ξ 0 is the subtends angle of the sun. Each mirror value (angle, distance, space) for; Q n = Q n 1 + W cos(θ n 1 ) + S 2 (2) S n = W sin(θ n 1 ) tan(2θ n + ξ 0 ) (3) θ 1 = 1 2 tan 1 [( Q 1+( W 2 ) cos(θ 1 ) f W )] (4) 2 sin(θ 1 ) Here; Qn is for mirror element located at a distance from the centre of aperture plane. W is width of mirror. θ is geometrical optics relations the skew of mirror element. Q 1 = W 2 + f tan(ξ 0) (5) Then; Q 1 = 305,2 mm 251 By considering the all above values the values of skew, shift and location can be calculated by EES the above equations for each mirror element using iteration method. Table 1. Admissible values Design Parameters Width of mirrors (W) 610 mm Focal distance from plane (f) 2500 mm The sun s the subtends angle (ξ0) 16 =0, radians Table 2. Reflector location parameters Mirror No Location (mm) Shift (mm) Skew (degree) 1 305,2 0 6, ,5 37,75 13, , ,5 24, ,5 27,95 II.2. Thermal Parameters Two different heat sources are considered when Linear Fresnel Reflector calculation process. The first thermal energy carried by the solar radiation reflected from the primary mirror, and the second is secondary reflector at the cavity part. In addition, outside of the reflector, the cavity region taken into direct radiation from the sun. Furthermore, there is heat loss from the cavity to the environment. Therefore, thermal analysis of cavity shows crucial importance for system performance. Table 3. Admissible values for thermal analysis Transmissivity of glass cover 0,95 Reflectivity of mirrors 0,92 Absorbivity of absorber tube 0,92 Emissivity of absorber 0,15 Ambient temperature 30 C Temperature of absorber tube 200 C Reflectivity of CPC 0,92 Inlet temperature of thermic fluid 25 C Glass covers temperature 50 C Heat transfer coefficient (convective) 1,445 W/m 2 K Mass flow rate of thermic fluid 0,394 kg/s Specific heat of thermic fluid at mean 2173 J/kg K operating temperature (100 C) Radius of absorber tube 0,0265 m Length of absorber tube 2 m The heat loss of cavity due to natural convection: q conv = h p a πd 0 ΔT (6) here, h p a = 1,445 W/m 2 K is considered. The heat loss of cavity due to radiation: q rad = σεπd 0 (ΔT 4 ) (7) here, σ is constant of Stefan-Boltzmann and e 8 Wm 2 K 4 is considered. ε is emissivity of absorber it is considered 0,15. Total heat loss We add convection and radiation losses in order to find the total heat loss. q total = q conv + q rad (8)

266 the heat loss per unit length of absorber: q L L = q total (9) here, L is length of absorber and it is 2 meters. 0,9 0,8 0,7 h The heat loss from glass cover to ambient is given by due to convection: q glass = h glass A(T glass T ambient ) (10) Herein, hglass is heat transfer coefficient caused by the wind through the glass covered tube. Wind speed is considered 5 m/s. h glass= V (11) The total heat quantity, the following equation is used to find: Q = (I b q reflector q CPC τ glass ) (12) Equation 13 is solved to find the net amount of heat. (Q q L q glass ) = m c p (T out T in ) (13) To find the efficiency of the system; h Tout 0,6 0,5 0, I b Figure 3. Efficiency changes depending on the radiation intensity T out η = (Q q total q glass Q) (14) III. Results and discussions Design parameters and thermal analysis are prepared with EES (Engineering Equation Solver) program code. The results of the analysis in Table 4 are shown. Table 4. Analysis Result I b η Q T out When the radiation intensity is increased on the system also heat and outlet temperature increased. Radiation intensity 300 to 600 increased, efficiency 0.61% heat 100 and outlet temperature increase of 0.24%. Depending on the intensity radiation efficiency curve changes in the Figure 3, changing in the outlet temperature graph shown in the Figure 4 is shown. Figure 4 is shown 300 to 1000 W/m 2 radiation values ranging, on the linear increase in temperature of the outlet water from the system I b Figure 4. Outlet temperature of water changes due to radiation intensity V. Conclusions In conclusion, this study's design and performance analysis of linear Fresnel reflectors are made. The values obtained from the analysis results show that a linear fresnel reflector applicability is possible. LFR systems, if possible in a closed system can reach high temperature. When the heat loss is analyzed that a significant heat loss in the cavity part. The isolation is required in order to prevent these losses. With the developing technology of this cavity should be made closer to real thermal modeling analysis. References Kalogirou S. A., Solar Energy Engineering: Processes and Systems, Academic Press, Cyprus, (2009) Tiwari G.N., Solar Energy Fundamentals, Design, Modelling and Applications, Alpha Science, Delhi, (2002). Kocer A., Atmaca I., Ertekin C., A Comparison of Flat Plate and Evacuated Tube Solar Collectors with F- Chart Method, Journal of Thermal Science and Technology, 35, 1,77-86 (2015). Itskhokinea D., Lècuillier P., Benmarrazea S., Rabut Q., Guillier L., Fresnel 1 Project: Design, Construction And Testing Of A Linear Fresnel Pilot 252

267 Plant in the Pyrenees, Retrieved 10st June, 2015, from (2015). Yanga F., Itskhokinea D., Benmarrazea S., Benmarrazea M., Hoferb A., Lecatc F., Ferrièrec A., Acceptance Testing Procedure for Linear Fresnel Reflector Solar Systems in Utility-Scale Solar Thermal Power Plants, Retrieved 10st June, 2015, from (2015). Hobeika S., Alexander B., Benmarraze S., Itskhokine D., Yang F., Benmarraze M., Case study: Linear Fresnel Reflectors (LFR) solar systems for industrial applications, Retrieved 10st June, 2015, from (2015). Rabl A., Active Solar Collectors and Their Applications, Oxford University Press, USA, (1985). Nixon, J. D., Davies, P. A. Cost-Exergy Optimisation of Linear Fresnel Reflectors, Solar Energy, 86 (2012), (2012). Gouthamraj, K., Rani, K. J., Satyanarayana, G. Design and Analysis of Rooftop Linear Fresnel Reflector Solar Concentrator, International Journal of Engineering and Innovative Technology (IJEIT), Vol. 2, Issue 11, pp: (2013). Dostuçok, İ., Selbaş, R., Şahin Şencan A. Experimental Investigation of A Linear Fresnel Reflector System Journal of Thermal Science and Technology, 34, 1, (2014). 253

268 Exergy Analysis of a Solar Photovoltaic Panel within Karabük Climate Conditions Mutlucan Bayat 1*, Mehmet Ozalp 2 1,2 Karabuk University, Engineering Faculty, Mechanical Engineering Department, Karabük, 78200, Turkey * mutlucanbayat@karabuk.edu.tr Abstract In this study, experimental and numerical investigations were made on silicon based a polycrystalline solar panel which has a nominal power of 130 Watt-peak (Wp). In order to carry out exergy analysis, a PV system was installed on the top of the Engineering Faculty in the Karabuk University, Karabük province, Turkey. To perform analysis, measurements on a PV module have been carried out between 9 am to 17 pm in a 30 minute intervals during November. For defining and evaluating exergy efficiency rates, electrical parameters of the module e.g. nominal current-voltage (Imax-Vmax), open circuit voltage (Voc), and short-circuit current (Isc) have been measured. Thanks to the obtained data, the daily maximum power point (Pmax) of the panel is defined for each day, and the variation of measured parameters associated with each others is daily examined. Besides, atmospheric parameters such as global solar radiation, the ambient and panel temperature as well as wind speed have been recorded since environmental conditions affect the real working performance the module. This effect of seasonal and climatic changes on the PV performance has been researched and exergy efficiency values have been compared on selected days. Besides efficiency rates, all exergy components of a PV including exergy input, exergy destruction, thermal exergy and electrical exergy was analysed. In order to illustrate the obtained results, MATLAB has been used and generated graphs are interpreted. Keywords: Exergy analysis, PV system, Poly-si, thermodynamic approach, MATLAB, climate condition I. Introduction Energy has always become a crucial factor for continuity of human life. Despite the increasing energy demand due to improved living standards all over the world, reducing dependence on foreign energy is therefore essential. For this purpose, developing installed electricity capacity with power generation technologies that are environmentally friendly and have high energy efficiency is needed. The solar power is affordable, inexhaustible and clean source of energy. Employing solar energy therefore brings enormous benefits considering enhance sustainability, decrease pollution and limiting global warming. It is not only a powerful option to reduce environmental concerns; it provides also an indigenous solution for diminishing use of fossil fuels and keeping clean power generation prices lower Luther (2013). The solar photovoltaic (PV) cell is an electrical device that converts the energy of light into direct-current electricity using semiconducting materials that exhibit the photovoltaic effect, which is called physical and chemical phenomenon. The working principle of cell starts with the absorption of light in which the electrons present in the valence band are being excited and become free. Thus, photons whose energy is equal to or greater than the band gap in semiconductor materials used in making cell constitute electron hole pairs or excitons at first. As the second step involves the separation of charge carriers of opposite types in the cell, separating extraction of those carriers to an external circuit takes place in the last phase. Consequently, PV cells as working a semiconductor diode convert carrying sunlight energy into directly electricity by utilizing internal photochemical reactions Fahrenbruch and Bube (1983). Since the process of direct conversion of sunlight occurs stationarily or motionlessly and it doesn t release environmental emissions during operation, PV applications have been more attractive and now used in industries for a few decades Bazilian et al. (2013). Besides, advances in PV technology has grown the market rapidly and drawn attention from policy makers. Therefore, power generation from solar PV has now been seen as one of the most promising clean and sustainable energy technology. With the increasing interest in solar energy, there have been many studies regarding both characterisation of cells and performance analysis with the thermodynamic approach. For this point of view; energy, entropy and exergy concepts as well as their importance and roles for thermodynamic systems have been discussed extensively by Dincer and Cengel (2001). Thermodynamic limitations in energy systems through solar energy conversion with the entropy balance has been investigated by Wurfel (2002), showing that an upper efficiency of 0,86 for maximally concentrated solar irradiation in any conversion process. Classification of radiation and calculation on exergy of the thermal radiation has been analysed by Petela (2003). Smestad (2004) studied evaluation of solar converter under the 254

269 photoelectric effect and the concept of carrying electrons that are being excited due to both heat and light during energy conversion. Exergy and sustainability has been linked by Dincer (2005), indicating that exergy method has significant impact on both environmental and sustainable development for assessing thermodynamic systems. Sahin et al. (2006, 2007) has evaluated the thermodynamic characteristics of wind energy systems and analysed solar PV cell systems from a perspective based on energy and exergy. Hepabsli (2008, 2015) conducted on dynamic exergy analysis, indicating sustainability index and improvement potential factor and also stated that assessment of renewable energy resources for a sustainable future. Joshi et al. (2009, 2011) has analysed performance characterisation of photovoltaic thermal system (PV/T) via PV simultaneously, in terms of exergy analysis and weather dependences. Several researchers Dubey and Tiwari (2010), Cuce and Cuce (2012), Saloux et al. (2013), Sobhnamayan et al. (2014) have also performed different significant research on the solar PV and PVT systems. They all have attemted to obtain the real performance of solar systems in various climatic conditions with the different perspective. In this present study, an experiment has been made to investigate following purposes stated below 1. applying an exergy analyses for the solar PV system, 2. developing a realistic model for predicting solar PV performance within a specified location, 3. defining the thermodynamic considerations based on exergy components (electrical and thermal exergy, exergy loses etc.), and 4. assisting to close the gaps in literature. II. System Description In this experimental study, a silicon based polycrystalline solar PV module was performed throughout November in Karabük province in Turkey. In order to carry out thermodynamic analysis, a PV system was installed on the top of the Engineering Faculty, Karabuk University, (41.12 N, W) Karabük, Turkey. There are 72 polycrystalline PV panels as seen in Fig. 1. Each has a rated power output of 130 Wp with a rated voltage of 21,9 V. The specific features and dimensions of each PV module have been given in Table 1. Here, the panel properties have been presented in standart test conditions including 1000 W/m 2 global solar radiation, 1,5 Air Mass (AM) and 25 o C ambient temperature. Fig.1: The installed PV sytem Tab. 1: The panel properties within standart test conditions IBC PolySol 130 GC Technical Data Nominal peak W p power Nominal voltage V 18.0 Nominal current A 7.23 Open circuit voltage V 21.9 Short-circuit current A 7.9 Temperature %/K coefficient of I sc Temperature mv/k coefficient of V OC Temperature %/K coefficient of P max Power conversion % efficiency Power Tolerance % ±2.5 Fill Factor (FF) Number of cell - 36 Length mm 1500 Width mm 670 Height mm 42 Weight kg 12.0 Effective Area m III. Experimental Setup To perform analysis, measurements on a PV module have been carried out between 9 am to 17 pm in a 30 minute intervals during November. For defining and evaluating exergy efficiency rates, electrical parameters of the module e.g. nominal current-voltage, open circuit voltage and short-circuit current have been measured. Besides, atmospheric parameters such as global solar radiation, the ambient and panel temperature as well as wind speed have been recorded since environmental conditions affect the real working performance the module. For measuring parameters mentioned above, some intruments have been utilized. For instance, the panel backside temperature measurement is performed with the calibrated digital thermocouple and thus, the temperature data has been provided to be taken in the middle of the module. These measurement instruments are clearly listed in Table 2. Besides, some devices in this list have given in Fig. 2, and the experimental setup of the panel backside as well as connection lines have been presented in Fig

270 Table 2: Parameter list and measurement instruments Parameter Notation Instrument Module T cell Lascar EL-USB-TC-LCD temperature digital thermocouple Ambient T amb Thomas traceable digital temperature thermometer Solar radiation S T MS-410 Pyranometer Wind velocity ϑ wind Delta OHM HD2303 digital anemometer Open circuit V oc 50 Ω wirewound rheostat voltage Short-circuit current I sc 50 Ω wirewound rheostat Nominal voltage V max UNI-T UT61B digital multimeter Nominal current I max MY-68 digital multimeter Fig.2: Some measurement instruments used in experiment Therefore, exergy is not only a significant feature for evaluating thermodynamics systems, but also a co-property of a system and the reference environment which must be specified for analysis Cengel (2012). The traditional method of assessing the energy disposition in either physical or chemical processes involving conversion of energy based on the first law of thermodynamics and this operation is by the completion of energy balance. This balance brings sufficient data for identifying heat loss and can be utilized for enhancing heat recovery by applying the system specification. However, it does not give details on the reason of energy degradation occurring in the process as well as useful part of energy and its quality in the system. Exergy analysis is a method that overcomes the limitations of the first law of thermodynamics. With this approach, the locations of energy degradation that may be enhanced with the advanced system or technology can be clearly indicated in the process. Besides, unlike the energy analysis, the quantity and quality of energy which is not conserved but is partly lost or destroyed in any real process can be measured Sahin et al. (2007). Thus, this analysis is seen as a highly effective and promising method, since the main objective of analysis is to identify the potential or the maximum amount of work produced by a system that destroys exergy, and to calculate exergy loses. Therefore, exergy analysis provides a better and finer understanding on system s behaviour that cannot be obtained from energy analysis alone. Moreover, it is important to highlight that exergy analysis can reveal whether or not and by how much it is possible to design more efficient energy systems by reducing the inefficiencies in existing systems Dincer and Cengel (2001). For exergy analysis, the state of reference environment must be identified, since the results of analysis are corresponding to the specific conditions that reference environment has. Therefore, in this paper, actual environment conditions during a month in Karabük, Turkey are taken as the reference state or environment properties. Fig.3: The experimental setup and connection line IV. Exergy Analysis The exergy of a system is zero when it is in equilibrium with the reference environment. Hence, exergy balance for a steady-state and steady-flow open system can be expressed as It is known that the science of thermodynamics based on the two basic principles, called the first and the second laws. The first law of thermodynamics is built primarily on the conservation of energy in which total amount of energy remains constant during interaction, even if the energy can be existed in different forms. Another aspect of energy is also tackled in the second law which asserts that the quality and quantity of energy should be considered together. This description brings out the exergy concept which based on both the first and the second law therorem. 256 in ex in ṁ in out ex out ṁ out + k Ėx Q Ėx W I = 0(1) where ṁ in and ṁ out refers mass input and output through the system, whereas ex in and ex out denotes specific exergy input and output respectively. Ėx Q and Ėx W (W) specify the exergy transfer occurring on the system boundry and the exergy transfer due to the work, correspondingly. Considering a PV system, since ṁ in = ṁ out = 0 for a closed system, Equation (1) simplify in that case to

271 Ėx Q k Ėx W I = 0 (2) Here, I (W) represents the exergy consumed due to irreversibilities during a process. Thus, the system exergy consumption I is greater than zero for an irreversible process and equal to zero for a reversible process. It is also worth to note that irreversibility accounts for the amount of exergy destroyed in a closed system, or in other words, the wasted work potential. For highly efficient systems, the value of I, is low, and vice versa. The equation to calculate the irreversibility of as closed system, as it relates to the exergy of that system, is as follows: I = T amb Ṡ gen (3) where T amb (K) denotes the ambient temperature and Ṡ gen (W/K) represents the entropy generation through the system and surround. Hence, by this definition, Ṡ gen can be calculated using physical exergy components associated with C p and Q loss which are the specific heat of the silicon used in PV cells and the heat losses on the PV surface under the sunlight, respectively. where V max and I max represent the voltage and current at the maximum power operation point, respectively. I sc is the short-circuit photocurrent and V oc is also the open-circuit voltage of solar cells. Here, I max is consitued by the effective number of photons that produce electron-hole pairs. Therefore, high number of photons that interact with the PN junction in the cell would let to higher photocurrent (i.e maximum or short-circuit) as being that I sc is slightly higher than I max, but close to it. Same situation also applies to the voltage where V oc is higher than V max as seen in Fig.4. Here, it is important to highlight that Equation (4-5) expresstions provides the convenience in calculation of both electrical exergy and exergy destruction considering irreversible effects such as photon solid angles, non-radiative recombination of charges (caused by impurities) and photon cooling vice versa occurring during interaction. Note that this expression of useful electrical exergy from the Sun is only valid for single-junction devices that have one kind of solar cell and one value of bandgap energy. IV.1. Electrical Exergy Calculations For exergetic calculations using solar energy parameters, identifying the exergy components of a solar PV system is important. Physical and chemical exergy expressions for such a system associated with enthalpy, entropy and energy conversion concepts under the Carnot limitations has been clarified Sahin et al. (2007). Therefore, in this paper, these concepts will not be broadly taken into consideration, but rather two major exergy components known as electrical and thermal exergy will be focused on for analysis. Since the electricity is generated by photovoltaic effect and it can be utilized for useful purposes via grid, produced electrical energy can be defined as electrical exergy of the PV. Based on this approach, for calculating the electrical exergy, Ėx electrical, there should be a princible assumption corresponding to the exergy content which is received by photovoltaic surface and completely usable for generating electricity. Namely, the useful exergy rate is taken as the maximum electrical power output given by the product of the current and voltage when it operates under the maximum power condition. For also exergy destruction, as it comes to the definition, the amount where the maximum electrical power output (WPV) is substracted from the maximum electrical exergy (Wmax) is considered as shown in Equation (4-5) Saloux et al. (2013). Ėx electrical = W PV = V max I max (4) Fig. 4: Representation of voltage and current points of a solar cell in a current-voltage (I-V) curve. The I-V curve represented in the figure above brings information on the photovoltaic performance of the solar cells because of being a distinctive feature. For exergy point of view, exergy components can be depicted on the figure. For exergy destruction, the area remains between the curve and rectangular line can be indicated. Electrical exergy can be also shown by rectangular area remaining under the maximum power point which is restricted by a term known as fill factor defined by FF = V maxi max V oc I sc (6) As it can be seen from above equation, there is a link between fill factor rates and exergy efficiency. The effect of fill factor on actual exergetic efficiencies has been disccussed by Rusirawan and Farkas (2010). However, it can be pointed out that higher fill factor values tend to be seen as a reason of higher exergy efficiencies. Ėx destruction = W max W PV = V oc I sc V max I max (5) 257

272 IV.2. Thermal Exergy and Efficiency Calculations In this section, thermal exergy referring to the heat radiated by the solar arrays is presented. When solar radiations fall on the solar cells, PV modules get heated during electricity production due to the thermal energy existing in solar radiation. Thus, since solar cells are manufactured by electronic circuits, they loss their efficiencies as a result of the heat and internal resistances occurring in the process. In order to obtain a better electrical efficiency of the system, the heat removal from the PV module is therefore essential. Thermal exergy is not used for useful purposes such as electrical exergy because it performs only on the PV surface. Therefore, it is called as heat loss to the ambient. With this perspective, the thermal exergy of the system Ėx thermal, consists of heat loss from the PV surface to the ambient and can be represented as Joshi (2009) Ėx thermal = (1 T ambient T cell ) Q (7) Here, Q represents the available thermal energy (W) and can be defined as Q = h ca A(T cell T ambient ) (8) where A (m 2 ) represents the effective area of PV surface. h ca (W/m 2 K) also refers to the convective and radiative heat transfer coefficient from the cell to the ambient. It can be calculated by considering wind velocity ( υ), density of air and the surrounding (ambient) conditions given by following equation Tiwari (2002) h ca = 5,7 + 3,8. υ (9) Since only two major exergy components occur in the PV system and thermal part wastes to the ambient as heat loss, total exergy expression of a PV system can be derived considering both electrical and thermal exergy expressions (4-9) given by Ėx PV = Ėx electrical Ėx thermal (10) And hence, it becomes Ėx PV = V max I max (1 T amb. ) [h T ca A(T cell T amb )] (11) cell For photovoltaic thermal systems (PV/T), electrical exergy and thermal exergy is summed up for total exergy calculation since thermal energy is not lost or destroyed in the PV/T systems. Therefore, positive sign in the expression shows that thermal energy recovery in the process as follows Ėx PV/T = V max I max + (1 T amb. T cell ) [h ca A(T cell T amb )] (12) For evaluating the exergy efficiency of solar cells, total solar irradiation is needed. The incident solar radiation with the direct and diffuse components 258 received on the PV surface affects the current and power output of the PV module. Since the magnitude of global solar radiation is not completely converted into electricity in practise, there should be mentioned on the Carnot limitations and atmospheric effects. With this approach, the exergy of solar irradiation, Ėx solar (also known as Ėx input ), depends on the intensity of solar irradiance and the area of the PV surface given as Santarelli and Macagno (2004) Ėx solar = (1 T amb. T sun ) S T A (13) Here, S T (W/m 2 ) defines the hourly measured global solar irradiance. T sun detones the effective temperature of the Sun and is estimated and taken as 5777 K in Holmberg s study Holmberg et al. (2006). For calculating the exergy efficiency, the fundamental efficiency expression based on the ratio of input and ouput parameters in any system is used. In this case, the total exergy of the PV is considered as the exergy output. For exergy input, the solar exergy is taken given as ψ = Ėx output Ėx input = Ėx PV Ėx solar (14) After substituting Eqs. (11) and (13) into Eq. (14) the final exergy efficiency expression for a PV system can be obtained as ψ PV = V T maxi max (1 amb. )[h T ca A(T cell T amb. )] cell (1 T amb. Tsun )S TA (15) Similarly, Equation (15) applies to the PV/T system with a difference that thermal exergy recovery carries out and thus, provides the positive sign in the same expression defined as ψ PV = V T maxi max +(1 amb. )[h T ca A(T cell T amb. )] cell (1 T amb. Tsun )S TA (16) The exergy efficiency can also be defined in terms of irreversibility or exergy destruction by organizing Equation (14). Since the exergy degradation occurs due to the both electrical and thermal destruction for the PV, and only the electrical destruction for the PV/T systems, the irreversibility can be shown as ψ = Ėx output Ėx input = Ėx input Ėx dest. Ėx solar = 1 I Ėx solar (17) If it is continued further, the irreversibility becomes as I = (1 ψ)ėx solar (18) For comparison purposes, the exergy efficiency can be evaluated with the power conversion efficiency of the PV together. For the solar cell, the efficiency expression η PV, can be defined as in terms of fill factor (FF) given in Equation (6), and the ratio of

273 actual electrical output as well as input solar energy incident on the PV surface area as follows η PV = FFV oci sc S T A (19) beginning to the end of November in general as being the highest point belong to the midday and the lowest belong to the morning. V. Results and Discussion In this section, the case study presented above is analyzed and discussed on selected days during November, 2015 from 9 am to 17 pm. Accordingly, exergy analysis results within seasonal and climatic changes in Karabuk province have been given in following figures. In order to investigate the variation of the parameters and prevent complexity in the graphs, 6 days have selected throughout the November and related figures have been interpreted. Fig. 5: Global Solar Radiation values (W/m 2 ) represented hourly during November First of all, the variation of global solar radiation values has been given in Fig. 5. According to the figure, the highest radiation values have been occurred in 5 th of November with W/m 2 in average, whereas the lowest one has taken place in 30 th with W/m 2. The factor influencing the intensity of solar radiance is due to the weather patterns for the selected days. For instance, in 5 th of November, it was a clear blue sky, but it was a windy and cloudy in 30 th of November. For a general assessment, solar radiation values have reached a peak during midday and decreased during both morning and afternoon as expected. Fig. 7: The changes in panel temperature values ( o C) on selected days Considering the ambient temperatures falling throuhout the month, panel temperatures have also shown similar tendencies in which diminish values are due to the insufficient solar radiation and ambient temperatures. Besides, it is seen that the panel temperatures have ranged from 22,9 o C to 37,4 o C in average in 30 th and 5 th of November, respectively. Additionally, this temperature interval has increased from the morning to the midday and then, gradually descreased untill 17:00 pm given in Fig. 7. The variation of the panel temperatures is crucial since the power conversion efficiency of the module and hence, generated current-voltage (I-V) curve can be directly affected by this parameter. Fig 8: The changes in electrical, thermal exergy and exergy output values on selected days (W) Fig. 6: The changes in ambient temperature values ( o C) on selected days With the global solar radiation, ambient temperature values are also another significant environmental property. According to the Fig. 6, the ambient temperatures have ranged from 7,2 o C (min) to 23,1 o C (max) during the month. Here, it is clearly seen that ambient temperatures have decreased from the 259 Under the environmental conditions illustrated in between Fig. 5 and Fig. 7, both electrical, thermal and exergy outputs using exergy expressions given in Section IV have been presented in Fig. 8. According to the figure, maximum electrical exergy has been seen in 5 th with the 72,3 W rate in average, whereas the minimum has been in 30 th with the 31,91 W rate. For thermal exergy, the maximum and minimum values have been seen in the same days with the 4,74 W and 1,69 W rates, respectively. In the figure, thermal exergy values have been given as 10 times in order to be obtained results better. Module exergy output has also been in parallel with electrical exergy during the month with the 30,2 W and 72,54 W rates. Thanks to the obtained data given in previous graphs, the exergy output of a PV, the exergy of solar radiation and also exergy loses during November has

274 been calculated and demonstrated in Fig. 9. The exergy output values have been given in Fig. 8. Solar exergy values have also ranged from 298,8 W to 466,4 W in average, whereas exergy loses have been between 268,6 W and 393,8 W in 30 th and 5 th of November, respectively. Here, it is observed that increasing thermal exergy of the PV due to the higher ambient and panel temperatures would lead to higher exergy loses. Therefore, since exergy loses depend on the weather pattern, low ambient temperature would provide the minimum loses, and hence higher exergy and power conversion efficiency. Fig 9: The changes in exergy input, output and also exergy loss values on selected days (W) VI. Conclusion In this study, the exergy efficiency of a polycrystalline solar PV has been observed within Karabuk climate conditions during November. The effect of environmental conditions and circumstances into the power conversion and exergy efficiency on a actual solar PV module has been investigated. Environmental properties contain global solar radiation, panel temperature, ambient temperature and also wind speed on the just above PV surface. The aim of the study is to develop an advanced and more realistic approach for the thermodynamic analysis and assessment of solar PV modules using exergy analysis. This method will provide a physical basis for understanding and predicting the variations in solar PV behavior. In this article, the exergy of the solar PV modules and also its components are discussed. Besides, the exergy analysis is formulated for the system and its components, and finally compared with the power conversion efficiency of the module. Carried out this exergy analysis of a solar PV in Karabuk during November has been a model to assessing more complex and bigger solar photovoltaic systems. Nomenclature Fig 10: The changes in exergy efficiency and power conversion efficiency values on selected days (W) The actual working performance of the module under environmental conditions has been given in Fig. 10. According to the graph, the lowest module efficiency, ηpce has occurred in the 30 th of November with the 10,38% rate in average when the weather pattern was windy and cloudy. Additionally, the highest ηpce has taken place in 5 th with the 16,03% ratio in which the solar PV could able to see the Sun directly and the humidty level was the minimum. Apart from these days, ηpce in 20 th of November was higher than 15 th with the rate of 13,66% and 13,58%, respectively in which the ambient temperatures could be sorted as T 15th > T 20th from the highest to the lowest. Besides, ηpce rate as average in 25 th was also bigger than 10 th with the 12,02% and 11,67% respectively, where ambient temperatures were inversely proportional at the same time. Exergy efficiency values have also ranged from 10,36% to15,90% in the 30 th and 5 th of November, respectively. Here, it can be argued that exergy efficiencies and power conversion efficiencies are close to the each other in general. S T : Global solar radiation (W/m 2 ) T amb : Ambient temperature ( o C) T cell : Panel temperature ( o C) θ wind : Wind velocity (m/s -1 ) I : Current (A) V : Voltage (V) Greek letters η : Power conversion efficiency ψ : Exergy efficiency Subscripts max : Maximum sc : Short circuit current oc : Open-circuit voltage References Abid, M. & Hepbasli, A., Dynamic Exergetic Analysis and Evaluation of Photovoltaic Modules. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 37(21), pp Available at: Cuce, P. M., E.C., A Novel Model of Photovoltaic Modules for Parameter Estimation and Thermodynamic Assessment. International Journal of Low-Carbon Technologies, 7(2), pp Çengel, Y.A., Mühendislik Yaklaşımıyla Termodinamik M.-H. L. Derbentli, T., ed., İstanbul: Güven Bilimsel. 260 Dincer, I., Thermodynamic aspects of renewables and sustainable development.

275 Renewable & Sustainable Energy Reviews, 9(2), pp Dincer, I., Cengel, Y., Energy, entropy concepts and their roles in thermal Engineering. Entropy (3)., Dubey, S., G.N.T., Energy and Exergy Analysis of Hybrid Photovoltaic/Thermal SolarWater Heater Considering with andwithoutwithdrawal from the Tank. Journal of Renewable and Sustainable Energy 2, 2(4), p Fahrenbruch, Alan L., Bube, R.H., Fundamentals of Solar Cells: Photovoltaic Solar Energy Conversion, New York, USA: Academic Press (May 28, 1983). Hepabsli, A., A Key Review on Exergetic Analysis and Assessment of Renewable Energy Resources for a Sustainable Future. Renewable and Sustainable Energy Reviews, 12, pp Holmberg, J., Flynn, C. & Portinari, L., The colours of the Sun. Monthly Notices of the Royal Astronomical Society, 367(2), pp Joshi, A.., Thermodynamic assessment of photovoltaic systems. Solar Energy, 83( ), pp.1 7. Joshi, A.S., Dincer, I. & Reddy, B. V., Analysis of energy and exergy efficiencies for hybrid PV/T systems. International Journal of Low-Carbon Technologies, 6(1), pp of photovoltaic (PV) and photovoltaic/thermal (PV/T) systems using the exergy method. Energy and Buildings, 67, pp Available at: Santarelli, M., A thermoeconomic analysis of a PVhydrogen system feeding the energy requests of a residential building in an isolated valley of the Alps. Energy Conversion and Management, 45(3), pp Smestad, G.P., Conversion of heat and light simultaneously using a vacuum photodiode and the thermionic and photoelectric effects. Solar Energy Materials and Solar Cells, 82(1-2), pp Sobhnamayan, F. et al., Optimization of a solar photovoltaic thermal ( PV / T ) water collector based on exergy concept. Renewable Energy, 68, pp Available at: Tiwari, G.N., Solar Energy: Fundamentals, Design, Modeling and Applications, Narosa Publishing House, New Delhi and CRC Press, Washington. Wurfel, P., Thermodynamic limitations to solar energy converstion. Physica E: Low- Dimensional Systems and Nanostructures, 14, p.18. Joshi, A.S., Dincer, I. & Reddy, B. V., Thermodynamic assessment of photovoltaic systems. Solar Energy, 83(8), pp Available at: Luther, J., World in transition towards a sustainable energy system. German Advisory Council on Global Change (WBGU), p.3. New, B. et al., Re-considering the E conomics of Photovoltaic Power Morgan Bazilian., pp Petela, R., E xergy of undiluted thermal radiation., 74(May), pp Rusirawan, D. & Farkas, I., Fill Factor Effects On Exergetic Efficiency Of Photovoltaic Modules., p Sahin, A.D., Thermodynamic analysis of solar photovoltaic cell systems. Solar Energy Materials&Solar Cells, 91(153). Sahin, A., Dincer, I., Rosen, M.A., Thermodynamic analysis of wind energy. International Journal of Energy Research, 30(8), pp Saloux, E., Teyssedou, A. & Sorin, M., Analysis 261

276 A Comparative and Experimental Study on Different Exergetic Efficiency Methods of a Solar Panel Mehmet Ozalp 1, Mutlucan Bayat 2* 1,2 Karabuk University, Engineering Faculty, Mechanical Engineering Department, Karabük, 78200, Turkey * mutlucanbayat@karabuk.edu.tr Abstract Exergy analysis is known by researches and engineers as an essential tool to assess of a PV system s performance with the thermodynamic approach. This analysis does not only demonstrate energy utilisation efficiency, it provides also varios usefull results corresponding exergy efficiency, which become a significant princible for comparing of solar panels. Therefore, there has been an enormous interest in exergetic calculations, considering irreversibility and energy delivery in the system, and many scientists have conducted theoretical and experimental study on this field, currently. To perform exergy analysis is based on second law of thermodynamic, however, some has presented different metedology to find out exergy efficiencies, for instances, (Zondag et al., 2002), (Petela, R., 2003), (Sahin et al., 2007), (Joshi et al., 2009) vice versa. For this reason, in this paper, different exergy efficiency perspectives has been studied and compared according to these approaches using experimental data applied to a solar panel during a month. Both electrical and environmental parameters of the panel have been measured and exergy efficiencies calculated. As a result of this investigation, exergy efficiencies have been simulated by using MATLAB. The differences between angle inclinations of graphs provided finer understanding and evulating of exergy efficiency values. Keywords: Exergy efficiencies, exergetic calculations, thermodynamic approach, MATLAB, Simulation I. Introduction Energy has always become a crucial factor for continuity of human life. Despite the increasing energy demand due to improved living standards all over the world, reducing dependence on foreign energy is therefore essential. For this purpose, developing installed electricity capacity with power generation technologies that are environmentally friendly and have high energy efficiency is needed. The solar power is affordable, inexhaustible and clean source of energy. Employing solar energy therefore brings enormous benefits considering enhance sustainability, decrease pollution and limiting global warming. It is not only a powerful option to reduce environmental concerns; it provides also an indigenous solution for diminishing use of fossil fuels and keeping clean power generation prices lower Luther (2013). The solar photovoltaic (PV) cell is an electrical device that converts the energy of light into direct-current electricity using semiconducting materials that exhibit the photovoltaic effect, which is called physical and chemical phenomenon. The working principle of cell starts with the absorption of light in which the electrons present in the valence band are being excited and become free. Thus, photons whose energy is equal to or greater than the band gap in semiconductor materials used in making cell constitute electron hole pairs or excitons at first. As the second step involves the separation of charge carriers of opposite types in the cell, separating extraction of those carriers to an external circuit takes place in the last phase. Consequently, PV cells as working a semiconductor diode convert carrying sunlight energy into directly electricity by utilizing internal photochemical reactions Fahrenbruch and Bube (1983). Since the process of direct conversion of sunlight occurs stationarily or motionlessly and it doesn t release environmental emissions during operation, PV applications have been more attractive and now used in industries for a few decades Bazilian et al. (2013). Besides, advances in PV technology has grown the market rapidly and drawn attention from policy makers. Therefore, power generation from solar PV has now been seen as one of the most promising clean and sustainable energy technology. With the increasing interest in solar energy, there have been many studies regarding both characterisation of cells and performance analysis with the thermodynamic approach. For this point of view, the variation of the specific heat capacity, C p according to the temperature in silicon solar cell has been investigated and the calculation formula has been proposed by Regel and Glazov (1980). For exergy efficiency calculation, the simple empirical expression based on the power conversion efficiency of a PV has been discussed by Zondag et al., (2002). For thermal exergy, the convective and radiative heat transfer coefficient depending on the wind velocity and density of air has been studied by Tiwari (2002). Petela (2003) presented a new approach for obtaining the exergy of solar irradiation depending on 262

277 the ambient temperature and the Sun. Besides, Santarelli and Macagno (2004) have analysed the solar exergy with the intensity of global solar irradiance and the area of the PV surface. Sahin et al., (2007) have conducted different methodology on exergy output of a PV using physical exergy expression including the enthalpy, entropy and chemical exergy components. Joshi et al., (2009) have also proposed exergy efficiency expression based on the electrical and thermal exergy of a PV indicating exergy destructions. II. Experimental Facility II.1. Setup In this study, a silicon based polycrystalline solar PV panel was performed throughout November in Karabük province in Turkey. In order to obtain changes in exergetic efficiencies, a PV system was installed on the top of the Engineering Faculty, Karabuk University, Karabük, Turkey. There are 72 polycrystalline PV panels as seen in Fig. 1. Each panel has 0.67 m 1.5 m dimensions as illustrated in technical drawing given in Fig. 2. Fig.1: The installed PV sytem Fig.2: Technical drawing of a PV panel In addition to the dimensions, the specific features of each module within standart test conditions (1000 W/m 2 global solar radiance, 1,5 AM and 25 o C ambient temperature) has been given in Table 1. Tab. 1: The panel properties within standart test conditions IBC PolySol 130 GC Technical Data Nominal peak W p power Nominal voltage V 18.0 Nominal current A 7.23 Open circuit voltage V 21.9 Short-circuit current A 7.9 Temperature %/K coefficient of I sc Temperature mv/k coefficient of V OC Temperature %/K coefficient of P max Power conversion % efficiency Power Tolerance % ±2.5 Fill Factor (FF) Number of cell - 36 Length mm 1500 Width mm 670 Height mm 42 Weight kg 12.0 Effective Area m II.2. Measurements Measurements on a PV module have been carried out between 9 am to 17 pm in a 30 minute intervals during November. For defining and evaluating exergy efficiency rates, electrical parameters of the module e.g. nominal current-voltage, open circuit voltage and short-circuit current have been measured. Besides, atmospheric parameters such as global solar radiation, the ambient and panel temperature as well as wind speed have been recorded since environmental conditions affect the real working performance the module. For measuring parameters mentioned above, some intruments have been utilized. For instance, the panel backside temperature measurement is performed with the calibrated digital thermocouple and thus, the temperature data has been provided to be taken in the middle of the module. Delta OHM HD2303 coded a digital anemometer is used to measure the ambient temperature and also the air flow on the just top of the PV surface. The global radiation (S T ) has been measured by using a MS-410 coded pyranometer. For obtaining current-voltage parameters, a wirewound rheostat which has 50 Ω resistance capacity was used. In order to define open circuit voltage (Voc) and short-circuit current (Isc), two points on the rheostat where low resistance and high resistance are located have been used. For also defining maximum curent and voltage, UNI-T UT61B and MY-68 coded two digital multimeters were utilized. All current-voltage data read by multimeters have been recorded into an excel sheet and maximum power points (Pmax) have been selected for half an hour by multiplying read the values. The components of the values indicating maximum power points were considered as maximum voltage (Vmax) and maximum current (Imax). Fig. 3 shows the some 263

278 measurement devices used in the experimental study. Tab. 2: Global solar radiation data in selected days Global Solar Radiation W/m 2 Days Average Value 1 st Nov th Nov th Nov th Nov th Nov th Nov 423,3 30 th Nov Tab. 3: Ambient temperature data in selected days Ambient Temperature o C Days Average Value 1 st Nov th Nov th Nov th Nov th Nov th Nov th Nov Fig. 3: Some measurement instruments used in experiment Using these intruments the data on the parameters has been collected day by day and how the PV module behaviour has changed under the different weather conditions has been analysed. In order to better obtain changes in exergetic efficiencies, 7 days have been selected. Environmental parameters which influence the exergy efficiency are given in average following tables. First of all, Table 2 defines the changes in global solar radiation in W/m 2 between 1 st of November and 30 th. Table 3 indicates the ambient temperature in o C in the same days. The panel temperature in o C has been given in Table 4 and the wind velocity has been also indicated in Table 5. For electrical parameters, Table 6 illustrates the current data dividing into short-circuit and maximum current values in average. Table 7 also express the voltage data given as average in both open circuit and maximum voltage values. These tables give sufficient information about PV behaviour in selected days by comparing both internal and external features. For instance, 5 th of November is the highest level of both the current and voltage where global radiation, ambient temperature and panel temperature values are maximum at the same day. In contrast, 30 th of November has the lowest electrical output where solar radiation, ambient and panel temperature is also decreased significantly. Although wind speed (θ wind ) is not directly taken into consideration for exergy analysis, velocity values are not also neglected since the convective and radiative heat transfer coefficient, hence thermal energy of the PV would change. In Table 5, the difference in wind speed in m/s has been presented. According to the table, average wind speed values have changed in random throughout the month. 264 Tab. 4: Panel temperature data in selected days Panel Temperature o C Days Average Value 1 st Nov th Nov th Nov th Nov th Nov th Nov th Nov Tab. 5: Wind speed data in selected days Wind Speed m/s Days Average Value 1 st Nov th Nov th Nov th Nov th Nov th Nov th Nov 0.21 Tab. 6: Current data in selected days Current Isc A Imax Days A. Value Days A. Value 1 st Nov st Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov 2.98 Tab. 7: Voltage data in selected days Voltage Voc V Vmax Days A. Value Days A. Value 1 st Nov st Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov th Nov 10.95

279 II. Exergy Efficiencies All the exergy efficiency calculation methods based on the fundamental efficiency expression depending on the ratio of input and ouput parameters in any system. For point of view, the exergy efficiency of a PV system can, in general, be given as ψ = Ėx output Ėx input = Ėx PV Ėx solar (1) where Ėx PV is the total exergy rate of the PV which is mainly electrical power output of the system considering exergy destruction. Thus, exergy of electrical energy can be defined as Saloux et al., (2013). Ėx electrical = V max I max (2) where V max and I max represent the voltage and current at the maximum power operation point, respectively. For Ėx PV calculation process, thermal exergy occurring due to the thermal energy gained by the system during electricity generation is also considered. The thermal exergy of the system Ėx thermal, consists of heat loss from the PV surface to the ambient and can be given as Joshi et al., (2009) Ėx thermal = (1 T ambient T cell ) Q (3) Here, T ambient refers to the ambient temperature and T cell denotes the temperature of solar cell. Q also represents the available thermal energy (W) and can be defined as Q = h ca A(T cell T ambient ) (4) where A (m 2 ) represents the effective area of PV surface. h ca (W/m 2 K) also refers to the convective and radiative heat transfer coefficient from the cell to the ambient. It can be calculated by considering wind velocity ( υ), density of air and the surrounding (ambient) conditions given by following equation Tiwari (2002) h ca = 5,7 + 3,8. υ (5) Here, it should be addressed that thermal part is subtracted from electrical exergy rate while calculating total exergy of a PV, since the thermal energy gained by the system is not desirable in the case of the PV. Hence, Ėx PV can be derived considering both electrical and thermal exergy expressions (1-5) given by Ėx PV = Ėx electrical Ėx thermal (6) And then, it becomes Ėx PV = V max I max (1 T amb. T cell ) [h ca A(T cell T amb )] (7) For evaluating the exergy efficiency of solar cells, total solar irradiation is also needed. The incident solar radiation with the direct and diffuse components received on the PV surface affects the current and power output of the PV module. With this approach, the exergy of solar irradiation, Ėx solar (also known as Ėx input ), depends on the intensity of solar irradiance and the area of the PV surface given as Santarelli and Macagno (2004) Ėx solar = (1 T amb. T sun ) S T A (8) Here, S T (W/m 2 ) defines the hourly measured global solar irradiance. T sun detones the effective temperature of the Sun and can be taken as 5777 K as in Holmberg s study Holmberg et al., (2006). In order to obtain the final exergy efficiency expression for a PV system according to Joshi et al., (2009), Eqs. (7) and (8) are substituted into Eq. (1) given as ψ Joshi = V T maxi max (1 amb. )[h T ca A(T cell T amb. )] cell (1 T amb. Tsun )S TA (9) For evaluating and defining exergy efficiencies, different metedologies have been presented in addition to the study of Joshi et al. For instance, Petela (2003) has proposed a different expression for calculating Ėx solar as follows Ėx solar = GA [ (T amb. T sun ) (T amb. T sun )] (10) where G denotes the intensity of solar irradiance. With this expression, the exergy efficiency rate of a PV can also be calculated after substituting Eqs (10) into Eqs (9) and using the relation proposed by Petela defined as ψ Petela = V T maxi max (1 amb. )[h T ca A(T cell T amb. )] cell GA[ (11) 3 (T amb. Tsun ) 4 3 (T amb. Tsun )] Sahin et al., (2007) has used similar expression with Santarelli and Macagno (2004) for Ėx solar indicated in equation (8). However, they demonstrated a different approach for Ėx PV expression which consists of physical exergy consept including enthalpy, entropy and additional chemical exergy components of a PV that are given below in sequential: H = C p (T cell T amb. ) (12) where H is the change in enthalpy (J/kg) and C p is the specific heat capacity of silicon solar cell and can be calculated from Regel and Glazov (1980) C p = 0, , T cell 1, (T cell ) 2 (13) 265

280 Besides, since the total entropy of the system can be written as S = S system + S surround (14) the working performance of the module under environmental conditions given in Table (2-5), the power conversion efficiency of the module (ηpce) is demonstrated in Figure (4). It becomes S = C p ln ( T cell T amb ) Q loss T cell (15) where Q loss = C p (T cell T amb. ) (16) Here S defines the change in entropy (J/kg.K) and Q loss denotes heat loses from the PV cell. Therefore, the total physical exergy for a PV cell system can be expressed using Eqs. (12)-(16) as follows Ėx physical = E gen + C p (T cell T amb. ) + T amb. (C p ln ( T cell T amb ) Q loss T cell ) (17) In the above equation, the first term of this expression, E gen represents the generated electricity at the highest energy concent of the electron Sahin et al (2007). Thus, it can be considered as maximum power point of the solar cell I-V curve (P max ) which is V max I max. Thus, physical exergy expression with the all enthalpy and entropy components becomes Ėx physical = V max I max + C p (T cell T amb. ) + C p T amb. (ln ( T cell ) (T cell T amb. ) ) (18) T amb T cell Besides, Ėx PV expression proposed by Sahin including physical exergy components indicated in Eq. (18) can be identified given as Fig. 4: The power conversion efficiency of the module on selected days According to the graph, the lowest module efficiency, ηpce has occurred in the 30 th of November with the 10,38% rate in average when the weather pattern was windy and cloudy. Additionally, the highest ηpce has taken place in 5 th with the 16,03% ratio in which the solar PV could able to see the Sun directly and the humidty level was the minimum. Apart from these days, ηpce in 20 th of November was higher than 15 th and 1 st with the rate of 13,66%, 13,58% and 13,49% respectively in which the ambient temperatures could be sorted as T 1st > T 15th > T 20th from the highest to the lowest. Besides, ηpce rate as average in 25 th was also bigger than 10 th with the 12,02% and 11,67% respectively, where ambient temperatures were inversely proportional at the same time. Using different exergy expressions given in Section II, all exergetic efficieny results proposed by Petela, Joshi, Zondag and Sahin have been illustrated in following Figure (5)-(11). Ėx PV = Ėx phs (1 T cell T sun ) [V oc I sc V max I max ] (19) Therefore, since solar exergy expression, Ėx solar is the same with Eq. (8) in the study mentioned above, the final exergy efficiency rate can be given as follows ψ Sahin = Ėx phs (1 T cell Tsun )[V oci sc V max I max ] (1 T amb. Tsun )S TA (20) Fig. 5: The exergetic efficiencies in 1 st of November Apart from these exergy efficiency expressions proposed by Joshi, Petela and Sahin, there is also another study which applies a simple equation for the efficiency rate Zondag (2002) given by ψ Zondag = n cell [1 0,0045(T cell 25)] (21) III. Results and Discussion Fig. 6: The exergetic efficiencies in 5 th of November In this section, the case study presented above is analyzed and discussed on selected days during November, 2015 from 9 am to 17 pm. For obtaining 266

281 of November with 8,47% rate, whereas the maxium ψ values have been seen in 5 th with 16,01% proportion. Fig. 7: The exergetic efficiencies in 10 th of November Fig. 8: The exergetic efficiencies in 15 th of November Fig. 9: The exergetic efficiencies in 20 th of November Fig. 10: The exergetic efficiencies in 25 th of November If we go into details for the first exergy graph, ψ values in 1st of November have changed from 11,50% (min) to 12,08% (max) in average according to Sahin and Petela s equations, respectively. Here, the efficiency order has been occured as ψ Petela > ψ Joshi > ψ Zondag > ψ Sahin for throughout the day. Same situation applies to the other days, except some instant measurements. Considering Figure (6), ψ values were in the range of between 14,81% and 16,01% in average. However, exergy efficiency rate significantly decreased from this range to an interval which is between 9,85% and 11,93% in the 10 th of November given in Figure (7). Here, it is observed that exergetic efficiency lines in Figure (6) had a certain gaps among each other, whereas exergy lines belong to Petela, Joshi and Zondag were much more in parallel in Figure (7). Comparing Figure (8) and Figure (9), exergy efficiency data has shown similar tendency during 15 th and 20 th of November. However, ψ value intervals have been between 12,60% and 13,81% along 20 th in which this range was a little more than 15th due to the atmospheric conditions. According to the exergy expression in Equation (20), ψ Sahin has been lower in both cases comparing to other exergy efficiency calculation methods. Here, it is thought that physical exergy components including enthalpy and entropy concepts have negatively affected both the electrical and thermal exergy proportion and hence, the total exergy output of PV has fallen as a reason of this result. Same condition has appeared also in Figure (10) and Figure (11). ψ values in 25 th were slightly higher than in 30 th of November where the exergy efficiency rates were the minimum during the month due to the weather condition mentioned before. Here, it is worth to note that these exergy lines have been more often and the closeness of each other has increased in 30 th comparing to the 25 th of November. IV. Conclusion Fig. 11: The exergetic efficiencies in 30 th of November A comprehensive thermodynamic investigation through using different exergetic efficiency calculation methods has been performed. ψ values corresponding to the expressions proposed by different authors have been illustrated for selected days and results have been indicated thanks to obtaining electrical parameters under the environmental conditions. For the selected days, it can be seen that exergy efficiencies (ψ) have ranged day by day. However, the critical point in here is that the variation of exergetic efficiency rates was similar to the difference in the power conversion efficiency of the PV. Namely, the lowest ψ values have been obtained in the 30 th 267 Considering all figures, it is seen that all exergetic efficiency results are realistic and feasible for assessment of a PV system. Besides, results are also close to the each other although ψ Sahin values are seem to be lower, and in contrast, ψ Petela values are also higher in general. It is because of some

282 components used in the exergy efficiency expressions. Here, it can be also argued that there is an inverse proportion between the ambient temperatures and exergy efficiencies, similar to the panel efficiency. Nomenclature S T : Global solar radiation (W/m 2 ) T amb : Ambient temperature ( o C) T cell : Panel temperature ( o C) θ wind : Wind velocity (m/s -1 ) I : Current (A) V : Voltage (V) Greek letters η : Power conversion efficiency ψ : Exergy efficiency Subscripts max : Maximum sc : Short circuit current oc : Open-circuit voltage 451. Tiwari, G.N., Solar Energy: Fundamentals, Design, Modeling and Applications, Narosa Publishing House, New Delhi and CRC Press, Washington. Zondag, H. A., De Vries, D. W., Van Helden, W. G. J., Van, Zolengen, R. J. C., and Van Steenhoven, A. A The thermal and electrical yield of a PV-thermal collector. Sol. Energy 72: References Bazilian et al., Re-considering the E conomics of Photovoltaic Power Morgan, pp Fahrenbruch, Alan L., Bube, R.H., Fundamentals of Solar Cells: Photovoltaic Solar Energy Conversion, New York, USA: Academic Press (May 28, 1983). Holmberg, J., Flynn, C. & Portinari, L., The colours of the Sun. Monthly Notices of the Royal Astronomical Society, 367(2), pp Joshi, A.., Thermodynamic assessment of photovoltaic systems. Solar Energy, 83( ), pp.1 7. Luther, J., World in transition towards a sustainable energy system. German Advisory Council on Global Change (WBGU), p.3. Petela, R., E xergy of undiluted thermal radiation., 74(May), pp Regel, A. R., and Glazov, V. M., Physical Properties of Electronic Melts, Nauka, Moscow, UK. Sahin, A.D., Thermodynamic analysis of solar photovoltaic cell systems. Solar Energy Materials&Solar Cells, 91(153). Saloux, E., Teyssedou, A. & Sorin, M., Analysis of photovoltaic (PV) and photovoltaic/thermal (PV/T) systems using the exergy method. Energy and Buildings, 67, pp Available at: Santarelli, M., A thermoeconomic analysis of a PVhydrogen system feeding the energy requests of a residential building in an isolated valley of the Alps. Energy Conversion and Management, 45(3), pp

283 Solar Assisted Multi-Generation System Using Nanofluids: A Comparative Analyzes Muhammad Abid 1*, Tahir A. H. Ratlamwala 2, Ugur Atikol 1 1 Eastern Mediterranean University, Faculty of Engineering, Department of Mechanical Engineering, Famagusta, North Cyprus via Mersin 10, Turkey 2 Shaheed Zulfikar Ali Bhutto Institute of Science and Technology, Clifton Campus, Karachi, Sindh, Pakistan * muhammad.abid@emu.edu.tr Abstract In this comparative study, a parabolic trough solar collector and a parabolic dish solar collector integrated with a Rankine cycle and an electrolyzer are analysed for power as well as hydrogen production. The absorption fluids used in the solar collectors are Al2O3 and Fe2O3 based nanofluids and molten salts of LiCl-RbCl and NaNO3-KNO3. Ambient temperature, initial temperature, and solar irradiance are varied to investigate their effects on heat rate and net power produced, the outlet temperature of the receiver, overall energy and exergy efficiencies and the rate of hydrogen produced. The results obtained show that the net power produced by the parabolic dish assisted power plant is higher (2.48 kw 8.17 kw) in comparison to parabolic trough (1 kw 6.23 kw). But the overall energy and exergy efficiencies are found to be higher for the former. It is observed that both aluminum oxide (Al2O3) and ferric oxide (Fe2O3) based nanofluids have better overall performance and generate higher net power as compared to the molten salts. An increase in inlet temperature is observed to decreasing the hydrogen production rate. The rate of hydrogen production is found to be higher using nanofluids as solar absorbers. Hydrogen production rate for parabolic dish thermal power plant and parabolic trough thermal power plant varies from g/s to g /s and from g/s to g/s, respectively. Keywords: Exergy, nanofluids, parabolic trough/dish, hydrogen production, multi-generation. I. Introduction Increase in energy demand and depletion of fossil fuels are two major concerns of the society. It is highly needed to meet the energy demand by offering clean and environment friendly energy, and it is only possible if fossil fuels would be replaced with renewable energy, as fossil fuels are harmful for the environment and as well as for human being. Solar energy which is clean and abundantly available renewable energy could be an option to be used to combat with global warming dilemma. Different techniques are available and are being used to harness solar energy. Parabolic trough and parabolic dish solar collectors PT/PDSC are already in use to produce electricity. The thermal performance of these kind of collectors is an area which is under consideration of many of the researchers. Their performance can be increased by using high heat transfer fluids, such as nanofluids and molten salts. PT/PDSC can also be used in integration with steam turbine and electrolyzer to produce power, heat and hydrogen simultaneously. A lot of research is been done on solar collectors. Their working temperature range is different from each other. As Kalogirou (2004) suggested the temperature range of different solar collectors, which are heliostat field ( C ), parabolic dish ( C ) and parabolic trough ( C ) respectively. Research done by many other researchers (Le-Roux et al., (2012); Garcia et al., (2011); Barigozzi et al., (2012); Huang et al., (2012); Kalogirou, (2004); Alkhamis and Sherif, (1997) found out that it is more useful and effective to use high temperature range solar collectors, because they have high power outputs as well as higher efficiencies. Nanofluids are considered to have better thermal properties than base fluids, such as water and oil. Researchers have conducted research on using nanofluids in solar collectors to see whether or not they can be used as heat transfer fluids. Li et al. (2009) explored that researchers have done experimental along with numerical work and have tried different preparation techniques to observe the effects of nanofluids on heat transfer properties of standard base fluids. Studies carried out by some other researchers (Choi et al., (2001); Natarajan and Sathish, (2009), Masuda et al., (1993), Grimm (1993), Thomas and Sobhan (2011) concluded that aluminum oxide-water nanofluids have much higher thermal properties as compared to base fluids. They also observed an enhancement of about 30% in thermal conductivity and proposed that the nanofluids enhance the efficiency of solar thermal collectors upon using them as the working fluids. Yousefi et al. (2012) carried out experimental studies on flat plate solar collectors by using aluminum oxide-water nanofluids. They found an enhancement in thermal efficiency of 28.3% at 0.2% weight fraction of nanoparticles dispersed in distilled water. The work done by Otanicar et al. (2010) showed the effect of nanofluids on the micro scale solar thermal collector. They observed an increment of 5% in thermal efficiency by using nanofluids as the working fluid. The authors proposed that the performance of solar 269

284 collectors can be enhanced by adding even a small amount of nanoparticles for instance, graphite, nanotubes and silver oxide. Tyagi et al. (2009) conducted research on direct absorption solar collectors (DASC) and have got very similar results with higher thermal properties. They used Al2O3-water nanofluids and obtained nine times higher increment in solar radiation in comparison to water. The researchers found the enhancement in efficiency of 10% higher in comparison to flat plate solar collectors upon using 0.1% to 0.5% volume percent of nanofluids. They also observed a sharp increase in thermal efficiency at low percentages. A comparative work done by Otanicar and Golden (2009) showed the environmental and economic traits of nanofluids on solar thermal collectors. Using hydrogen as a fuel has extraordinary benefits which can lead hydrogen to become a prominent energy carrier in near future for environment friendly and viable development. The main advantage of using hydrogen as a fuel is that it s by products are water and heat only. As clean and environmentally benign energies are turning it to the be the need of the day, it is predicted that the hydrogen will play an important role in fulfilling the future energy needs (Turner, (2004), Gadalla et al. (2010), Abuadala and dincer, (2012), Ratlamwala et al., (2012), Muradov and Veziroglu, (2008). In earlier discussion it is said that the hydrogen has the potential to be the main energy carrier of the future as an eco-friendly and most efficient source of power generation. The main difficulty of using hydrogen as a fuel is that it is not easily available in the atmosphere. Producing hydrogen by methane is the conventional way and it has the disadvantages in the form of producing harmful gasses. The emissions of harmful gasses are due to the presence of carbon contents in methane gas which leads to the same problem of global warming. Producing hydrogen by methane gas is not environment friendly and not even clean from life cycle viewpoints (Ratlamwala et al. (2011). Thus some other ways need to be explored to use hydrogen as energy source. Using solar energy could be the way to produce hydrogen by electrolysis. Solar energy is renewable and vastly available energy in nature and water is another abundantly available source. So producing hydrogen from solar energy would be an alternative to have clean and serene environment. Hydrogen can also be produced via photo-catalysis, water electrolysis, thermochemical and water by using solar energy as the source of energy (Ni et al. (2007) and Ni et al. (2006). Amongst these techniques of producing hydrogen, water electrolysis would be considered as most efficient way of producing hydrogen, and it can also be used to produce hydrogen on a large scale by Ni et al. (2006). Performance of the integrated system can only be evaluated by conducting the energy along with exergy analysis. Conduction both exergetic and energetic analysis simultaneously would provide the clear idea about the real performance of the system. 270 Research conducted by (Ratlamwala et al. (2012); and Kaushik et al. (2000) pinpoints that the researchers have done numerous work on exergy analysis of solar thermal power plants. Dincer and Rosen (2007) have used the approach of second law of thermodynamics to do the exergy analysis of parabolic trough solar collectors. They claimed that irreversibilities between the ambient air and solar collector and between solar collector and sun affect the exergetic performance of solar collectors. Some other researchers Rosen et al. (2008) and Abid et al. (2015) performed exergetic along with environmental analysis of solar collectors. They concluded that it is necessary to do the exergetic analysis because it helps to understand the environmental aspects of solar collectors. Ahmadi et al. (2012) have done exergo-environmental analysis of an integrated organic Rankine cycle (ORC) for electricity, cooling and heating production simultaneously. In the present study, an integrated system containing a parabolic trough and a parabolic dish solar collector (PT/PDSC) a Rankine cycle, and an electrolyzer unit is established for the purpose of hydrogen production. Working parameters such as solar irradiation, ambient temperature and inlet temperature are varied to comprehend their effects on energy efficiency, exergy efficiency, heat rate produced, net power produced and hydrogen produced by solar thermal power plant running on four different fluids. Two of the fluids are nanofluids (Ferric Oxide (Fe2O3) and Aluminum Oxide (Al2O3)) and the other two are salts of LiCl-RbCl and NaNO3-KNO3 in molten state. II. System Description Parabolic trough/parabolic dish solar collector incorporated with Reheat Rankine Cycle is described in fig. 1. Parabolic trough/dish solar collector reflects the solar rays onto the solar receiver. The receiver then transfers the collected energy to the absorption fluids flowing through it. Solar absorption fluids used are Aluminum Oxide (Al2O3), Ferric Oxide (Fe2O3), LiCl-RbCl and NaNO3-KNO3. Aluminum Oxide and Ferric Oxide are nanoparticles mixed in distilled water. The other two fluids are salts in a molten state. Solar absorbers receive the solar energy from the receiver and leave the receiver at comparatively high temperature. Hot fluid leaves for boiler of the steam cycle at state 16 where it exchanges heat with the fluid coming from state 8, and goes back to the receiver at relatively low temperature and enters to the solar collector at state 15 to get heated again by completing the loop. High pressure and high temperature steam produced in the boiler is directed for high pressure turbine at state 9. It provides mechanical energy in the turbine at the expense of losing pressure and temperature. A stream of intermediate temperature and pressure taken from the turbine is directed towards the boiler to be reheated again at state 11. At state 10 a portion of steam is taken out for the closed feed-water heater (CFWH). Steam at state 12 gets reheated at the same temperature as it was at state 9 and heads for

285 the low pressure turbine. It produces power yet again and leaves for the condenser as saturated liquid at state 14. Moderately low pressure and low temperature mixture then enters into the condenser to get cooled and further directed towards the pump at state 1 as compressed liquid. Its pressure gets increased somewhat by pump work and guided towards open feed-water (OFWH) at state 2. Some relatively medium pressure steam is taken from low pressure turbine for the open feed-water heater (OFWH) at state 13. Both streams from state 2 and state 13 get mixed in OFWH and enter into the pump 2 at state 3. Its pressure increases to a high level by flowing through the pump and enters into the CFWH at state 4. The feed-water exchanges heat from the steam coming from state 10 and leaves the CFWH at a relatively high temperature at state 7. Steam coming from high pressure turbine at state 10 loses its energy in CFWH and enters into the pump 3 at state 5. The pressure increases further at state 6 and it gets mixed with feed-water at state 7 and advances towards the boiler of the steam cycle as high pressure fluid at state 8. It gets heated to a high temperature in the boiler and directed towards the turbine to produce power yet again by completing the cycle. Some portion of the power produced of the integrated system is provided to electrolyzer for hydrogen and oxygen production. The remaining power is connected to the grid to be used for domestic proposes. Fig. 1: schematic diagram of the integrated system III. Governing equations This section covers the formulations and equations used to do energy and exergy analysis of parabolic dish and parabolic trough solar collectors. Ambient temperature of 300 K and an atmospheric pressure of 100 kpa is considered for the analysis. III.1. Parabolic Trough Solar Collector The model used for parabolic trough collector is modified form of the model presented by (Kalogirou, 2009, Al-Sulaiman 2014 and Abid et al. 2016). Some important equations for parabolic trough are presented below. The area of the solar receiver is defined as A r = π. D r,0. L (1) where, Dr,o and L represent outer diameter of the receiver and length of collector. The aperture area of the solar collector is determined as A ap = (W D c,0 ). L (2) In order to find the flow regime, we need to calculate the Reynolds number and it is calculated as R e = ρ.v.d c,o μ (3) The Nusselt number based on the Reynold s number is defined as N u = 0.3. R e 0.6 (4) The specific heat capacity of nanofluids used as heat transfer fluids is determined as C pfe2o3 = [ T 3 ri T 2 ri + ] (5a) T ri C pri = X Fe2O3. C pfe2o3 + (1 X Fe2O3 ). C pwater (5b) where T ri and X Fe2O3 represent inlet temperature of receiver and percentage of ferric oxide nanoparticles 271

286 in the water. The heat removal factor is calculated as F r = m r. Cp r A r. U L. [1 exp ( A r. U L. F 1 )] (6) m r. Cp r wherecp r, F1 and UL represent average specific heat capacity of the fluid, the ratio between overall heat loss and solar collector heat loss coefficients and solar collector heat loss coefficient. The useful energy gain can be estimated using the idea of absorbed solar radiation Q g = F r. [S. A ap A r. U L. (T r,i T 0 )] (7) where S and Aap represent absorbed solar radiation and aperture area of the collector. The energy and exergy of PTSC is determined as collector is defined as Q u = F r A a. [S A r A a U L. (T in T 0 )] (12) Heat removal factor is calculated as F r = m C p A r U L [1 exp ( A r.u L.F m C p )] (13) Where, F is the ratio between U0 and UL. The thermal efficiency of the collector is the ratio between the useful energy delivered and the energy of the solar. η en = Q u Q s (14) The exergy efficiency of PDSC calculated based on the equation suggested by Ratlamwala et al. (2013) as η en,pstc = F r. [η r U L. ( T r,i T 0 )] (8a) G b C η ex = E total E solar (15) η ex,pstc = E x col E x solar (8b) where ηr, Gb and C represent receiver efficiency, solar beam radiation per unit area of concentrator and concentration ratio between aperture area and receiver area. III.2. Parabolic Dish Solar Collector III.3. Rankine Cycle In order to find the required parameters to analyze the Rankine cycle, it is necessary to first find the enthalpy values at all states. Energy efficiency of the high pressure turbine is calculated as. The PDSC model adopted in our analysis is the revised version of the model used by (Lloyd C. Ngo, CRSES; University of Pretoria). The total heat loss of the solar receiver needs to be calculated to find the thermal efficiency of the receiver. It is determined as Q l = U l. A r (T r T 0 ) (9) Where UL is the overall heat transfer coefficient, it can be calculated from the same equation used for parabolic trough. Tr is temperature of the receiver. The net solar heat of the sun which falls on the solar receiver is calculated as Q s = G b. A ap (10) The absorbed solar radiation is found using the total beam radiation. S = η 0. G b (11) where η0 represents the optical efficiency of the PDSC and it is assumed to be 0.85 taken from Wu et al. (2010) The Hottel-Whillier equation proposed by Wu et al. (2010) for actual heat gain of concentrating solar 272 η hpt = h 9 h 10 h 9 h s,10 (16) where h9 and h10 represents enthalpy values at state 9 and 10. The same approach can be used to find the efficiency of low pressure turbine. The heat input provided to the boiler and heat rejected by the condenser are determined as q in = h 9 h 8 + z. (h 12 h 11 ) q out = n. (h 14 h 1 ) (17a) (17b) where z, n are fractions of steam taken from the high and low pressure turbines. The energy efficiency of the steam is calculated as η en,st = W net Q b (18) where Q b and W net are heat rate produced of the boiler and net power produced by the overall system. The overall energy efficiency of the PT/PD solar thermal power plant is determined as η en,ov = W net Q solar (19)

287 III.4. Exergy Equations mass flow rate of hydrogen as Exergy values are calculated at all the states of the system to find the exergy destroyed and exergy efficiency of the steam cycle. Exergy at state 1 can be determined as E x 1 = m 1. (h 1 h 0 ) T 0. (S 1 S 0 ) (20) wherem 1, h1, S1 and T0 represent mass flow rate at state 1, enthalpy at state 1, entropy at state 1 and ambient temperature respectively. The exergy at other states of the system can also be calculated using the same approach. The exergy destruction rate of the boiler and the condenser is calculated as E x th,b = [1 T 0 T b ]. Q b E x th,c = [1 T 0 T c ]. Q c (21a) (21b) where Tb and Q b represent temperature of boiler and heat rate produced by boiler. The overall exergy efficiency of the PT/PDSTPP is be calculated as η ex,ov = W net E x solar (22) where E x solar III.5. Electrolyzer exergy of solar. The equations used to analyze the electrolyzer are presented in this section. The exergy of hydrogen needs to be calculated to determine the exergy rate of hydrogen. The exergy of hydrogen is composed of chemical and physical exergies and it is calculated as E x,h2 = E x,ch + E x,ph (23) The chemical and physical exergies are determined as E x,ch = ( ) M H 2 (24) E x,ph = [(h H2 h 0 ) T 0 (S H2 S 0 )] (25) where M H2 represents molar mass of hydrogen The mass flow rate of hydrogen produced is calculated as m H 2 = η elec W net HHV (26) where η elec and HHV are electrical efficiency and high heating value of hydrogen. The exergy rate of hydrogen is calculated using the 273 E x,h 2 = m H 2 E x,h2 (27) Finally the net power produced by electrolyzer is determined as W net,elect = W net 0.2 (28) To solve the mathematical models and the required equations, the Engineering Equation Solver (EES) software is deployed Klein (1975). IV. Results and Discussions A steam turbine integrated with parabolic trough/dish solar collector is presented in this study. Four different absorption fluids are used in both parabolic trough and parabolic dish solar collectors. Parametric studies are done by varying solar irradiance and ambient temperature to observe their effects on outlet temperature of the solar collector, heat rate produced, net power produced, overall energy and exergy efficiencies and the rate of hydrogen produced of the PT/PD solar thermal power plant. The parabolic trough model used in our study is the modified version of the model presented by (Kalogirou, 2009). Under the proposed assumptions, the energy efficiency of parabolic trough solar collector used by Kalogirou is 72.5% and the energy efficiency of the model used in this study is 73.12%, 73.09%, 69.95% and 72.08% respectively for four different fluids. From the efficiency point of view, it is validated that the model used in our analysis is as precise as presented by Kalogirou S. (2009). IV.1. Effect of Solar Irradiation Solar irradiation is considered as the main parameter to affect the performance of solar collectors. It is the amount of energy that is transferred to the circulating fluid in the form of temperature. As the solar irradiation increases, the outlet temperature of the collector increases proportionally. It can be seen in fig. 2 that the outlet temperature of a parabolic dish and parabolic trough solar collector (PD/PTSC) increases with increase in solar irradiance. The outlet temperature of parabolic dish solar collector is found to be increasing from K to 562 K, K to K, K to K and K to K respectively, and for parabolic trough it increases from K to 547 K, K to K, K to K and K to K by varying the solar irradiance from 400 W/m 2 to 1100 W/m 2. The outlet temperature of parabolic dish collector is higher than parabolic trough for all the absorbers tested. Because higher outlet temperature for the parabolic dish is that the concentration ratio (ratio of aperture area over the receiver area) is very high in comparison with parabolic trough where it is relatively small. The higher concentration ratio of parabolic dish makes it the winner in outlet temperature of solar collectors. Higher outlet temperature would be expected because of higher solar intensity reflected

288 onto the receiver. The higher exit temperature of the solar collector results in higher heat rate produced by steam cycle boiler. As the heat, rate produced increases, the power produced by solar thermal power plant grows too. Fig. 3 illustrates the heat rate generated by the boiler increases from 630 W to 2289 W, 631 W to 2292 W, 589 W to 2138 W and 619 W to 2248 W respectively parabolic dish collector, and for parabolic trough it increases from 382 W to 1942 W, 382 W to 1946 W, 320 W to 1558 W and 363 W to 1827 W respectively by varying the solar irradiance in between 400 W/m 2 to 1100 W/m 2. It is clearly seen that the parabolic dish solar collector produces more heat as compared to parabolic trough solar collector. Ferric oxide nanofluid has the highest rate of heat production among all the solar absorbers used. The overall energy efficiency of parabolic dish and trough solar thermal power plant for four different absorbers is shown in fig. 4. The overall energy efficiency of the system using parabolic dish collectors increases from 16.05% to 21.17%, 16.07% to 21.2%, 14.99% to 19.77% and 15.76% to 20.89% respectively, and for parabolic trough collector it increases from 11.18% to 21.08%, 11.19% to 21.09%, 10.69% to 20.18% and 11.03% to 20.79% respectively, by increasing the solar irradiation. It was expected for the overall energy efficiency to increase because increased solar irradiation would result in transferring more energy to the absorption fluid that increases the net power produced and would increase the performance of the system. The overall exergy efficiency follows the footsteps of overall energy efficiency and increases by increasing the solar irradiance. It was observed that the overall exergy efficiency of parabolic trough/dish solar thermal power plant increased proportionally with an increase in solar irradiation as illustrated in fig. 5. The overall exergy efficiency of the incorporated system for parabolic dish as the solar collector increases from 17.28% to 22.83%, 17.3% to 22.83%, 16.14% to 21.29% and 16.97% to 22.39%, respectively and when parabolic trough was used then the overall energy efficiency was found to be increasing from 11.82% to 22.28%, 11.82% to %, 11.3% to % and 11.66% to 21.97% respectively, by varying solar irradiance from 400 W/m 2 to 1100 W/m 2. It can be seen that nanofluids have the higher overall energy and exergy efficiencies for both parabolic dish and parabolic trough collectors. The net power produced which can be considered the most important parameter for solar driven thermal power plants. T out [K] T r,o,pd,al2o3 T r,o,pd,fe2o3 T r,o,pd,liclrbcl T r,o,pd,nano3kno3 T r,o,pt,al2o3 T r,o,pt,fe2o3 T r,o,pt,liclrbcl T r,o,pt,nano3kno G b [W/m 2 ] Fig. 2: Effect of solar irradiation on outlet temperature of solar collectors. Q produced [W] hen,ov [%] hex,ov [%] Q PD,Al2O3,produced Q PD,Fe2O3,produced Q PD,LiClRbCl,produced Q PD,NaNo3KNo3,produced Q PT,Al2O3,produced Q PT,Fe2O3,produced Q PT,LiClRbCl,produced Q PT,NaNo3KNo3,produced G b [W/m 2 ] Fig. 3: Effect of solar irradiation on heat rate produced. 0.2 hen,ov,pd,al2o3 hen,ov,pd,fe2o3 hen,ov,pd,liclrbcl hen,ov,pd,nano3kno3 hen,ov,pt,al2o3 hen,ov,pt,fe2o3 hen,ov,pt,liclrbcl hen,ov,pt,nano3kno G b [W/m 2 ] Fig. 4: Effect of solar irradiation on overall energy efficiency hex,ov,pd,al2o3 hex,ov,pd,fe2o3 hex,ov,pd,liclrbcl hex,ov,pd,nano3kno3 hex,ov,pt,al2o3 hex,ov,pt,fe2o3 hex,ov,pt,liclrbcl hex,ov,pt,nano3kno G b [W/m 2 ] Fig. 5: Effect of solar irradiation on overall exergy efficiency. 274

289 W net [W] m H2 [g/s] G b [W/m 2 ] Fig. 6: Effect of solar irradiation on the net power produced W net,pd,al2o3 W net,pd,fe2o3 W net,pd,liclrbcl W net,pd,nano3kno3 W net,pt,al2o3 W net,pt,fe2o3 W net,pt,liclrbcl W net,pt,nano3kno3 m H2,PD,Al2O3 m H2,PD,Fe2O3 m H2,PD,LiClRbCl m H2,PD,NaNO3KNO3 m H2,PT,Al2O3 m H2,PT,Fe2O3 m H2,PT,LiClRbCl m H2,PT,NaNO3KNO G b [W/m 2 ] Fig. 7: Effect of solar irradiation on the rate of hydrogen produced. It is being shown in fig. 6 that the net power produced by using parabolic dish is higher as compared to parabolic trough. It increases by linearly increasing the solar irradiance. The net power produced by solar thermal power plant for parabolic dish increases from W to W, W to W, W to W and 206 W to W, respectively and for parabolic trough it increases from W to W, W to W, W to W and 102 W to 513 W respectively, with rise in solar irradiance from 400 W/m 2 to 1100 W/m 2. It is very clear that as the boiler heat rate increase, the power produced by the thermal power plant increase as well. Fig. 7 shows the relation between the solar irradiation and the rate of hydrogen produced. The rate of hydrogen produced increases with increase in solar irradiation. The rate of hydrogen produced by PDSTPP increases from g/s to g/s and for PTSTPP it increases from g/s to g/s. the rate of hydrogen production for PDSTPP is 8.1% higher in comparison to PTSTPP K to 553 K, K to K, K to K, and K to K respectively, and for parabolic trough it increases from K to K, K to K, K to K, and K to K by varying the ambient temperature from 275 K to 325 K. It was observed that the LiCl-RbCl had the highest outlet temperature followed by NaNo3KNo3 for both the parabolic trough and for parabolic dish solar collectors. The higher exit temperature fluid would have higher energy among the other absorption fluids. The hotter ambient environment also plays a crucial role in enhancing the performance of solar collectors. The rate of heat production increases by increasing the ambient temperature. It can be seen from the fig. 9 that heat rate produced by a boiler of the steam cycle increases proportionally with an increase in ambient temperature. The heat rate produced using parabolic dish increases from 1990 W to 2105 W, 2002 W to 2108 W, 1870 W to 1970 W and 1960 W to 2070 W respectively, and for parabolic dish it increases from 1620 W to 1820 W, 1630 W to 1830 W, 1300 W to 1480 W and 1530 W to 1720 W respectively by varying the ambient temperature range from 275 K to 325 K. It is clearly seen that the heat rate produced by parabolic dish is relatively higher than parabolic trough that was expected because the outlet temperature of the collector was higher which would result in higher heat transfer rate of the parabolic dish collector. The ambient temperature does affect the overall energy efficiency of solar collectors. The overall energy efficiency of solar thermal power plant working on parabolic dish/trough is shown in fig. 10. It is seen from the illustration that the overall energy efficiency of parabolic dish collector is higher at low ambient temperatures, but it starts to level off with parabolic trough when the temperature got increased. The overall energy efficiency of solar thermal power plant using parabolic dish as the solar collector increases from 20.34% to 21.42%, 20.37% to 21.45%, 19% to 20% and 19.99% to 21.03% respectively, and for parabolic trough it varies between 19% to 21.75% by varying the ambient temperature range from 275 K to 325 K. The overall exergy efficiency shown in fig. 11 followed different trend and it increases linearly with increase with ambient temperature for both parabolic dish and parabolic trough. IV.2. Effect of Ambient Temperature Ambient temperature is another important parameter that affects the performance of solar thermal systems. It can be seen in fig. 7 that the increase in ambient temperature increases the outlet temperature of the solar collector. The exit temperature of the parabolic dish collector for the four tested fluids increases from 275

290 T out [K] Q produced [W] hen,ov [%] T r,o,pd,al2o3 T r,o,pd,fe2o3 T r,o,pd,liclrbcl T r,o,pd,nano3kno3 T r,o,pt,al2o3 T r,o,pt,fe2o3 T r,o,pt,liclrbcl T r,o,pt,nano3kno T 0 [K] Fig. 8: Effect of ambient temperature on outlet temperature of the receiver Q PD,Al2O3,produced Q PD,Fe2O3,produced Q PD,LiClRbCl,produced Q PD,NaNo3KNo3,produced 1400 Q PT,Al2O3,produced Q PT,Fe2O3,produced Q PT,LiClRbCl,produced Q PT,NaNo3KNo3,produced T 0 [K] Fig. 9: Effect of ambient temperature on heat rate produced hen,ov,pd,al2o3 hen,ov,pd,fe2o3 hen,ov,pd,liclrbcl hen,ov,pd,nano3kno3 hen,ov,pt,al2o3 hen,ov,pt,fe2o3 hen,ov,pt,liclrbcl hen,ov,pt,nano3kno T 0 [K] Fig. 10: Effect of ambient temperature on the overall energy efficiency hex,ov,pd,al2o3 hex,ov,pd,fe2o3 hex,ov,pd,liclrbcl hex,ov,pd,nano3kno3 The overall exergy efficiency of parabolic dish driven thermal power plant varies from 20.33% to 23.25% and for parabolic trough it is in between 19.29% to 23.09%, respectively with increase in ambient temperature from 275 K to 325 K. The net power produced by solar thermal power plant increases with increase in ambient temperature. But the net power produced by using dish solar collector is much higher as compared to parabolic trough solar collector. The reason for the higher power output for parabolic dish can be attributed to higher concentration ratio, because higher concentration ratio would result in higher outlet temperature of the solar collector and finally the higher net power produced by STPP. The net power produced by STPP increases from W to 700 W, W to 701 W, 621 W to W and W to W using parabolic dish collector, and for parabolic trough it increases from W to W, W to W, 366 W to 415 W and W to W respectively, with rise in ambient temperature from 275 K to 325 K. It can be seen again from the fig. 12 that the both aluminum oxide and ferric oxide nanofluids have higher net power produced in comparison with molten salts. Fig. 13 shows the relation between the ambient temperature and the rate of hydrogen produced. It can be clearly seen that the increase in ambient temperature increases the hydrogen production rate. Hydrogen production rate for parabolic dish thermal power plant and parabolic trough thermal power plant varies from g/s to g/s and from g/s to g/s, respectively with increase in ambient temperature. W net [W] T 0 [K] Fig. 12: Effect of ambient temperature on the net power produced W net,pd,al2o3 W net,pd,fe2o3 W net,pd,liclrbcl W net,pd,nano3kno3 W net,pt,al2o3 W net,pt,fe2o3 W net,pt,liclrbcl W net,pt,nano3kno3 hex,ov [%] hex,ov,pt,al2o3 hex,ov,pt,fe2o3 hex,ov,pt,liclrbcl hex,ov,pt,nano3kno3 m H2 [g/s] m H2,PD,Al2O3 m H2,PD,Fe2O3 m H2,PD,LiClRbCl m H2,PD,NaNO3KNO3 m H2,PT,Al2O3 m H2,PT,Fe2O3 m H2,PT,LiClRbCl m H2,PT,NaNO3KNO T 0 [K] Fig. 11: Effect of ambient temperature on the overall exergy efficiency G b [W/m 2 ] Fig. 13: Effect of ambient temperature on the rate of hydrogen produced.

291 IV.3. Effect of Inlet Temperature Inlet temperature is another important parameter which affects the performance of the solar collectors in great deal. Fig. 14 describes the relationships between the inlet temperature and the outlet temperature of solar collectors. Outlet temperature increases linearly with increase in inlet temperature. The outlet temperature of PDSC increases from K to K and for PTSC it increases from K to K with increase in inlet temperature from 425 K to 525 K. Fig. 15 illustrates that the overall energy efficiency decreases with increase in inlet temperature of solar collectors. The same trend can be seen in fig. 16 where the overall exergy efficiency decreases proportionally with increase in inlet temperature of solar collectors. T out [K] hen,ov [%] hex,ov [%] T r,o,pd,al2o3 T r,o,pd,fe2o3 T r,o,pd,liclrbcl T r,o,pd,nano3kno3 550 T r,o,pt,al2o3 T r,o,pt,fe2o3 T r,o,pt,liclrbcl T r,o,pt,nano3kno T in [K] Fig. 14: Effect of inlet temperature on outlet temperature of solar collectors hen,ov,pd,al2o3 hen,ov,pd,fe2o3 hen,ov,pd,liclrbcl hen,ov,pd,nano3kno3 hen,ov,pt,al2o3 hen,ov,pt,fe2o3 hen,ov,pt,liclrbcl hen,ov,pt,nano3kno T in [K] Fig. 15: Effect of inlet temperature on the overall energy efficiency hex,ov,pd,al2o3 hex,ov,pd,fe2o3 hex,ov,pd,liclrbcl hex,ov,pd,nano3kno3 hex,ov,pt,al2o3 hex,ov,pt,fe2o3 hex,ov,pt,liclrbcl hex,ov,pt,nano3kno T in [K] Fig. 16: Effect of inlet temperature on the overall exergy efficiency. 277 The rate of hydrogen production decreases in the same fashion as overall energy and exergy efficiencies. The rate of hydrogen production decreases with increase in inlet temperature of solar collectors. It is shown in fig. 17 that the hydrogen production rate decreases from g/s to g/s and g/s to g/s for both PDSC and PTSC with increase in inlet temperature from 425 K to 525 K. m H2 [g/s] T in [K] Fig. 17: Effect of inlet temperature on the rate of hydrogen produced. V. Conclusions m H2,PD,Al2O3 m H2,PD,Fe2O3 m H2,PD,LiClRbCl m H2,PD,NaNO3KNO3 m H2,PT,Al2O3 m H2,PT,Fe2O3 m H2,PT,LiClRbCl m H2,PT,NaNO3KNO This comparative study has been conducted for energy and exergy analyses of a parabolic trough and parabolic dish solar thermal power plant. Four different solar absorption fluids are used from which, two of them are nanofluids (Aluminum Oxide and ferric Oxide) and others are LiCl-RbCl and NaNO3-KNO3 molten salts. The operated parameters such as solar irradiation, ambient temperature and inlet temperature are varied to comprehend their effects on outlet temperature of the receiver, heat rate produced, net power produced, overall energy efficiency, overall exergy efficiency and rate of hydrogen produced of parabolic dish and parabolic trough assisted solar thermal power plant. It is observed that the performance parameters such as outlet temperature of solar receiver, net power output, overall energy efficiency, overall exergy efficiency and hydrogen production rate are found to be increasing function of solar irradiation and ambient temperature. But it is found to be the opposite with increase in inlet temperature of the solar collectors. The findings show that the outlet temperature of solar receiver increases from K to 562 K, K to K, K to K and K to K respectively, for parabolic dish collector, and for parabolic trough collector it increases from K to 547 K, K to K, K to K and K to K respectively, by varying the solar irradiance from 400 W/m 2 to 1100 W/m 2. The rate of hydrogen production is found to be increases from g/s to g/s for PDSTPP and for PTSTPP it varies between g/s to g/s. It can be seen that the parabolic dish solar collector has higher outlet temperature of the receiver in comparison with

292 parabolic trough. The higher exit temperature would be attributed to high concentration ratio of parabolic dish collector. The overall exergy efficiency of parabolic dish driven solar thermal power plant for four solar absorbers increases from 21.77% to 23.21%, 21.8% to 23.25%, 20.33% to 21.68%, and 21.38% to 22.79%, respectively, and for parabolic trough it increases from 20.3% to 23.08%, 20.31% to 23.09%, 19.29% to 22.24%, and 19.98% to 22.81%, with increase in ambient temperature from 275 K to 325K. Findings of the conducted study show that both nanofluids have higher exergetic and energetic efficiencies as compared to molten salts. Their performance even gets better on using them as solar absorbers in parabolic dish collector. Net power produced by parabolic dish and parabolic trough solar thermal power plant is also observed to be higher with nanofluids for all three operated parameters. The rate of hydrogen production by ferric oxide nanofluid is 4.28% higher than LiCl-RbCl salt for PDSTPP. Nomenclature Aap : aperture area (m 2 ) Ar : receiver area (m 2 ) C : concentration ratio Cp : specific heat capacity (J / g C) Dr,o : outer diameter of receiver (m) Dc,o : outer diameter of cover (m) Ex : Exergy rate (kw) Fr : heat removal factor Gb : solar irradiation (w/m 2 ) h : heat transfer coefficient h : enthalpy (kj/kg) hc,ca : convective heat transfer coefficient between ambient and glass cover hr,cr : convective heat transfer coefficient between glass cover and receiver tube Kr : thermal conductivity of receiver tube L : length of the solar receiver (m) ṁ : mass flow rate (kg/s) Nu : Nusselt number UL : overall heat loss (W/m 2.C) Q : Heat rate (kw) r : receiver Re : Reynolds number S : absorbed solar radiation (w/m2) U0 : overall heat transfer coefficient W : Work rate (kw) V : velocity (m/s) T0 : ambient temperature (K) Tc : glass cover temperature (K) Greek letters ϵ : surface emissivity ρ : density (kg/m3) η : efficiency µ : dynamic viscosity Subscripts abs : absorbed avg : average b : boiler 278 con : condenser cond : conduction col : collector dest : destroyed en : energy ex : exergy i : inner o : outer p : pump r : receiver sol : solar s : ideal st : steam 0 16 : state numbers Acronyms CFWH DASC LiCl-RbCl NaNO3-KNO3 OFWH ORC PT/PD PTSTPP References : closed feed water heater : direct absorption solar collector : Lithium chloride-rubidium chloride : Sodium nitrate-potassium nitrate : open feed water heater : organic Rankine cycle : parabolic trough/parabolic dish : parabolic trough solar thermal power plant Abuiadala, A., Dincer, I. A review on biomass-based hydrogen production and potential applications. Int. J. Energy Res. 36, (2012). Ahmadi P., I. Dincer, Marc A. Rosen. Exergo-environmental analysis of an integrated organic Rankine cycle for trigeneration Energy Conversion and Management 64, (2012). Alkhamis, A.I., Sherif, S.A. Feasibility study of a solar-assisted heating/cooling system for an aquatic center in hot and humid climates. Int. J. Energy Res. 21, (1997). Al-Sulaiman F.A. Exergy analysis of parabolic trough solar collectors integrated with combined steam and organic Rankine cycles, Energy Conver. Manag (2014). Bakos G.C., I. Ioannidis, N.F. Tsagas, I. Seftelis. Design optimization and conversion-efficiency determination of a line-focus parabolic-trough solar collector (PTC), Applied Energy (2001). Barigozzi, G., Bonetti, G., Perdichizzi, F.A., Ravelli, S. Thermal performance prediction of a solar hybrid gas turbine. Sol. Energy 86, (2012). Choi S., J. A. Eastman, Enhanced Heat Transfer Using Nanofluids, U.S. Patent. 6221,275 (2001). Dincer I., Rosen MA. Exergy, energy, environment and sustainable development. Oxford: Elsevier; (2007).

293 Dincer I., and Ratlamwala T. A.H. Solar Thermal Power Systems. Earth Systems and Environmental Sciences, (2013) Gadalla, M.A., Ratlamwala, T.A.H., Dincer, I. Energy and exergy analyses of an integrated fuel cell and absorption cooling system. Int. J. Exergy 7, (2010). Garcia, I.L., Alvarez, J.L., Blanco, D. Performance model for parabolic trough solar thermal power plants with thermal storage: comparison to operating plant data. Sol. Energy 85, (2011). Grimm A. Powdered aluminum-containing heat transfer fluids, German Patent DE A1 (1993). Huang, W., Hu, P., Chen, Z. Performance simulation of a parabolic trough solar collector. Sol. Energy 86, (2012). Kalogirou, S.A. Solar thermal collectors and applications. Prog. Energy Combust. Sci. 30, (2004). Kalogirou S. A. Solar energy engineering: processes and systems. London: Elsevier Inc, (2009). KALOGIROU, S.A. STC and Applications. Progress in Energy & Combustion Science, 30(3), pp (2004). KALOGIROU, S., A. Exergy Analysis and Genetic Algorithms for the Optimization of Flat Plate Solar Collectors, The 25th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems (2012). Kalogirou S. A. Solar energy engineering: processes and systems. London: Elsevier Inc, KAR, K., A., Exergy Efficiency and Optimum Operation of Solar Collectors. Applied Energy, 21, pp (2009). Kaushik SC, Misra RD, Singh N. Second law analysis of a solar thermal power system. Int J Solar Energy 20: (2000). Kaushik S.C., M.K. Singhal, S.K. Tyagi. Solar collector technologies for power generation and space air conditioning applications: A state of the art, Internal Report, CES, Indian Institute of Technology, Delhi, India. (2001). Klein S. A. F-Chart Software; engineering equation solver. (1975). Le-Roux, W.G., Bello-Ochende, T., Meyer, J.P. Thermodynamic optimization of the integrated design of a small-scale solar thermal Brayton cycle. Int. J. Energy Res. 36, (2012). Li Y., J. Zhou, S. Tung, E. Schneider, S. Xi, A review 279 on development of nanofluid preparation and characterization, Powder Technol (2009). Lloyd C. Ngo, Exergetic Analysis and Optimization of a Parabolic Dish Collector for Low Power Application, Centre for Renewable and Sustainable Energy Studies, University of Pretoria (2013). Masuda H, Ebata A, Teramae K, Hishinuma N. Alteration of thermal conductivity and viscosity of liquid by dispersing ultra-fine particles (dispersion of g- Al2O3, SiO2 and TiO2 ultra-fine particles). Netsu Bussei (Japan) 7: (1993). M. Abid, T.A.H. Ratlamwala, U. Atikol. Performance Assessment of parabolic dish and parabolic trough solar thermal power plant using nanofluids and molten salts. Int. J. Energy Res. 40: (2016). M. Abid, T.A.H. Ratlamwala, U. Atikol. Comparative Energy, Exergy and Environmental Analyses of Parabolic Trough Solar Thermal Power Plant Using Nanofluids. IEEES-7. 27th - 30th April. ENSIAME-LAMIH, France. (2015). Muradov, N.Z., Veziroglu, T.N. Green path from fossil-based to hydrogen economy: an overview of carbon-neutral technologies. Int. J. Hydrogen Energy 33, (2008). Natarajan E, Sathish R. Role of nanofluids in solar water heater. Int J Adv Manuf Technol. doi: /s (2009). Ni M, Leung MKH, Leung DYC. A review and recent developments in photo catalytic water splitting using TiO2 for hydrogen production. Renewable and Sustainable Energy Reviews.11: (2007). Ngo LC. Exergetic Analysis and Optimization of a Parabolic Dish Collector for Low Power Application, Centre for Renewable and Sustainable Energy Studies. University of Pretoria, (2013). Ni M, Leung MKH, Leung DYC. An electrochemical model of a solid oxide steam electrolyzer for hydrogen production. Chemical Engineering & Technology. 29: (2006). Otanicar T., P.E. Phelan, R.S. Prasher, G. Rosengarten, R.A. Taylor. Nanofluid based direct absorption solar collector, J. Renew. Sust. Energy (2010). Otanicar T., P.E. Phelan, R.S. Prasher, G. Rosengarten, R.A. Taylor, Nanofluid based direct absorption solar collector, J. Renew. Sust. Energy (2010). Otanicar T., J. Golden, Comparative environmental and economic analysis of conventional and nanofluid solar hot water technologies, Environ. Sci. Technol (2009).

294 Ratlamwala TAH, Dincer I, Gadalla MA. Energy and exergy analyses of an integrated solar based desalination quadruple effect absorption system for fresh water and cooling production. International Journal of Energy Research 37: (2013). Ratlamwala, T.A.H., Dincer, I., M. Aydin. Energy, exergy analyses and optimization study of an integrated solar heliostat field system for hydrogen production. International journal of hydrogen energy (2012). Ratlamwala TAH, Gadalla MA, Dincer I. Performance assessment of an integrated PV/T and triple effect cooling system for hydrogen and cooling production. Int J Hydrogen Energy 36: (2011). Rosen, M.A., Dincer, I., Kanoglu, M., Role of exergy in increasing efficiency and sustainability and reducing environmental impact. Energy Policy 36 (1), (2008). Singh N., S.C Kaushik. Technology assessment and economic evaluation solar thermal power systems: A state of art, Report, Centre for Energy Studies, Indian Institute of Technology, Delhi, India (1993). Thomas S., C. Sobhan, A review of experimental investigations on thermal phenomena in nanofluids, Nanoscale Res. Lett (2011). Tyagi H., P. Phelan, R. Prasher, Predicted efficiency of a low-temperature nanofluid based direct absorption solar collector, J. Sol. Energy. Eng (2009). TYAGI, S.K., WANG, S., SINGHAL, M.K., KAUSHIK, S.C. and PARK, S.R. Exergy analysis and parametric study of concentrating type solar collectors. International Journal of Thermal Sciences, 46(12), pp (2007). Turner, J.A. Sustainable hydrogen production. Science 305, (2004). Wu S, Xiao L, Cao Y, Li Y. A parabolic dish/amtec solar thermal power system and its performance evaluation. Applied Energy 87(2): (2010). Yousefi T, Veysi F, Shojaeizadeh E, Zinadini S. An experimental investigation on the effect of Al2O3 H2O nanofluid on the efficiency of flat-plate solar collectors. Renew Energy; 39:293 8 (2012). 280

295 New Climate Zone Definitions of Turkey by Using Typical Meteorological Year Data Serpil Yilmaz 1*, İsmail Ekmekci 2 1 Gedik University, Cumhuriyet Mah, İlkbahar Sk , Kartal/İstanbul, Turkey 2 İstanbul Commerce University, Küçükyalı E5 Kavşağı İnönü Cad. No: 4 Küçükyalı İstanbul, 34840, Turkey * serpilbozkurtyilmaz@gmail.com ; iekmekci@ticaret.edu.tr Abstract Climate has a major impact on the energy use of residential buildings, and energy codes. Climate zone map is so important to the construction industry and standards rely on a clear definition of climate zones to convey requirements to builders. In Turkey there are four climate zones for 81 cities. For these four zones, the current regulations and building design parameters are explained in TS 825 Standard. Experiences show that the existing climate zoning for Turkey applies only for heating conditions and needs to be modified. In this study, a new climate zoning was established according to the format recommended by the American Society of Heating, Refrigerating and AirConditioning Engineers Inc. (ASHRAE). Heating degree day (HDD) with base temperature 18 C and cooling degree day (CDD) s with base temperature 10 C values were taken as basis. All zoning data were calculated according to Typical Meteorological Year data. For this purpose, the hourly measured weather data of a 23-year period ( ) were analysed for 81 cities of Turkey. The results were shown in tabulated and graphical form. This study results 12 climate zones which indicates that a more detailed analysis should be performed and new building design parameters must be produced for Turkey. Keywords: Climate, Energy, Climate Zone, Heating Degree Day (HDD), Cooling Degree Day (CDD), Typical Meteorological Year (TMY) I. Introduction Energy consumption is one of the most important factors indicating the economic and industrial levels of countries; however energy protection policy has become an important research subject in developed and developing countries, following the global energy crisis in The building sector (commercial, residential, and industrial) consumes 30 50% of the total energy requirements of a society. For sustainable development, a reduction in energy demand is essential. This could be achieved through improving energy efficiency, effective energy conservation and management. The weather conditions of a given region are the most important considerations for the proper design of space AC systems. Estimations of the future energy consumption of buildings are becoming increasingly important as a basis for energy management, energy renovation, investment planning, and for determining the feasibility of technologies and designs. Future weather scenarios, where the outdoor climate is usually represented by future weather files, are needed for estimating the future energy consumption. In many cases, however, the practitioner s ability to conveniently provide an estimate of the future energy consumption is hindered by the lack of easily available future weather files. Energy performance calculations are so important to be able to identify the buildings in the meaning of efficiency. Useful results have been obtained in recent years, thanks to these studies. Climate is very important factor for calculating the energy use of residential buildings. For simplicity, climate classification is a common way to group cities which have similar characteristics. In Turkey, there are four degree-day regions which are used in building heating energy need calculations. And building design parameters are suggested for these four climate zones. This classification is so rough. In this paper we tried to produce a new climate zone approach which is based on degree-hour data derived from Typical Meteorological Year data. II. Methodology II.1. Typical Meteorological Year The design of buildings, in terms of energy consumption and thermal comfort, requires appropriate weather data. The generation of a typical meteorological year is important factor for calculations concerning many applications in the field of thermal engineering and building energy simulations. A representative database for year duration is known as a Typical Meteorological Year (TMY), a term mainly used in the USA, consists of individual months of meteorological data sets selected from different years over the available data period which is called long-term measured data series. One of the most common methodologies for generating a TMY is the one proposed by Hall et al. using the Filkenstein Schafer (FS) statistical method. A TMY consists of the months selected from the individual years and sorted to form a complete year. 281

296 In the literature, there are many attempts to produce weather databases for different locations. The main objective of these methods is to select representative months from the multi-year database.in this study, the hourly measured weather data of a 23-year period ( ) were analysed for 81 cities of Turkey and typical meteorological years were calculated every city of Turkey by SQL programmıng language. II.2. Degree-Days Method One of significant meteorological variables that relate to building and residential energy consumption is heating (HDD) and cooling degree-days (CDD). Both of them are basic quantities for preliminarily estimating energy consumption of a building. Unfortunately the degree-days are not commonly used in the past few years because of advanced communication systems that allows one to be able to emotely access data storages or servers to directly retrieve data. The data could be yearly, daily, hourly, or even subhourly whichever matches the users requirements. Another reason is that effects of latent loads are not accounted into the degree-days. Nevertheless HDD and CDD are still useful in preliminary energy audits to estimate savings of energy conservation measures (ECM) that depend on seasonal and outdoor temperature for which simple methods are more appropriate than complex and time-consuming methods. Thus there still are several researches in recent years to study and predict variations and trends of Essentially degree-days are a summation of the differences between the outdoor temperature and some reference temperature over a specified time period. The reference temperature is known as the base temperature which, for buildings, is a balance point temperature, i.e. the outdoor temperature at which the heating (or cooling) systems do not need to run in order to maintain comfort conditions. Degree-days provide a significant advantage over other simplified methods that use mean outdoor temperatures to calculate energy demand. Because degree-days account for fluctuations in the outdoor temperature, and eliminate those periods when heating (or cooling) systems do not need to operate, they can capture extreme conditions in a way that mean temperature methods cannot. This makes them more reliable in estimating energy consumption, particularly in the milder months, but also inthose periods with extreme cold snaps where they capture both magnitude and duration of an event. It is well-established tool for energy analysis if the buildings use, the efficiency of HVAC equipment, indoor temperature and internal gains are relatively constant. The two main uses for degree-days in buildings are: cooling for new build and major refurbishments for on-going energy monitoring and analysis of existing buildings based on historical data It must be stressed that, particularly for estimation purposes, degree-day techniques can only provide approximate results since there are a number of simplifying assumptions that need to be made. These assumptions particularly relate to the use of average conditions (internal temperatures, casual gains, air infiltration rates etc), and that these can be used in conjunction with each other to provide a good approximation of building response. The advantage to their use, therefore, lies in their relative ease and speed of use, and all of the information required Heating and cooling degree-days are defined as the sum of the differences between daily average temperatures and the base temperature. For example, the number of heating degreedays in a month, HDD is calculated as HDD = (1) where N is the number of days in the month, T b is the base temperature to which the degree-days are calculated, and is the mean daily temperature. The + superscript indicates that only positive values of the bracketed quantity are taken into account in the sum. Similarly, monthly cooling degreedays CDD are calculated as [10] CDD = N T b T + i=1 i N T i T + i=1 b (2) In this study, heating degree day (HDD) with base temperature 18 C and cooling degree day (CDD) s with base temperature 10 C values were taken as basis III. Climate Zone Definition The climate of a given place / region / area is the total composition of many factors defining the state of the atmosphere at that place. Such factors include temperature, humidity (wetness / dryness), wind (speed, direction), atmospheric clarity (or dustiness) etc. Some of the major factors influencing climate on a global scale will be further explained below. The sun is the major factor influencing climates. Almost all of the energy reaching the earth comes from the sun in the form of radiation. In this study, a new climate zoning was established according to the format recommended by the American Society of Heating, Refrigerating and AirConditioning Engineers Inc. (ASHRAE) as shown in Table 1. to estimate energy consumption and carbon dioxide emissions due to space heating and 282

297 Tab. 1: Ashrae Climate Zone Definition Tab. 2: Climate Zones for Turkey CITY NAME ZONE NUMBER ZONE EXPLANATION BARTIN 4A AND 4B MIXED-HUMID (4A),DRY(4B) ZONGULDAK 5A AND 5B COOL-HUMID (5A),DRY(5B) SİNOP 4A AND 4B MIXED-HUMID (4A),DRY(4B) SAMSUN 3C WARM-MARINE (3C) ORDU 4A AND 4B MIXED-HUMID (4A),DRY(4B) GİRESUN 4A AND 4B MIXED-HUMID (4A),DRY(4B) TRABZON 3C WARM-MARINE (3C) RİZE 4A AND 4B MIXED-HUMID (4A),DRY(4B) ARTVİN 5A AND 5B COOL-HUMID (5A),DRY(5B) ARDAHAN 7 VERY COLD(7) EDİRNE 2 HOT-HUMID (2A),DRY(2B) KIRKLARELİ 5A AND 5B COOL-HUMID (5A),DRY(5B) TEKİRDAĞ 5A AND 5B COOL-HUMID (5A),DRY(5B) İSTANBUL 3C WARM-MARINE (3C) KOCAELİ 2 HOT-HUMID (2A),DRY(2B) SAKARYA 4A AND 4B MIXED-HUMID (4A),DRY(4B) BOLU 6A AND 6B COLD-HUMID (6A),DRY(6B) DÜZCE 5A AND 5B COOL-HUMID (5A),DRY(5B) KASTAMONU 6A AND 6B COLD-HUMID (6A),DRY(6B) KARABÜK 6A AND 6B COLD-HUMID (6A),DRY(6B) ÇANKIRI 5A AND 5B COOL-HUMID (5A),DRY(5B) ÇORUM 5A AND 5B COOL-HUMID (5A),DRY(5B) AMASYA 4A AND 4B MIXED-HUMID (4A),DRY(4B) TOKAT 5A AND 5B COOL-HUMID (5A),DRY(5B) GÜMÜŞHANE 6A AND 6B COLD-HUMID (6A),DRY(6B) BAYBURT 7 VERY COLD(7) SİVAS 7 VERY COLD(7) ERZİNCAN 6A AND 6B COLD-HUMID (6A),DRY(6B) ERZURUM 7 VERY COLD(7) KARS 7 VERY COLD(7) AĞRI 8 SUBARCTIC/ARCTIC(8) IĞDIR 6A AND 6B COLD-HUMID (6A),DRY(6B) ÇANAKKALE 2 HOT-HUMID (2A),DRY(2B) BURSA 4A AND 4B MIXED-HUMID (4A),DRY(4B) YALOVA 4A AND 4B MIXED-HUMID (4A),DRY(4B) BİLECİK 5A AND 5B COOL-HUMID (5A),DRY(5B) ESKİŞEHİR 7 VERY COLD(7) ANKARA 5A AND 5B COOL-HUMID (5A),DRY(5B) KIRIKKALE 5A AND 5B COOL-HUMID (5A),DRY(5B) YOZGAT 7 VERY COLD(7) BALIKESİR 4A AND 4B MIXED-HUMID (4A),DRY(4B) 283 CITY NAME ZONE NUMBER ZONE EXPLANATION KÜTAHYA 6A AND 6B COLD-HUMID (6A),DRY(6B) KIRŞEHİR 5A AND 5B COOL-HUMID (5A),DRY(5B) TUNCELİ 6A AND 6B COLD-HUMID (6A),DRY(6B) VAN 6A AND 6B COLD-HUMID (6A),DRY(6B) MANİSA 4A AND 4B MIXED-HUMID (4A),DRY(4B) UŞAK 5A AND 5B COOL-HUMID (5A),DRY(5B) AFYONKARAHISAR 5A AND 5B COOL-HUMID (5A),DRY(5B) AKSARAY 2 HOT-HUMID (2A),DRY(2B) NEVŞEHİR 6A AND 6B COLD-HUMID (6A),DRY(6B) KAYSERİ 6A AND 6B COLD-HUMID (6A),DRY(6B) MALATYA 5A AND 5B COOL-HUMID (5A),DRY(5B) ELAZIĞ 5A AND 5B COOL-HUMID (5A),DRY(5B) BİNGÖL 6A AND 6B COLD-HUMID (6A),DRY(6B) MUŞ 6A AND 6B COLD-HUMID (6A),DRY(6B) BİTLİS 7 VERY COLD(7) SİİRT 2 HOT-HUMID (2A),DRY(2B) İZMİR 1 VERY HOT-HUMID (1A),DRY(1 AYDIN 2 HOT-HUMID (2A),DRY(2B) DENİZLİ 2 HOT-HUMID (2A),DRY(2B) BURDUR 5A AND 5B COOL-HUMID (5A),DRY(5B) ISPARTA 5A AND 5B COOL-HUMID (5A),DRY(5B) KONYA 5A AND 5B COOL-HUMID (5A),DRY(5B) KARAMAN 6A AND 6B COLD-HUMID (6A),DRY(6B) NİĞDE 5A AND 5B COOL-HUMID (5A),DRY(5B) KAHRAMANMARAŞ 2 HOT-HUMID (2A),DRY(2B) GAZİANTEP 4A AND 4B MIXED-HUMID (4A),DRY(4B) KİLİS 3A AND 3B WARM-HUMID (3A),DRY(3B) ADIYAMAN 4A AND 4B MIXED-HUMID (4A),DRY(4B) ŞANLIURFA 1 VERY HOT-HUMID (1A),DRY(1 MARDİN 2 HOT-HUMID (2A),DRY(2B) DİYARBAKIR 2 HOT-HUMID (2A),DRY(2B) BATMAN 1 VERY HOT-HUMID (1A),DRY(1 HAKKARİ 6A AND 6B COLD-HUMID (6A),DRY(6B) ŞIRNAK 2 HOT-HUMID (2A),DRY(2B) MUĞLA 4A AND 4B MIXED-HUMID (4A),DRY(4B) ANTALYA 3A AND 3B WARM-HUMID (3A),DRY(3B) MERSİN 1 VERY HOT-HUMID (1A),DRY(1B) ADANA 1 VERY HOT-HUMID (1A),DRY(1B) OSMANİYE 2 HOT-HUMID (2A),DRY(2B) HATAY 1 VERY HOT-HUMID (1A),DRY(1B)

298 Fig. 1: The Contour Map of Climate Zones for Turkey IV. Conclusions In this study, a new climate zoning was created according to the format recommended by the American Society of Heating, Refrigerating and AirConditioning Engineers Inc. (ASHRAE 2001). Heating degree day (HDD) with base temperature 18 C and cooling degree day (CDD) s with base temperature 10 C values were taken as basis. 81 city of Turkey was simply divided into nine climatic regions. While comparing with current regulations and TS 825, this study shows that a more detailed analysis should be performed for all the cities, and the new building design parameters must be produced. Acknowledgements I would like to thank my advisor Assist. Prof. Mustafa Yilmaz and MSc advisor Prof. Dr. Birol Kılkış for his guidance and help. I would also like to thank Research & Technical Services Manager of Ashrae, Mr Michael Vaughn and TC 4.2 member Mr Didier Thenevard for very valuable support in USA. 284 References ASHRAE, Ashrae handbook Fundamentals, D. Thevenard, Methods for estimating heating and cooling degree-days to any base temperature, ASHRAE Transactions, vol. 117,pp. 1, pp , Besharat F, Deghan A,Faghih A. Emprical models for estimating global solar radiation : A review and case study. Renewable and Sustainable Enegy Reviews 21(2013) Besharat F,Dehghan A,Faghih A.Emprical models for estimating global solar radiation: A review and case study.renewable and Sustainable Enegy Reviews 21(2013) Bulut H, Büyükalaca O, Yılmaz T. Analysis of variable-base heating and cooling degree-days for Turkey. Applied Energy 69(2001) Bulut H, Büyükalaca O, Yılmaz T. Türkiye için ısıtma ve soğutma derece-gün bölgeleri In: 16. National heat science and technique congress; 2007 [ in Turkish ] Butler D. Architects of a low-energy future. Nature

299 2008;452(3):520e3. F. Jiang, X. Li, B. Wei, R. Hu, and Z. Li, Observed trends of heating and cooling degree-days in Xinjiang province, China, Theoretical &Applied Climatology, vol. 97, pp , G. Krese, M. Prek, and V. Butala, Incorporation of latent loads into the cooling degree days concept, Energy and Buildings, vol. 43, pp , M. Kadioglu, Z. Sen, and L. Gultekin, Variations and trends in Turkish seasonal heating and cooling degree-days, Climate Change,vol. 9, pp , M. Kolokotroni, Y. Zhang, and R. Giridharan, Heating and cooling degree day prediction within the London urban heat island area, Building Services Engineering Research and Technology, vol. 30, no. 3,pp , Merter U, Arif I. Typical weather data of main Turkish cities for energy applications. Int J. Energy Res 2000; 24(8) : O. J. Adigun and J. A. Olorunmaiye, Modeling of cooling degree-days for Southern Nigeria using beta distribution, Journal of Engineeringand Technology, vol. 3, no. 2, pp. 1-7, 2005 Perez-Lombard L, Ortiz J, Pout C. A review on buildings energy consumption information. Energy Build 2008;40(3):394e8. Pusat S, Ekmekçi İ, Akkoyunlu T. Generation of typical meteorological year for different climates of Turkey. Pusat S, Ekmekçi İ. A study on climatic zones of Turkey, In:IX. İnternational HVAC + R technology Symposium :

300 Exergetic and Energetic Performance Evaluation of a Flat Plate Solar Collector in Dynamic Behavior Hamed Mouna 1*, Ben Brahim Ammar 2 1,2 University of Gabes, National Engineering School, Applied Thermodynamics Research Unit (UR:11ES80), 6072 Gabes, Tunisia * Hamedm@hotmail.fr Abstract This work aims to evaluate the energy and the exergy performance of a flat plate solar collector in transient conditions. A theoretical model based on the first and the second laws of thermodynamics is developed to predict the thermal behavior of the system. Exergy analysis is accomplished to determine the location, type, and magnitude of exergy destroyed and losses over the collector. The mathematical model obtained is solved numerically using Matlab computational program and the software Stat Ease Expert Design. A detailed parametric study is done to assess the effect of various parameters such as: mass flow rates, inlet water temperature, absorber emissivity, tubes number and pipes diameter on the energy and exergy efficiencies and on the exergy destroyed and losses. The results show that the values of exergy efficiency are very low compared to those energetic. Further, the main causes of exergy destruction result from absorption of radiation by the absorber plate. The mass flow rate, the inlet water temperature and the absorber emissivity are critical parameters for a solar collector and should be chosen carefully. However, the tubes number and the pipes diameter have a negligeable effects on the collector performance. Thus, more accurate results and beneficial applications of the exergy method in transient behaviors to design the solar collectors are obtained. Keywords: Exergy, energy, efficiency, performance, dynamic I. Introduction In view of the world s depleting fuel reserves, which provide the major source of energy, the development of non conventional renewable energy sources has received an impetus for heating cooling, drying and power generation applications. Solar energy is one kind of important resource for clean and renewable energy and is widely investigated in many fields. There are two main methods to tap solar energy, i.e. with PV cells and with solar collectors (Subiantoro and Tiw, 2013). Among various types of solar collectors, flat plate solar collector (FPSC), based on its simpler technology and lower costs has globally become the mostly used one. Hence, the optimal performance of the solar collector is highly important. The fundamental operational problem with solar collectors is the collection and delivery of solar energy to users with minimum losses. The optimum operating conditions for solar collectors can be investigated using different modes of performance. The common aim is to optimize the thermal efficiency of any collector. During the past decade, many researchers have conducted on the importance of increasing energy efficiency (Sӧzen et al., 2008; Rodriguez-Hidalgo et al., 2011; Chen et al., 2012; Fudholi et al., 2013). The conventional energy analysis, based on the first law of thermodynamic, does not give the qualitative assessment of the various losses occurring in the components. For that, the exergy analysis, based on the second law of thermodynamic, is used to get a clear picture of the various losses quantitatively as well as qualitatively (Saidur et al., 2010). The exergy analysis is a useful method to complement, not to replace, the energy analysis. However, exergy data are more practical and realistic in comparison to the respective energy values. Thus, the exergy analysis provides a more realistic view of process, sometimes dramatically different in comparison to standard energy analyses (Petela, 2008). In the literature, several researches have undertaken many studies under steady state condition covering the exergy analysis of FPSC. An experimental analysis was developed by Gunerhan and Hepbasli (2007). They showed that the exergy efficiency of the solar water heater was around 4%. Farahat et al. (2009) presented a numerical study based on the exergy analysis in order to determine the optimal efficiency and design parameters of FPSC. The performance of the collector is evaluated considering the inlet temperature equal to ambient temperature. They concluded that the results found are in good agreement with the experimental measurements noted in the previous literature. Hazami et al. (2010) experimentally investigated the thermal performance of a collector with a thin layer concrete as absorption plate. Jafarkazemi and Ahmadifar (2013) performed a numerical study of a FPSC in order to determine the effect of design parameters on the energy and exergy efficiencies. 286

301 In all the aforementioned researches, the performance of the collector evaluated either assuming a steady state 287ehaviour.The difficulties in achieving steady state test conditions have led to the development of transient test procedures for rating the thermal performance of collectors. In this study, we improve the model by recognizing that collectors must operate under time varying conditions promoted by the daily insolation cycle. This dynamic model based on energy and exergy analyses is aimed to identify the principal sources of irreversibility and to minimize the entropy generation in the solar system. We study in detail the effect of time history on the different exergy compounds. Furthermore, the effects of various parameters on energy and exergy efficiencies and on exergy destroyed and losses are examined in order to maximize the performance of the solar heater. II. Thermodynamic Analysis II.1. Energy Analysis The collector under consederation is a flat plate solar water collector consisting of a: transparent cover, air gap, absorber plate, tubes, insulations and container. The selected specification and design conditions of this solar heater are shown in Tab.1. The air draped between the transparent cover and the absorber is assumed to be stagnant and transparent; The sky is assimilated as a blackbody. On the basis of the cited assumptions, the energy balance equations for the various elements of the heater may be written as follows: (i) Heat balance for the transparent cover dtc MCC C =h r,c-sky SC Tsky-T C +hcv,c-ambsc Tamb -TC dt +h r,c-abs SC Tabs -T C +h cv,c-a SC Ta -TC αcτc 1-αabs +Uloss,eSC,e Tamb -T C +αcscg+scg 1-1-α C abs (1) hcv,c-amb and hcv,c-a are the heat transfer coefficients by convection, calculated as follows (Duffie and Beckman, 1974): h =3.9 v (2) h cv,c-amb Nu wind a cv,c-a (3) ea The Nusselt number (Nu) is given by (Aranovitch, 1981): Tab. 1. Specification and design of the solar collector. Dimensions Transparent cover Absorber Tubes Insulation 1941 x 1027 x 88 (mm) Glass without iron Thickness 3.2 mm Transmissivity (τ) = 0.9 Emissivity=0.89 Copper Surface 1.87 m 2 _Black selective Absorptivity (α) = 0.95 Emissivity = 0.05 Copper Number of tubes13_diameter = 8.81mm Bottom Edge Glass wool Thickness = 40mm Polyester wool Thickness = 28mm By writing energy conservation balance for all the components of the solar heater such as: transparent cover, air gap, absorber plate and working fluid inside the duct a mathematical model, comprises a set of temperature dependent equations, is obtained. To simplify the analysis, the following assumptions are applied: The potential and the kinetic energy effects are negligible; The physical properties of the solids are assumed to be constant; The surfaces of heat exchange by radiation are assumed gray and flat; The transparent cover is opaque to infrared radiation; The fluid velocity is uniform; The losses by radiation to the bottom and sides surfaces of the insulation are assumed negligible; 287 β Nu = Gr Gr β Nu = Gr 10 Gr Gr is the Grashoff number, expressed as fellows: Gr = 3 g T - T e abs C a 2 Ta (4) (5) hr,c-sky and hr,c-abs are the heat transfer coefficients by radiation, calculated respectively, as follows: 2 2 h = ε σ T + T T + T (6) h r,c-sky C C sky C sky r,c-abs = σ T 2 2 C + Tabs T C + Tabs ε ε C abs (7) where Tsky is the sky temperature, given by (Shiv et al., 2010): T = T (8) sky 1.5 amb (ii) Heat balance for the air gap dt ρ V C = h S T - T +h S T - T dt + U S T - T a a a a cv,c-a C C a cv,abs-a abs abs a loss,e a,e amb a (9)

302 hcv,abs-a is the heat transfer coefficient by convection between the absorber plate and the air gap, calculated using Eq.(3). (iii) Heat balance for the absorber dtabs MabsC abs = h r,c-abs Sabs TC -T abs + hcv,abs-asabs Ta - T abs dt + h S T -T + U S T - T cv,abs-f exch,f f abs loss,b abs amb abs C abs + Uloss,eSabs,e Tamb - T abs SabsG 1- C 1-α abs τα (10) hcv,abs-f is the convection heat exchange coefficient estimated using the correlation proposed by Incorpora et al. (2006): Nu = Re Pr (11) Uloss,b and Uloss,e are the bottom and the edge losses coefficients, calculated by: τα (16) C abs η= 0 1- C 1-α abs The exergy destructed during the heat transfer process from absorber plate to working fluid is given by:.. Edes,abs-f n T Sgen (17) tube amb where Ṡgen is the rate of entropy generation over the length of duct, calculed as:.. L δq 1 1 abs-f S gen = dx 0 dx Tf Tabs (18) Exergy losses are the exergy leakages rates out to the surroundings due to optical error and heat transfer to ambient, are expressed respectively as: 1 1 e 1 e 1 ins,b ins,e U loss,b + and U loss,e + λins,b hcv,wind λins,e hcv,wind (iv) Heat balance for the transfer fluid (12). T amb Eloss,opt SG 1 1 η0 Tsun. Se T amb Eloss,amb U loss,t+u loss,b+uloss,e STabs -Tamb 1- S Tabs (19) (20) D T T ρ C + v = h T - T 4 dt dx int-tube f f f f f cv,abs-f abs f (13) The efficiency of a collector is defined as the ratio of energy achieved to incident solar energy for the same period of time, is calculated as fellows: η. m C T -T f f f,out f,in en (14) SG II.2. Exergy Analysis Exergy analysis is a technique that uses the mass conservation and energy conservation together with the second law of thermodynamics. One of the main objectives of this analysis is to locate and characterize the causes of exergy destruction and exergy losses due to irreversibilities in any real process. In a solar collector, the heat transfer process from the sun to the collector s working fluid consists of two main parts, absorbing the solar radiation by absorber plate and heat transfer from absorber plate to working fluid. Exergy destructed during absorbing process is due to temperature difference between the absorber plate and the sun. It is calculated as follows (Altfeld et al., 1988; Akpinar et al., 2010; Ahmad et al., 2013): The top loss coefficient is expressed as: 1 U loss,t = h 1 r,c-sky +hcv,wind h h cv,c-a cv,abs-a +h r,c-abs (21) The exergy efficiency of the solar collector is calculated by dividing the useful exergy gain to the exergy of solar radiation, so (Jafarkazemi and Ahmadifar, 2013):. T mf Cf Tf,out -Tf,in -Tamb Ln T η ex = T amb SG 1- Tsun f,out f,in III. Modelling of Solar Radiation (22) The prediction of collector performance requires information on the solar energy radiation. The global solar radiation absorbed by different components of the collector is calculated by Perrin Brinchambaut model (2008). It proposes the following correlations to estimate global solar radiation on inclined surface.the direct irradiance is calculated as:. T T amb amb Edes,sun-abs η0sg - Tabs Tsun (15) 1 I = A cos ( i) exp- B sin (h + 2) (23) where η0 is given by: The diffuse irradiance is given by: 288

303 1+ cos (β) 1- cos (β) D = D h + G 2 2 Dh and Gh are expressed as: h 0.4 B'' h h (24) D = A' sin(h) and G = A'' sin(h) (25) B, A', A'' and B'' are constants which depend to the atmospheric state. IV. Results and Discussions The mathematical model obtained in the previous sections has transposed into a MATLAB computational program. The results are presented only for the summer solstice (21 June). The parameters that are maintained initially fixed are: the ambient temperature Tamb=31 C, the inlet water temperature Tf,in = 31 C, the wind speed vwind = 1.5 m/s and the mass flow rate ṁf = kg/s. Fig.1 shows the hourly variations of global solar radiation predicted and measured on a horizontal surface. Climatic data using in this current investigation measured in Gabes (33 51N, 10 03E), Tunisia. It is seen a good agreement between experimental and theoretical results. The maximum global solar radiation intensity attained is approximately 963 W/m 2. It is observed that the exergy efficiency curves tend to follow the solar intensity, it increases in the sunrise until it reaches a maximum around 4.24% at solar noon after it decreases, whereas the energy efficiency curves are independent of the insolation pattern. The energy efficiency increases gradually during the first hours of day to reach a value of 78.44% at afternoon this is may be attribueted to the collector thermal inertia. In fact, the absorber plate and tubes accumulate heat during hight solar intensity and realese this heat when the solar flow decreases. Results indicate also that the exergy efficiency values are very low compared to those energetic, this is due to the important amount of exergy destroyed and losses in the solar collector, as shown in Figs.3 and 4. As depicted by Fig.3, the exergy components Ėdes,sun-abs and Ėdes,abs-f depend strongly on the solar intensity. The highest values obtained for exergy destruction during absorption and heat transfer processes are up to 1201,3 W and 86,83 W, respectively. It is seen that the main cause of exergy destruction in the solar collector results from absorption of radiation by the absorber plate Ėdes,sun-abs. This is explained by the hight difference of temperature between the sun (4500 C) and the the absorber plate (82.37 C) E des,sun-abs E des,abs-f Brinchambaut model Experimental results E des,sun-abs (w) E des,abs-f (w) Solar intensity (W/m 2 ) Time (h) Fig.3: Variation of exergy destruction vs solar time Time (h) Fig.1: Hourly variations of measured and predicted global solar radiation in 21 st June. Fig.2 presents the variation of energy and exergy efficiencies versus solar time. Exergie (W) E loss-opt E loss-t E loss-b E loss-e Energy efficiency Exergy efficiency Energy efficiency Time (h) Fig.2: Variation of energy exergy efficiencies vs solar time Exergy efficiency Time(h) Fig.4: Variation of exergy losses vs solar time. From Fig.4, it is observed that the exergy losses due to optical error Ėloss,opt is the highest. Then in a decreasing order, we have the top and the bottom exergy losses Ėloss,abs-amb,t and Ėloss,abs-amb,b. This is explained by the fact that the bottom heat loss coefficient is smaller than the top heat loss coefficient, since the solar collector is very isolated in the back. Given the low value of the edge heat loss coefficient because the sides surfaces of the collector are very 289

304 small and isolated the variation of the edge exergy losses versus solar time is the lowest. The maximum values reached at midday are approximately 234,46 W, 66 W, 15,05 W et 5,14 W for optical, top, bottom and edge exergy losses, respectively. This section presents a parametric study for which the objective is to estimate and compare the direct effects of the investigated parameters. There are many plans that can be used to achieve this objective. Here, the full factorial design is employed: it is the simplest and most widely used. Five parameters have been identified as most relevant. These parameters are: mass flow rates, inlet water temperature, absorber emissivity, tubes number, and pipes diameter. The preliminary tests have also allowed to identify, for each parameter studied, the range of values (high and low), used. Tab.2 summarizes the selected parameters and their levels used. The energy and the exergy efficiencies of the collector are selected as the responses parameters. The number of tests of the full factorial plan 2 5 is 32. To determine whether or not the studied parameters have a significant effect on the efficiencies an analysis of variance is performed using the software Stat Ease Expert Design. Symbol Tab.2: Parameters and their levels. Factors Level Low (-1) High (+1) A Inlet temperature ( C) B Mass flow rates (kg/s) C Absorber emissivity D Tubes number 2 20 E Tubes diameter (m) The Pareto chart (Fig.5) provides the influential factors in decreasing order of contribution. The horizontal line in this diagram indicates the statistical significance at 95% of confidence level and separates factors that are significant to those that are not. Any effect that extends beyond this line is potentially important. The results show the predominance of mass flow rates (B) for the tow reponses. Next, in order of importance is the absorber emissivity (C) and the inlet temperature (A) for the energy efficiency and conversely for the exergy effiency. Wheras, the tubes number (E) and the pipes diameter (E) have a negligeable effects for both reponses. In the next section, we will study only the effect of the significant parameters. Moreover, the optical exergy losses are not considered in the parametric study since they are independent to these parameters. Fig.5: Pareto chart of: (a) energy efficiency and (b) exergy efficiency. Fig.6 show the effect of the mass flow rates on the energy and exergy efficiencies, exergy destruction and exergy losses. The results are presented for mass flow rates from kg/s to 0.1kg/s. The energy efficieny increases asymptotically with the mass flow rates while the exergy efficiency decreases. This trend is not surprising since high mass flow rates will induce higher velocities and Reynolds number and which in turn enhance the convective heat transfer from the absorber plate and the transfer fluid and consequently reduce the heat losses. It is seen from Fig.6.b that the curves of exergy destruction by absorption and heat transfer are similar to the profiles of energy and exergy efficiencies, respectively. In fact, an increase in the mass flow causes a decrease in absorber temperture. From Eq.(15), it is obvious that the absorber plate temperature is the most effective parameter in exergy destroyed by absorption. Therefore, a decrease of this temperature leads to an increase in this exergy. It observed from Fig.6.c that the top, the bottom and the edge exergy losses decrease exponentially with the mass flow rates until to cancel since a rise in mass flow rate causes a decrease in absorber temperature which in turn results a decrease in overall heat loss coefficients and therefore a decrease in exergy losses to the environment. 290 For a mass flow rate of 0.01 kg/s, the outlet water temperature, the energy and exergy efficiencies are approximately C, 78% and 4.24 %, respectively. So, we consider this flow rate value as the most appropriate.

305 (a) Energy efficiency Exergy efficiency (b) E des,sun-abs E des,abs-f Energy efficiency Exergy efficiency E des,sun-abs (w) E des,abs-f (w) (b) Mass flow rates (kg/s) E des,sun-abs E des,abs-f (c) Inlet temperature ( C) E loss-t E loss-b 12 E loss-e E des,sun-abs (w) E des,abs-f (w) Exergy (W) (c) Exergy (W) Mass flow rates (kg/s) E loss-t E loss-b E loss-e Inlet temperature ( C) Fig.7: Effect of inlet temperature on: (a) energy and exergy efficiencies, (b) exergy destruction and (c) exergy losses. As the inlet temperature rises from 31 C to 45 C the energy efficiency decreases from 73.04% to 64% wheras the exergy efficiency increases from 4.11% to 6.88% (Fig.7.a). So, a lower value of inlet temperature is desirable to improve the solar thermal efficiency Mass flow rates(kg/s) Fig.6: Effect of mass flow rates on: (a) energy and exergy efficiencies, (b) exergy destruction and (c) exergy losses. Fig.7 illustrates the effect of varying inlet water temperature on the energy and exergy efficiencies and the exergy destruction and losses. (a) Energy efficiency Energy efficiency Exergy efficiency Inlet temperature ( C) Exergy efficiency The exergy destruction Ėdes,sun-abs caused by the temperature difference between the absorber and the sun decreases with increasing inlet water temperature because the rise in this temperature increases the average absorber temperature and consequently reduces the exergy destruction by absorption of heat. Also, the exergy destructed during the heat transfer process Ėdes,abs-f from absorber plate to working fluid decreases with the rise in inlet fluid temperature. Whereas, the exergy losses caused by heat leakage from the absorber plate to the environment increase with the inlet temperature since a rise in this temperature increases the average absorber temperature witch in turn leades a rise in heat losses from the collector to the surrounding. Effect of absorber emissivity on the performance of solar collector is predicted in Fig. 8. It can be seen that for lower values of emissivity the energy and exergy efficiencies are higher. The energy efficiency varies from 77.92% to 67.77% and exergy efficiency varies from 4.24% to 3.24%, for absorber emissivity ranging from to Numerical results show that the exergy destruction by absoption and the top exergy losses rise with the absorber emissivity. This may be due that an 291

306 increase in emissivity causes a rise in radiative thermal losses between absorber plate and taransparent cover and therefore a decrease in absorber temperature and a rise in top thermal losses. Furthermore, the exergy destruction during the heat transfer process Ėdes,abs-f and the bottom and the edge exergy losses decrease with the absorber emissivity. In fact, a rise in emissivity from to 0.95 results a drop of 15.08W, 1.56W and 0.53W for the Ėdes,abs-f, Ėloss,abs-amb,b and Ėloss,abs-amb,e, respectively. (a) (b) Energy efficiency E des,sun-abs (w) (c) Exergy (W) Fig.8: Effect of emissivity on: (a) energy and exergy efficiencies, (b) exergy destruction and (c) exergy losses. V. Conclusions Absorber emissivity Energy efficiency Exergy efficiency Absorber Emissivity E des,sun-abs E des,abs-f Absorber emissivity E loss-t E loss-b E loss-e In this study, a theoretical model based in the first and the second laws of thermodynamics is presented to prdict the performance of flat palte solar collector in transient conditions. Results show that the mass flow rates, the inlet water temperature and the absorber emissivity have a grat influence on energy and exergy efficiencies, while the tubes number and the pipes diameter have a a negligeable effects. Analysis indicates that the main cause of exergy destruction in E des,abs-f (w) Exergy efficiency 292 the collector results from absorption of solar radiation by absorber plate. A decrease in mass flow rates and absorber emissivity and a rise in inlet water temperature can be effective to decrease this exergy destroyed. Moreover, in order to enhance the thermal efficiency of the solar collector system, inlet water temperature and absorber emissivity should be as low as possible. Nomenclature C : heat capacity (J/kg.K) D : tube diameter (m) Ė : exergy (J/s) e : thickness (m) g : acceleration gravity (m/s 2 ) G : solar global radiation (W/m 2 ) Gr : Grashoff number h : exchange coefficient (W/m 2.K) L : length of tube (m) M : mass (kg) ṁ : mass flow rate (kg/s) Nu : Nusselt number Pr : Prandlt number Re : Reynold number S : area (m 2 ) Ṡ : entropy generation rate (W/K) T : temperature (K) t : time (h) U : heat loss coefficient (W/m 2.K) v : velocity (m/s) Greek letters µ : dynamic viscosity (Pa.s) α : absorption coefficient β : tilt angle ( ) ε : emissivity η : efficiency θ : expansion coefficient (K -1 ) λ : thermal conductivity (W/ m.k) ρ : albedo ρ : density (kg/m 3 ) σ : Stefan-Boltzman constant (W/m 2 K 4 ) τ : transmission coefficient ϕ : reflection coefficient Subscripts a : trapped air abs : absorber amb : ambient b : bottom C : glass cover cv : convection e : edge en : energy ex : exergy exch : exchange f : transfer fluid in : inlet int : internel loss : loss opt : optical out : outlet r : radiation sky : sky

307 sun Tube : sun : tube Petela R., An approach to the exergy analysis of photosynthesis, Solar Energy, 82, (2008). References Akpinar E.K, Kocyigit F., Energy and exergy analysis of new flat plate solar collector having different obstacles on absorber plates, Appllied Energy, 87, (2010). Altfeld K., Leiner W., Fiebig M., Second law Optimization of flat-plate solar air heaters: Part l: The concept of net exergy flow and the modeling of solar air heaters, Solar Energy, 41, (1988). Aranovitch E., Heat transfer processes in solar collectors, Energy Buildings, 3, (1981). Bekkouche S.M.A., Modélisation du comportement thermique de quelques dispositifs solaires, PhD. Thesis, University ABOU-BAKAR BELKAID, Tlemcen (2008). Chen Z., Furbo S., Perers B., Fan J., Anderse E., Efficiencies of flat plate solar collectors at different flow rate, Energy Procedia, 30, (2012). Duffie J.A., Beckman W.A., Solar energy thermal process, Wiley Interscience, New York, (1974). Farahat S., Sarhaddi F., Ajam H., Exergetic optimization of flat plate solar collectors, Renewable Energy, 34, (2009). Rodriguez-Hidalgo M.C, Rodriguer-Aumente P.A., Lecuona A., Gutierrez-Urueta G.L., Ventas R., Flat plate thermal solar collector efficiency: transient behavior under working conditions, Part I: Model description and experimental validation, Applied Thermal Engineering, 31, (2011). Saidur R., Ahamed J.U., Masjuk H.H., Energy, exergy and economic analysis of industrial boilers, Energy Policy 38, (2010). Shiv K., Tiwari G.N., Gaur M.K., Development of empirical relation to evaluate the heat transfer coefficients and fractional energy in basin type hybrid (PV/T) active solar still, Desalination, 250, (2010). Subiantoro A., Tiw O.K., Analytical models for the computation and optimization of single and double glazing flat plate solar collectors with normal and small air gap spacing, Applied Energy, 104, (2013). Sӧzen A., Menlik T., Ünvar S., Determination of efficiency of flat-plate solar collectors using neural network approach, Expert Systems with Applications, 35, (2008). Fudholi A., Sopian K., Othman M.Y., Ruslan M.H., Bakhtyar B., Energy analysis and improvement potential of finned double-pass solar collector, Energy Conversion and Management 75, (2013). Fudholi F., Sopian K., Othman M.Y., Ruslan M.H., Bakhtyar B., Energy analysis and improvement potential of finned double-pass solar collector, Energy Conversion and Management, 75, (2013). Gunerhan H., Hepbasli A., Exergetic modeling and performance evaluation of solar water heating systems for building applications, Energy Buildings, 39, (2007). Hazami M., Kooli S., Lazâar M., Farhat A., Belghith A., Energetic and exergetic performances of an economical and available integrated solar storage collector based on a concrete matrix, Energy Conversion and Management; 51, (2010). Incropera F.P., Dewitt D.P., Bergman T.L., Lavine V.S., Fundamentals of heat and mass transfer, 6th edition (2006). Jafarkazemi F., Ahmadifar E., Energetic and exergetic evaluation of flat plate solar collectors, Renewable Energy, 56, (2013). 293

308 Optimization of Tilt Angles of PV Arrays for Different Seasons Ahmet Senpinar* College of Technical Sciences, Department of Electronics Technology, Firat University, 23100, Elazig, Turkey * Abstract The earth s position relative to the sun changes due to its axis and movement around the solar system. This state varies the length of days in a year and causes the different seasons. Turkey has significant solar energy potential; however optimal exploitation must account for different insolation levels across the country. As the incidence angle of sunlight varies in time, an important parameter affecting the efficiency of a PV array is its tilt angle to the horizontal. This paper presents a mathematical model for determining the optimum tilt angle and orientation for PV arrays in some cities across Turkey. The model provides a means for calculating a monthly optimum tilt angle that accounts for seasonal and geographic variations. Keywords: Optimum tilt angle, PV system efficiency, solar angles, solar radiation. I. Introduction The energy obtained from the sun on earth per unit of time is called the solar constant and is represented with GSC. The value of the solar constant as adopted by the World Radiation Center (WRC) is 1367 W/m 2 (1.96 cal/cm 2 min) (Beckman and Duffie, 1991). The sun is a gaseous body, with a mass of about kg, and a diameter of 1.39*10 9 m. The distance from the sun to earth is about 1.49*10 11 m (Cheremisinoff and Dickinson, 1980). Solar radiation has many advantages as a renewable energy source, particularly in terms of its abundance and being pollution free. It is utilized for both generating thermal energy and electricity directly using photovoltaic cells. Thus solar radiation data is an important parameter for the design and calibration of solar energy applications (Wu et al. 2007, Almorox and Hontoria, 2003, Daut et al. 2011). For example, calculations of beam and diffuse solar radiation on horizontal surfaces are widely used for simulations, modeling and sizing of solar processes. Data on solar radiation, which is measured hourly and daily, has significant importance for PV system designs, meteorology, solar maps, and engineering applications. Some researchers also use sky clearness to measure surface global solar irradiance (Gueymard, 1993, Alam et al. 2009). PV systems are employed in a variety of ways and the application of this technology is expanding throughout the world (Chambouleyron, 1996, Senpinar, 2005). Current applications range from simple domestic PV energy systems to ones for street illumination, cooling, pumping water, traffic signaling, plants for PV electricity generation, hybrid (sun+wind) systems, space systems, and telecommunication systems, etc (Green, et al. 2001, Kuwano, 1998). Clearly the use of solar energy has 294 an important place among energy sources in meeting future energy needs. During the last few years, photovoltaic solar systems have become one of the most popular renewable energy sources in Europe (Peragón et al. 2011, Montoya et al. 2014). A key aspect of the efficiency of a solar array is its tilt angle to the horizontal. Tilt angle varies according to season and location. Generally, PV systems in the northern hemisphere are mounted facing due south with a certain angle. A variety of different values for the tilt angle have been suggested. Some studies have used angles calculated as, Ø + 20 ( Hottel, 1954),Ø + (1030) (Lof, 1973), Ø + 10 ( Kern and Harris, 1975), Ø-10 (Hyewood, 1971), whereas other researchers suggest two values for the tilt angle, such as Ø ± 20 (Yellott, 1973), Ø ± 8 (Lewis, 1987), Ø ±5 (Garg and Gupta, 1978), where Ø is the latitude angle of the region, + for winter, and - for summer. For optimal performance on any given day, a fixed array should be mounted on the ground to have a horizontal angle of (Ø-δ) (Messenger and Ventre, 2000). Here, δ is the declination angle known as the angle between the direction of the sun and equator plane. This study recognizes the importance of seasonally adjusted tilt angles to the optimal application of PV arrays in different locations across Turkey. The tilt angles of some cities are presented and calculations made to determine the optimum tilt angle and orientation for PV arrays. Graphical results associated with each city are presented using Matlab software. II. Geographical location and insolation level Turkey is located geographically in the northern hemisphere between the (N) latitudes and (E) longitudes (TSMS, 2015). There is a 19 longitude-difference between locations at the eastern-most and western-most ends of the country.

309 So, the sun rises and sets earlier in a location in the east due to that difference but rises and sets later in a location in the west. While Turkey has good insolation potential, levels vary between different locations. The annual average total insolation duration in Turkey has been measured as 2640 hours-year (7,2 hours/day), and the total average annual solar radiation as 1311 kwh/m²-year (3,6 kwh/m²-day, Table1) (General Directorate of Electrical Power Resources Survey and Development Administration, 2015). Tab.1: Monthly average solar potential of Turkey Monthly Total Solar Sunshine Months Energy Duration (Kcal/cm 2 - (kwh/m 2 - (hours / month) month) month) January 4,45 51,75 103,0 February 5,44 63,27 115,0 March 8,31 96,65 165,0 April 10,51 122,23 197,0 May 13,23 153,86 273,0 June 14,51 168,75 325,0 July 15,08 175,38 365,0 August 13,62 158,40 343,0 September 10,60 123,28 280,0 October 7,73 89,90 214,0 November 5,23 60,82 157,0 December 4,03 46,87 103,0 TOTAL 112, (per year) AVERAGE 308,0 cal/cm 2 - day 3,6 kwh/m 2 - day 7,2 hours/day England, and is called the prime meridian. Points east of the prime meridian have negative longitudes, and points west of it have positive ones. The angle Ø on the earth s surface measured north or south of the equator to a point is its latitude. Latitude values increase toward the poles, with the North Pole being 90, and the South Pole -90. As latitude increases, the curvature of the earth has the effect of tilting the observer away from the sun. An array is tilted toward the equator to compensate for this effect. The earth revolves around the sun once a year in an elliptical orbit that is almost circular, the earth-sun distance varying by about 3 percent from a mean distance of 150 million km. The earth is closest to the sun in the summer and farthest away in the winter because the rotational axis of the earth is inclined at 23,44 to the axis of the orbital plane. Thus, in the winter, the earth is tilted with the northern hemisphere away from the sun, and in summer, the northern hemisphere is tilted toward the sun. This phenomenon is referred to as the tilt or declination angle δ of the axis relative to the sunearth line. Some of angles are indicated in Figure 1. As these angles constantly change with the seasons relative to the position of a location, it is necessary to calculate the optimum tilt angle 12 times in a year to provide monthly optimum tilt angles that can also be used to calculate seasonal angles if necessary. Some of the angles to consider when calculating an optimum are as follows (Beckman and Duffie, 1991, Senpinar, 2005); III. Mathematical model III.1.Calculation of solar angles PV arrays can be mounted to track the sun but fixed systems need to be maintained at a certain angle to the horizontal to make full use of the available sunlight at the location. If this tilt angle is determined well, the amount of insolation and the energy that is generated increases. To maximize energy amount, solar panels, such as photovoltaic modules, are usually oriented towards the equator with an optimal tilt-angle from the horizon which depends on climatic conditions and site latitude (Tang and Wu, 2004, Gopinathan et al., 2007, Gunerhan and Hepbasli, 2007, Benghanem, 2011). Tilt angle and location are also important considerations as energy demand dictates the design and operation of a stand-alone PV cell system, and the number of modules and batteries to be used. Thus the performance of a system is subject to load, isolation level, and module characteristics. Tilt angle can be determined using the meridians of longitude and latitude for any location. The 0 meridian longitude passes through the former site of the Royal Astronomical Observatory in Greenwich, Fig. 1: Some angles for a tilted surface. III.1.1. Latitude angle (Ø) Latitude is the angular location north or south of the equator and it changes north positive, -90º Ø 90º. III.1.2. Declination angle (δ) The declination angle is the angular position of the sun at solar noon with respect to the plane of the equator, north positive; -23,45º δ 23,45º (Fig.2.). 295

310 The declination δ can be found from the equation of Cooper; δ = 23,45.sin( ( 360.(284 n)) )º (1) 365 where, n represent the day of the year (n=1, for 1 January) ( Beckman and Duffie, 1991). Table2 represent zenith angle, slope, solar azimuth angle, surface azimuth angle, and solar altitude angle for a tilted surface. Declination angle (degree) Days of Year (20 March March 2016) Fig. 2: Changes in the declination angle as annual Tab. 2: Recommended average days for Months and values of n by Months Months n for ith Day of Month Date n, Day of year δ, Declination January i ,03 February 31+i ,58 March 59+i ,80 April 90+i ,53 May 120+i ,78 June 151+i ,01 July 181+i ,10 August 212+i ,36 September 243+i ,21 October 273+i ,10 November 304+i ,91 December 334+i ,12 III.1.3. Zenith angle (θz) The zenith angle (θz) is the angle between the vertical and the line to the sun and is calculated as follows (Beckman and Duffie, 1991); cosθz = cosδ.cosø.cosω+ sinδ.sinø (2) where, ω is the solar hour angle and it is determined by (Senpinar) formula as follows (Senpinar, 2005); ω=((hour.60 + minute)-720)/4 º (3) III.1.4. Solar altitude angle (αs) The solar altitude angle is the angle between the horizontal and the line to the sun and it is complementary to the zenith angle and calculated as follows; αs+ θz= 90º (4) III.1.5. Incidence angle (θ) The incidence angle (θ) is the angle between the beam radiation on a surface and the normal angle to that surface and is calculated as follows (Beckman and Duffie, 1991); cosθ = cosθz.cosβ + sinθz.sinβ.cos(γs-γ) (5) where, γ represent the surface azimuth angle. III.1.6. Solar azimuth angle (γs) The solar azimuth angle (γs) is the angular displacement from south of the projection of beam radiation on the horizontal plane. Displacements east of south are negative and west of south are positive, the solar azimuth angle changes in the range of - 180º to 180º. For north or south latitudes between 23,45º and 66,45º, γs will be between 90º and -90º. To calculate γs, we need know the sun s position (Beckman and Duffie, 1991). A general formula for γs, from Braun and Mitchell (1983), is conveniently written in terms of γs, a pseudo surface azimuth angle in the first or fourth quadrant; γs = a1a2γs a3((1- a1a2)/2) (6) where; sin γs = ((sinω.cosδ)/sinθz) (7) where, a1, a2 and a3 are constants related with the sunrise and sunset. Solar altitude angle (degree) Zenith angle (degree) morning (05:00) Solar azimuth angle (degree) evening (19:30) morning (05:00) Solar azimuth angle (degree) evening (19:30) Fig. 3: Solar azimuth angle, zenith angle, and change in solar altitude angle γ represent suface azimuth angle, the deviation of the projection on a horizontal plane of the normal to the 296

311 surface from the local meridian, -180º γ 180º. Figure 3 shows changing of the zenith angle and solar altitude angle according to solar azimuth angle. III.1.7. Tilt angle (β) The tilt angle (β) is the angle between the plane of the surface in question and the horizontal and the value of it changes 0º β 180º (Beckman and Duffie, 1991). tanβ=tanθz cos γs (8) IV. Results and discussion IV.1.Applications of optimum tilt angles for different seasons Solar radiation data are used in several forms and for a variety of purposes. The most detailed information available is beam and diffuse solar radiation on a horizontal surface, by hours, which is useful in simulations of solar process. Daily data are often available and hourly radiation can be estimated from daily data. Data for monthly total solar radiation on a horizontal surface is used in some process design methods. Some solar collectors track the sun by moving in prescribed ways to minimize the angle of incidence of beam radiation on their surfaces and thus maximize the incident beam radiation. Installation and operation of these collectors requires relevant data on the angles of incidence and the surface azimuth angles. Tracking PV systems are classified by their motions. Rotation can be about a single axis (horizontal east-west, horizontal north-south, or parallel to the earth s axis) or it can be about two axes. For a plane rotating about a horizontal eastwest axis with a single daily adjustment so that the beam radiation is normal to the surface at noon each day (Beckman and Duffie, 1991); solar radiation available at that location (Tariku and Tassew, 2015). The tilt of an array can also be seasonally adjusted. In summer or winter, the tilt angle of an array is different from the other seasons; however the tilt of an array at any time of year can be set as an optimum value for the rest of the season. For the best average tilt for summer, winter and optimum annual performance an array should be mounted at (Ø-15), (Ø+15) respectively and with a tilt angle of (0,9.Ø) (Messenger and Ventre, 2000). Daily and monthly average tilt angles can be calculated using the Matlab software program.for optimal seasonal performance; one simply chooses the average value of tilt angle for the season. Using monthly average tilt angles, the annual average optimum tilt angle can be determined. Periodic adjustment of the tilt angle can be economically advantageous at higher latitudes and help to maximize the performance and generating efficiency of fixed systems. The amount of isolation received in different locations across Turkey varies according to geographical position and local climatic conditions. Thus the researcher calculated an optimum tilt angle for 13 cities across Turkey using data on insolation levels and meteorological records from The meteorological data for the 13 cities is shown in Tables 3 and 4 along with the average monthly and seasonal optimum tilt angles. First, monthly average values for each city were calculated. Then, the annual average tilt angle value (0,9.Ø) was calculated using the monthly average values (Messenger and Ventre, 2000). Figure 4 (a-d) presents the average values of optimum tilt angles for some cities in Turkey. Figure 5 presents seasonal average values. cosθ= sin 2 δ+cos 2.cosω (9) the tilt of this surface can be calculated as follows; β = Ø-δ (10) For a plane rotated about a horizontal east-west axis with continuous adjustment to minimize the angle of incidence (Beckman and Duffie, 1991); cosθ = (1-cos 2 δ.sin 2 ω)½ (11) the tilt of this surface is; tan β = tanθz cos γs (12) where, γs is the solar azimuth angle. If the tilt of an array is set to θz angle to the horizontal, the radiation on the array is normal to the surface at noon. Thus, an array would be exposed to the maximum level of 297 Tilt Angles (degree) Tilt Angles (degree) Months (for Adana) Months (for istanbul) (a) Graphics of Adana and Istanbul

312 Spring Tilt Angles (degree) Months (for Gaziantep) Autumn Winter Summer Tilt Angles (degree) Months (for Igdir) Ankara Elazığ İstanbul Malatya Hatay(Dortyol) Igdır Canakkale Adana Samsun Sivas Kayseri Gaziantep Mugla (b) Graphics of Gaziantep and Igdir Fig. 5: Seasonal average optimum tilt angles for 13 cities in Turkey. V. Conclusion Tilt Angles (degree) Tilt Angles (degree) Tilt Angles (degree) Tilt Angles (degree) Months (for Samsun) Months (for Hatay Dortyol) (c) Graphics of Samsun and Hatay Months (for Ankara) Months (for Mugla) (d) Graphics of Ankara and Mugla Conventional, fossil fuel energy sources have negative effects on the environment caused by emissions of pollutants and greenhouse gases. Nuclear energy generation has also produced serious environmental challenges that mitigate against its use in many locations. A growing number of studies reflect the considerable interest being shown in renewable energy sources, also known as the future s energy sources. These include solar energy systems, the efficiency of which is subject to many factors including accurate prediction of the optimum tilt angle of an array at different times of the year. The prediction of solar radiation is quite important for many solar applications and is affected by geographic location and climatic conditions. Table 2 and 3 show how optimum tilt angles vary according to geographical locations across Turkey. Monthly, seasonal and annual values are shown. As winter and summer values vary considerably in Turkey, the efficiency of PV arrays can be maximized if they are mounted according to these seasonal variations. It is thus economically beneficial to make monthly adjustment to the tilt angle of a PV system in latitudes such as those found in Turkey, and calculated according to seasonal values. Fig. 4: Average optimum tilt angles for some cities in Turkey (a-d). 298

313 Tab.3: Seasonal and annual average values of optimum tilt angles for 13 different cities in Turkey Tab. 4: Monthly average values of optimum tilt angles for 13 different cities in Turkey 299

314 Acknowledgement Author received his Ph.D. in Solar Energy Systems in 2005 from Firat University, Elazig, Turkey. Currently, he is assistant professor at College of Technical Sciences, Department of Electronics Technology, Firat University and is also chairman of Department of Electronics and Automation. His research interests are in the areas of solar energy, solar angles, tracking systems and PV systems. References Alam, S., Kaushik, S.C., Garg, S.N. Assessment of Diffuse Solar Energy Under General Sky Condition Using Artificial Neural Network. Appl Energy 86, , (2009). Almorox, J., Hontoria, C.Global solar radiation estimation using sunshine duration in Spain.Energ Convers Manage 45, , (2003). Beckman, William A. and Duffie, John A, Solar Engineering of Thermal Processes, Second Edition, Canada, John Wiley and Sons Inc., (1991). Benghanem M. Optimization of tilt angle for solar panel: case study for Madinah, Saudi Arabia. Appl Energy 88, , (2011). Braun JE, Mitchell JC. Sol Energy 31, 439 [Solar Geometry for Fixed and Tracking Surfaces], (1983). Chambouleyron, I. Photovoltaics in the developing World. Energy Vol.21,No.5, , (1996). Cheremisinoff, Paul N. and Dickinson, William C., Solar Energy Technology Hanbook, New York, Marcel Dekker, Inc., (1980). Daut I., Irwanto M., Irwan Y.M., Gomesh N., Ahmad N.S. Combination of Hargreaves method and linear regression as a new method to estimate solar radiation in Perlis, Northern Malaysia. Sol Energy 85, , (2011). Garg HP, Gupta GL. In: Proceedings of the international solar energy society, congress. New Delhi 1134, (1978). General Directorate of Electrical Power Resources Survey and Development Administration (EIE), Turkey, ( ml) (2015). Development Southern Africa Vol:18, No.1, 19-30, (2001). Gueymard, C. Critical analysis and performance assessment of clear solar sky irradiance models using theoretical and measured data.sol Energy 51, , (1993). Gunerhan H, Hepbasli A. Determination of the optimum tilt angles of solar collectors for building applications. Build Environ 42,779 83, (2007). Hottel HC. Performance of flat-plate energy collectors. In: Space heating with solar energy. In: Proceedings of the Course Symposium. Cambridge: MIT Press, (1954). Hyewood H. Operating experience with solar water heating. J Inst Heat Vent Eng 39 (63), 9, (1971). Kern J, Harris I. On the optimum tilt of a solar collector. Sol Energy 17, , (1975). Kuwano Yukinori. Progress of photovoltaic system for houses and buildings in Japan. Renew Energy Vol:15, (1998). Lewis G. Optimum tilt of a solar collector. Sol Wind Technol 4, 407, (1987). Lof GOG, Taybout RA. Cost of house heating with solar energy. Sol Energy 14, 253, (1973). Messenger R., Ventre J. Photovoltaic Systems Engineering. p385, Florida, Crc Pres Llc., (2000). Montoya F.G., Montoya M.G. Gómez J., Manzano- Agugliaro F., Alameda-Hernández E.,The research on energy in Spain: a scientometric approach. Renew Sustain Energy Rev 29, , (2014). Peragón F.Cruz-, Peláez P.J. Casanova, Díaz F.A., García R. López, Palomar J.M.. An approach to evaluate the energy advantage of two axes solar tracking systems in Spain. Appl Energy 88, 12, , (2011). Senpinar A. The control of the stand-alone photovoltaic cell systems by computer. Firat University Graduate School of Natural and Applied Sciences, PhD Thesis, Elazig,(2005). Tang Runsheng, Wu Tong. Optimal tilt-angles for solar collectors used in China, Appl Energy 79, , (2004). Gopinathan KK, Maliehe NB, Mpholo M. Study on the intercepted insulation as a function of slope and azimuth of the surface. Energy 32, , (2007). Green, J.M., Wilson, M., Cawood, W. Maphephethe rural electrification (photovoltaic) programme: the constraints on the adoption of solar home systems. 300 Tariku Negash, Tassew Tadiwose. Experimental Investigation of the Effect of Tilt Angle on the Dust Photovoltaic Module. International Journal of Energy and Power Engineering 4 (4), , (2015).

315 Turkish State Meteorological Service, Republic of Turkey Ministry of Environment and Forestry (2015). Wu, G., Lin, Y., Wang, T.Method and strategy for modeling daily global solar radiation with measured meteorological data-a case study in Nanchang station China. Energ Convers Manage 48:p , (2007). Yellott H.Utilization of sun and sky radiation for heating cooling of buildings. ASHRAE J 15,31, (1973). 301

316 Key Factors for the Operation of a Solar Air Collector: A Parametric Study Ahmet Caglar *, Mustafa Burak Bahadir Akdeniz University, Faculty of Engineering, Department of Mechanical Engineering, Dumlupınar Boulevard Engineering Faculty B-309 Campus, Antalya, 07058, Turkey * acaglar@akdeniz.edu.tr Abstract The effect of various operational conditions on the outlet temperature and efficiency of a solar air colllector is investigated. A parametric study including operational parameters of collector inlet temperature, mass flow rate of air, solar irradiation and ambient temperature is performed to predict the outlet temperature of the air and efficiency of the collector. The parametric study is simulated for a v-groove type solar air collector. Results show that the increase in the collector inlet temperature, solar irradiation and ambient temperature have possitive effects on the collector outlet temperature whereas the mass flow rate of air has a negative effect. Regarding the collector efficiency, the mass flow rate of air and ambient temperature have a possitive effect whereas the collector inlet temperature has a negative effect. Furthermore, the collector effciency increases rapidly with increasing solar irradiation till a certain value of solar irradiation (about 300 W/m 2 ). After this value, the effect of the solar irradiation on the collector efficiency is not significant. Keywords: Solar air collector, operational parameters, outlet temperature, efficiency I. Introduction The demand of energy is increasing day by day and this energy requirement is likely to continue in the future worldwide. Fossil fuels used as energy source are both expensive and harmful to the environment. The global energy crisis and the destructive effects of greenhouse gases on the environment because of the burning fossil fuels are directed the governments to develop and apply new energy sources and technologies. Furthermore, as world population continues to grow and the limited amount of fossil fuels begins to diminish, it may not be possible to provide the amount of energy demanded by the world by only using fossil fuels. Therefore, countries must take action to promote a greater use of renewable energy resources, such as solar, geothermal or wind energy, so that renewable energies can provide major part of the solution. A great effort is currently made in this direction all over the world. Solar energy is one of the most important renewable energy sources and has a great potential for low temperature applications, particularly for air heating applications (Akpinar et al., 2004; Close, 1963; Kalogirou, 2004; Saxena and El-Sebaii, 2015; Turhan, 2006). Solar air collectors are used to heat air transferring the energy from the sun to the air passing through the collector. Many researchers have performed analytical studies of solar air collectors defining the mathematical model of the problem in detail (Duffie and Beckman, 1980; Garg and Prakash, 1997; Tchinda, 2009). Solar air collectors are manufactured in different types in terms of construction. Considered from the viewpoint of absorber plate, commonly used types are v-groove, finned and flat plate. Wazed et al. (2010) proposed flat plate collectors for low-temperature applications due to their low cost and ease of use. An experimental study on solar air heaters with and without fins has been reported by Indrajit et al. (1985). El-Sebaii et al. (2011) proposed an analytical model for flat and v-corrugated plate solar air heaters. The results showed that the double pass v-corrugated plate solar air heater is 11-14% more efficient compared to the double pass flat plate solar air heater. Another design criteria is the number of pass of air through the collector: solar air collectors with a single pass (Njomo and Daguenet, 2006; Paisarn and Kongtragool, 2003; Zhai et al., 2005) and double pass (Aldabbagh et al., 2010; Languri et al., 2011; Silvina et al., 2014). Alta et al. (2010) have investigated experimentally the solar air heaters with and without fin, with single and double glass cover. Several researchers are investigated the solar air heaters for drying processes (Ekechukwu and Norton, 1999; Koyuncu, 2006; Sreekumar, 2010). The thermal performance of solar air collectors is significantly influenced by operational conditions as well as design parameters. Kurtbaş et al. (2004) pointed out from the experimental study of the new-design collector that the efficiency of the collector increases with increasing mass flow rates due to an enhanced heat transfer to the air flow. Effects of operating variables on the thermal performance have been reported comparing v-corrugated and flat plate collectors by Karim et al. (2006). The results show that the temperature of the air at the collector outlet decreases with mass flow rate of air yielding an increase in the efficiency due to lower thermal losses. Sancar and Bulut (2015) studied experimentally on the change of collector efficiency at various operational conditions. They 302

317 investigated the effects of collector inlet temperature, solar irradiation and air velocity on the collector efficiency. Gülçimen et al. (2009) reported the linear dependence between the mass flow rate and collector efficiency in their experimental study. Özkaya et al. (2007) performed an experimental study to explore the variation of collector efficiency and outlet temperature for different flow rates. Results show that collector efficiency is higher at higher flow rates of air. Theoretical and experimental thermal performance of cross-corrugated solar air heater is studied for different design and operational conditions (Gao et al., 2007). In this study, a parametric study has been performed to investigate the effect of various operational conditions on the efficiency of a v-groove type solar air collector. Air outlet temperature of the collector is also estimated. The operational parameters varied in the simulations are the collector inlet temperature, mass flow rate of air, solar irradiation and ambient temperature. The results presented are to be a guideline in the application of the solar air collector for operational conditions to be set. This study is also expected to be beneficial to the new solar air collector designers. II. System Description The solar air collector used in this study is taken from a previous study performed by Karim and Hawlader (2004). V-groove type solar air collector consists of glass cover, absorber plate, collector frame and the back insulation in the bottom of the collector. The configuration of a v-groove air collector is shown in Fig. 1. V-groove type solar air collector has 50 mm v-height, leading to a flow passage area of m 2. As shown from the figure, the bottom of v-groove absorber is located on the aluminum plate of the collector and the glass cover is mounted on the top of the solar air collector. Furthermore, the insulation is made bottom of the solar air collector. The air is passing through the gap between the absorber plate and aluminum plate of the solar air collector. 1 (Karim and Hawlader, 2004): Tab. 1: Efficiency factor, loss factor and efficiency equations at different mass flow rates of air for the solar air collector. Flow Rate F R (τα) F R U L Efficiency Equations (kg/s) y= x y= x y= x y= x y= x In Table 1, y represent the efficiency and x represents x=(t i-t a)/i. III. Results and Discussions A parametric study including the collector inlet temperature, mass flow rate of air, solar irradiation and ambient temperature has been performed to investigate their effects on the collector outlet temperature and efficiency. III.1. Effect of the Collector Inlet Temperature Fig. 2 shows the effect of collector inlet temperature on the collector outlet temperature and collector efficiency for a v-groove solar air collector. The chosen values for the ambient temperature, mass flow rate of air and solar irradiation are 15 o C, kg/s and 600 W/m 2, respectively. The outlet temperature varies from 17.9 o C to 37.6 o C while the inlet temperature varies from 5 o C to 30 o C, respectively. It is obvious that the collector outlet temperature increases with the increasing inlet temperature. The efficiency varies from to while the collector inlet temperature varies from 5 o C to 30 o C, respectively. The higher air inlet temperatures yield the smaller efficiencies. This is due to the decrease in the temperature difference between the plate temperature and air inlet temperature yielding a decrease in the useful energy transferred from the plate to the air. Fig. 1: Schematic representation of the v-groove solar air collector (Karim and Hawlader, 2004). Collector efficiency can be expressed theoretically (Tyagi et al., 2012): η th = q u I (T = F R [(τα) e U i T a ) L ] (1) I or experimentally η th = m c p T A c I (2) The typical efficiency curves of the v-groove type solar air collector derived for different mass flow rates is obtained from the previous study as shown in Table 303 Fig. 2: Effect of collector inlet temperature on collector outlet temperature and collector efficiency.

318 III.2. Effect of the Mass Flow Rate of Air Fig. 3 shows the effect of mass flow rate of air on the collector outlet temperature and collector efficiency for a v-groove solar air collector. Simulated values for the collector inlet and ambient temperatures and the solar irradiation are 15 o C, 10 o C and 600 W/m 2, respectively. The outlet temperature varies from 32.6 o C to 23.3 o C while the mass flow rate of air varies from kg/s to kg/s, respectively. The collector outlet temperature decreases with the increase in the mass flow rate of air. The reason for this decrease is that the air inside the collector with lower mass flow rate is contacted to hot absorber plate for a longer period. On the other hand, the collector efficiency varies between As seen from Fig. 3, the collector efficiency increases when the air mass flow rate increases. The reason for this is that positive effect of the increase in the mass flow rate on the collector efficiency is larger than the negative effect of the decrease in the collector outlet temperature. For a prescribed collector outlet temperature according to purpose of the application, an optimization should be performed for the choice of the proper mass flow rate. The similar result is represented in (Karim and Hawlader, 2004) due to the use of same efficiency curve and the same conditions overlap. However, the increase in temperature gradient is getting smaller and smaller and the effect of temperature gradient has a similar impact with the solar irradiation on the collector efficiency. Fig. 4: Effect of solar irradiation on collector outlet temperature and collector efficiency. III.4. Effect of the Ambient Temperature The ambient temperature is a significant parameter for the solar air collector. Fig. 5 shows the variation of the efficiency and the collector outlet temperature for different ambient temperatures. The collector outlet temperature and efficiency increase with ambient temperature. Simulated values for the collector inlet temperatures, mass flow rate and the solar irradiation are 15 o C, kg/s and 600 W/m 2, respectively. The collector outlet temperature and efficiency varies from o C to o C, 0.57 to 0.85, while the ambient temperature varies from 0 o C to 25 o C, respectively. Fig. 3: Effect of mass flow rate of air on collector outlet temperature and collector efficiency. III.3. Effect of the Solar Irradiation Fig. 4 shows the effect of the solar irradiation on the collector outlet temperature and collector efficiency. Collector outlet temperature increases as the solar irradiation increases the change of collector outlet temperature with the solar irradiation shows a linear behavior. The collector efficiency also increases with the increase in the solar irradiation. This increase is sharp until the solar irradiation reaches about 300 W/m 2. After that point, the effect of solar irradiation on the collector efficiency is weak. This probably due to that at low solar irradiation values temperature gradient increases as well as solar irradiation. 304 Fig. 5: Effect of ambient temperature on collector outlet temperature and collector efficiency. IV. Conclusions Operational conditions have a significant effect on the efficiency of a solar air collector. In the current research, a parametric study is performed and the effects of different operational parameters including the collector inlet temperature, mass flow rate of air,

319 solar irradiation and ambient temperature on the collector outlet temperature and efficiency have been investigated. In the investigation of the effect of one of parameters, only the value of corresponding parameter is changed while the others are kept constant. Results show that the collector efficiency increases with mass flow rate of air, solar irradiation and ambient temperature, but decreases with the collector inlet temperature. From viewpoint of collector outlet temperature, collector inlet temperature, solar irradiation and ambient temperature have a positive effect. However, increasing mass flow rate of air has a negative effect on the collector outlet temperature. Nomenclature A c : Surface area of absorber plate (m 2 ) c p : Specific heat of air (J/kg. o C) F R : Heat removal factor I : Intensity of solar radiation (W/m 2 ) q u : Useful heat flux (W/m 2 ) T i : Inlet temperature ( o C) T a : Ambient temperature ( o C) m : Mass flow rate (kg/s) U L : Overall heat loss coefficient (W/m 2. o C) α : Absorptance of inner surface of evacuated tube collector η th : Collector efficiency T : Outlet-inlet temperature difference ( o C) References Akpinar E.K., Sarsilmaz C., Yildiz C., Mathematical modeling of a thin layer drying of apricots in a solar energized rotary dryer, International Journal of Energy Research, 28, (2004). Aldabbagh L.B.Y., Egelioglu F., Ilkan M., Single and double pass solar air heaters with wire mesh as packing bed, Energy, 35, (2010). Alta D., Bilgili E., Ertekin C., Yaldiz O., Experimental investigation of three different solar air heater: energy and exergy analysis, Appl. Energy, 87, (2010). Close D.J., Solar air heaters for low and moderate temperature applications, Solar Energy, 7(3), (1963). Duffie J.A., Beckman W.A., Solar engineering of thermal processes, Wiley, New York (1980). Ekechukwu O.V., Norton B., Review of solar-energy drying systems II: an overview of solar drying technology, Energy Conversion & Management, 40(6), (1999). El-Sebaii A.A., Aboul-Enein S., Ramadan M.R.I., Shalaby S.M., Moharram B.M., Investigation of thermal performance of-double pass-flat and v-corrugated plate solar air heaters, Energy, 36, (2011). (2007). Garg H.P., Prakash J., Solar energy fundamentals and applications, McGraw-Hill, New Delhi: Tata (1997). Gülçimen F., Durmuş A., Durmuş A., Kurutmada kullanmak için havalı kollektörler tasarlanması, V. Yeni ve Yenilenebilir Enerji Kaynakları Sempozyumu, TMMOB Makine Mühendisleri Odası Kayseri Şubesi, Sayfa , Ekim Indrajit, Bansal N.K., Garg H.P., An experimental study on a finned type and nonporous type solar air heater with a solar simulator, Energy Conversion and Management, 25(2), (1985). Kalogirou S.A., Solar thermal collectors and applications, Progress in Energy and Combustion Science, 30, (2004). Karim M.A., Hawlader M.N.A., Development of solar air collectors for drying applications, Energy Conversion and Management, 45, (2004). Karim M.A., Hawlader M.N.A., Performance evaluation of a v-groove solar air collector for drying applications, Applied Thermal Engineering, 26, (2006). Koyuncu T., Performance of various design of solar air heaters for crop drying applications, Renewable Energy, 31, (2006). Kurtbas İ., Durmuş A., Efficiency and exergy analysis of a new solar air heater, Renewable Energy, 29, (2004). Languri E.M., Taherian H., Hooman K., Reisel J., Enhanced double-pass solar air heater with and without porous medium, Int J. Green Energy, 8, (2011). Njomo D., Daguenet M., Sensitivity analysis of thermal performances of flat plate solar air heaters, Heat Mass Transfer, 42, (2006). Özkaya M.G., Kırbaş İ., İncili V., Havalı Güneş Kollektörünün Performansının Deneysel Olarak İncelenmesi, Politeknik Dergisi, Cilt:10 Sayı: 3, s (2007). Paisarn N., Kongtragool B., Theoretical study on heat transfer characteristics and performance of the flat-plate solar air heaters, International Communications in Heat and Mass Transfer, 30(8), (2003). Sancar İ., Bulut H., Mahal ısıtmasında kullanılan havalı güneş kollektörleri ve Adıyaman şartlarında performansının incelenmesi, Adıyaman Üniversitesi Bilim, Kültür ve Sanat Sempozyumu-II, Adıyaman, 2-3 Nisan Saxena A., El-Sebaii A.A., A thermodynamic review of solar air heater, Renew. Sustain Energy Rev., 43, (2015). Silvina M.G., Silvana F.L., Alejandro H., Graciela L., Thermal evaluation and modeling of a double-pass solar collector for air heating, Energy Procedia, 57, (2014). Gao W., Lin W., Liu T., Xia C., Analytical and experimental studies on the thermal performance of cross-corrugated and flat-plate solar air heaters, Applied Energy, 84, Sreekumar A., Techno-economic analysis of a roof-integrated solar air heating system for drying fruit and vegetables, Energy Conversion and Management, 51,

320 (2010). Tchinda R., A review of the mathematical models for predicting solar air heaters systems, Renewable and Sustainable Energy Reviews, 13, (2009). Turhan K., Performance of various design of solar air heaters for crop drying applications, Renewable Energy, 31, (2006). Tyagi V.V., Panwar N.L., Rahim N.A., Kothari R., Review on solar air heating system with and without thermal energy storage system, Renewable and Sustainable Energy Reviews,16, (2012). Wazed M.A., Nukman Y., Islam M.T., Design and fabrication of a cost effective solar air heater for Bangladesh, Applied Energy, 87(10), (2010). Zhai X.Q., Dai Y.J., Wang R.Z., Comparison of heating and natural ventilation in a solar house induced by two roof solar collectors, Applied Thermal Engineering, 25, (2005). 306

321 Experimental Analysis of Solar Space Heater Performance Guvenc Umur Alpaydin *, Serhan Kucuka Dokuz Eylul University, Department of Mechanical Engineering, Izmir, Turkey * guvencalpaydin@ogr.deu.edu.tr Abstract Energy consumption in buildings can be reduced important amount using equipments supported by solar energy. Solar radiation falling on the surface can be transferred to indoor with collectors which to be placed on the building exterior. Thus a part of the heat load can be compensated. The integrated system which made experimental investigation, contain an exterior unit, an interior unit and thermosyphon cycle which provide the one-way heat flux from outside to inside. On the exterior unit, glass absorber tubes with vacuum pipes used as collectors. Interior unit contained a radiator. Experimental studies of integrated system were performed in outdoor where is a container house in April. Results of the study show that, this integrated system can be used different research interests like greenhouse area and it gain advantage from energy consumption. Keywords: Solar space heater, thermal diode, all-glass solar vacuum tubes, I. Introduction Nowadays, it is widely used in solar-assisted heating technologies can be classified as active or passive systems. Active systems are work forced convection but passive systems work with natural convection. There are many applications of different passive systems. In Trombe wall, a thin flow region is formed with help of coated glass on the outer wall. The effect of solar radiation energy is stored at the heated wall. A hot air flow is provided with help of the natural convection in the cavity to the indoor. Trombe wall, although it is easily applied due to a simple structure having building, depending on the operation conditions of heat transfer can be adversely affected. Chan et al. (2010), summarized the negative side of the Trombe wall. On days when solar radiation is low, inverse heat transfer can be happened to outwards from the hot interior. Oscillations in the wall temperature occured because of the variability of the time-dependent of solar energy. thus, the desired heating load may not be obtained Sizing is important because the heat transfer in the cavity will affect the optimum geometric dimension. Classic trombe wall have a low thermal resistance and it leads to significant heat loss in cold time. Michel Trombe-composite wall has been developed to solve this problem. In addition to classic trombe wall, at Michel Trombe-composite wall, a insulated wall put on. Air flow and convection took place between these two walls. Shen et al. (2007) presented as an alternative method to classical in general, double glazing can be used instead of a single glazing to prevent heat loss in the walls. This method not only prevents heat loss in winter, but also contributes to passive cooling in summer. At summer cooling application, inner surface of the storage wall can be insulated to maximize the rate of ventilation. Gan et al. (1998). Another solar assisted heating system is solar chimney. In this system, A solar chimney applications allow that air which is heated by collectors and reduced density can be transfered to external environment with natural convection. The first model of a solar chimney project designed observed for one year and 2 models are designed according to feedback. One of two solar-powered, including ground level and on the other on the roof is available in natural convection air heaters. In the daytime, air is absorbed from the inside by the rooftop solar air heater fan and it serves as excreted out. Other heaters are used to heat the fresh air supplied to the room. Two rectangular air spaces were opened in bottom of the wall and the top of vertical collector's corner due to it was intended to ensure air circulation by Raman et al. (2001). The slope of the solar chimney s facing south portion was investigated by Harris et al. (2007). They concluded that the inclination of 67.5 degrees is %11 more efficient than the vertical position. %10 higher efficiencies were obtained with the wall surface with a low emissivity. Passive building heating systems listed above requires total or partial replacement of the facade of the building. As an alternative to these designs that solar thermal energy is transferred into a one-way diode, has been studied by different researchers. Unidirectionally transferring heat inside is provided with a heat pipe or water heater cycles. Pioneering research on this issue was started by Chun, Chan and their teams. The thermal performance of the Bayonet-type thermal-diode is investigated experimentally by Chen et al. (1998). In this design, 307

322 the fluid reservoir there was a narrow channel between two sloping and warming fluid moves to the reservoir at the top because of buoyancy. Chun et al. (1996) emphasized that solar energy can be imported indoor with using thermal diode system. Time dependent flow model was established and it was compared with experimental results. Chun and Chen (2002) a new thermal-diode was developed. In this design the direction of heat transfer can be changed mechanically. The horizontal pipes that comprise the thermosiphon loop designed as which open the external are moving part, neighbor to internal are rotary joint. With the vertical movement of the surface of the collector, natural convection can be reversed. Thus, the system can be used for both heating and cooling purposes. Experimental measurements were occured for solar radiation between 400 and 800 W/m 2 at all heat loads, for natural convection to occur, it is stated that mobility should take approximately 15 minutes. Sytem can be steady state at 2 hours. In another study conducted by Chun et al. (2009), the effect of working fluids on the thermal diode performance was investigated. Five liquids that have different thermal conductivity and viscosity, has been tested with the help of solar simulator. Experiments showed that water and low-viscosity silicone oil have high heat transfer rate. Fang and Xia (2010), were divided by a vertical partition of a 30 cm wide water reservoir. with a plug right or left to adjust the direction of flow are provided. Therefore, in winter period, heated water on the surface, which receives incident radiation, transferred from upper chamber to the inner part by natural convection. In summer period, it was aimed to provide passive cooling by inverse water circulation at nights. They are designed temperature controlled test room. One surface of this room is water reservoir. It was determined the optimum localization of part in the reservoir and amount of transferred heat. Pratical way, evacuated tube solar collectors are used in passive solar system. Budihardjo et al. (2007), examined the effect of natural convection of water circulation with the actual numerical and experimental. In the study, correlations were developed to determine the amount of movement in the tube and heat transfer coefficients. Another study of Budihardjo et al. (2009), are examined the annual performance of evacuated tube solar collectors in different cities in Australia and compared with flat plate the performance of solar collectors. Tang et al. (2009), aimed to determine the optimum angle of inclination. The inclination and azimuth of the changes due to radiation from the collector is given. An analytical model has been developed by Ma et al. (2010) due to the calculation of solar collector s heat loss to the environment. A U-pipe was placed in solar collector. They showed that surface temperature of solar collector is important parameter for heat loss. In the literature have been numerous studies that various methods associated with the use of solar energy for heating. Some of this work involves the transmission of solar energy with various systems like Trombe wall, or storage of solar energy at a heat 308 tank. However, these systems require structural architectural arrangements. In some other studies, heat pipe, bayonet and thermosyphon systems have been investigated to heat transfer from the exterior to interior of the solar energy. These systems are simpler form because these are not connected to heating system of the building In this study, components of system are vacuum solar collectors, radiator and an electirical heater to found water flow at natural circulation. When vacuum solar collectors that was transferred thermal energy with help of thermosyphon loop, was placed on the building exterior, radiator that was in the interior, transferred the thermal energy to room throughout the daytime. Thus, reducing heating costs with this system is considered as an adjunct to conventional heating systems are proposed. II. Experimental Facility II.1. Setup Experiments are performed in Solaris building where is located DEU Tinaztepe Campus in Izmir/Turkey. Building consists two room and test room is almost 10 m 2 and two wall of this room are outer walls. One of them is south-facing wall that was used for placing experimental setup. Outer unit of experimental unit, that considered fifteen evacuated tube solar collectors, a manifold was placed vertically on the outer wall. External unit is placed to the 10 º south-east location. Fig. 1. Exterior unit of system Basically, when a portion of solar energy that comes through the outer surface is absorbed, a portion of solar energy is reflected. There is heating loss from outer unit to environment. Remain energy is transferred to the inner unit with thermo-syphon cycle. In the inner unit, energy is transferred to the room with help of radiator. For natural circulation, collectors is placed below the radiator. Evacuated tube solar collectors started circulation by heating the water. Properties of Evacuated Tube Solar Collectors which are manufactured by Solarsan

323 INC, are given at table 1. Inlet and outlet temperatures of manifold were mesaured with help of thermocouples. Tab. 1. Properties of evacuated tube solar collectors Length 1000 mm External tube diameter Internal tube diameter Material Surface Coating 58 mm 47 mm Borosilicate glass AL-N\AL sellektif Reflection of %7 (100 o C ) Vacuum P<0.005 Pa Heat Loss Coefficient <0.8W/ ( m^2 o C ) System needed the open expansion tank due to the increase in the volume of heated water. Water reinforcement to system is made from the expansion tank. The instantaneous global irradiance on a vertical surface was measured at intervals of 1 min using an Eppley radiometer, model 8-48, serial number The millivolt output of the radiometer was read with a datalogger. The voltmeter readings were converted to radiant flux units using a conversion factor of 1 W/m^2=10.72 µv, supplied by the radiometer manufacturer. Interior unit consists of single panel single convector radiator (1000*600mm) and an electrical heater. Heater was used to calculate the flow rate and it is connected to adjustable power supply. Input and output temperatures of the radiator and electrical heater measured with help of thermocouples as shown in figure 2. Furthermore, temperature difference was measured as a voltage difference and Voltage differences are converted to temperature differences with the help of table 2 which given below. Temperatures were recorded with (T-type) copper-constantan thermocouples The reason of making the voltage difference of the measurement was to achieve more sensitive measurement. (b) Right view Fig. 2. Interior unit of the system Table2. Type T thermocouple reference table Although all pipes are insolated, heat loss is occured. Heat loss must be calculated with equation set before the flow rate calculations for this reason. (1) (2) and (3) (4) After heat loss calculated with help of previosly equations, mass flow rate could determined equation 5. III. Results and discussions (5) (a) Front view Experiments were performed in two sunny days (21 st and 30 th of April) which are similar to in terms of weather conditions. Temperatures were recorded at 60 second intervals, between 7.00 and local time.time periods started 7 a.m and finished 7 p.m with two hours interval. 309

324 During the experiment the amount of solar radiation from the surface were measured with a pyranometer which is placed above the manifold. III.1. Case 1 Average ambient temperatures of 21 st of April are given in Table 3 Table 3. Average temperature and insolation for two hours periods of first experiment Time Period Time Interval Ambient Temperature ( o C) Insolation (W) 1 st nd rd th th th Inlet and outlet temperatures of manifold were measured during the experiment. These results are given at Figure 3. During the daytime, for the most insolation ranges (4 th and 5 th time periods), the temperature of both ends of the manifold is obvioused that the increase in parallel to the incident radiation. It showed that cold side (manifold inlet and radiator outlet) and hot side (manifold outlet and radiator inlet) temperatures are similar. The hottes value was almost 55 o C and 40 o C for hot and cold side respectively. III.2. Case 2 In second experiment, room temperature was fixed at 20 o C for observing the effect of low room temperature to natural convection speed. Table 4. Average temperature and insolation for two hours periods of second experiment Time Period Time Interval Ambient Temperature ( o C) Insolation (W) 1 st nd rd th th th Although it is similar to the previous experiment with to have the insolation, the temperature drop has been obtained because of low room temperature. It observed that temperatures of radiator and manifold decreased about 6-7 o C Fig 3. Inlet and outlet temperatures of manifold Inlet and outlet temperatures of the radiator are shown Figure 4 for first experiment. According to Figure 4 room temperature increased with radiator temperature. At the beginning (1 st time period), significant amount of increase was not observed at room temperature. The room temperature increased with increasing amount of insolation. In the last period, due to reduced amount of radiation, ambient temperature and radiator inlet-outlet temperatures decreased. Fig. 5. Inlet and outlet temperatures of manifold Fig. 6. Variation of room temperature according to input and output temperatures of radiator III.3. Comparative results Heat loss from the pipe which is connected the electrical heater, is given in the Figure 3 according to the pipe temperature. It observed that heat loss increased linearly with pipe temperature. Fig 4. Variation of room temperature according to input and output temperatures of radiator 310

325 Figure 9. Position of Sun based on time (a) First experiment Table 4. Average of power and efficiency values for each time period Time period Power Efficiency 1st period 44.0 W nd period 46.7 W rd period 44.2 W V. Conclusions (b) Second experiment Fig. 7 Pipe Temperature and Heat Loss When the room temperature is fixed at 20 o C, water mass flow into the radiator increased as expected. It can be concluded that if the room temperature and ambient temperature difference between enhance, system will operate at higher mass flow rates. Calculated mass flow rate is given in figure 8 for both cases. In 1st time period the efficiency reached the highest value. In 2nd period, the highest power output is obtained. But the efficiency is lower than the 1st time period. The reason of this, incident radiation is reaches highest value during the day time, in 2nd period. Acknowledgements The author gratefully thanks to Solarsan INC for collaborating and special thanks to DEU Solaris Team Nomenclature Fig. 8 Pipe temperature and heat loss It clearly seen in Figure 8, in case 1,mass flow rate decreased with time more dramatically than in case 2. The reason of this, room temperature increased during daytime and differences of the radiator temperature and room temperature decreased in case 1 but in case 2 because of fixing the room temperature at 20 o C, it was not observed the amount of mass flow rate drop like case 1. Calculated mass flow rate is used for determined to collectors power and efficiency. Pr Nu D : Reynold number : Prandtl number : Nusselt number : density (kg/m 3 ) : Flow rate (kg m 1 s 1 ) : Diameter of pipe (m) : Diameter of pipe (m) : Length of pipe (m) : Heat transfer coefficient (W/(m 2 K)) : thermal conductivity (W/(m K)) Q : Heat loss from pipe (W) r L h k loss Q : Given heat from heater (W) given T i T : Pipe Temperature ( C) : Room Temperature ( C) : Mass Flow Rate (g/s) : Specific heat of Water (J/gK) References Budihardjo I., Morrison G. L., Behnia M Natural circulation flow through water-in-glass 311

326 evacuated tube solar collectors, Solar Energy,81, Budihardjo I., Morrison G. L., Behnia M Performance of water-in-glass evacuated tube solar water heaters, Solar Energy,83, Chan, H.Y., Riffat, S.B., Zhu J Review of passive solar heating and cooling technologies, Renewable and Sustainable Energy Reviews, 14, Chen K., Chailapo P., Chun W., Kim S., Lee J. L., The dynamic behavior of a bayonet-type thermal diode, Sol. Energy, 64 (4 6) (1998), pp Chun W., Lee Y. J., Lee J. Y., Chen K., Kim H. T., Lee T. K Application of The Thermal Diode Concept For The Utilization of Solar Energy, Energy Conversion Engineering Conference, Chun W., Chen K Test Results of a Bi-Directional Thermodiode System For Solar Energy Utilization, Solar Energy,73, Chun W., Koa Y.J., Lee H.J.,Han H., Kim J.T., Chen K Effects of working fluids on the performance of a bi-directional thermodiode for solar energy utilization in buildings,solar Energy,83, Fang, X., Xia L Heating performance investigation of a bidirectional partition fluid thermal diode, Renewable Energy, 35, Gan, G A Parametric Study of Trombe Walls for Passive Cooling of Buildings, Energy and Buildings, 27, Harris, D.J., Helwig, N Solar Chimney and Building ventilation, Applied Energy, 84, Ma L., Lu Z., Zhang J., Liang R Thermal performance analysis of the glass evacuated tube solar collector with U-tube, Building and Environment, 45, Raman, P., Mande, S., Kishore N.V.V A Passive Solar System for Thermal Comfort Conditioning of Buildings in Composite limates, Solar Energy, 70, Shen, J., Lassue, S., Zalewski, L., Huang,D Numerical Study on Thermal Behavior of Classical or Composite Trombe Solar Walls, Energy and Building, 39, Tang R., Gao W., Yu Y., Chen H Optimal tilt-angles of all-glass evacuated tube solar collectors, Energy,34,

327 SUSTAINABLE AND RENEWABLE ENERGY DEVELOPMENT 313

328 Abstract Load Side Management in Smart Grids using a Global Optimizer Abdelmadjid Recioui *, Mossaab Djehaiche, Abderrahim Boumezrag Laboratory signals and systems, Institite of electrical and electronic engineering, University of Boumerdes Avenue de l indépendance, 35000, Boumerdes, Algeria. * rec79dz2002@yahoo.com In this paper, a genetic algorithm is used in conjunction with demand side management techniques to find the optimal planning of energy consumption inside N buildings in a neighborhood. The issue is formulated as multiobjective optimization problem aiming at reducing the peak load as well as minimizing the energy cost. The effectiveness of the approach is confirmed by simulation results carried out on a residential area with a variety of electrical appliances. The simulations reveal that the adopted strategy is able to plan the daily energy consumptions of a great number of electrical devices with good performance in terms of computational cost. Keywords: Smart grids, energy management, load side management, optimization, genetic algorithms. I. Introduction In power grids, generation capacity is required to meet peak-hour load demand plus a security margin. However, according to recent studies, the average utilization of the generation capacity is below 55% (US energy information, 2014). This leads to inefficient operation of power grids because a portion of generation plants is largely unused or underutilized, but must still be maintained and supervised to guarantee its reliability. On the other hand, as energy demand and peak load demand as well, continue increasing, additional generation capacity will be needed to accommodate future load demand, which requires a large investment and might lead to even lower utilization. Recently, the smart grid (SG) has been proposed as a new type of electrical grid to modernize current power grids to efficiently deliver reliable, economic, and sustainable electricity services. One of the key features of the SG is the replacement of conventional mechanical meters with smart meters to enable two-way communications between users and grid operators. Using the communication infrastructure of the SG, it is possible to shape load demand curves of the users by means of demand side management (DSM) programs (Goudazi et al., 2011 and Agnetis et al., 2013) Commonly, demand side management is term used by electric utilities to describe programs developed to influence the electricity usage patterns of customers to control the energy consumption at the consumer/meter side and many approaches are introduced to solve it as an optimization problem with proper objective function (Barbato et al., 2011). The DSM is an opportunity to avoid or delay the need to construct new generating capacity by reduction or shift the energy of consumers. Also, for domestic or industrial consumer, DSM can be considered an opportunity to save money by reducing their electricity bill taking the advantage of financial incentive provided by utility. In the global energy scenario the demand management is an important function of the smart grid, which ensures the grid sustainability and reliability. Demand management is not entirely new for electric grid, but it is moving a bit towards a customer driven activity in the near future (Agnetis at al., 2011; Zhao at al., 2013 and Guo et al., 2012). Demand management mechanisms can be designed to control the electric resource of individual users (Strbac, 2008). Two different approaches are proposed in the literature to address the DSM problem: optimization and game theory (Barbato and Capone, 2014). Game theory is practically used since it can model complex interactions among the independent rational players of the power grid (Saad et al., 2012). The extension of demand management mechanisms for communities of users is represented by techniques designed for micro-grids, which are small-scale versions of the electricity systems that locally generate and distribute electricity to consumers. These grids are an ideal way to integrate renewable resources at the community level and allow for customer participation in the electricity market (Liang and Zhuang, 2014; Prabaakaran et al., 2014; Ravichandran, 2013). Optimization methods for demand-side management can be classified based on three main Characteristics (Barbato and Capone, 2014): Firstly, DSM systems can be designed to optimize the usage of electric resources of either individual users or a community of cooperative consumers. In the first case, users are individually managed, while in the second case, consumers collaborate in defining their operating plans and DSM methods are used to optimize a shared utility function. A further classification can be obtained based on whether deterministic or stochastic techniques are utilized to design the demand management 314

329 mechanism. Finally, DSM systems can be classified based on the time scale used to manage the resources of customers: day-ahead and real-time. In the day-ahead stage, the operating plan of electric resources of users is defined over the next 24-h time period (or a different time horizon). Various dynamic and effective schemes for autonomous DSM in smart girds have been proposed in literature. Examples of pioneering works include the one of (Mohsenian-Rad et al.,2010) who proposed an autonomous load scheduling algorithm based on cooperative game theory, where each user is a player and their load schedules are the strategies. (Agarwal and Cui, 2011) proposed a load scheduling no cooperative game among users that can be reduced to a congestion game. In both studies, the single optimization objective is to minimize the electric bill of the users, while the reduction of the peak-hour consumption is considered as a desirable secondary effect. (Samadi et al., 2011) proposed an auction based scheme where users provide their utility functions and energy constraints to the utility company, who then replies with a set of prices that maximizes the utility functions of users. A similar auction scheme is also proposed by (Li et al., 2011). We can notice that previous studies mostly aim at a single objective, e.g., to minimize the cost of the users. In this paper, we introduce the concept of intelligent home energy management in a neighborhood and develop an energy consumption planning system of the daily tasks of a set of household users. The planning strategy aims at reducing the peak load as well as minimizing the energy cost. The maximization of the peak-to-average ratio of the total energy demand is considered as a desired objective function for the utility and the minimization of the energy cost is considered for consumer. The optimal solution of this multi-objective planning problem is found using a Genetic Algorithm (GA). II. Energy management and DSM The term Energy Management can be considered as the wisdom and effective use of energy to maximize profits (minimize costs) and enhance competitive positions (Samadi et al., 2011). Energy management is also the strategy of adjusting and optimizing energy using systems and procedures to reduce energy requirements per unit of output while holding constant or reducing total costs of producing the output from these systems. In fact, one is mainly concerned with the ones that relate to saving energy businesses, public-sector, government organizations, and home satisfaction. Energy management can be categorized into two main classes: II.1. Supply Side Management Supply-side management (SSM) refers to actions 315 taken to ensure the generation, transmission and distribution of energy are conducted efficiently. This has become especially important with the deregulation of the electricity of industry in many countries, where the efficient use of available energy sources becomes essential to remain competitive. The electricity generated should be utilized efficiently to meet the demand of countries. This improves the reliability of the power supply system. SSM is used primarily with reference to electricity but it can also be applied to actions concerning the supply of other energy resources such as fossil fuels and renewable resources. Energy users will normally focus their efforts on demand-side management methods (DSM) but some will consider the supply side too. For example, they may look at on-site generation alternatives (including cogeneration) or consider diversifying to alternative fuel sources such as natural gas, solar, wind and biomass. II.2. Demand Side Management Energy demand management, also known as demand side management (DSM), is the modification of consumer demand for energy through various methods such as financial incentives (Wei-Yu et al., 2013) and education. Usually, the goal of demand side management is to encourage the consumer to use less energy during peak hours, or to move the time of energy use to off-peak times such as nighttime and weekends. Peak demand management does not necessarily decrease total energy consumption, but could be expected to reduce the need for investments in networks and/or power plants for meeting peak demands. An example is the use of energy storage units to store energy during off-peak hours and discharge them during peak hours (Wei-Yu et al., 2012). Energy demand management can be classified into three types: Energy Efficiency: Using less power to perform the same tasks. Demand Response: Demand Response refers to a wide range of actions which can be taken at the customer side of the electricity meter in response to particular conditions within the electricity system (such as peak period network congestion or high prices) (Torriti et al., 2010). Dynamic Demand: Advance or delay appliance operating cycles by a few seconds to increase the diversity factor of the set of loads. II.3. Demand Side Management A DSM program is a program used to control the load profile indirectly in order to achieve the utility objectives. These objectives are: To have the load factor as close as possible to 1.0 To have the peak load within the proper margin. By achieving the previous objectives, the utility would get the maximum possible energy from the installed units, thus maximizing the total profit and minimizing

330 the average cost per KWh. There exist five programs in literature and can be described as follows (El-Sobki, 1996) : 1. Valley Filling In this program, the main objective is to increase the demand during the off peak periods while having the same load peak (Fig. 1). This could be achieved by encouraging the consumers to increase their demand. 2. Load Shifting In this program, it is required to shift part of the demand at the peak period to the off peak periods (Fig. 2). This program could be used in case that the installed capacity is not enough during the peak load. Fig. 4 Energy conservation program effect 3. Peak Clipping This program is used to decrease the demand during the peak load periods (Fig. 3). Also, these loads can t be shifted to the off peak periods. This could be due to lack of installed capacity during these periods. This program could be achieved be indirectly forcing the consumers to decrease their loads by the use of miniatures on their supply points. 4. Energy Conservation This program is used when it is required to decrease the energy consumption all over load period (Fig. 4). This could be achieved by using high efficiency components. Fig. 1 Valley filling program effect 5. Load Building This program is used when it is required to increase the energy consumption (Fig. 5). This could be very beneficial in case of surplus capacity as because the average cost per KWh will decrease. Fig. 2 Load shifting program effect Fig. 5 Load building program effect III. Optimal formulation of DSM programs Fig. 3 Peak clipping program effect 316 DSM has a major role of utility planning and operation. In this section, an optimal based formulation is developed to simulate the implementation process of the DSM program to assess its technical and financial impacts for both utility and users. The objective function is formulated either to control the use of the supply side resources subject to end user demand for power and energy without loss of production or comfort, or to improve system performance by increasing load factor and enhance the customer service quality. The mathematical construction model for any optimization problem is generally determined by clarifying the

331 following questions: What does the model seek to determine? What are the objectives (goals) needed to be achieved to determine the best solution? What are the variables of the problem? What constraints must be imposed on variables to simulate properly actual variables? Generally, two objective functions exist, either to maximize the system load factor for the utility, or to minimize the total cost of the bill for the customer. The imposed constraints on the demand type at different time intervals (control variables) differ from a technique to another and depend, also, on the load peculiarities and the power system. Depending on the desired objective, a DSM program seeks to optimize either of the following objective functions: N J P( i, j). t( j) i1 j1 J t( j) j1 Max L. F. (1) N P( i, k) Max L. F. J j1 i1 PTO( j). t( j) J t( j) j1 PTO( k) (2) N J N J Min C P( i, j). t( j). ce( i, j) P( i, j). cd( i, j) (3) i1 j1 i1 j1 IV. Simulation Results and Discussions This section considers a dynamic demand management for the residential sector and formulates the energy consumption-scheduling problem as a multi- objective optimization problem, addressed with a heuristic approach. The adopted planning strategy aims at reducing the peak load as well as minimizing the energy cost. It has to be noted that the considered optimization objectives are mostly conflicting and non-commensurable. Therefore the optimal solution of this multi-objective planning problem is found using a Multi-Objective Genetic Algorithm (MO-GA). IV. 1 Problem Formulation The increasing number of automation system and electrical appliances in residential sector makes the enhancing of electrical efficiency of commercial and domestic buildings highly desirable. On the other hand, comfort of user and quality of life must be preserved. This part of project addresses this challenging issue as a constrained multi-objective optimization problem. The aim is the balancing of the energy consumption in a residential district to avoid the concentration of simultaneous electricity request on the same time. This has to be done by saving the cost and by shifting loads from on-peak to off-peak periods. We considered a neighborhood of N buildings, whose U users program the set of daily tasks to be done (basic electrical appliances). It is assumed that each electrical appliance of the buildings is equipped with a terminal unit controller (TUC), which collects and transmit consumption to a building controller, connected to the energy consumption-planning system (ECPS) of the residential area. The TUC turn on and off the appliances according to the scheduling pattern planned by ECPS. The ECPS schedule the tasks at times multiple of t in a discrete time setting. Where: L.F.: is the system load factor. P (i,j): is the demand of load type i at time interval number j. N: is the total number of load demand types. J: is the total number of time intervals. PTO(j): is the total demand for all the loads types from j=1 to j=j over the time interval number j. k: is the number of time interval at which the maximum demand for all the load types numbers from i=1, N over all the time duration from j=1, J occurs. C: is the total cost of the electrical demand and energy consumption. ce(i,j) : is the cost of energy for load type i at time interval number j. cd(i,j) : is the cost of demand for load type i at time interval number j. 317 The consumption scheduling system shifts in time the execution of the tasks according to the minimization of the two objective functions, associated with the electrical load profile and the energy cost. The first function is a measure of the maximum load factor to reduce the peak load power consumption by the following equation (1). The second objective function takes into account the energy price as in equation (3). IV. 2 Multi-Objective Genetic Algorithms GA belongs to the techniques that are based on the concept of natural evolution, i.e., in natural environment, only the strong individuals survive and have the greater opportunity to pass their genes to next generation. Over the generation, the individuals with the correct combination of genes in their chromosomes become dominant. In GA the individuals represent the candidate solutions of the multi-criteria optimization problem. Their chromosomes are parameters univocally associated

332 with a solution through a mapping mechanism. Otherwise, the values of objective functions define a point in a multi-dimensional space (objective space). To create the new population from current individuals three main operators are used in GA: crossover, mutation, and selection. The crossover combines the chromosomes of two individuals, called parents, to generate new individuals, called offspring. Otherwise, the mutation introduces random changes in the chromosomes of the individual. The crossover is responsible for the convergence of population to the no dominated solutions, while the mutation provides the escape from local minima of optimization functions. The selection operator emphasizes the better solutions and ensures their survival to next generation. In GA these operators are iteratively applied until a stopping condition is satisfied and an approximation of the Pareto optimal set is found. IV. 3 Case study and simulation results consumer to construct the total electrical load profile of that home with the earliest starting time specified by the same consumer. Finally, the total power consumption of that home is shown in Fig. 6. The total power is not uniformly consumed; which causes a peak load. power consupmtion (kw) power consumption specified by the consumer To demonstrate the effectiveness of the proposed heuristic strategy for optimal planning of daily electrical consumptions, we considered a case study of a residential area with 100 smart homes that have several basic electrical appliances. We supposed that each domestic unit has at least five electrical devices (fridge, washing machine, dishwasher, electrical car and interior lighting) and up to twelve appliances. So, there are at least 1000 electrical devices over the given residential area with different consumption patterns specified in (table.1). Tab. 1 Simulated home load Scenario Tasks Earliest Latest Duration Power starting finishing (hours) (kw/h) time time Dish washer Washing machine Spin dryer Cooker hob Cooker oven Cooker microwave Interior Lighting Laptop Desktop Vacuum cleaner Fridge Electrical car Other tasks The planning system schedules the tasks according to the given earliest starting times, the latest finishing times, the durations, the power requirements and the start times specified by the user. Taking a single house as a first step, we started the work with the electricity consumption and work for each task with the earliest possible starting time specified by the time Fig. 6 daily home load profile So, The TCU will fix that by turning on and off the appliances according to the scheduling pattern planned by ECPS. The ECPS schedule the tasks at times multiple of t in a discrete time setting. (Table. 2) indicates the new electrical load profile of that home with the best starting times. Tab. 2 New scheduled load profile with best starting times Time(h) power (kw) Time(h) power (kw) power (kw) Time(h) power (kw) power consupmtion (kw) old power consumption new power consumption time Fig. 7 old and new home power consumption It is clear that the overall power consumption of the newly scheduled activities is less than the old profile

333 (Fig. 7). This can be accomplished by selectively shifting some loads (as suggested by Genetic algorithms). As for the user, the shifting can be done either manually (user implication) or automatically through the smart meter. One main characteristic of the scheduled activities is that the main (high consuming power) activities are shifted to be executed in the off-peak periods. This work describes an efficient multi-objective planning system to manage the electricity demand in smart residential area. The proposed strategy is based on a heuristic approach using the Genetic Optimization Algorithm. The obtained results show that the scheduler not only decreases the pick load and reduces the utility bills but also preserves the user satisfaction. V. Conclusions This paper highlighted the benefits of smart grids and energy management on managing the electricity consumption, the utility bills and preserving the user satisfaction. The optimal usage times were found using Genetic Algorithms. We succeeded in achieving the best optimal solution of rescheduling the total power consumption of the selected residential area by clipping the total peak demand (a consistent reduction of approximately 20 % of peak load) and minimizing the cost. Moreover, the 1 st objective minimizing the total peak- was tested through three DSM programs which are: peak clipping, valley filling and load shifting (each one subjected to specific constrains). Finally, it can be concluded that load shift technique has more accuracy in minimizing the total peak demand and maximizing the system s load factor without affecting the total energy consumption (same amount of energy consumption before and after rescheduling). References Agnetis, A.; de Pascale, G.; Detti, P.; Vicino, A. Load Scheduling for Household Energy Consumption Optimization. IEEE Trans. Smart Grid 2013, 4, Agnetis, A.; Dellino, G.; Detti, P.; Innocenti, G.; de Pascale, G.; Vicino, A. Appliance operation scheduling for electricity consumption optimization. In Proceedings of the 50th IEEE Conference on Decision and Control and European Control Conference (CDC-ECC), Orlando, FL, USA, December 2011; pp Agarwal T. and Cui S., Noncooperative games for autonomous con-sumer load balancing over smart grid, CoRR, 2011 [Online]. Avail-able: Antimo Barbato and Antonio Capone, Optimization Models and Methods for Demand-Side Management of Residential Users: A Survey, Energies 2014, 7, ; doi: /en , pp: Barbato, A.; Capone, A.; Carello, G.; Delfanti, M.; 319 Merlo, M.; Zaminga, A. House energy demand optimization in single and multi-user scenarios. In Proceedings of the IEEE International Conference on Smart Grid Communications (SmartGridComm), Brussels, Belgium, October 2011; pp Elsobki M.S., 1996, DSM Strategy Options An Optimal Based formulation, Proceedings AUPTDECIRED International Symposium, pp , 3-6 June, Amman, Jordan. Goudarzi, H.; Hatami, S.; Pedram, M. Demand-side load scheduling incentivized by dynamic energy prices. In Proceedings of the IEEE International Conference on Smart Grid Communications (SmartGridComm), Brussels, Belgium, October 2011; pp Guo, Y.; Pan, M.; Fang, Y. Optimal power management of residential customers in the smart grid. IEEE Trans. Parallel Distrib. Syst. 2012, 23, Li D., Jayaweera S., and Naseri A., Auctioning game based demand-response scheduling in smart grid, in Online Conf. Green Commun.(GreenCom),Sep.2011 Liang, H.; Zhuang, W. Stochastic modeling and optimization in a microgrid: A survey. Energies 2014, 7, Mohsenian-Rad A., Wong V., Jatskevich J., Schober R., and Leon-Garcia A., Autonomous demand-side management based on game-theo-retic energy consumption scheduling for the future smart grid, IEEE Trans. Smart Grid, vol. 1, no. 3, pp , Dec Office of Energy.Government of Western Australia "Demand Management.» n.d. Web. 30 Nov Prabaakaran, K.; Chitra, N.; Kumar, A.S. Power quality enhancement in microgrid A survey. In Proceedings of the IEEE International Conference on Circuits, Power and Computing Technologies (ICCPCT), Nagercoil, India, March 2013; pp Ravichandran, A.; Malysz, P.; Sirouspour, S.; Emadi, A. The critical role of microgrids in transition to a smarter grid: A technical review. In Proceedings of the IEEE Transportation Electrification Conference and Expo (ITEC), Detroit, MI, USA, June 2013; pp Saad, W.; Han, Z.; Poor, H.V.; Basar, T. Game-theoretic methods for the smart grid: An overview of microgrid systems, demand-side management, and smart grid communications. IEEE Signal Process. Mag. 2012, 29, Samadi P., Schober R., and Wong V., Optimal energy consumptionscheduling using mechanism

334 design for the future smart grid, in Proc. IEEE Int. Conf. Smart Grid Commun. (SmartGridComm), Brussels, Belgium, Oct Strbac, G. Demand side management: Benefits and challenges. Energy Policy 2008, 36, Torriti et al (2010) Demand response experience in Europe: policies, programmes and implementation. Energy 35 (4) U.S. Energy Information Administration, Independent Statistics and Analysis; U.S. Energy Information Administration, Wei-Yu Chiu Hongjian Sun H.V. Poor, "Energy Imbalance Management Using a Robust Pricing Scheme," IEEE Transactions on Smart Grid, vol.4, no.2, pp , June Wei-Yu Chiu; Hongjian Sun; H.V. Poor, "Demand-side energy storage system management in smart grid," 2012 IEEE Third International Conference on Smart Grid Communications (SmartGridComm), pp.73,78, 5-8 Nov Zhao, Z.; Lee, W.C.; Shin, Y.; Song, K.B. An optimal power scheduling method for demand response in home energy management system. IEEE Trans. Smart Grid 2013, 4,

335 Performance Assessment of Various Greenhouse Heating Systems; a Case Study in Antalya M. Tolga Balta 1, Fatih Yilmaz 2 *, Resat Selbas 3 1 Department of Mechanical Engineering, Faculty of Engineering, Aksaray University, 68100, Aksaray, Turkey 2 Department of Electrical and Energy, Vocational Schools of Technical Sciences Aksaray University, 68100, Aksaray, Turkey 3 Department of Energy Systems Engineering, Faculty of Technology, Suleyman Demirel University, 32100, Isparta, Turkey * fatiyilmaz7@gmail.com Abstract: In this his study deal with performance analysis of the greenhouse for heating system. For the heating applications, three options are studied with (a) solar collector vacuum tube, (b) a heat pump, and (c) a wood biomass boiler. The greenhouse in Antalya from Turkey, volume 5000 m 3 and a net floor area 1000 m 2 are consider as case study with indoor temperature is 25 o C and exterior air temperature is 10 o C. An energy and exergy analysis are employed to assess their performances and compare them through energy and exergy efficiencies and sustainability index. Energy and exergy flows are investigated and illustrated. Also, the energetic and exergetic renewability ratios are utilized here along with sustainability index the results show that solar collector-based heating system gives the highest efficiency and sustainability index values. The total exergy efficiencies of the solar collector, heat pump and wood biomass heating systems are 34.37%, 1.72 % and 14.69%, respectively. Keywords: Efficiency, energy, exergy, greenhouse. I. Introduction Many of the future environmental concerns, such as global warming, that come with an extended usage of energy, has increased the importance of energy saving measure and the necessity for an increased efficiency in all forms of energy utilization. According to latest reports on climate change and the need for reduction in carbon dioxide emission must be made efforts in the future to conserve high quality, or primary energy sources (Annex 49, 2007). Energy use reductions or efficiency energy usage, can be achieved by minimizing the energy demand, by rational energy use etc. The reduction of effect environmental impact of any building at sustainable, the residual energy demand must be covered with renewable energy. In this theme integral concepts for buildings with both excellent indoor environment control and sustainable environmental impact are presented (Omer, 2008). Fossil fuels and low efficient equipment has been extensively used in many developed countries. Therefore, energy utilization in an efficient way for space heating and cooling is very important for the development of the energy systems. Furthermore the excess usage of fossil fuels give rise to several environmental problems (Balta et al, 2011). Recently sustainable buildings has been a discussion topic. In order to find further potentials energy utilization, exergy can be effective. The low exergy approach is the main object to constitute a sustainable built environment. Future buildings should be planned to use sustainable energy sources for heating and cooling. One characteristic of these energy sources is that only a relatively moderate temperature level can be reached, if reasonably efficient systems are desired (Schmidt and Juusela 2004). Although there is still considerable energy saving potential in building stock, the results of the finished IEA ECBCS Annex 37, Low Exergy Systems for Heating and Cooling of Buildings, show that there is an equal or greater potential in exergy management (Schmidt, 2009; Annex 37, 2016) The agricultural production, processing and distribution, energy usage is substantially high. Sufficient supply of the right amount of energy along with its effective and efficient utilization is necessary for an improved agricultural production. Also crop yields and food supplies are directly contact to energy usage (Mohammadi and Omid, 2010). Greenhouse technology is mainly to create suitable for crop growth of light, temperature, water, soil and other environmental factors. Today to meet heating and cooling of greenhouses, a variety of heating systems such as steam or hot water radiation system, have been used. The systems are generally heating requirement but the temperature distribution patterns within the greenhouse associated with such systems are readily influenced by the outdoor weather conditions. The heat pump system for greenhouse air-conditioning great potential in employed (Chou et al, 2004). Greenhouses also have important economic potential in Turkey s agriculture. In addition to solar energy gain, greenhouses should be heated during 321

336 nights and cold days. In order to establish optimum growth conditions in greenhouses, renewable energy sources should be utilized as much as possible (Ozgener and Hepbaslı, 2005a) In the last few years, also due to the increasing interest in low temperature heating and high temperature cooling systems, a research cooperation in a working group of the International Energy Agency (IEA) has been formed within the Energy Conservation in Buildings and Community Systems Programme (ECBCSP): Low Exergy Systems for Heating and Cooling of Buildings (IEA, 2016; Hepbaslı, 2001). Recently, a considerable number of studies have been carried out on the exergetic analysis of Lowex heating and cooling systems. Schmidt and Juusela (2004), studied on design, optimization and performance assessment of the buildings using low exergy concept. Balta et al. (2010) performed heating applications and they assessed their performances and compare to them through energy and exergy efficiencies and sustainability index. Also they concluded that solar collector-based heating system gives the highest efficiency and sustainability index values. LowEx approach for comparing energy and exergy flow in a building system heat demand of the building is supplied by fossil fuel plant, ground source and air source heat pump system conducted by Lohani and Schmidt (2010). Ozgener and Hepbaslı, (2005b) studied experimental investigation of the performance of a solar-assisted ground-source heat pump system for greenhouse heating. They calculated that the heating coefficient of performance of the heat pump (COPHP) is about 2.13 at the end of a cloudy day, while it is about 2.84 at the end of sunny day and fluctuates between these values in other times. Greenhouse heating using heat pumps with a high coefficient of performance (COP) analyzed by Tong et al, When the inside air temperature was kept at about 16 o C and the outside air temperature (Tout) ranged between -5 o C and 6 o C, the average hourly COP was 4.0, with a highest value of 5.8. Hepbaslı, (2012), performed low exergy (LowEx) heating and cooling systems for sustainable buildings and societies. The study comprehensively reviews the studies conducted on LowEx heating and cooling systems for establishing the sustainable buildings. The exergy efficiency values of the LowEx heating and cooling systems for buildings are obtained to range from 0.40% to 25.3% while those for greenhouses vary between 0.11% and 11.5%. Studies on LowEx heating and cooling systems have been conducted by several authors (Yucer and Hepbasli, 2011; Shukuya and Hammache, 2002; Shukuya, 2009). In this paper presented performance analysis of the greenhouse for heating system. For the heating applications, three options are studied with (a) solar 322 collector, (b) a heat pump, and (c) a wood biomass boiler. The greenhouse in Antalya from Turkey, volume 5000 m 3 and a net floor area 1000 m 2 are consider as case study with indoor temperature is 25 o C and exterior air temperature is 10 o C. Also the systems of energy and exergy efficiency are determined. II. System description In this study, a greenhouse with a volume of 5000 m 3 and net floor area of 1000 m 2 is consider a case study. For the tomato growing in greenhouse ideal indoor temperature from 20 to 30 o C (South Dakota Extension Fact Sheet 915, 2003). The greenhouse in Antalya from Turkey, indoor temperature is 25 o C and exterior air temperature is 10 o C. The construction materials consist of glass and iron. Fig. 1 illustrates a schematic of green-house for heating in Antalya and Fig 2. shown that the greenhouse front view. Fig. 1 illustrates a schematic of green-house Greenhouse tomato production has attracted much attention in recent years, partly because of a new wave of interest in alternative crops. The attraction is based on the perception that greenhouse tomatoes may be more profitable than the more conventional agronomic or horticultural crops. The popularity may also be due to misconceptions about how easily this crop can be grown. Fig 2. The greenhouse front view For the heating applications, three options are studied with (a) solar collector, (b) a heat pump, and (c) a wood biomass boiler. For the Case A, we consider greenhouse is heated by solar collector system. Its efficiency is assume to be In the Case B, a ground heat pump system (water/water) is used for heat production. And last case, a wood biomass boiler is used for heat production.

337 III. Analysis In this study, the methodology and relations used are based on a pre-design for an exergy optimized building IEA-ECBCS Annex 37 Version 2.3 analysis tool. It main objective were to better understanding exergy flows in building and carried out possible energy usage improvements in building (LowEx, 2015;Hepbaslı, 2012) Firstly the greenhouse project data and boundary conditions are reviewed. The volume of the greenhouse and net floor area V and A N. T 0 is outdoor temperature and T i is the indoor temperature in design conditions. The outdoor temperature is taken reference temperature of Antalya for analysis purpose. The total heat loss of the greenhouse from all surface can be calculated as; Q T = (U i. A i. F xi ). (T i T 0 ) (1) where Q T is the transmission heat loss rate and U i is the thermal transmittance in i surface F xi is their specific temperature correction factor. The greenhouse in not any ventilation. The solar heat rate obtain is calculated as below; Q s = (I s,j. (1 F f ). A w,j. F s,h. g j. F no ) (2) where Q s is the solar heat gain rate, I s,j is the solar radiation, F f is the window frame fraction, A w,j is all window areas, g j is the total energy transmittance of the glazing, F s,h is the possible shading effects of other surrounding buildings and the correction for non-orthogonal radiation on the windowpanes F no. Both are estimated to be 0.9 for all cases. The internal heat obtain can be calculated as; Q 0 = Q n 0. no 0 (3) and Q c = Q c n. A N (4) The greenhouse not used electricity such as for artificial lighting and ventilation. All heat flows, heat loses and internal obtains, occurring inside the greenhouse have to be summed up to create the following energy balance which refers to the first law of thermodynamics. In this study, F p and F q,s are estimated to be 3 and 1 for the heat pump system, respectively. Heat demand rate =overall heat losses rate-overall heat obtain rate Q h=q T-(Q c + Q c + Q s) (5) The heat demands rate is generally state in specific number for be able to compare different greenhouse with each other Q " h= Q h (6) A N The quality factor of greenhouse temperature is calculated as; indoor air T q,air = 1 T 0 T i (7) The exergy load rate can be given by; Eẋ air = T q,air. Q h (8) " T heat = T heat (9) Heater surface a new quality factor for using this temperature can be calculated by; F q,heat = 1 T ref T " (10) heat The exergy load rate at the heater is; Eẋ heat = F q,heat. Q h (11) The distribution system of energy efficiency is not 100%. Heat loss rate energy load calculation given by; Q loss,hs = Q h( 1 η HS 1) (12) Keeping the derivation of the exergy demand rate of the heating system as calculated from; Q loss,dis = (Q h + Q loss,hs).( 1 1) (13) η D The heat loss rate of the distribution system calculated by equation (13). Where η D is the energy efficiency of distribution system. In order to seasonal storage is integrated system, thermal solar power system some of the required heat cover solar fraction F s. The required energy for generator given below; Q GE = (Q h + Q loss,hs + Q loss,dis).( 1 η GE ). (1 F s ) (14) The demand rate on auxiliary energy of the generation system to drive pumps and fans is given by; P aux,ge = p aux,ge. (Q h + Q loss,hs + Q loss,dis) (15) The exergy load rate of the generation is calculated by; Eẋ GE = F q,dis. Q GE (16) For the fossil or non-renewable part of the primary 323

338 energy, the result becomes; E p,tot = Q GE. F p + (P aux,ge + P aux,dis + P aux,hs ) (17) The additional renewable energy load rate is; E R = Q GE. F R + E env (18) The total exergy load rate of the greenhouse; E x tot = Q GE. F p. F q,s + (P aux,ge + P aux,dis + P aux,hs )+ E R. F q,r (19) The details about to analyses maybe found in the ref. (Balta et. al, 2010; Hepbaslı, 2012; Hepbaslı, 2011; Balta et. al, 2011; Ozgener and Hepbasli, 2005a; Ozgener and Hepbasli, 2005b) ψ = E out (20) E in ψ = 1 1 (21) SI The relation between exergy efficiency (ψ) and the sustainability index (SI) as given in eq. (21). IV. Results and discussion In this study, some actual data are selected carefully as: the volume is 5000m 3, the net floor area is 1000m2, the net windows are 50 m 2 south and north sides, and the door are 15m 2 east side entrance greenhouse. While the greenhouse indoor and outdoor air temperatures are 25 and 10 o C, respectively, using eq. (1) and data total transmission loss rate is calculated to be W The calculated results of the analysis are illustrated in Table. 1 and 2, where losses occurred are indicated. In the Table 1. illustrated that energy flows rate heating proses. Total energy input of the considered cases requires W, W and W. The exergy flow rate of heating system is illustrated that given in Table 2. Total energy input of the considered cases requires W, W and W. As mentioned above, the difference between exergy demand of the room and the exergy flow rate of after the greenhouse comes from solar and internal heat gains. Cases Input (W) Table 1. Energy flow rates in W in the subsystems of the three cases studied. After After After After After primarily boiler distribution heating room transmission(w) (W) (W) system (W) (W) After envelope (W) A B C Cases Table 2. Exergy flow rates in W in the subsystems of the three cases studied. Input After After boiler After After After (W) primarily (W) distribution heating room transmission(w) (W) system (W) (W) A B C After envelope (W) It was observed that, while the greenhouse indoor temperature increases from 20 to 34 o C, the total exergy efficiency rate increases. We assumed the outdoor temperature of Antalya is constant 10 o C. The total exergy efficiency of Case A, Case B and Case C are calculated 27.95, and %, respectively. The variation of exergy efficiency of all case with indoor temperature is given in Fig 3. It was observed that with, while the greenhouse indoor temperature increases from 20 to o C, 34 o C the total exergy efficiency of the all case rate increases. The figure is shown total exergy efficiency are calculated to be Case A > Case C > Case C. As a result of this figure the total exergy efficiency that the most effective system Case A. 324 Fig 3. Variation of exergy efficiency of all case with indoor temperature Total energy input are calculated for all cases and illustrated in Fig 4. As a result from this figure when the greenhouse indoor temperature increases from 20 to o C, 24 o C the total exergy efficiency of the all case rate increases.

339 V. Conclusions In this paper has been conducted a study energy and exergy analyses of three case which a solar collector vacuum tube (a), heat pump (b) and wood biomass (c). This cases as driven solar energy and fossil fuels for heating greenhouse with a net area 1000 m 2 and compare their performance exergy efficiency. Some conclusion drawn from result this study present as follows; Fig 4. Variation of total energy input envelope with indoor temperature In this study we also investigated how the exergy efficiencies for three cases with the reference temperature. The influence of changing reference temperature on exergy efficiencies as shown below in Fig 5. The highest exergy efficiency of 34.37%, is obtained for Case A, while the indoor temperature of greenhouse is kept constant at 30 o C. Also it is shown that exergy efficiencies decrease with the reference temperature increases from 10 to 25 o C. The energy demand rate the greenhouse is W. The total exergy efficiencies of the solar collector, heat pump and wood biomass heating systems are 34.37%, 1.72 % and 14.69%, respectively. The sustainability index values for three case with solar collector, heat pump and wood biomass heating systems are calculated as 1.38, 1.01 and 1.14, respectively The highest and smallest total exergy input of heating systems are W W for Cases B and A respectively. While the greenhouse indoor temperature increases from 20 to o C, 34 o C the total exergy efficiency of the all case rate increases. The sustainability index values of the all cases decrease with the increase in the reference environment temperature. Nomenclature Fig 5. Variation of overall exergy efficiencies with reference temperature The sustainably ındex for all cases are calculated and illustrated in Fig 6. which includes the effects of changing reference temperatures on the sustainability index values. As can be seen from this figure, sustainability index of the all cases decrease with the increase in the reference environment temperature from 10 o C to 25 o C. Fig 6. Variation of sustainably index with reference temperature A area (m 2 ) c p specific heat at constant pressure (kj/kg K) E energy rate (W) E x exergy rate (W) F factor f approximation factor g total transmittance I radiation intensity (W/m 2 ) N percentage of equipment resistance n d air exchange rate (1/h) n 0 number Q heat transfer rate (kw) R pressure drop of the pipe (Pa/m) R R renewability ratio SI sustainability index T temperature (K) U thermal transmittance (W/m 2 K) E volumetric flow rate η energy efficiency ψ exergy efficiency Ge generation N netto SI sustainability index q guality p primary energy s source 325

340 References Annex 49, Energy Conservation in Buildings and Community Systems Low Exergy Systems for High Performance Buildings and Communities, homepage: (2007). Annex 37. Energy conservation in buildings and community systems low exergy systems for heating and cooling of buildings. < annex37/> [accessed ]. Balta, M. T. Dincer, I. Hepbasli, A. Development of sustainable energy options for buildings in a sustainable society, Sustainable Cities and Society, 1, 72 80, (2011). Balta, M. T. Dincer, I. Hepbasli, A. Performance and sustainability assessment of energy options for building HVAC applications. Energy and Buildings, 42(8), (2010). Chou SK, Chua KJ, Ho JC, Ooi CL. On the study of an energy-efficient greenhouse for heating, cooling and dehumidification applications. Appl Energy, 77: (2004). IEA, Low exergy heating and cooling of buildings Annex 37. < [ ]. Hepbaslı, A. A comparative investigation of various greenhouse heating options using exergy analysis method, Applied Energy, 88; , (2011) Hepbaslı, A. Low exergy (LowEx) heating and cooling systems for sustainable buildings and societies, Renewable and Sustainable Energy Reviews 16, , (2012). Shukuya, M. Hammache, A. Introduction to the Concept of Exergy For a Better Understanding of Low-Temperature-Heating and High-Temperature- Cooling Systems, VTT research notes 2158, Espoo, Finland, Shukuya, M. Exergy concept and its application to the built environment, Building and Environment 44 (7) (2009). South Dakota Extension Fact Sheet 915, Growing tomatoes in the Home Garden, 2003 ( ticles/fs915.pdf) Tong, Y. Kozai, T. Nishioka,N. Ohyama, K., Greenhouse heating using heat pumps with a high coefficient of performance (COP), Biosystems Engineering 106, , (2010) Ozgener O, Hepbasli A. Experimental investigation of the performance of a solar assisted ground-source heat pump system for greenhouse heating. Int J. Energy Res.29(2):217 31(2005a). Ozgener O, Hepbasli A. Performance analysis of a solar assisted ground-source heat pump system for greenhouse heating: an experimental study. Build Environ 40(8): (2005b). Omer AM. Renewable building energy systems and passive human comfort solutions. Renew Sustain Energy Rev;12: (2008). Yucer, C.T, Hepbasli, A Thermodynamic analysis of a building using exergy analysis method, Energy and Buildings 43 (2 3) (2011). Lohani, S.P., Schmidt, D. Comparison of energy and exergy analysis of fossil plant, ground and air source heat pump building heating system. Renewable Energy, 35, (2010). LowEx., LowEx.Net, Network of International Society for Low Exergy Systemsin Buildings (access date: ) Mohammadi, A. Omid, M. Economical analysis and relation between energy inputs and yield of greenhouse cucumber production in Iran. Appl Energy, 87:191 6 (2010). Schmidt, D. Juusela, M.A. Low-exergy systems for heating and cooling of buildings. In Proceedings of the 21st Conference on Passive and Low Energy Architecture, Eindhoven, The Netherlands, Schmidt D. Low exergy systems for highperformance buildings and communities. Energy Build 2009;41:

341 Cost Risk Modeling for Hybrid Power Generation from Geothermal, Biomass Resources and CSP in Turkey - Southeastern Anatolia and Eastern Anatolia Region Yildirim Ismail Tosun Sirnak University, Faculty of Engineering, Department of Mining Engineering, Sirnak, Turkey * yildirimismailtosun@gmail.com Abstract Crust formed by the heat existing in various depths, hot water containing chemicals, vapors and gases. These resources they contain high amounts of heat energy generated by the formation. To detect thermal energy sources is an issue requiring expertise; geological, geochemical, mineralogical, geological, geophysical surveys are carried out evaluation applied together. Usually hot springs spa and evaluation of district heating Southeastern Anatolia and Eastern Anatolia between 45 C and 125 C, less dense the population is not economically in the region. Uncertainties are taken into account by adding a contingency factor. This approach is simple and it is advantageous to be close to the real data. However, variable and uncertain application of parametric variables feet makes possible reliable risk analysis. This research has examined various risk models for energy production. The most appropriate model is determined by comparison. Drilling wells reports data were analyzed as a probability function were the main data source for this task but also gives equipment are used. Fed with data obtained by comparing the model results and the actual date has been confirmed. The overall objective of the geothermal energy is the presence of a geothermal system can be produced economically. Geothermal energy was started exploration in the vast area to be searched, amended as a result of the research data, the field is narrowed down to investigate to direct heating of regional area. At the same time studies in the cost-benefit criterion it has been considered, and thus became the economic research work. Keywords: risk assessment, stochastic cost estimation, simulation, direct heating simulation. I. Introduction In the past five years about several deep geothermal wells were drilled in the southern Anatolian region. Some of them planned for geothermal technology, resulting in an advantageous transportation and green energy solutions for industrial project options in future (IEA, 2013). One of the main tasks of geothermal energy consultants also cost planning and risk estimate is for the construction process. These estimates of the total wells to be constructed for energy production facility until construction time and unit cost must be based on risk analysis parameters (Lentsch and Schubert A, 2013). applied to the input, just as a human masters the processing of fuzzy control system similar to control it. So people like fuzzy logic and decisions of machine operations can be achieved by using fuzzy sets. In the current applications, so smart grid electricity (intelligent) systems began to record a rapid development. After Italy and Iceland, Turkey ranks third in Europe for the realization of installations and activities in geothermal power plants (Figure 1) (IEA, 2013). Uncertainties are taken into account by adding a contingency factor. This approach is simple and it is advantageous to be close to the real data. However, variable and uncertain application of parametric variables feet makes possible reliable risk analysis (Liu, 1997). Probablistic approach makes easier in use power generation data of geological sources and biomass resources. (IEA, 2007) After the groove of important steps were being taken in this area, so that the control system has come up today in most areas where the application of fuzzy logic (Lin and Lee,1996). Unlike conventional control systems their clients are, without the need for mathematical models of the system, is only set to give the desired output signal 327 Fig. 1: Distribution of geothermal energy investments in Europe This research has examined various risk models for energy production. The most appropriate model is determined by comparison. Drilling wells reports data were analyzed as a probability function were the main

342 data source for this task but also gives equipment are used. Fed with data obtained by comparing the model results and the actual data have been confirmed. Model data trends observed in the average of approaches were examined and evaluated as a very well-parametric analysis of costs. It has been expanded by recent model costs. This risk analysis and plant construction, investors can provide insurance companies for risk assessment and decision-makers geothermal wells. Thus, the risk analysis will help in calculating the correct budgeting and insurance premiums. I.1. South Eastern and Eastern Anatolian Geothermal Sources Depending on the volcanic and tectonic activity in Eastern and Southeastern Anatolia there are several geothermal areas to be considered as the direct heating or energy production purposes (MTA,1987); In Van-Ercis field the temperature of the water is the about C, in the Diyarbakır-Cermik field the temperature is 51 C, and in the Urfa-Karaali the temperature is 49 C as the shallow reservoirs. Agri-Diyadin (78 C), the Bitlis-Nemrut field (59 C) are the geothermal hot springs recorded in the Eastern Anatolian Region. The hottest springs recorded in the Southeastern Anatolia is in the south Diyarbakir, Cermik located in the geothermal areas and shallow m depth in 51 C, flow rate of 21 (l/s) used as central heating and pumped for irrigation. A part of water cooled is piped to Dicle University Physical Therapy and Rehabilitation Center for spa facilities and it utilizes warm water. Mardin Germav water supply is 63.5 C and 15 (l/s) with the flow rate. Hot water from two pools of private management, are used as medicinal water. There are 5 pieces of Siirt Billoris geothermal sources. The temperature of the water from wells 40 - total flow rate is between 55 C (l/s). Sanliurfa Karaali 7 drilling results conducted in geothermal field, 5 wells have passed activities C and the flow of resources is arasınn (l/s) it varies. Batman's Kozluk-Taşlıdere Holi geothermal spa water temperature is 83 C and in the flow of 16 (l/sec) and is evaluated as thermal springs from the source and greenhouse heating. Geothermal field in the province of Sirnak Güçlükonak field, and flow at 73.5 C 12 (l / s) is used as the water source in the spa treatment. II. Capital Investment Cost Risk Modeling II.1. Geothermal Drilling Drilling needs feasibility study and performed the last exploration on where the drilling will be opened and greatly followthe data and the data result can be boring, drilling logging index, lost time parameters and the investment needs the processing logging, bore placing and cementing of the well of coatings. Well head, the construction of the heel should be carried out without interruption, depending on the time and depth. The construction of the borehole takes place in 328 two parts with a variety of applications. This Figure 2 varies depending on the time and depth as seen in the graph as shown horizontally and the model does not include the waiting time, this linear relationship. As given in Fig 3, the investment cost model vs depth is considered to be time-dependent progression throughout the entire process. Exponentially increased cost was calculated due to the boring difficulties under 3000m below. Drilling Capital Cost, MillionTL Yatırım Capital Cost Maliyeti y = e x R² = Drilling Well Depth,m Fig. 2: Drilling depth chart and Geothermal Well Investment Risk Drilling Capital Cost, MilyonTL y = 0.005x R² = NormalDağılım Distribution 0.3 of Capital Cost Drilling Well Depth,m Fig. 3: Drilling depends on the well depth chart and Normal Distribution curve for Capital Investment Risk II.2. Cost Risk Modelling for Agricultural Biomass Waste Potential of Turkey Considerable research on coal combustion has been conducted over the years, but the waste combustion results are widely dispersed because of the complex chemistry of waste (TKİ 2009, TTK 2009). Time related coal combustion modeling assumes basically first-order kinetic equations, or less sensitive for heating rate (Bell et al, 2011, Kajitani et al 2011). It is basically depend on the coal properties but also cover to some extent, the effect of heat-and-mass transfer phenomena (Jess et al,2010, Schultz et al, 2011). Fluidized bed combustion is preferred for clean

343 emissions in the unit (EIA, 2007). The clean emission from biowaste and coal co-combustion could be managed in NOx and SOx due to low combustion temperature (Fig 4). The potential biowastes projected in Southeastern Anatolian region was mainly the maize slush and animal manure, the digested biowaste. The proximate analysis and calorific values are given in Table 1. Fig. 4: Fluidized Bed Combustion of Coal and Biomass for power generation. Tab. 1: Combustible bio-waste proximate analysis Weight(%) Wood Waste Trash Cow Waste Poultry Waste Corn Waste Moisture Ash Fixed Carbon Volatile Matter Calorific Value (kcal/kg) Combustion Weight, TGA,% y = ln(x) y = ln(x) y = ln(x) y = ln(x) Time, s 10mm mm Şırnak Şırnak Municipal ÇöpBiyoKütle Waste, Biomass 5 5 mm Cow SığırWaste Küspe 5 5 mm mm TavukPoultry Waste Küspe 5 Mısır mm Maize Sapı Slush Fig. 5: TGA combustion weight of Biowaste types and combustion rate change. Alfa Makine offered semi-mobile municipal waste incinerator for electricity regarding even the biowastes as shown in Fig 6. The calorific values of bio wastes changed with the moisture content of the waste type. The most proposed bio waste was the corn waste, having a calorific value of 3780kcal/kg. III. Projected Results and Discussion 50 g samples were dried waste is subjected to combustion in the laboratory TGA analysis. The test results are shown in Figure 7. In Figure 7, the reactor temperature above 900 C with respect to the amount of combustion is after pyrolysis. This increase in the burning rate of Sirnak landfill waste 28%/min cow pulp 52%/min and chicken waste at 53%/min in the corn stalks were 68%/min. Burning fuel as coal dust and combustion kinetics of Şırnak asphaltites used 10% sample weight ratio of the mixture is reduced by 25%. The combustion experiments stoker boiler is used for and obtained similar results. Biomass waste and coal types are co fired in stoker or fluidized bed at 900 C and toxic gas emissions are secondly fired inthe secondary chamber by gas at 1000 C and even alkali matter are added into the combustion chamber (Fig 4). While the lime addition into the chamber at weight rate of 10% at 850 C combustion rate values are shown in Fig Fig. 6: Integrated CSP and mobile Biowaste and biogas combustion units and ORC power generation. The capital cost values of units in waste combustion and power plant for both mobile plant at the capacity of tons/year and the integrated plant at the capacity of tons/year are determined by firm's offer and calculations. The cost values of combustion and power plant are given below Tab 2. For Integrated facility the capital investment cost of 500 thousand tons/year capacity was 51 million$, while 1 million tons/year capacity for exit doubled. Already region for high-capacity incinerators are not considered due to the impossibility of obtaining funds is not feasible. For mobile tons/year capacity plant, depending on the companies' unit costs was determined as 10 million$ (as given in Table 2). Tab. 2: The Capital costs of Mobile and Integrated Biogas power generation

344 Unit Cost, $ Mobile tons/y Integrated tons/y Biowaste bins: Trash bin: $ Waste mix bins: Pressed trash bin: $ Coal fine bin: $ Feeder Stoker Belt: $ CoalBrulors 2: $ Biowaste Auger feeder: $ Biowaste drying chamber: : $ ALFA KAZAN combustion stoker mm: $ Secondary combustion Brulors : $ Secondary Combustion Chamber (ALFA KAZAN) : $ Ash Auger 2: 2*50 000$ Bio waste shredder -10 mm Adet: $ Gaz Cyclones 4: $ Ash Dispose Belts: 12 $ Centrifuge Dust Separator 2 : * $ Combustion Fan Filter bag units : 12* $ Dust Collector Units: 3* $ Alkali reactor 6 : 6* $ Alkali ponds 3: 3* $ Alkali pumps 4: : 4* $ CAT Excavator 2: 2* $ FORD Lorry 30 TON 3: * $ Automation Control System Field Cost Engineering Project Power Plant TOTAL :$ Mobile plant and integrated plant operating costs were calculated based on the present prices. As Table 2 also given mobile plant labor, it will provide advantages in terms of reactive maintenance. Mobile plant operating cost approximately 25 TL/ton is defined as garbage. This integrated facility cost rose to 63TL/ton with landfills. Mobile plant and integrated plant operating costs and energy production (70% and 60% thermal efficiency fuel efficiency) was calculated to be connected. mobile plant as given in Figure 7, while in a period of their capital investment in 22 months, after a period of 36 months will generate more revenue for the integrated plant operating costs will be advantageous investment capital back to paying (Figure 7). IV. Investment Risk Modeling of Power Generation from Geothermal and Biomass Model data trends observed in the average of approaches were examined and evaluated as a very well-parametric analysis of costs. It has been expanded by recent model costs. This risk analysis and plant construction, investors can provide insurance companies for risk assessment and decision-makers geothermal wells. Thus, the risk analysis will help in calculating the correct budgeting and insurance premiums. The installed capacity of the planned plant was about 2 million kwh/year and flow rate of water in the entire unit in energy production was 220 l/s. Drilling cost risk values calculated regarding 4 well drills at averagely different rock types and depths are given in Table 3. Capital Cost, Revenue of Bio Power Plant.TL Total Revenue Total Profit Capital Cost Month Fig. 7: Change of the capital Cost and Revenue of Mobile/Integrated Biomass Power Plant vs month. Tab. 3: Correlation and variable values in Drilling depth with the cost of investment. RİSK Point Weak Rocks Mid Rock Hard Rock 500m 1500m 2500m Depth,m Advance Rate Drilling Period Investments Risk Risk Error Correlation Coefficient The cost calculation of the plant, Calculation of unit cost of the facility, Calculation of the investment costs of the facility at which it will go into production, Plant operating costs and the calculation of the annual income, TV C is the total cost, T x is tax, F is the interest, O m&o is maintanance cost, D is share rate, cm is capacity factor, K is the unit capacity. As given below; TV C = T x + F + O m&o + D (1) Q(n) = 8760xCF(n)xK (2) The cost need to be calculated in three stages. R(n) = Q(n)xP(n) M m=1 1/(1 r) M m (3) E(0) = (1 f) M m=1 Cm(1 + r) M m (4) 330

345 L(0) = f M m=1 Cm(1 + r) M m (5) u(x; t; Ө) = n i=0 u(x, t) + ɸ(x; t; Ө). e tiθ (6) where R is the revenue, Q is the capacity, P is the sale price, r is the interest rate, m is month, n is the integer of month, E is investment cost, f is debt rate, cm is capacity factor, L is the debt, u cost function, t is time, Ө is the hybrid unit parameter. Gaussian normal distribution of risk probability values defines the value of the investment in data-intensive midpoint. Drilling cost estimation is obtained as given by the following equation (Table 3). u(x; t; ϼ) = N n u(x,t) u(x; t) i=0 (7) u(x,t) generation, so that every units may not decrease enthalpy yields in series generation, using geothermal, biomass and biogas combustion and CSP solar units. The system basicly is shown in Fig 8. Especially low heat sources may not be feasible in power generation, but hybrid parallel operation shoul be advantageous in the Southeastern Anatolian region. The projected Batman and Siirt case plants were considered regarding the potentials of biomass/csp and geothermal sources/csp, and the ORC proposed plant parameters using hot oil (or R112 liquids) are given in Table 4. For Batman and Siirt case potentials of biomass/csp and geothermal sources/csp, the cost values of proposed 35 MW hybrid power plants are given in Table 5. The ORC plant has planned for hybrid parallel power Fig. 8: ORC Use for Low Heat Geothermal and Biomass sources in energy and risks of Capital Investment costs in Turkey Tab. 4: The planned values of the variables of projected Organic Rankine Cycle for the Batman and Siirt Geothermal and Biomass energy sources Organic Rankine Cycle Variables Batman Siirt Batman Siirt Biomass Geothermal Geothermal Biomass Geothermal hot water temperature, o C 120 o C 95 o C 135 o C 145 o C Geothermal hot water flow rate in kg / s Organic condenser exit temperature, 92 o C 74 o C 92 o C 94 o C Organic mass flow rate kg / s Organic return rate, kg/saat,% Mass flow rate of water consumption kg/h Organic Turbine Output pressure drop,bar Power conversion efficiency,net h Condenser total energy MWh Power cycle / TES pump power MWh Gross electricity production MWh Net Electricity production MWh Organic Thermal Power Generation MWh Thermal Power Generation MWH Total pipe heat loss MWh Return cold geothermal waste heat loss MWh Organic spin cycle heat loss MWh Total Thermal Loss MWh External heat consumption MWh

346 Tab. 5: Organic Rankine Cycle Variables for Geothermal and Biomass energy capital cost risk Projected Cost and Revenues Batman Geothermal Siirt Geothermal Batman Biomass Siirt Biomass Cost Risk Batman Cost Risk Siirt Net Electricity kwh 137,000, ,000, ,000, ,000, Average Annual 0,26 0,26 0,26 0, Sale TL Production Cost 0,21 0,21 0,21 0, nominal Production Cost 0,11 0,11 0,11 0, actual Return rate,% Annual Net Profit 22,000,000.TL 21,000,000.TL 22,000,000.TL 21,000,000.TL 3 3 Calculated Sale price change,% Calculated debt rate,% Capacity factor Land cost 1,6 1,6 1,6 1,6 2 2 System performance factor Toatal field, acre Cogeneration Sellective Sellective Sellective Sellective 1 1 Average Risk 3 3 Reservoir characteristics of low heat geothermal resources provided over 6 point in risk analysis, the presence of hybrid biomass and biogas combustion became a great support in power generation even waste sources evaluated. However, the hybrid power plants need more capital and complex power generation units due to heat recovery and absorption; so that specific oils or liquids high heat conductive materials are preferred. Aditionally, possible heat sources, storing, availability of these resources and logistics, price should be determined prior to parametric cost analysis. Assessment made after the Table 5 investment cost values and energy revenue and energy cost analysis vs steam flowrate are shown in Fig 9 for hybrid plants. Cost values per electricity kwh increased with CSP hybrid plants over 2$/kwh with probability approach, and the best approach cost risk analysis hybrid plant returns in 90 months are very critical in terms of interest rates and taxes. The use of the ORC unit outcomes high cost of energy production and increase the capital cost of every hybrid unit. Fig. 9: The capital investment for ORC Use for Low Heat Geothermal and Biomass sources in energy and cost risks of Hybrid power plant in Turkey 332

347 V. Conclusions Reservoir characteristics of geothermal resources in addition risk analysis, the presence of natural mineral waters, the investigation of possible heat sources, development, protection, to be eligible on these resources and rights transfer, are discussed in parametric cost analysis management in the most efficient manner compatible with the environment. Assessment made after the necessary cost risk analysis (the number of wells, depths, locations determined costs may be produced suitable ORC power with geophysical work. Cost parameters probability approach, Gaussian, Markovian, and the best approach cost risk analysis are discussed. The use of the ORC unit outcomes high cost of energy production and increase the risk of opening the analysis of deep drilling. In addition, it is another parameter that increases the cost of the environmental risk of water contamination. Benefaction from biowastes in the various parametric combustion systems, in order to receive clean energy and higher enthalpy yield could be generated in low temperature combustion. It is also advised that the high amount of formation of flue gas will be managed at higher combustion temperatures over 700 o C and extracts more environmental friendly gaseous products. Biomass combustion carried out with Şırnak asphaltite in 30mm size distribution showed sufficient enthalpy yields from corn biowaste between to o C and even other type of biowastes showed similar trend, the higher enthalpy yields of % at lower combustion temperatures. In the research works production of clean energy with the design of the addition of high-quality coal biowaste mixtures are processed and biomass fuels could be an alternative clean fuel sources. Clean energy sources may be supplied in South East Anatolian region in Turkey. Hence, those clean alternative resources will further enhance the industrial development in the region. References Akpınar, N, Şen, M, Kentsel katı atıklardan enerji üretimi, Enerji Enstitüsü Anonymous a, 2015, Mobile incinerators, p, ATİ Şirketi Anonymous b, 2015, Yakma Kazanları, Alfa Kazan ve Makine AŞ,Ankara Anonymous c, 2015, Kalina Cycle, enerji.com.tr, İmparator Enerji, GeoPower, İstanbul Anonymous d, 2015, Akışkan Yataklı Yakma Kazanı, Mimsan A.Ş., İstanbul Anonymous e, 2015, Anonymous f, 2015, Anonymous g, 2015, systems/ mobile_systems.htm Bell D.A. Towler B.F., Fan M., 2011, Coal Gasification and Applications, ISBN: , Elsevier Inc., Oxford Cherubini, F. Bargigli, S. Ulgiati, S. 2009, Life cycle assessment (LCA) of waste management strategies: landfilling, sorting plant and incineration, Energy, 34, pp Çakal, G.Ö. H. Yücel, A.G. Gürüz, 2007, Physical and chemical properties of selected Turkish lignites and their pyrolysis and gasification rates determined by thermogravimetric analysis, Journal of Analytical and Applied Pyrolysis, Volume 80, Issue 1, Donskoi, E.& McElwain, D.L.S., 1999, Approximate modelling of coal pyrolysis, Fuel, 78, pp IEA, 2007, IEA Coal Research Ltd, Clean Coal Technology Report, ( A.J. Minchener and J.T. McMullan) IEA, 2012, World Energy Outlook Jess A, Andresen A-K. Influence of mass transfer on thermogravimetric analysis of combustion and gasification reactivity of coke. Fuel.; doi: /j.fuel Kajitani S, Suzuki N, Ashizawa M, et al. CO2 gasification rate analysis of coal char in entrained flow coal gasifier. Fuel. 2006;85: Kajitani S, Suzuki N, Ashizawa M, et al. CO2 gasification rate analysis of coal char in entrained flow coal gasifier. Fuel. 2006;85: Karakaya, İ.,2008, İstanbul için stratejik kentsel katı atık Yönetimi yaklaşımı, Yüksek LisansTezi, İTÜ FBE Çevre Müh.Böl. Kreith, F Tchobanoglous, G 2002, Handbook of Solid Waste Management Lentsch D, Schubert A,, 2013 "Risk Assesmment for Geothermal Wells- A probabilistic Approach to Time and Cost Estimation CRC Transactions, Vol 37, p Lin C.T. and Lee, C.S.G., 1996 Neural Fuzzy Systems: A Neuro-Fuzzy Synergism to Intelligent Systems, Prentice Hall Liu, C., 1997 Intelligent system applications to power systems, IEEE Computer Applications in Power, Vol.10, No.4, pp , October. 333

348 Liu, G., Benyon, P., Benfell, K.E., Bryant, G.W., Tate, A.G., Boyd R.K., 2002, The porous structure of bituminous coal chars and its influence on combustion and gasification under chemically-controlled conditions, Fuel, 79, pp Mahmoudi, S.. Gholamian, E. Zare V, 2015, Exergy analysis of a new configuration of trigeneration system based on biomass gasifier, ECRES2015, Antalya Richard A. Denison, John Ruston, 1990, Recycling and Incineration: Evaluating the Choices TTK, 2009, The Turkish Ministry of Energy, Energy, Dept., Hard Coal Report Wei-Biao F., Quing-Hua, W. 2001, A general relationship between the kinetic parameters for the gasification of coal chars with CO2 and coal type, Fuel Processing Technology, 72, pp Wiktorsson L.P., W. Wanzl, 2000, Kinetic parameters for coal pyrolysis at low and high heating rates a comparison of data from different laboratory equipment, Fuel, 79, pp. 701 Ron Isaacson, 1990, Methane from Community Wastes (Elsevier Applied Biotechnology Series) Schora F.B., 1967, Fuel Gasification, 152 nd Meeting of American Chemical Society, New York Schurtz R, Fletcher TH. Pyrolysis and gasification of a sub-bituminous coal at high heating rates, 26th Annual Int Pittsburgh Coal Conf, Sept , Shadle LJ, Monazam ER, Swanson ML. Coal gasification in a transport reactor. Ind Eng Chem Res. 2001;40: Shadle LJ, Monazam ER, Swanson ML. Coal gasification in a transport reactor. Ind Eng Chem Res. 2001;40: Sharma A, Saito I, Takanohashi T. Catalytic steam gasification reactivity of hypercoals produced from different rank of coals at o C. Energy & Fuels. 2008;22: TAM, 2014,Tarım ve Köy İşleri Bakanlığı İstatistikleri, TEFM, 2008, Orman biyokütlesinden yakıt ve enerji üretimi, (Kahveci, O) TC. Çevre ve Orman Bakanlığı Orman genel müdürlüğü TEFM, 2009, Orman Genel Müdürlüğü nde Biyoenerji Konusunda Yapılan Çalışmalar, Orman Genel Müdürlüğü, Biyoenerji Çalışma Grubu, Orman ve Enerji, Ankara, TKI, 2009, The Turkish Ministry of Energy, Energy, Dept., Lignite Coal Report Tosun YI, 2012, Semi-fused Salt-Caustic Mixture Leaching of Turkish Lignites - Sorel Cement Use for Desulfurization, Proeedings of XIIIth International Mieral Processing Symposium, Bodrum, Turkey. Tosun YI, 2012, Semi-fused Salt-Caustic Mixture Leaching of Turkish Lignites - Sorel Cement Use for Desulfurization, Proeedings of XIIIth International Mieral Processing Symposium, Bodrum, Turkey. TSI,2014, Türkiye İstatistik Kurumu Verileri, 2014, 334

349 Sustainable Re Use of Dairy Cow Manure as Bedding and Compost: Nutrients, Pathogens and Self-Heating Potential from Increased Residence Time in a Tumbling Composter Joe Ackerman 1*, Ehsan Khafipour 2, Nazim Cicek 1 1,2 University of Manitoba, Department of Biosystems Engineering, E2 376 EITC, 75 Chancellor s Circle, Winnipeg, R3T 5V6, Canada * joe.ackerman@gmail.com Abstract Dairy farm operations rely on a continuous supply of bedding material for cow comfort and hygiene. The re use of manure for this purpose has become possible after solids separation of the nutrient rich liquid stream and 24 hr processing of the solids through a tumbling drum composter. The finished solids produce superior bedding to fresh straw but to export these solids off-farm is only possible if certain pathogens are destroyed. In addition, the compost must be finished to prevent re-heating if bagged or piled. These two aspects were investigated on a Canadian dairy farm by extending the residence time of solids in the tumbler from 24 to 38 hrs and persistent living pathogens in the finished solids as well as the self-heating potential were assessed. Results indicated a significant change in pathogen densities as well as a change in surviving microbial communities. Of pathogens harmful to humans, none of the 132 composted samples tested positive for Salmonella and only 1 composed sample was positive for E. coli O157:H7. Although the composting process did not impact the total number of E. coli, the number of pathogenic E. coli in the recovered bedding was numerically reduced. Depending on the methodology used, 1.5% to 9.8% of samples were identified as Mycobacterium avium subspecies paratuberculosi (MAP) positive, (a pathogen harmful to dairy cattle), indicating the composting process did not reduce the prevalence of MAP. Increased retention time did not reduce self heating temperature whereas curing compost in piles for 4 weeks reduced compost reheating from 22.5 C above ambient to 7.7 C. Changes in the nutrients of manure solids from 24 hr of tumbling were confined to an increase in all soluble forms (ammonium, soluble phosphate, soluble potassium) except nitrate, which decreased. Keywords: Compost, nutrients, dairy bedding, pathogen reduction I. Introduction The research was driven by interest in the quality of compost produced from a bedding recovery unit (BRU) processing dairy manure for re-use as bedding by dairy cows. Aspects of inquiry related to the thermal stability of solids under normal and extended BRU residence times, the pathogen load presence and reduction from both 24 hr and 38 hr residence times and the effects of letting the solids cure in piles under normal farm conditions for several months. The underlying concern with the reheating potential of digested solids is the ability for them to be stored safely without risk of overheating (combustion hazard) or continuing to give off gasses. The BRU tumbles recovered solids for approximately 24 hrs during which time it reaches temperatures over 55 C, indicating a high level of microbial activity. Most compost systems maintain these temperatures of >60 C for 3 days or longer and then undergo a finishing stage of storage for a month or more to produce compost with stabilized solids. Resolution to these questions are important if the processed manure solids were to be packaged as compost and awareness of the pathogen load is important if the product is handled by the general public or exposed to other dairy herds. The nutrient content was also analyzed to determine possible changes during manure separation and composting. The BRU separates dairy manure into solid and liquid streams, the liquid is stored to fertilize farm fields and the solids are tumbled in a large drum (3 m di and 12 m long) where temperatures normally reach 55 to 70 C. The residence time in the drum is normally 24 hrs but can be extended somewhat without impeding the operation of the entire system. It was determined that 38 hrs was the maximum BRU residence time without overfilling the tumbler of causing a manure backlog. After tumbling, it is common for compost operations to set piles of solids to cure for several weeks or months to allow equilibration of microbial communities and possibly further pathogen destruction. Piles of processed BRU solids were created and stored for 4 to 8 weeks in winter and summer conditions. Three replicate piles were created and stored in 1) an unused barn in February, 2) in a covered 3 sided shed in March, and 3) on an open concrete pad in June. Piles were approximately 1 cubic meter in volume except for the single June pile which was 3 cubic meters in size. II. Setup The dairy manure system consists of a raw manure collection tank from where mixed manure is pumped 335

350 into a screw press that separates the liquid and solid streams. The solid stream is conveyed into the drum dryer/composter where it is tumbled for 24 or 38 hours, depending on the chosen retention time. The liquid stream is pumped into a separate tank where it is collected before pumping into an outdoor lagoon. Both tanks had internal mixers that were activated for 5 minutes before triplicates samples were taken with a 12 foot extension handle dipper. Solids were collected as they exited the screw press and also the end of the tumbler drum. Samples were well mixed in a pail before they were sub-sampled (triplicates) and either flash frozen in liquid nitrogen for microbial community composition or chilled on ice for pathogen loading and nutrient analysis. Freshly tumbled solids and piles of cured solids were collected in separate 40 L tubs; the cured solids were sampled from various depths and locations and a composite sample of at least 10 samples were combined, mixed and subsampled for both pathogen analysis and selfheating potential. II. 1. Solids Reheating Potential Stability as it relates to compost is not clearly defined but there is agreement that stability is based on the amount of potential biological activity in the composted material. There are three ways to observe and measure this activity and many methods have been developed to do each of these: measuring a compost tendency to self-heat (self-heating test), measuring the evolution of carbon dioxide (CO2), and measuring the rate of oxygen uptake. Standards have been developed in Canada by the Canadian Council of Ministers of the Environment (CCME), and in the United States by the United States Environmental Protection Agency (U.S. EPA) to determine compost stability based on measurement of these three variables. The CCME describes a compost as mature and stable if, The temperature rise of the compost above ambient temperature is less than 8 C, while the U.S. EPA has a three tiered maturity index that describes compost having a net rise of 7-8 o C as very mature, a net rise of o C as mature, and a net rise above over 20 o C as immature. We used the CCME self-heating guidelines (<8 C) as an indicator of the bedding material s stability. II. 2. Self-Heating Test After initial tests indicated poor reproducibility in 20 cm cubic Styrofoam containers, insulated 1.5 L stainless steel wide-mouthed thermoses were equipped with a minimum/maximum thermometer and loosely packed with processed BRU solids and operated without lids to ensure aerobic conditions (Fig. 1). The solids moisture content was adjusted if total solids were higher than 40%. Total weight of solids was adjusted to ensure equal density between replicate dewars. Daily maximum temperature was recorded inside as well as ambient room temperature. The experiment was conducted over a period of 4 to 10 days, or until the maximum temperature inside the 336 flask declined for two days in a row. Fig. 1: Replicate dewars for conducting compost selfheating tests. III. Nutrient analysis Total solids (TS) are a measure of dry mass content of manure while volatile solids (VS) is a measure of combustible material within that dry mass. Twenty to 40 grams of sample were dried (24hrs) at 103 C to determine moisture loss and percent TS. The dried material was fired at 550 C for 2 hrs and cooled in a desiccator to determine percent VS. Total P and K were determined by digestion with nitric acid/microwave and analysis with an inductively coupled plasma optical emission spectrometer (ICP- OES). Soluble P and K were determined by water addition and filtration of raw sample and analysis with ICP-OES. Total nitrogen was determined with the Kjeldahl process and the automated distillation method. Ammonium nitrogen and nitrate nitrogen were extracted with potassium chloride and the phentate method and segmented flow cadmium reduction method. All results were collected on a wet sample basis and converted to a dry weight basis as required for meaningful comparison of carbon: nitrogen ratio (C:N) to literature standards for compost. Optimum compost C:N ratio satisfies the protein requirement of the growing microbial community (nitrogen) as well as the energy requirement (carbon). A good C:N ratio for an initial compost mix is 30 and a finished compost is lower (10 to 15), due to the consumption of carbon in the composting process. Calculation of this ratio for samples in this experiment were based on the determination of total organic carbon from VS (Jimenez & Garcia 1992; Cornell compost website, see Appendix for formulae) and conversion of TKN data to a dry wt basis. III. 1. Microbial pathogen load The effect of composting on the microbiota profile and pathogen load of recovered bedding material (RBM) was evaluated. DNA was extracted from manure (MAN), liquid stream (LS), solid stream undigested (SSU) and solid stream digested (SSD) samples (total

351 n=132) and consequently subjected to bacterial 16S rrna gene sequencing for microbial community analysis. The dynamic of community from winter and spring into summer were compared by combining samples from available seasonal data collected during this study. III. 2. Library Construction and Illumina Sequencing Library construction and Illumina sequencing were performed as described by Derakhshani et al. (2014). In brief, the V4 region of 16S rrna gene was targeted for PCR amplification using modified F515/R806 primers (Caporaso et al., 2012). The reverse PCR primer was indexed with 12-base Golay barcodes allowing for multiplexing of samples. PCR reaction for each sample was performed in duplicate and contained 1.0 µl of pre-normalized DNA, 1.0 µl of each forward and reverse primers (10 µm), 12 µl HPLC grade water (Fisher Scientific, ON, Canada) and 10 µl 5 Prime Hot MasterMix (5 Prime, Inc., Gaithersburg, USA). Reactions consisted of an initial denaturing step at 94 C for 3 min followed by 35 amplification cycles at 94 C for 45 sec, 50 C for 60 sec, and 72 C for 90 sec; finalized by an extension step at 72 C for 10 min in an Eppendorf Mastercycler pro (Eppendorf, Hamburg, Germany). PCR products were then purified using ZR-96 DNA Clean-up Kit (ZYMO Research, CA, USA) to remove primers, dntps and reaction components. The V4 library was then generated by pooling 200 ng of each sample, quantified by Picogreen dsdna (Invitrogen, NY, USA). This was followed by multiple dilution steps using prechilled hybridization buffer (HT1) (Illumina, CA, USA) to bring the pooled amplicons to a final concentration of 5 pm, measured by Qubit 2.0 Fluorometer (Life technologies, ON, Canada). Finally, 15% of PhiX control library was spiked into the amplicon pool to improve the unbalanced and biased base composition, a known characteristic of low diversity 16S rrna libraries. Customized sequencing primers for read1 (5 -TATGGTAATTGTGTGCCAGCMGCCGCGGTAA- 3 ), read2 (5 - AGTCAGTCAGCCGGACTACHVGGGTWTCTAAT- 3 ) and index read (5 - ATTAGAWACCCBDGTAGTCCGGCTGACTGACT- 3 ) were synthesized and purified by polyacrylamide gel electrophoresis (Integrated DNA Technologies, IA, USA) and added to the MiSeq Reagent Kit V2 (300- cycle) (Illumina, CA, USA). The 150 paired-end sequencing reaction was performed on a MiSeq platform (Illumina, CA, USA) at the Gut Microbiome and Large Animal Biosecurity Laboratories, Department of Animal Science, University of Manitoba, Canada. The pathogen load was evaluated by qpcr for absolute quantification of 10 pathogens that are considered important in public health (Escherichia coli O157:H7, Salmonella enterica, and Listeria monocytogenes), are associated with environmental mastitis (generic Escherichia coli, Streptococcus uberis, Klebsiella pneumoniae) and contagious 337 mastitis (Staphylococcus aureus, Streptococcus agalactiae, Arcanobacteruim pyogenes), or are important indicator of dairy herd health (Mycobacterium avium subspecies paratuberculosi (MAP)). The required bacterial isolates were either present at the Gut Microbiome and Large Animal Biosecurity Laboratory of the Department of Animal Science, University of Manitoba, or obtained from Diagnostic Services of Manitoba, Winnipeg, Manitoba. Frozen stocks of pathogenic isolates were placed onto specific mediums for each strain, and cultured aerobically or anaerobically for specific period of time as required for the growth of each strain. Five colonies were then isolated and further inoculated into a specific liquid medium for each strain. For bacterial cfu (colony forming units) counts, ten-fold serial dilutions were made to obtain the desired concentration of bacteria for spiking. DNA was extracted from certain dilutions and used to generate the standard curves for absolute quantification of strains of interest using qpcr. The qpcr was performed using primers and probes previously described and published in the literature. IV. Results and discussions Stability of separated solids The use of standardized insulated flasks for the self heating tests gave more reliable results than the Styrofoam boxes used in the past, however, high variability was found between batches and sometimes between dewars using the same batch of solids. Tested solids always underwent self-heating, although the maximum temperature reached and the time taken to reach that temperature differed. Generally speaking, for solids with a 24 hr retention time in the BRU, maximum self-heating occurred after 50 to 140 hrs in the dewar and temperature increase ranged from 8.1 to 36.6 C. Although graphs are usually easier to view, the data has been presented in a table so that the variation between trials and between replicates can be clearly seen (Table 1). Most trials consisted of replicate dewars of a single batch, however in some cases (i.e., 24 hr Uncured) three batches were tested and yielded a large range of maximum temperature levels. It is recommended that multiple dewars be used to test multiple batches to increase confidence of the true mean temperature increase. Due to delays in BRU adjustment to 38 hrs and complications in manure management at that retention time, only one batch was available with extended retention time. Piles created for curing were inadvertently removed after 4 and 8 weeks, preventing testing of longer term curing. Of the data available it should be noted a significant decline in self heating temperature occurred after 4 weeks of curing, whether this was done in the winter or summer. A reduction from a mean of 22.5 C temperature increase (mean of 8 replicates of 3 trials) to 9.5 C and less was observed. The 4 week curing resulted in a 7.7 and 8.5 C maximum temp increase while the 8

352 week curing had a 9.5 C max temp increase. The fact the 8 week curing resulted in a higher temperature than the 4 week curing is likely due to a low sample size and large variance. More replicates would more closely identify the temperature range to be expected with 8 weeks of curing. The natural variability found in compost batches and the delayed or lack of heating response sometimes observed in compost microbial communities was also evident with the 4 week pile cured in the summer: the compost took 5 to 10 times longer to reach maximum temperature as fresh BRU solids (Fig. 2 shows the mean of replicates for three separate batches). Extending the retention time in the BRU rotating drum from 24 hrs to 38 hrs did not have the expected results, as maximum temperature increase was 34.6 C, which was at the upper temperature range of the trials with a 24 hr retention time. Four week curing of these solids showed a significant drop (17 C) in temperature increase. Tab. 1: Results of self heating tests on Bedding Recovery Unit solids with 24 and 38 hr retention time. The results of self heating tests of piles of solids stored for 4 and 8 weeks are also included. A temperature rise of less than 8 C classifies solids as mature compost (CCME). 24 hr retention time Max temp C Hours of incubation 38 hr retention time Max temp C Hours of incubation Uncured Mean week curing (winter) Mean week curing (winter) Mean weeks curing (summer) Mean Fig. 2: Comparison of self heating time response for solids from the Bedding Recovery Unit. The July sample was dry (66% TS) and water was added before incubation. Total Solids and Volatile solids Analysis of the manure, liquid and solid streams and separated solids gave general indication of solids content (TS) and volatile solids (VS), revealing natural variation in manure and equipment performance (Table 2). Sample replicates indicated low variance at sampling events but higher variance between events indicating homogeneous conditions through each process but variance between events, probably due to machine adjustment, filter screen condition and natural variance in manure. The average raw manure TS was 7.5%, with a range between sampling events of 7.11% TS and 8.4% TS. The screw press removed a large portion of high carbon solids reducing VS in the manure from 84% to 77% in the liquid stream. Total solids in the liquid stream were quite variable ranging from 3.7% to 6.3%. The variability of these results are directly resultant from frequency of screen cleaning as showed by variable levels of solids diverted to the liquid stream. The BRU received solids with 36.7% TS (ranging from 31.9 to 42.0%) and had minimal drying effect during the 24 hr processing time, producing finished bedding solids of 37.8% (ranging from 32.7 to 40.7% TS). Tab. 2: Total Solids and volatile solids of manure components of the Bedding Recovery Unit values are mean with standard deviation in parenthesis. %TS %VS Manure 7.5 (0.4) 84.2 (0.7) Liquid Stream 4.6 (1.0) 77.1 (4.8) Sep Solids Undigested 36.7 (3.1) 93.0 (0.9) Sep Solids Digested 37.8 (2.7) 92.3 (0.8) Storage of solids in curing piles had a drying effect, raising total solids content from 36.7% to 40.6% in 4 weeks and to 45.7% in 8 weeks. The 24 hr BRU retention (winter) pile was kept in an unheated empty barn and was found to be frozen at each of the sampling intervals due to air temperatures below -20 C. Freeze drying was the likely cause of moisture loss. Similar drying effect was found on the 38 hr solids, reduced from 37.9% TS to 45.9% TS after 4 weeks of storage under an open air, three sided shed. Both sets of solids were reconstituted to 38% TS by the addition of water before self heating testing commenced. The open air curing of the third pile for 4 weeks in June rendered it with TS of 66% and this extreme drying

353 may have been responsible for the slow rate of maximum heating even though the moisture content was also corrected before testing (Fig. 1). Nutrient analysis The manure, liquid, and solid fractions all showed adequate levels of N, P, and K for use as a fertilizer or compost (Table 3). The liquid stream maintained all of the ammonium-n and the trace amounts of nitrate N present in the raw manure. Total N was marginally lower in the liquid due to a portion of organic N being removed with the solid stream. Soluble P and K in the liquid also had no significant change from raw manure. In terms of ratios of N to P, there was a slight concentration of P (6.6 in raw manure and 6.4 in the liquid stream). Tab. 3: Nutrient analysis of raw dairy manure, liquid and separated solid streams. Values are the mean of five sample events with the standard deviation in parenthesis. TNK NH4 NO3 TK mg/kg mg/kg mg/kg mg/kg Manure (2) 2425 (286) (94) (28) Liquid (3) 2523 Stream (151) (131) (105) Sep Solids (2) 2167 (39) (27) (51) Sep Solids (2) 2435 (after (435) (143) (90) BRU) Soluble K TP Soluble mg/kg mg/kg P mg/kg 2697 (96) (43) (52) (29) (133) (31) (138) (19) (131) 2209 (76) (83) (87) The nutrient content and value of the raw manure, liquid stream and separated solids indicate they are similar (Table 4). Plant available N ranged from 1.8 to 2.3 kg/tonne (wet weight) for the four manure components tested. Available N was calculated as NH4 + available organic N (TKN minus NH4) and assuming a 25% organic N availability in the first year. Plant available P ranged between 1 and 1.9 kg/tonne (wet weight) and was determined by conversion of TP to P2O5 (TP/0.43) with an assumption of 100% plant availability. By these calculations, the manure and compost derived from it appear to be low in nitrogen for most agricultural crops. Tab. 4: Plant available macronutrients in raw manure, separated liquid and separated solids calculated on a wet weight basis. Plant avail N TP as P2O5 TK as K2O kg/tonne kg/tonne Manure Liquid Stream Separated Solids (digest) Calculation of the C:N ratios for the separated solids gave a ratio of 32.4 for the undigested solids, 31.7 after they passed through the digester and after curing. The lower ratio in the cured solids indicates continued carbon digestion during storage. Microbial community of recovered solids On average, 29,982 of high-quality sequences were generated per sample. The alpha-diversity of SSD (solid stream digested) microbial communities was 339 found to be lower (p<0.001) than other treatment groups, with the MAN (manure) samples showing the most diverse microbiota followed by LS (liquid stream) and SSU (solid stream undigested) (Fig. 3). The beta-diversity of microbiota showed clear separation between groups. The MAN and LS have similar microbial community regardless of season. In contrast SSU and SSD had each a distinct microbial profile regardless of season. The lower alphadiversity and distinct beta-diversity profile of SSD samples indicate that only specific member of the community can survive the composting process. The proportion of phylum Firmicutes (members of the family Bacillaceae) was greater (p<0.05) in SSD group, and the relative abundances of several members of the phyla Proteobacteria (family Moraxellaceae and genus Acinetobacter) and Bacteroidetes (family Sphingobacteriaceae) were found to be significantly (p<0.05) higher in SSU compared to other groups. The proportions of several members of the phylum Proteobacteria were also found to be higher in SSU compared to LS, suggesting that, in the absence of composting process, separation of the solid and liquid part of the RBM increases the proportion of the opportunistic microorganism in the solid fraction. The number of pathogenic microorganisms significantly varied among samples. Below is a breakdown of the results for the 10 pathogens that were quantified in this work. Escherichia coli is considered a natural inhabitant of the intestinal tract of mammals and birds, and thus large numbers of this bacterium are shed in the feces. However, only specific strains of E. coli, such as O157:H7/STEC, are pathogenic to humans and thus are of concern for public health. Dairy and beef cattle are one of the most important reservoirs of E. coli O157:H7, and large numbers of this pathogen can be found in bovine feces. Although O157:H7 has no harm for the dairy animals, the shedding and cyclic spread of this bacterium through recovered bedding material can be of concern if used as garden fertilizer by users who are unaware of pathogenic potential of the composted manure. To assess the pathogenic risks of E. coli O157/STEC, the bacterium should be quantifiable in a broad range of concentrations. In this study, we used a combination of target genes find in generic E. coli and O157:H7/STEC to quantify both pathogenic and non-pathogenic strains of this bacterium. As expected, all of the samples were positive for generic E. coli. E. coli O157:H7/STEC was identified in a number of MAN, SSU and LS samples. However, only 1 SSD sample was positive for O157:H7, 4 for stx1 and 9 for stx2, which are indicator genes for STEC. Although, the composting process did not impact the total number of E. coli, the number of pathogenic E. coli in the recovered bedding was numerically reduced.

354 p< p< p<0.9 Fig. 3: Alpha-diversities of microbial communities among treatment groups (MAN, LS, SSU and SSD) in the summer Salmonella can survive for several months in feces, which makes this pathogen a constant concern for poultry and livestock operations. In this study, all 132 samples were negative for Salmonella. Listeria monocytogenes causes bacteremia and meningitis in humans. The pathogen was present only in limited number of MAN samples and the number was numerically reduced in the composted samples (SSD). Streptococcus uberis, Klebsiella pneumonia, Staphylococcus aureus, Streptococcus agalactiae, and Arcanobacteruim pyogenes, which are involved in environmental or contagious mastitis, were more prevalent among all samples. Streptococcus uberis was present in 18.1% of samples. Staphylococcus aureus was present in 6.8% of samples. Streptococcus agalactiae was present in 6.1% of samples. In contrast, majority of 132 samples were positive for Arcanobacteruim pyogenes (81.8%) and Klebsiella pneumonia (98%). The compositing process only significantly reduced the number of Arcanobacteruim pyogenes in the SSD samples. Different primers vary in their sensitivity to detect Mycobacterium avium subspecies paratuberculosi (MAP) in the sample. We used a combination of culture based which is the gold standard for detection of MAP as well as three target genes that are sensitive to high or low numbers of the bacterium. Depending on the methodology used, 1.5% to 9.8% of samples were identified as MAP positive. The composting process 340 didn't reduce the prevalence of MAP in the SSD samples. It is important to note that the abovementioned percentages are not reflective of within-herd seroprevalence of MAP, and although the dairy operation is MAP positive, the seroprevalence among animals might be low. Tab. 5: Incidence of select pathogens and reduction from 24 hr residence time in BRU. Pathogen Incidence in SS (%) Health concern Result of BRU process Escherichia coli O157:H7 Human health Human health Numerically reduced c Salmonella enterica 0 Human Not found health Listeria monocytogenes Human health Numerically reduced generic Escherichia coli 100 Mastitis No change 100% Streptococcus uberis 18% 18 Mastitis No change Klebsiella pneumoniae 98% 98 Mastitis No change Klebsiella pneumoniae 98% 98 Mastitis No change Staphylococcus aureus 7% 7 Mastitis No change Streptococcus agalactiae 6% 6 Mastitis No change Arcanobacteruim pyogenes 81% 81 Mastitis Significant reduction Mycobacterium avium 2-10 Johne s No change subspecies paratuberculosi disease (MAP) 2-10% V. Conclusions The purpose of this study was to determine the potential uses for BRU product streams whether as off farm bedding, compost or liquid fertilizer. The viability of these uses was determined by persistence of pathogenic bacteria, nutrient content, and the maturity of compost as tested by thermal stability. For

355 the most part, the treated solids were not changed significantly from longer retention time in the tumbler but curing reduced self heating potential. Analysis of nutrient content found little difference in the raw manure and the liquid stream. On a wet weight basis, use of either of these products as fertilizer are practically the same for agronomic purposes. The solid stream however, had 25% more plant available nitrogen and 80% more P than the liquid stream or raw manure. Even though the separation process increased P concentration for most field applications additional nitrogen would be required in addition to either the manure or BRU outputs. Extending the residence time in the BRU from 24 to 38 hrs did not reduced the self heating potential of the solids and results showed self heating to be in the upper temperature range after 38 hrs. This result was from a single experimental batch and further tests would be useful to determine if this was an outlier or not. Due to variability in results of incubated self heating tests with different batches of solids, it is recommended that replicate dewars of several batches be conducted. Increasing the residence caused some stress in manure management of the farm in this study because the entire system was designed on a 24 hr retention time and increasing it without increasing upstream holding capacities created a bottleneck in the system. Storing piles of processed solids to cure for 4 weeks caused a definite reduction in maximum temperature increase during self heating tests. Maximum temperature was decreased by half after 4 weeks of storage of either 24 hr (22.5 C to < 9.5 C) or 38 hr (34.6 C to 17 C) retention solids. Curing for 8 weeks did not result in a lower maximum temperature increase, but only one test was completed and more would increase certainty. It is interesting to note that if the EPA standards were used instead of the CCME standards, the reduction would be considered more favourable, going from immature to mature compost. Below 8 C is considered very mature. In terms of microbial community and pathogen load, our data showed that composting process significantly reduced the diversity of microbial community indicating only specific members of the community can tolerate and survive the high temperatures generated during the process. Diversity of microbial communities in manure samples was greater in the summer compared to the winter. As such, the microbial diversity of SSD samples was higher in the summer compared to the winter, thus, the chance of pathogen transfer through the recovered bedding is greater in the summer compared to winter. Of the 10 pathogens that were monitored, the bacterial load of the Escherichia coli O157:H7, Salmonella enterica, and Listeria monocytogenes that are considered public health concern were low in SSD samples suggesting that selling bedding material as an organic fertilizer to gardens has lower risk for transmission to human end users. Although it should 341 be noted that the pathogenicity of these three bacteria depends on different factors including dose and virulence factors present in each microorganism. MAP was only identified in a limited number of samples but its prevalence was not changed in SSD samples. On the other hand, the prevalence of bacteria that are associated with either environmental or contagious mastitis (Streptococcus uberis, Staphylococcus aureus, Streptococcus agalactiae) ranged from 6% to 18% in the samples, with the exception of Klebsiella pneumonia, Arcanobacteruim pyogenes, and generic E. coli that were 81.8%, 98%, and 100%, respectively. The pathogen load was not significantly impacted by the composting process with the exception of A. pyogenes, which was reduced in SSD samples. Taking into account that in our analysis pathogen load was calculated on the wet weight of the samples as opposed to their dry matter content, and considering MAN and LS had the least dry matter followed by SSU and SSD samples, the pathogen load is expected to be lower in the SSD samples if dry matter content of samples was considered. Acknowledgements This research was supported by the Manitoba government through the Agri-Food Research and Development Initiative. References Anderson M. PERMANOVA: a FORTRAN computer program for permutational multivariate analysis of variance. Department of Statistics, University of Auckland, New Zealand. (2005) Chao A. Nonparametric estimation of the number of classes in a population. Scandinavian Journal of statistics (1984) Canadian Council of Ministers of the Environment: _e.pdf Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N, et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. The ISME journal. 6: (2012. Cornell University Compost website: Derakhshani H, Khazanehei HR, Tun HM, Alqarni S, Cardoso FC, Plaizier JC, Khafipour E, Loor JJ. The microbiome composition of the rumen is altered during the peripartal period in dairy cattle. PloS One. In press. (2014) US Environmental Protection Agency

356 Cost Modeling for Thermal Energy Storage in Hybrid Power Generation from CSP and Biomass Resources in Turkey - Southeastern Anatolia and Eastern Anatolia Region Yıldırım Ismail Tosun 1* 1 Sirnak University, Faculty of Engineering, Department of Mining Engineering, Sirnak, Turkey * yildirimismailtosun@gmail.com Abstract Usually hot springs spa and evaluation of district heating Southeastern Anatolia and Eastern Anatolia between 45 C and 125 C, less dense the population is not economically in the region. There is a great need for storage of excess energy which is cleaner than the energy produced from the fossil fuel power plants. This storage has facilitated the increasing penetration of storage technologies and primarily from renewable energy sources (RES). The efficienrt use of excess energy sources in today s electricity networks may minimize the threat of global warming and climate change. However, the cost of power output and storage of these energy sources should be considerable and easy to adjust to changing demand cycles as the current and future energy storage technologies used for energy. Model data trends observed in the average of approaches were examined and evaluated as a very well-parametric analysis of costs. It has been expanded by recent model costs. This risk analysis and plant construction, investors can provide insurance companies for risk assessment and decision-makers geothermal wells. Thus, the risk analysis will help in calculating the correct budgeting and insurance premiums. The overall objective of hybrid the parabolic dish CSP solar panels and geothermal energy is the presence of a hybrid geothermal system can be produced economically. Geothermal energy was started in the vast area to be searched, amended as a result of the research data, the field is narrowed down to investigate to direct heating of regional area. At the same time studies in the cost-benefit criterion it has been considered, and thus became the economic research work. Keywords: energy storage, risk assessment, stochastic cost estimation, simulation, direct heating simulation I. Introduction Lignite consumption in energy production is increasing in parallel with growing energy needs today. In terms of reserve and production quantities of high quality lignite, natural resources are limited. The significant amount of electricity is produced primarily from coal in the world (TKİ 2009,TTK 2009). The almost 211TWh total electricity in 2011, Turkey were produced primarily from imported natural gas and domestic coal (Fig 1) (EIA, 2013). The total amount of asphaltite resource in reserves and production in Şırnak City are over 82 million tons of available asphaltite reserve and 400 thousand tons per year, respectively (MTA, 1987). Turkish coal industry needs noble gasification technologies and high gasification performances at lower cost with various types of local coals regarding researches. The most effective and cost-effective combustion or gasification technologies are needed for clean coal products in today's modern technologies (Bell et al, 2011, Shadle et al, 2001, Sharma et al, 2008, Jess et al, 2009, Kajitani et al, 2006). Turkish coal industry needs specific tests (Çakal et al, 2007) in order to measure gasification performances of various types of local coals regarding standard qualification tests. There are lots of signs for the waste yield of biomass in industrial many fields even using regular high capacity biomass of cellulosic wastes (Anonymous, 2009, Karakaya, 2008, TAM 2009,TSI 2009). Processing technologies using biomass should be under contribution to the fuel side producing high value cleaned products, pyrolysis and gasification of lignite. Fig.1: Primary Electricity production vs resources in Turkey regarding II. Biomass Waste Potential of Turkey In Turkey, the Ministry of Agriculture and Rural Affairs declared the statistics that the amount of waste generated from annual and perennial crops received from local authorities; the amount of production and acreage of each product is calculated using the data for In our country, agricultural production waste is commonly left in the field. Cereal straw is used for various purposes, for example, used as animal feed, and fertilizer. The main waste in the field following the agricultural sawed products, plantations is allowed to rest. These kinds of waste cotton stalks, corn stalks, sunflower stalks, hay and tobacco stalks are waste, etc. Total amount of waste products are 342

357 divided into theoretical and actual values given in the Table 1. The total annual production of field crops in Turkey and waste quantities are given in Table 2. Table 1. Total amount of Municipal waste divided into actual values in Turkey and Eastern Anatolian Region in Waste Type Heat Value, kj/kg Waste Statistics Country, Actual million ton/year Textile,Rubber, ,6 2,1 Plastics Wood, ,4 1,6 Cardboard, Paper Organic ,2 29 Municipal Waste Animal Waste ,9 21 Forestry and ,8 63 Agricultural Biomass Total ,9 116,7 Eastern Anatolian Region Actual, 1000tons/year Table 2. The total annual production of biomass waste in Şırnak and Eastern Anatolian Region. Waste Type Waste Statistics Heat Value, kj/kg Eastern Anatolian Region Actual 1000ton/year Plastic ,1 1,3 Agricultural ,8 1,6 waste Cow, Sheep Poultry Wastes Forest Waste Total ,9 46,9 Şırnak Actual, 1000ton/year Total heat value of approximately 16,8 kj/kg for corn stalk and 20,2 kj/kg for peanut shell waste. According to the total thermal value, basic products are corn 33.4%, wheat 27.6% and cotton 16.1 %, respectively. In Table 1, the total annual production of horticultural crops waste in Turkey is given. Its total thermal value is approximately 21.5 kj/kg, respectively. Within the total calorific value of the product with the biggest 56.3% nut shell and olive seed 25.2%. According to the number of animals in Şırnak, the calorific value of the amount of waste and animal waste are as given in Table 2, about 13, 30 and 26,5 million in Turkey for cows, sheep and poultry, and approximately the amount of annual waste capacity are 0.2, 0.1, million tons, respectively. The total annual amount of forest, bush and wood waste, are 6, 0.6 and 0.49 million tones, respectively. The total available 65 % solids content of 3% and 99% values were determined by the availability of forest, bush and wood waste, respectively TSI, 2013, TAM, Thus, Şırnak's cows, sheep and poultry waste calorific value of the annual total was found to be of about 48, were 3 and 0.7 MJ, respectively. Today in Şırnak, including biomass in terms of research and development in renewable energy incentive researchs are carried out and the agricultural and forest wastes in cellulosic form are investigated in power generation. In recent times the output of the power of free licence laws before energy prices, energy-producing companies buy the energy of the states were determined through interviews. Now renewable energy prices will be forced to comply with market conditions. This situation of renewable energy sources in order to better compete and market additional policy tools need to emphasize that EU energy policy and law by examining the conclusions drawn from Turkey according to the potential policy instruments include: the country, the purpose specified deviation from the guarantees to domestic targets, including renewable energy sources, given up to one year, domestic gross energy consumption a certain share have to target (about 10%) should be determining policy and legislation. These are obtained from renewable resources and electric power for heating may comprise separate but integrated objectives. These policies and laws only for biomass separate, but can also include an integrated target. All use of renewable sources in the EU target of achieving 12% market share for the biomass should be increased up to 300%. The country, the purpose specified deviation from the guarantees to domestic objectives including, biologically-based fuels, specific to a year, used in transport petrol and diesel fuel market, a certain share to have a target for determining the policies or laws should be removed. Regard to the energy in Turkey appropriate potential market instruments include: biomass -generated electricity, heating and / or used in transport fuels, tax reductions or exemptions, flexible loans, low interest loans, loan guarantees, property first operating subsidies and / or grants and related service for consumers willing to use discounts as well as other financial support mechanisms. A potential market instrument of state support is not required to continue forever. They just won and sustaining investor confidence during development and may be important. III. Combustion of Biowaste, Oil and Char in Fluidized Bed Considerable research on coal pyrolysis and gasification has been conducted over the years, but the pyrolysis results are widely dispersed because of the complex chemistry of coal (Wei-Biao et al, 2001, Liu et al 2002). Time related coal-pyrolysis modeling assumes basically first-order kinetic equations, or less sensitive for heating rate (Shale et al 2001, Sharma et al 2008). The other distributed activation model is dependent on the heating rate. The last two more advanced models need three and four constants, respectively, which basically depend on the coal properties but also cover to some extent, the effect of heat-and-mass transfer phenomena (Donskoi et al, 1999, Wictorsson et al, 2000). 343

358 Fig. 2: Fluidized Bed Combustion of Coal and Biomass for combustion Table 3. The Proximate Analysis of Şırnak Biomass Types. Weight(%) Wood Cow Poultry Corn Trash Waste Waste Waste Waste Moisture Ash Fixed Carbon Volatile Matter Calorific Value (kcal/kg) emerging applications in power storage like wind farm power stabilization, steam, hydro, wind, etc. (Anonymous f,h, 2013, Ibrahim et al 2007) The figures of statistical potential of storage in the world is defined as seen in Fig 3. The use of different techniques in energy storage are classified as seen in Fig 4 regarding discharge time. Table 4. The total annual production of animal waste in Şırnak and waste quantities. Waste Statistics Waste Type Heat Value, kj/kg Theoretical, 1000ton/year Actual, ton/year Cow ,4 12,3 Sheep ,3 1,6 Poultry ,7 0,4 Total ,4 13,3 That is the reason for the different values of the activation energy and pre-exponential factor cited in the literature and the lack of generally valid data. The same situation exists in the case of coal-char gasification. The reaction rate of char is influenced mainly by chemical and physical factors, which include coal type, catalytic effect of the ash and the specific surface area of biochar, which changes during the reaction course with the development of internal pores. Fig. 3: The Compressed air and Thermal Storage Use Distribution in Energy Storage. IV. The Energy Storage The energy storage provide the main support to the renewable energy production. The commercial successes in electric power storage Technologies mentioned and its ability were described some of the 344 Fig. 4: The Discharge rates of Different Types of Energy Storage Units vs Power.

359 Some cost evaluations covering security of supply and environmental impacts, climate change evaluations, and technical and economic analysis, may be disussed in energy planning and activities. Natural power output cannot be controlled. The power from these generators varies according to the availability of the resources (for example in case of Solar heat or wind turbine, it depends on the availability of wind or sun) (Anonymous a,b,c, 2013, Ibrahim et al 2007). During the non-availability of the power the storage technologies manage a vital role in improving the overall stability and reliability of power system (islolated/grid connected/systems with large share of renewable sources) and improve the controlled electricity distribution capacity to meet ever growing power demand. The internal grid circulation with the energy storage in renewable sources may stabilize power level in heap and 50-60% yield recoveries were provided at the end of year. As seen in Fig 5 the power consumption can be supplied by stored earlier day time. (Anonymous d,e, 2013, Poullikkas et al 2007) entering the cavern and used for adiabatic expansion extracting heat from the thermal storage system. Fig. 6: CAES Method The only two existing CAES plants in Huntorf, Germany, and in McIntosh, Alabama, USA, as well as all the new plants being planned regarding the diabatic method. The compression stage normally uses up about 2/3 of the turbine capacity, the CAES turbine unhindered by the compression work can generate 3 times the output for the same natural gas input. This reduces the specific gas consumption and cuts the associated CO2 emissions by around 40 to 60%, depending on whether the waste heat is used to warm up the air in a recuperator. The power-to-power efficiency is approx. 42% without, and 55% with waste heat utilization. Fig. 5: The power demand of California on daily period in summer IV.1. Compressed Air Storage The special thing about compressed air storage is that the air heats up strongly when being compressed from atmospheric pressure to a storage pressure of approx. 1,015 psia (70 bar). Standard multistage air compressors use inter and after coolers to reduce discharge temperatures to 300/350 F (149/177 C) and cavern injection air temperature reduced to 110/120 F (43/49 C) (Anonymous g,h, 2013, Ibrahim et al 2007). The heat of compression therefore is extracted during the compression process or removed by an intermediate cooler. The loss of this heat energy then has be compensated for during the expansion turbine power production phase by heating the high pressure air in combustors using biogas, fuel. Alternatively using the waste heat of a combustion gas turbine exhaust in a recuperator is able to heat the incoming air before the expansion. Even the heat of compression can be thermally stored before 345 IV.2. Hydrogen Storage Electricity can be converted into hydrogen by electrolysis. The hydrogen can be then stored and eventually re-electrified. The performans today is as low as 30 to 40% but could increase up to 50% if more efficient technologies are developed. Despite this low efficiency the interest in hydrogen storage is growing due to the much higher storage ability compared to batteries (small scale) or pumped hydro and CAES (large scale). (Anonymous f,h, 2013, Ibrahim et al 2007) Small amounts of hydrogen (up to a few MWh) can be stored in pressurized vessels at 100~300 bar or liquefied at 20.3K (-423 deg F). Alternatively, solid metal hydrides or nanotubes can store hydrogen with a very high density. Very large amounts of hydrogen can be stored in man made underground salt caverns of up to 500,000 m3 at 200 bar (2,900 psi), corresponding to a storage capacity of 167 GWh hydrogen (100 GWh electricity). In this way, longer periods of flaws or of excess wind / PV energy production can be leveled. Even balancing seasonal variations might be possible.

360 IV.3. Liquid Air Energy Storage Liquid Air Energy Storage (LAES) is sometimes referred to as Cryogenic Energy Storage (CES). The word cryogenic refers to a gas in a liquid state at very low temperatures. The working fluid is Liquefied Air or Liquid Nitrogen (78% of air). The systems share similar performance characteristics to pumped hydro and can harness industrial low-grade waste heat/waste cold from co-located processes, converting it to power. Size range extends from around 5MW/15MWh to >50MW/250MWh and with capacity and energy being de-coupled, the systems are very well suited to long duration applications. Fig. 7: Linde Liquid Air Storage Tanks. IV.4. Projected Tests for Hybrid Power and Capital Model for Energy Storage A fluidized bed reactor was used in coal combustion heated till 600 o C with a rate 7-10 o C/min by fuel. The process was tested at a scale of 2 3 kg/h; collecting operational and design data to build an industrial installation. A technological diagram of the coal gasification-pyrolysis process developed unit is made. Thermal destruction almost observed at temperature Heat exchanger oil temperature increase from 350 C to 400 C with storage and performance of 60-70% and also simultaneous dilution of waste oil products by condenser distillate may be used. it is necessary to optimize the storage conditions of oil circulation without the heat loss transported to heat exchanger, where the average heat conductivity per oil amounts changes kw/m 3, so enhancing ORC heat transfers. IV.5. Cost Model for Energy Storage Compresse air storage for large scale installation of coal combustion was planned simply. Porous shale beds in Southeastern Anatolian region, in Batman was planned with fine coal burned with oil and biowaste at 0,2 m/s. In order to achieve this, it is necessary to create gaseous conditions of internal circulation without the transported coal and char in the fluidized bed, where the average concentration of solids amounts to m 3 /m 3, i.e. the conditions for residence time are long enough for the thermal oil storage of combustion units (Dinçer and Rosen 2011, Cabeza, 2015) and intensive gas firing so enhancing mass and heat transfers. The cost of underground compressed air storage will greatly depend on drill cost. The parameters are defined in Table 5 below. Combustion Weight, TGA,% Time, s Fig. 8: The rates of Combustion of biowastes used in coal combustion at 600 o C in fluidized bed. Thermal oil storage may greatly concerned by combustion using power plant systems, due to heat loss in echanger and recupator use in the heat recovery. Tablo 5. Correlation and variable values in Drilling depth with the cost of investment. RİSK Point Weak Mid Hard Rocks Rock Rock 500m 1500m 2500m Depth,m Advance Rate Drilling Period Investments Risk Risk Error Correlation Coefficient Compressed air use was investigated in the projected study due to the regional potential of Southeastern Anatolian region of Turkey. Espeially oil resoirvoirs was thought as storage field, in the mean time the natural gas using power plants may advantageously evaluate the compressed air or liquidified air. However liquidified air was not considered in ths projected study due to costly compression units was needed. Gaussian normal distribution of risk probability values defines the value of the compresse air plant investment in data-intensive midpoint. It is obtained as given below section. As given in Tab 6 the compressed air storage unit cost reached high levels. Time became the main parameter in the evaluation. u(x; t; ϼ) = N n u(x,t) u(x; t) i=0 (1) u(x,t) 10mm mm Şırnak Municipal Waste, ÇöpBiyoKütle Biomass 5 mm mm Cow Sığır Waste Küspe 5 mm Poultry Tavuk Waste Küspe 5 mm Maize Slush Mısır Sapı y = ln(x) y = ln(x) y = ln(x)

361 Table 6. The Capital costs of Compresse air storage in the region into the used oil reservoirs Unit Cost, $ m3/y m3/y cost function, σ is the variance,k is hybrid distribution parameter, Ө is the tim parameter, x is the flow rate by the following equation. Compression units: 4 stage Surface Storage Tanks: $ Drilling well Water oil seperators Combustion Fan Filter bag units : $ Dust Collector Units: $ Automation Control System Field Cost Engineering Project Power Plant TOTAL :$ TV C is the total cost, T x is tax, F is the interest, O m&o is maintanance cost, D is share rate, cm is capacity factor, K is the unit capacity. As given below; For the integrated plant, the capital investment cost of 500 thousand tons/year capacity was determined as 53 million $, while 1 million tons/year capacity the investment value was doubled. Already region for high-capacity incinerators could not considered due to the impossibility of sources and the obtaining funds and logistics were not feasible. Mobile tons/year capacity plant, depending on the companies' unit costs is determined as 11 million $ (as given in Table 6). Mobile plant and integrated plant operating costs were calculated based on the present prices. As Table 6 also given mobile plant labor, it will provide advantages in terms of reactive maintenance. Mobile plant operating cost was calculated as approximately 25 TL/ton for defined biowaste. This integrated facility cost rose to 63TL/ton. Mobile plant and integrated plant operating costs and energy production (70% and 60% thermal efficiency fuel efficiency) was calculated to be connected. mobile plant as given in Figure 8, while in a period of their capital investment in 22 months, after a period of 36 months will generate more revenue for the integrated plant operating costs will be advantageous investment capital back to paying (Figure 8). IV.6. Investment Modeling of Compressed Air Storage from Coal and Biomass Combustion Thus, the installed capacity of the planned plant was about 2 million kwh/year and the entire unit energy yield was 22 MWh/year and 15% storage capacity was used in cost estimations. The costs of drilling equipment such as maintenance costs. General Pareto distribution of the cumulative risk the possibility of failure depth is closer to the discovery and exploration of geothermal resources available well data. Figure 9, as determined in the MATLAB software risk probability value is above 5% probability of u(x, σ, c) = (1/σ) (1 + k/σ (x Ө)) 1 1/k (2) while exponential distribution k = 0, where u is the 347 Fig. 9: Change of the Cost of Drilling for Compressed Air Storage TV C = T x + F + O m&o + D (3) Q(n) = 8760xCF(n)xK (4) The cost calculation of the plant, Calculation of unit cost of the facility, Calculation of the investment costs of the facility at which it will go into production, Plant operating costs and the calculation of the annual income, R(n) = Q(n)xP(n) E(0) = (1 f) M m=1 (5) 1/(1 r) M m M m=1 (6) Cm(1 + r) M m L(0) = f M m=1 Cm(1 + r) M m (7) The capital cost for compressed air storage and thermal oil storage may greatly affected by the ost of storage material and process units complecity.the proposed cost of investment depenting on flow rate of stored material and conversion to heat or energy as given below. u(x; t; Ө) = n i=0 u(x, t) + ɸ(x; t; Ө). e tiθ (8) where R is the revenue, Q is the capacity, P is the sale price, r is the interest rate, m is month, n is the integer of month, E is investment cost, f is debt rate, cm is capacity factor, L is the debt, u cost function, t is time, Ө is the time unit parameter.

362 As given Tab 7 the cost distribution changed by time of storage well drilling and time. Risk point was similar to geothermal well drilling. As shown in Fig 10 risk parameters occurred in the projected cost risk analysis. As seen in Fig 11 the capital cost was ranged up to 180 million $ for m 3 /y compressed air storage. Table 7. Normal distribution Cost variable values in Compressed air Storage Well Drilling with the cost of investment regarding depth. RİSK Point Weak Mid Rocks Rock Hard 500m 1500m 2500m Rock Depth,m Advance Rate Drilling Period Investments Risk Fig. 10: ORC Use for Low Heat Geothermal and Biomass sources in energy and risks of Capital Investment costs in Turkey 348 Fig. 11: The Investment rates of Compressed Air Storage in Combustion Power Plant. Table 8. Compresse Air Storage Cost Variables on Coal and Biomass energy capital cost risk Projected Cost and Revenues Batman Without Storage Siirt Without Storage Batman With Storage Siirt With Storage Cost Risk Batman Cost Risk Siirt Net Electricity kwh 137,000, ,000, ,000, ,000, Average Annual Sale TL 0,26 0,26 0,26 0, Production Cost nominal 0,21 0,21 0,26 0, Production Cost actual 0,11 0,11 0,15 0, Return rate,% Annual Net Profit 22,000,000.TL 21,000,000.TL 12,000,000.TL 11,000,000.TL 6 6 Calculated Sale price change,% Calculated debt rate,% Capacity factor Land cost 1,6 1,6 1,6 1,6 2 2 System performance factor Toatal field, acre Cogeneration Sellective Sellective Sellective Sellective 1 1 Average Risk 6 6 V. Conclusions This project approach assumes basically that the thermal oil storage units are much feasible instead of compressed air units, so that is a decisive factor for the path of the storage method and heat exchanger units. Therefore a model of hot oil storage may be supported by biogas combustion, too. The gas temperature and. fluid bed combustion and pyrolysis oil use for coals and biomass may provide clean combustion and oil storage. Combustion rates were lower higher at the high temperatures over 900 C so that heat conduction loss to boiler materials might be fundamentally low enthalpy yield. The main conclusions are as follows: Time uncertainties are taken into account by adding a contingency factor. This approach is simple and it is advantageous to be close to the real projected cost data. Storage by compressed air was efficient but also requires more equipment for compression needed. The liquidified air storage system needs subsequent high cost compression unıts. However, time variable and uncertain application of parametric variables makes possible reliable risk analysis. This research has examined various risk models for energy production. The most appropriate model is determined by comparison. The compressed air was efficient in natural gas combustion turbines but also requires optimization. The liquidified air storage system needs high cost stage by stage compression unıts. The used oil fields will be suitable for compressed air

363 storage. The capital cost of thermal oil storage will be suitable for industrial combustion systems at small scale and shorter time concern. References Anonymous a, 2013, Anonymous b, 2013, Anonymous c, 2013, Anonymous d, 2013, Anonymous e, 2013, Anonymous f, 2013, Anonymous h, 2013, Anonymous a, 2015, Mobile incinerators, p, ATİ Şirketi Anonymous b, 2015, Yakma Kazanları, Alfa Kazan ve Makine AŞ,Ankara Anonymous c, 2015, Akışkan Yataklı Yakma Kazanı, Mimsan A.Ş., İstanbul Anonymous d, 2015, Anonymous e, 2015, Anonymous f, 2015, _systems/mobile_systems.htm Akpınar, N, Şen, M, 1987, Kentsel katı atıklardan enerji üretimi, Enerji Enstitüsü Bell D.A. Towler B.F., Fan M., 2011, Coal Gasification and Applications, ISBN: , Elsevier Inc., Oxford Cabeza, L. F. (Ed)2015, Advances in thermal energy storage systems, Elsevier, Woodhead Publishing Series in Energy ISBN: Çakal, G.Ö. H. Yücel, A.G. Gürüz, 2007, Physical and chemical properties of selected Turkish lignites and their pyrolysis and gasification rates determined by thermogravimetric analysis, Journal of Analytical and Applied Pyrolysis, Volume 80, Issue 1, Cherubini, F. Bargigli, S. Ulgiati, S. 2009, Life cycle assessment (LCA) of waste management strategies: landfilling, sorting plant and incineration, Energy, 34, pp Dinçer, İ. Rosen. M. A., 2011, Thermal energy storage: systems and applications, second edition, John Wiley & Sons, Ltd,ISBN: Online ISBN: , DOI: / Donskoi, E.& McElwain, D.L.S., 1999, Approximate modelling of coal pyrolysis, Fuel, 78, pp Ibrahim H, Ilinca A, Perron J. Energy storage systems characteristics and comparisons. Renewable and Sustainable Energy Reviews 2007;12: IEA, 2007,,A.J. Minchener and J.T. McMullan IEA Coal Research Ltd, Clean Coal Technology IEA, 2013, World Energy Outlook Jess A, Andresen A-K. Influence of mass transfer on thermogravimetric analysis of combustion and gasification reactivity of coke. Fuel.; doi: /j.fuel Kajitani S, Suzuki N, Ashizawa M, et al. CO2 gasification rate analysis of coal char in entrained flow coal gasifier. Fuel. 2006;85: Kajitani S, Suzuki N, Ashizawa M, et al. CO2 gasification rate analysis of coal char in entrained flow coal gasifier. Fuel. 2006;85: Karakaya, İ.,2008, İstanbul için stratejik kentsel katı atık Yönetimi yaklaşımı, Yüksek LisansTezi, İTÜ FBE Çevre Müh.Böl. Kreith, F Tchobanoglous, G, 2002, Handbook of Solid Waste Management L.P. Wiktorsson, W. Wanzl, 2000, Kinetic parameters for coal pyrolysis at low and high heating rates a comparison of data from different laboratory equipment, Fuel, 79, pp Liu, G., Benyon, P., Benfell, K.E., Bryant, G.W., Tate, A.G., Boyd R.K., 2002, The porous structure of bituminous coal chars and its influence on combustion and gasification under chemically-controlled conditions, Fuel, 79, pp Poullikkas A. Implementation of distributed generation technologies in isolated power systems. Renewable and Sustainable Energy Reviews 2007;11: Richard A. Denison, J R, 1990, Recycling and Incineration: Evaluating the Choices Ron Isaacson, 1990, Methane from Community Wastes (Elsevier Applied Biotechnology Series) 349

364 Schurtz R, Fletcher TH. Pyrolysis and gasification of a sub-bituminous coal at high heating rates, 26th Annual Int Pittsburgh Coal Conf, Sept , Shadle LJ, Monazam ER, Swanson ML. Coal gasification in a transport reactor. Ind Eng Chem Res. 2001;40: Shadle LJ, Monazam ER, Swanson ML. Coal gasification in a transport reactor. Ind Eng Chem Res. 2001;40: Sharma A, Saito I, Takanohashi T. Catalytic steam gasification reactivity of hypercoals produced from different rank of coals at o C. Energy & Fuels. 2008;22: TAM, 2014,Tarım ve Köy İşleri Bakanlığı İstatistikleri, TEFM, 2008, Orman biyokütlesinden yakıt ve enerji üretimi, (Kahveci, O) TC. Çevre ve Orman Bakanlığı Orman genel müdürlüğü TEFM, 2009, Orman Genel Müdürlüğü nde Biyoenerji Konusunda Yapılan Çalışmalar, Orman Genel Müdürlüğü, Biyoenerji Çalışma Grubu, Orman ve Enerji, Ankara, TKI, 2009, The Turkish Ministry of Energy, Energy, Dept., Lignite Coal Report Tosun YI, 2012, Semi-fused Salt-Caustic Mixture Leaching of Turkish Lignites - Sorel Cement Use for Desulfurization, Proeedings of XIIIth International Mieral Processing Symposium, Bodrum, Turkey. TSI,2014, Türkiye İstatistik Kurumu Verileri, 2014, TTK, 2009, The Turkish Ministry of Energy, Energy, Dept., Hard Coal Report Wei-Biao, F.. Quing-Hua, W 2001, A general relationship between the kinetic parameters for the gasification of coal chars with CO2 and coal type, Fuel Processing Technology, 72, pp

365 Vacuum Stripping Membrane Desalination for Marmara Sea-Water Filiz Ugur Nigiz 1*, Nilufer Durmaz Hilmioglu 1 1 Kocaeli University, Engineering Faculty, Department of Chemical Engineering, Kocaeli, 41380, TURKEY * filiz.ugur@kocaeli.edu.tr Abstract Sea water desalination is an attractive research area for acedemical and industrial researcher. Because of the irrigation and drinking water absence in the world, it is need to treat non-utilisable brackish or seawater. Indeed desalination by using thermal treatment system has long been used in many years. Because of the huge energy requirements, these systems lose their attraction. In some countries with limited fresh water, clean water demand is supplied from seawater. When the water future is considered, each country needs to develop own desalination technique. In Turkey, there are several reverse osmosis desalination plants that are settled around the Marmara, Aegean and Mediterranean Sea. In order to increase the number and quality of these facilities, investigations should be concentrated on new technique development that leads to a reduction of energy usage and an increase in water quality. Several years, pervaporative desalination of seawater becomes an age as energy intensive, cost effective and efficient process. It is possible to achieve above 99.9% of pure drinking water from seawater that contains millions of contaminatings such as heavy metals, microorganism, industrial pollutants and ions. So this study focused on the model solution and Marmara seawater desalination with this technique by using clay incorporated carboxymethyl cellulose natural membrane. Effects of salt concentration in the feed mixture and operation temperature were investigated as function of water flux and rejection factor. At the end of the six hours above 99 % pure water was obtained at low temperature. Keywords: Seawater desalination, cellulose membrane I. Introduction Water and energy demand are the main problems of current world. Many of alternative technologies have been developed for energy production from a wide variety of source. However, it is needed to find natural solution for fresh water that cannot be produced from any resource. Researchers have predicted that the biggest problem and crisis will rise caused by water absence in the world of future. Therefore, usage of natural water source by treated or converted to available water can solve the major part of problem. Recycling of industrial wastewater, reusing of rain and domestic wastewater are examples of these solutions. More importantly, approximately 96 % of water source of earth such as brackish and seawater cannot be used directly and must be converted clean water by using some treatment methods such as thermal treatment and membrane desalination. Desalination means the obtaining fresh water by removing the ions, salt heavy metals and other pollutants from sea water. The concentration of these impurities and salt are changed according to the regional climatic conditions, precipitation period in year. Hence, effectiveness of the system alters from one country to another. Desalination techniques are more abundant in countries that have limited fresh water (Latteman et al. 2008). Most common desalination technique is the thermal distillation techniques such as multi flash distillation and multi effect distillation. These systems based on the vaporization of sea water into a distillation unit and obtaining ultra pure water. The advantage of flash distillation is the vacuum applying to reduce the boiling point and saving the energy. The energy of system is provided from industrial waste-heat. However, large volume of distillation units increases the capital and investment cost (Bart et al. 2002; Akili et al., 2008; Xing et al. 2013). Reverse osmosis (RO) is another desalination technique. Porous membrane that has small pore diameter is employed in RO system and separation occurs according to the pressure difference between the upstream and downstream sides of membrane. Different membrane materials (polymeric, inorganic or composite) can be preferred in RO (Nur et al, 2016; Misdan et al. 2010; Avlonitis et al. 2012). The efficiency of system depends on the membrane productivity. Membrane fouling, concentration polarization, unstable membrane performance and energy consuming for desired pressure are the major drawbacks of the system. Pervaporation (PV) is another suggested desalination system for deep purification of water. Pervaporative purification of industrial water has long been used to separate organic solvents, heavy metals, hazardous chemicals, ions from industrial wastewater (Dali et al. 1995; Dabrowski et al. 2004; Schafer et al. 2001). Several years, pervaporative desalination technique has been studied by the different researchers from all around the world. Because of the low operation conditions such as low 351

366 temperature and atmospheric pressure, it can be defined as energy saving, economical and clean technique. The driving force in PV is the chemical potential gradient and it controlled by vacuum or inert purge gas pressure on the permeate side of membrane. Alter from the reverse osmosis, non-porous membrane is used in PV and selective character of membrane comes into prominence. Separation model can be explained by three main steps. Dissolution of component on the surface of membrane, diffusing through the membrane and desorption to downstream side. Selected component passes through the membrane among the molecular spaces of membrane material. In case of polymeric membrane, the component to be separated diffuses through polymeric chain spaces. Same as the RO, polymeric (Korngold et al. 1996) inorganic (Poul et al. 2012; Churl et al. 2011; Martin et al. 2012) or composite (Bin et al. 2015, Xie et al. 2012; Elma et al. 2012) membrane materials can be used according to the mixture to be purified. Due to the selective separation capability of PV, ultra pure water can be obtained. However, low flux values restrict the commercial availability of PV. Hence, researchers focus on the produce high performance membrane. It is important to achieve high salt retention associated with high flux. Polymeric membranes have large free volume and exhibit high flux value at relatively high temperature and water concentration. Because of the swelling tendency, high flux causes low salt rejection. Although the preparation of polymeric membrane is quite simple, they show poor mechanical, chemical and thermal stability. If a polymeric membrane is supported by an inorganic material such as zeolite and clays, the structural stability of them increases. Inorganic incorporated polymeric membrane or mixed matrix membranes exhibit unique separation performance with long period lifetime as well. In this study synthetic salt solution in which contains K +, Cl -, Na + and Ca +2 ions was selectively purified by pervaporation. Bio-based carboxy methylcellulose (CMC) was used as polymeric material, pristine and sodium montmorillonite (Na + MMT) loaded membranes were prepared as selective membrane. Membrane and system performance were evaluated as function of salt retention and water flux. Effects of salt concentration and temperature on PV performance were concluded. KCl and CaCl2 concentrations were kept (wt. 2 % of salt) stable in the mixture and NaCl concentration was changed in order to obtain effect of salt concentration on separation performance. After optimum operation conditions had been determined, Marmara Sea Water which was obtained from the Gulf of Izmit was purified with that condition. II. Experimental II.1 Materials Carboxymethyl cellulose (CMC) was purchased from Aldrich Chemicals. Glutaraldehyde (GA), hydrochloric acid (HCl), acetone were supplied from Merck Chemicals in Turkey. II.2 Membrane Preparation For the preparation of pristine membrane; 1.5 wt. % CMC-water solution was prepared and stirred for 10 hours. After a homogeneous solution had been obtained, that was cast onto a poly (methyl methacrylate) plate. Membrane was dried for 2-3 days at room temperature. And it was cross-linked in a bath that was consisted of glutaraldehyde, HCl, acetone and water. In case of clay incorporated mixed matrix membrane, 5, 10 and 15 wt. % of clay was added to the 1.5 wt. % CMC-water solution and sonicated for 30 min. Membrane forming procedure was same as the pristine membrane. II.3. Membrane Characterization Clay distribution in CMC matrix was analysed by using a JEOL JSM-6335 F Field Emission Scanning Electron Microscope. Clay distribution was also analyzed by Microscope. Cross-linking performance was analyzed by using Perkin Elmer Pyris 1 FTIR spectrophotometer. II.4. Swelling Experiments Degree of swelling (DS) test was done to determine membrane affinity to seawater. Membranes were immersed in salt solution separately. Swelling degrees were calculated from the weight difference of swollen (Ws) and dry (Wd) membrane as seen in Eq. 1; DS (%) = (Ws-Wd/Wd) *100 (1) II.5. Pervaporation Experiments Pervaporation tests were carried out at with different model solutions (1 wt. %,3 wt. %,5 wt. %, 7 wt. % NaCl and constant KCl and CaCl2 concentrations at four different temperatures (30,40,50,60 ºC) by using Na+MMT incorporated mixed matrix membrane. The effective membrane area was 19.6 cm 2, the cell capacity was 250 ml. Upstream of the membrane was kept at atmospheric pressure and downstream of membrane was 5 mbar. Salt solution was continuously circulated to membrane cell. The separation performance of the PV was determined as a function of flux (J) and salt rejection (R) or rejection factor. J= Wp /A.t (2) 352

367 R= (Cf Cp / Cf )*100 (3) Wp represents the total permeate mass of mixtures, A is the effective area of membrane and t is the time. Cf and Cp represent the salt concentration in feed and permeate mixture respectively III. Results and Discussion III.1. Characterization Results FTIR test was done to prove the cross-linking reaction between the glutaraldehyde and CMC and the peak was observed. The reaction between the carboxyl group of CMC and GA was seen as C-O-C formation peak at 1050 cm -1 region. voids may cause an uncontrolled passage of salt ions and rejection decreases. As it was proved from the close shots of cross-sectional micrographs a successful membrane formation was done. III.2. Swelling Results Degree of swelling is a measurement to determine the membrane affinity to component to be selected. It should be a certain value within limitations. While excess swelling degree can cause non-selective ion passage through the membrane, an insufficient swelling gives very low flux results. Additionally, it gives an idea about the durability of membrane during the long term operation period. Figure 3 showed the swelling degrees of uncross-linked CMC membrane, pristine and 5 wt. % Na + MMT loaded membranes. Fig.1. FTIR spectra of CMC membrane Surface and cross-sectional SEM observations were shown in Figure 2. Figure 2a and 2b indicated the homogeneous distribution of clay particles on the membrane surface. Fig.3. Swelling results of uncross-linked, pristine and clay loaded membrane in model solution (40 C, 5 wt. % NaCI) In this study cross-linking was applied to prevent dissolving of membrane in water. As it was proved from the Figure 3 that the uncross-linked membrane solved in model solution within three hours. It was pointed out from the Figure 3 that the clay incorporation restricted the excess swelling degree. III.3. Effect of Temperature on Separation Performance Fig.2. SEM micrographs of NaMMT loaded CMC membrane Figure 2c and 2d indicated the interfacial contact area between the clay and polymer. In order to achieve selective separation in pervaporation and for retaining the hydrated ions in feed side, there should not be any interfacial voids between the polymer and clay. These The relationship between the flux and temperature has been studied almost in all pervaporation research. Figure 4 showed the effect of temperature on flux for different membranes. Increasing feed temperature increased the partial vapour pressure of water and water transfer rate through the membrane was accelerated. Additionally, temperature dependent diffusion coefficient of water increased and flux enhanced. 353

368 temperature as it is shown in Figure 5. III.4. Effect of Salt Concentration on Separation Performance Figure 6 shows the flux change attributed to salinity of model solution at 40 ºC temperature. When concentration of hydrated ions increases in feed solution, vapor pressure of bulk solution decreases (Liang et al. 2015). Because of the chemical potential driven properties of PV, flux decreased with reducing water concentration. Fig.4. Effect of temperature on flux (5 wt. % NaCI) Polymeric membrane matrix is directly affected from the temperature change. According to the free volume theory of polymers, with rasing temperature the chain spaces become larger and flux increases (Xie 2012). Relatively low flux increasing was the evidence of this situation. Clay particles prevented the extent swelling degree and the total free volume enlargement was restricted by clay incorporation. Figure 5 showed the relationship between the salt rejection factor and temperature at constant water salinity. Fig.6. Effect of NaCl concentration on flux (40 ºC) Figure 7 shows the salt rejection changes with NaCl concentration. Contrary to flux, there were no significant changes in rejection at constant temperature so it could be conclueded that the concentration of sea water was not so important to achieve good separation by using PV method. In case of composite membrane, above 99 % rejection was achieved and membranes exhibited stable trend. Fig.5. Effect of temperature on salt rejection factor (5 wt. % NaCI) Due to the solid nature of salt, they can solve on membrane surface but cannot evaporate to permeate side. Hence, very high rejection can be obtained in PV (Cho et al. 2012). Yet, a little change in rejection caused by temperature was observed in Figure 5. The factors enhance the flux, showed adverse effects on rejection. Although the kinetic diameters of salt ions were higher than that of water, they could be drifted with water molecule through the enlarged free volume of polymer. Therefore, salt rejection decreased with temperature increment by the pristine membrane particularly. Because of the rigid and crystalline structure of clay, composite membranes showed good rejection performance and did not much affected from the 354 Fig.7. Effect of NaCl concentration on salt rejection factor (40 ºC) III.5. Separation Performance of Marmara Seawater After the separation results of pristine and composite

369 membrane were evaluated, Marmara seawater desalination was done by using 10 wt. % clay loaded CMC membrane at 40 ºC. At the end of the three hours, 99.9 % purity water with 1.85 kg/m 2.h flux was achieved. Pervaporative desalination test was applied for three times at same conditions with same membrane. After three times, membrane decomposed. Thereby, it was evaluated that the membrane worked perfectly but it was needed to improve durability. Joao C.D.C.,Long term pervaporation desalination of tubular MFI zeolite membranes, Journal of Membrane Science, 415, , (2012). Bin L., Wu Z., Genggeng Q., Sensen L., Qian N., Yuxuan L., Bing C., Kai P., High performance graphene oxide/poly acrylonitrile composite pervaporation membranes for desalination applications, J. Mater. Chem. A, 3, , (2015). IV. Conclusions In present study, selective desalination capability of cellulose based pristine and composite membranes were investigated. The results showed that the clay incorporation improved both flux value and salt rejection in all conditions. At low temperature, above 99.8 % rejection was achieved by using 10 wt. % Na + MMT loaded CMC membrane accompanied by 2.1 kg/m 2.h flux value. Rejection was not affected from the salinity of feed solution. Marmara seawater was desalinated to pure water with 99.9% excellent rejection performance. Hence, it was conclueded that the pervaporation process is a very good candidate for seawater desalination and composite cellulose membrane could be used commercially after some improvements were performed. Acknowledgements This research was supported by the Scientific Research Project Center of Kocaeli University. Nomenclature C : Concentration CMC : Carboxy methylcellulose DS : Degree of swelling (%) GA : Gluteraldehyde J : Flux (kg/m 2.h) Na + MMT: Sodium montmorillonite NaCl : Sodium chloride PV : Pervaporation R : Rejection W : Weight (kg) References Paul S., Brenden T., Ankita G., Weizhu A., Steven M. K.i., Pervaporative desalination of water using natural zeolite membranes, Desalination 285, 68 72, (2012). Komgold E., Korin E., Ladizhensky I., Water desalination by pervaporation with hollow fiber membranes, Desalination 107, , (1996). Churl H.C., Ka Y.O., Si K.K., Jeong G.Y., Pankaj S., Pervaporative seawater desalination using NaA zeolite membrane: Mechanisms of high water flux and high salt rejection, Journal of Membrane Science, 37, , (2011). Martin D., Christelle Y., Julius M., Anne J., Liping D. 355 Xie Z., Hybrid Organic-Inorganic Pervaporation Membranes For Desalination, PhD Thesis, Victoria University, Melbourne, (2012). Muthia E., Christelle Y., David K.W., Simon S., Joao C. D. C., Microporous Silica Based Membranes for Desalination, Water, 4, , (2012). Bart V.B., Carlo V., Distillation vs. membrane filtration: overview of process evolutions in seawater desalination, Desalination,143, , (2002). Akili D. K., Ibrahim K. K., Jong M.W., Advances in seawater desalination technologies, Desalination, 221, 47 69, (2008). Nur M.M.,Dimitar P., Andrew G. L., Energy consumption for desalination A comparison of forward osmosis with reverse osmosis, and the potential for perfect membranes, Desalination, 377, , (2016). Misdan N., Lau W.J., Ismail, A.F., Seawater Reverse Osmosis (SWRO) desalination by thin-film composite membrane Current development, challenges and future prospects, Desalination, 287, , (2012). Sabine L., Thomas H., Environmental impact and impact assessment of seawater desalination, Desalination, 220, 1 15, (2008). Avlonitis S. A., Avlonitis,D. A., Panagiotidis T., Experimental study of the specific energy consumption for brackish water desalination by reverse osmosis, Journal of Energy Research, 36, 36 45, (2012). Xing Y.,Hiroki N.,Hiroyuki S., Yukio T., Kazuhiko K., Ryutaro H. Study of an incrementally loaded multistage flash desalination system for optimum use of sensible waste heat from nuclear power plant, Journal of Energy Research, 37, , (2013). Schafer T., Carla M. R., Carlos A. M. A., Joao G. C., Selective recovery of solutes from ionic liquids by pervaporation a novel approach for purification and green processing, Chem. Commun., , (2001). Dabrowski A., Hubicki Z., Podkoscielny P., Robens E., Selective removal of the heavy metal ions from waters and industrial wastewaters by ion-exchange method, Chemosphere, 56, , (2004).

370 Dali Y., Sudipto M., Suphan K., Kamalesh K. S., Hollow fiber contained liquid membrane pervaporation system for the removal of toxic volatile organics from wastewater, Journal of Membrane Science 103, , (1995). 356

371 Architecture in the Net Zero Houses of the Future Okay Gonulol, Ayca Tokuc * Dokuz Eylul University, Faculty of Architecture, Department of Architecture, Tinaztepe Yerleskesi, İzmir, 35160, Turkey * ayca.tokuc@deu.edu.tr Abstract To live in a healthy, comfortable and safe environment is a right for all human beings, yet the sustainability of this habitat requires proper utilization of natural resources. There are many concepts in architecture that take into account this basic assumption including; nearly zero energy, solar, green, sustainable, integrated, intelligent and zero carbon. In addition, there are many national and international directives and initiatives that promote and require constructions with a lower impact on the environment; however this is not the only issue in architectural design. Some of the other essential topics in architecture are structural soundness, affordability, aesthetics, ease of construction, the ability to deal with natural disasters etc. Therefore the design and construction of a building requires a team of professionals and customers to work together. This paper investigates the state of the art architectural technologies in net zero energy houses of the future by looking at the built examples from the Decathlon 2015 Competition, which are designed to be net zero energy as well as dealing with all the aforementioned topics. The methodology involves determination of architectural characteristics, other energy efficiency solutions and comparison studies between houses. The architectural characteristics, building elements, passive strategies, active strategies, solar energy, and other strategies are detailed for six competition houses that took the highest points in the Energy Balance category by measured results. Furthermore, state of the art technologies, which were integrated into the buildings, are also discussed in light of the measured energy efficiency of the houses. In conclusion, the common characteristics and individual technologies of the successful and not so successful net zero houses are given so that they can help light the way to more energy efficient constructions. Keywords: Net zero house, architecture, competition, energy balance, energy strategies. I. Introduction The building sector is a large energy end-use sector that accounts for a larger proportion than both the industry and transportation sectors in many developed countries in terms of total energy consumption (Yang et.al, 2014). Many countries are aware of the essential role of the building sector in energy consumption and take precautions against it in their energy policies, one of which is the European Union s Nearly Zero Energy Buildings Directive that requires all new buildings to be designed as nearly zero energy by the end of Even though this target is ambitious, it is possible to significantly decrease the energy requirements of buildings by the use of appropriate design and technologies (Cabeza et.al., 2014). Residential buildings constitute 75 percent of the total building stock on a European average (Economidou, 2011). Due to their quantity they are the leader in both energy consumption and CO2 emissions, 61% of CO2 gas emissions is related to residential buildings (Voss & Musali, 2011). The concern for energy use in residential buildings started after the energy crisis in the 1970 s and led to actions and programmes aiming to rationalize the energy consumption of dwellings (Xing, Hewitt & Griffiths, 2011). Chandel et.al. (2016) provide a review of energy regulations of 17 countries energy efficiency initiatives and regulations for residential buildings. One of the most significant strategies that are used to spread and accustom energy efficient practices and make the future architects and engineers learn energy efficient design practices is architectural student competition such as the Solar Dechatlon. In the Solar Decathlon 2015 Competition, state of the art architectural technologies were used to build net zero energy houses. This paper aims to investigate the integration of these technologies into the houses while creating high quality environments that simultaneously deal with various other concerns such as structural soundness, affordability, aesthetics, ease of construction, natural disasters etc. The teams of Texas/Germany, U at Buffalo, Mass/Central America, UC Davis and Team Orange County and NY City Tech were ranked in first five places of the Energy Balance section according to in place measurements. These examples are studied and compared according to their architectural characteristics, building elements, passive strategies, active strategies, solar energy and other strategies. In addition how young architects and engineers look at these problems is observed. 357

372 II. Solar Decathlon 2015 The U.S. Department of Energy Solar Decathlon is a program which challenges teams cooperated by universities, to design, build, and operate solar-powered houses that are cost-effective, energy-efficient, and attractive at the same time. The winner team is the team that blends affordability, consumer appeal, and design excellence with optimal energy production and maximum efficiency by the best way (Solar Dechatlon, 2015). Solar Decathlon was held in 2002 first time. Since 2005 till 2015, it has been organized biannually. The contestant houses are built on an area that is defined by competition organizators and they are exhibited for 2-3 weeks for the visitors who want to explore them. According to a research report in 2009, 500,000 visitors have visited the exhibition houses from 2002 to The exhibition served different goals and effected various experiences for homeowners, participating-student and solar-related researchers. The Decathlon also contributed some non-attending fields like market acceleration and public awareness (Martin, 2012). The latest Solar Decathlon, held in 2015, was organised at Orange County Great Park, Irvine in California State on 8-18 October, This contest has 10 sub contests divided into two categories, which are measured and juried. Juried sub-contest are examined and scored by juries. They are; Architecture, Market Appeal, Engineering, Communications and Affordability. Measured sub-contests are graded according to measurements taken from the built houses on the days of the exhibition and have some minimum limits; if teams can t reach these limits they are eliminated. The sub-contests are; Comfort Zone, Appliances, Home Life, Commuting and Energy balance. In the Solar Decathlon 2015, out of the 17 teams selected to build a house for the exhibition, only 14 were ranked. III. Energy balance The Energy Balance Sub-Contest of the U.S. Department of Energy in the Solar Decathlon 2015 was divided into two sections, which are energy production and energy consumption, each of which was worth 50 points. balance competition and their overall standing are shown on Tab.1. It can be clearly noticed that the standings of teams on energy balance are not depended on, and are usually compeletely different from, the overall standing of a team. This data shows that only energy balance competition leadership is not enough to get first standing on overall standing, thus teams need to give importance to points from other areas of competition. The first five ranked teams on energy balance competition are investigated in this paper. These teams are; Texas/Germany, U at Buffalo, Mass/Central America, UC Davis, Team Orange County and NY Tech. Since the first place is shared by two teams, there are six teams in the first five places. A short description is given about each of these six teams regarding their energy performance and relationship with architectural and technological solutions. Tab. 1: Energy balance and overall standings of teams Teams Energy Balance Competition Overall Standing Texas/Germany T U at Buffalo T Mass/Central America UC Davis Team Orange County NY City Tech Stevens Missouri S&T Cal Poly Crowder/Drury West Virginia/Rome Team NY Alfred Clemson Sacramento State Austin, Texas is a fast growing city and this growth is putting a strain on its infrastructure, especially water and electricity. The University of Texas has partnered with Muenchen University for this competition and designed NexusHaus. The house provides a model for accessory housing in Austin, which aims to collect most of its own water in a closed loop aquaponics system. For energy production, a team would receive full points for producing at least as much energy as its house needed, thus achieving a net energy consumption of zero during the competition. Points would be reduced for a net electrical energy balance between -50 kwh and 0 kwh. For energy consumption, a team whould receive full points for using 175 kwh of energy or less throughout the competition. Points would be reduced for consumption between 175 kwh and 300 kwh. (Solar Dechatlon, 2015) Points of contestant teams taken from energy 358 Grow home is designed by University at Buffalo and shares the first place with The University of Texas on energy balance standing while it has taken second place on general standing. Grow stands for the first letters of Garden Relax or Work. In Grow home, urban living and self sufficiency concepts are reevaluated and dynamic, seasonally changing elements are used. Western New England University, Universidad Tecnológica de Panamá, and Universidad Tecnológica Centroamericana have teamed up to design EASI house, which has taken 14th place on

373 general standing and second place on energy balance standing. EASI stands for Efficient, Affordable Solar Innovation. The house was designed for New England climate. UC Davis has an agricultural tradition; therefore it aimed at creating sustainable space at an affordable price, mainly for underserved farmworkers and other communities. The vibrant ethnic diversity of the New York City and the effects of global warming led students of the New York City College of Technology to design the DURA house, short for Diverse, Urban, Resilient, and Adaptable. The house was designed for a lot that was damaged by the Hurricane Sandy and aims to mitigate disaster damage. Since the area is a dense urban environment, their design can be stacked up to four stories. The University of California, Irvine; Chapman University; Irvine Valley College; and Saddleback College have teamed up and designed Casa del Sol that has taken 9th place on general standing and fourth place on energy balance standing. III.1. Architectural characteristics NexusHaus is a prototype modular home that consists of two rectangular living modules (a module for day use and one for night use) and a nexus, which connects the two modules. In the day module, the living, kitchen and dining space are combined. A thick wall on the west provides space for seating and cooking elements. The night module has two bedrooms and a shared bathroom. A glass wall brings the outdoors in. The textile roof on the deck creates a flexible outdoor area for living. Sliding doors in the nexus allows for variations in seasonal living. The house can provide outdoor living for three seasons of the year. Grow home consists of dry and wet modules, which compose a T shape. The wet module incorporates bathroom, kitchen and mechanical room, while the dry module has the bedroom, living room and growlarium. A canopy made of steel frames covers the roof of the Grow Home. That canopy also makes up the Growlarium part of the house. collects water while reducing the gutters. This water collection plays a prominent role for the active strategies of the house. The DURA house has flexibility as its main defining characteristic. The rooms can be configured to accommodate a number of different living styles and functions. A room is created by two movable walls and adaptable furniture within the room can be converted from a desk to a bunk bed. The building is elevated by footings so that water would pass without obstruction in case of a hurricane or rising sea levels. The structure is durable against gale force winds and seismic activity. Casa del Sol has a square-like shape, its living spaces haven t exact quantity and borders because of this house s flexible design. The house has two bathrooms and one kitchen, which also do not have exact borders and can be extended by its users. The young is the target client therefore users can have large empty spaces when they have parties and have extra rooms when they have visitors. Tab. 2: Spatial room distributions and flexibility. Teams Plan Type Flexibility Texas / Germany - NexusHaus U at Buffalo Grow Mass/ Central America - EASI UC Davis - AggieSol Team Orange County - DURA Rectangular shaped EASI House has two bedroms, a kitchen, a bathroom, a technical room and a living area. Kitchen is at the centre and is intimately linked with living space. By doing this living space can be heated by taking advantage of heat comes from kitchen. Bedrooms of the house are located end spaces of rectangle. One of the most prominent problems of farmworkers is the accumulation of debris inside their living space, to which AggieSol proposed a separate entry named the cleansing room with bath, shower and double sided lockers and air purifying plants inside. The house has a butterfly shaped roof that drains and 359 NY City Tech Casa del Sol Living and kitchen Bedroom Bath Semi Open Technical Hall Flexibility Vegetation From the six houses, only Mass/Central America and NY City Tech teams prefered a rectangular plan shape. The other four teams prefered sqare-like plan shapes (Tab. 2). The square plan allows for more compactness, thus reducing winter heat losses and summer heat gains. This compact form is usually broken by using semi-open areas and flexibility;

374 therefore natural ventilation would flow through the house. III.2. Building elements The NexusHaus has a steel structure with a wooden deck, yet its outlook is mostly of wood, glass and greenery. The canopy is covered by textile. In Grow home, roof and walls of wet module and dry module except Growlarium are structural insulated panels (SIP). SIP used on the wall and roof have nearly 26 cm thickness and 0.34 U value. SIP walls consist of EPS insulation between of OSB boards. Windows and sliding doors of the Growlarium are highly insulated and have triple tempered, air filled, low-e glazings (Bohm, 2015). EASI house has a wooden frame which is made of SPF studs, R-19 fiber glass insulation and OSB boards. House s exterior wall thickness is 12.4 cm. Insulation on exterior walls make the U value of the walls Same insulation material with greater thickness has been used on roof. Windows of the home have low-e glazings (Lee, 2015). AggieSol uses balloon framing with 24 axial stud placement instead of the traditional 16, therefore it uses 10% less lumber while making space for 5% more insulation. Additionally aligning the roof, wall and floor joists allows for 15% less material and 10% less labor thus reduces related costs. The materials used such as zero VOC linoleum flooring were selected with care (Good, 2015). The students of NY Tech have designed a highly-insulated, air-tight, yet vapor permeable wall section for the DURA house. The placement of the openings were carefully simulated and optimised, coupled with careful construction the envelope showed high performance during the competition. Exterior walls of roof of Casa del Sol have 20.3 cm thickness (without exterior fiber cement panels). These walls include 14 cm mineral wool and nearly 5 cm EPS insulation. Each side was covered with plywood sheeting. Window and sliding doors have low-e glazing filled with argon gas (McDonald, 2015). III.3. Passive strategies The solar heat gain in summer is a huge problem for Texas, therefore NexusHaus has a large deck on the south side that is covered against the sun. In addition the nexus can be opened for ventilation. During the winter, the nexus is closed for both privacy and thermal buffer zone. The thermal/water storage system is integrated into the house. Growlarium of the Grow home is enclosed with glazings and sliding glasses. Glasses on the side surfaces of Growlarium are highly insulated and have a high U-value, meaning they don t cause air leaks. In summer times, shaders, placed on the roof of the 360 Growlarium helps shading inside and doesn t let overheating. Sliding glasses helps the natural ventilation by the help of little openings that are placed on the kitchen wall. In the winter time, Growlarium turns into a buffer zone for living space by closing insulated, low-e glazings and removing the shades that are placed on the roof. Sliding doors that separate the living room from Growlarium help to control the buffer zone. Another passive system of the house is its canopy, which encloses the house, shades the house and the deck to reduce the cooling loads (Bohm, 2015). AggieSol utilises low window to wall ratio, operatable exterior window shades, light pipes, light colored exterior, and thermal mass. Thermal mass is present in both the granite countertops and the gypsum concrete radiant floor. Flexible reflective tube from the roof leads sunlight into the house without accompanying solar heat gains. The thermal mass in the radiant slab is not very useful for reducing heating load but it would help to decrease the cooling load by 11.4% (Good, 2015). DURA haouse has a highly insulated building envelope and passive ventilation. The designers used the house design guidelines from Milne et.al, 2009 and incorporated various design elements into the house including thermal insulation and internal heat gain, passive solar direct gain for low mass, sun shading of windows, wind protection of outdoor spaces, high thermal mass interior surfaces, and a heat chimney. The climatic conditions analysis showed that the outdoor conditions were comfortable only 10.7% of the total yearly hours and the use of passive strategies increased this ratio to 46% of yearly comfortable hours (Aptekar, 2015). Casa Del Sol s passive solar system is named The California State flower by the design team because it opens and closes as it adjusts to the light of the sun. The outdoor living room is connected to the surrounding neighborhood by pivot panels. Being in the northern hemisphere, the sun hangs in the southern sky. To protect residents from the sun and moderate solar heat gain, southern shading elements have been placed over windows and living spaces such as in the tensile structure above. Unsheltered windows on the western side of open to and invite in prevailing cool ocean winds. Automated windows serve to naturally ventilate living spaces. A brise soleil is built up on the eastern side to protect occupants against warm, violent Santa Ana winds, which often strike in October; this system is also attached on the other side of the detached studio. III.4. Active strategies The modules in NexusHaus are air conditioned. Thermal storage for a district cooling system is proposed with a nearly 4m water tower that is cooled down at night by electricity (Garrison, 2015). Ductless ventilators precondition air coming into the modules. In the early morning, the air conditioned air passes

375 through the water that was cooled through the night and is piped throughout the house. In NexusHaus, the solar power, thermal/water storage system and a smart home management system are integrated to the house. The smart system visualises how the occupant behaviour would affect the energy and water use in the house. The homeowner can read information, visualise different scenarios and control the environmental conditions inside the home. Grow House has an air-to-air heating ventilation air conditioning (HVAC) system, which has four zone variable air volume. This system ensures maximum energy savings by only conditioning and distributing air to the needed spaces at needed time within the house. Heating and cooling demand of EASI house is solved by a ductless heating and cooling system that is 9,000 Btu/h, 1 kw and is centrally located to provide maximum comfort throughout the house. This system can be moved during extreme weather conditions for increased comfort in individual areas (Lee, 2015). The primary heating and cooling system of the AggieSol house is the radiant floor system that uses water from a large rainwater reservoir. Throughout the night, the stored water is sprinkled to the roof, cooled by radiant/sky radiation, recollected, and filtered. In the day it is pumped through the slab to cool down the house. The water can also be heated and repumped through the slab for radiative heating. BEopt and Excel simulations suggest that the combined radiant/night sky system would reduce space cooling by around 3500 kwh. A greywater heat recovery system extracts energy from the outgoing water, which would reduce the cooling load further, around 1400 kwh. It would also be used to preheat the potable water for the domestic supply. The active systems of the DURA house include a hydronic heating system that uses thermal energy from the solar panels, a super energy-efficient energy recovery ventilation triple-action air filtration and minisplits. Thermal electric panels convert waste heat from the photovoltaic panels into electricity. Temperature, humidity and lighting levels are monitored and regulated by the home automation system. The Casa del Sol has a low entropy ventilator to reach HVAC requirements. A home automation system monitors various criteria such as temperature, humidity, light and motion. Therefore it regulates the ventilator, façade elements, lighting and shading for better energy balance. III.5. Solar energy The electricity provided by the solar panels on the NexusHaus in the afternoon is fed to the grid and then energy from the grid is used to cool down the water tower for air conditioning. The other systems in the house reduce peak electricity consumption by 361 almost 80% against a traditional Austin house (Garrison, 2015). Grow home has 24 photovoltaic panels with 7 kw energy production power. The system s rated efficiency is 17.2%. It is planned to produce 9 MWh of electrical energy per year when it is installed in Buffalo after the competition. That means the solar system will produce nearly twice the energy the house will consume (Bohm, 2015). The solar electric system set up in EASI house can produce 5kW energy with twenty panels, each with 250W power, and is mounted on the roof. Water heating is supplied by an 11kW tankless water heater that heats on-demand to decrease the electricity load. AggieSol uses 42.5% less energy than a conventional home and is 1.9% energy positive throughout the year. The appliance selection was done by an appliance scoring equation that evaluates efficiency, cost, availability and aesthetics, thus reducing the yearly energy load by 400 kwh. The sprinkler cooling system that operates during the night also cleans the photovoltaic panels so that they would perform without related performance losses (Good, 2015). In the DURA house, the photovoltaic panels are used as façade elements that emphasize the use of active technologies to the outsiders. The seasonal optimisation of the solar panels is done by building integrated tracking solar shutters. Moreover, manual individualised operation in accordance with the users lifestyles is possible. Casa Del Sol has twentyfour photovoltaic panels, each with 305W production power. A sub wet-bulb evaporative chiller is used to chill water that runs in the home s ceilings and a heat pump provides air conditioning. To increase efficiency, the heat pump rejects its waste heat into the hot water tank. A water to water heat pump works with the solar thermal collectors to provide hot and cold water for the radiant ceiling and domestic water systems. The water-to-water heat pump uses electricity to transfer the heat from the water to the house during heating, and transfer heat from the house to the water during cooling, all without generating its own heat. Casa del Sol also has solar thermal collectors that help to provide heat for its dryer besides providing the house s hot water. (McDonald, 2015). III.6. Other strategies NexusHaus aims for being a net zero water house i.e. a water independent home in addition to being a net zero energy house. The house is designed to run only on water from direct rainfall captured by the shading structure and the greywater produced on site. The home aquapodonics system uses 90% less water than traditional food production and the landscape elements around the house are selected

376 such that they can be used as food. In addition, an underfloor purification and storage system provides potable water for indoor use. The house requires connection to the central water system only for times of long dry spells (DOE Solar Decathlon, 2015). Grow house does not only use the sun to produce energy but also to save the energy when basic needed vegetables are grown and delivered to user from the greenhouse in the Growlarium. In this part users can grow their own vegetables like onions, pepper, tomatoes etc. on portable tables. The house also has solar collectors to produce hot water. AggieSol is designed to be easily constructed in a factory and shipped in place by two trucks, one for the north and one for the south part of the building. The manufacture and shipping cost was calculated as 150$ per square foot. In DURA, there are interior and exterior green walls that help to purify air, provide evaporative cooling and humidity. Greywater is used for toilets and harvested rainwater is stored in tanks for watering purposes. However in extreme cases such as a natural disaster, greywater could be filtered for watering or drinking purposes. When it rains, rainwater from the roof of Casa Del Sol is caught and diverted into multiple rainwater tanks installed on the side of Casa del Sol, allowing the water to be stored for future use, including gray water systems, emergency potable water or flushing toilets. IV. Conclusion The Solar Decathlon competition addresses the issues of sustainable and affordable urban housing within a net zero house concept. All of the examined houses were designed to deal with various issues and their architectural characteristics, building elements, passive strategies, active strategies, solar energy, and other strategies all work together towards the same goals. At the same time, exhibition houses have the capacity to catalyse change in the housing industry towards more sustainable practices (Martin, 2012). Many students from construction, architecture and engineering were involved in a project that required all of their input and insights, which is an invaluable experience. From Fig.1, it is clearly seen that neither being the most energy efficient nor having the best design is enough to win this competition. All sub-goals should be achieved to reach the upper ranks. The best example of that stuation is Mass Central. They are in the third place in terms of energy balance yet when it comes to overall standing they are only on the 14 th place. This rating style does not let the teams to care about only one part of the architectural design process but also requires them to care about other criteria such as aestehics and user satisfication. Fig. 1: Overall standing of Solar Dechatlon 2015 showing points from sub-categories These examples contain the best practices; well designed, sustainable and economical. It could be more meaningful to add houses whom points are lower on previous works. In addition, some other strategies can be investigated. These projects are affordable and buildable, and they show that sustainability is not an idea for the future; it can be available for everyone. Yet the way to accomplish this is not clear however the Solar Decathlon Student Competition, which prepares the future generations for such a world, is a step in the right direction. References Aptekar A., Project manual, Team NYCCT. U.S. Department of Energy Solar Decathlon (2015). Bohm M., Grow Buffalo, As buılt project manual. U.S. Department of Energy Solar Decathlon (2015). Cabeza L.F., Rincón L., Vilariño V., Pérez G., Castell A., Life cycle assessment (LCA) and life cycle energy analysis (LCEA) of buildings and the building sector: A review, Renewable and Sustainable Energy Reviews, 29, (2014). Chandel S.S., Sharma A., Marwaha B.M., Review of energy efficiency initiatives and regulations for residential buildings in India, Renewable and Sustainable Energy Reviews, 54, (2016). Design Strategies derived from a series of studies by Murray Milne, Robin Liggett, Andrew Benson, and Yasmin Bhattacharya UCLA Department of Architecture and Urban Design Climate Consultant 4.0 Develops Design Guidelines for Each Unique Climate ASES09-Milne.pdf DOE Solar Decathlon, Texas/Germany Audiovisual Presentation SolarDecathlon th January 2016 (2015). Economidou M., Europe's buildings under the microscope. Buildings Performance Institute Europe (2011). 362

377 Garrison M., Project manual, Nexushaus. U.S. Department of Energy Solar Decathlon (2015). Good R., AggieSol, Project Manual. U.S. Department of Energy Solar Decathlon (2015). Lee K., EASI House, Project Manual. U.S. Department of Energy Solar Decathlon (2015). Martin L., Impact Evaluation of the U.S. Department of Energy s Solar Decathlon Program, DOE Office of Energy Efficiency and Renewable Energy (2012). McDonald A., Casa Del Sol, Project Manual. U.S. Department of Energy Solar Decathlon (2015). Solar Dechatlon, The U.S. Department of Energy Solar Decathlon. 1th January 2016 (2015). Voss K., & Musali E., Net zero energy buildings: international comparison of carbon-neutral lifestyle. Munich: Birkhauser (2011). Xing Y., Hewitt N., & Griffiths P., Zero carbon buildings refurbishment A Hierarchical pathway. Cardif University (2011). Yang L., Yan H., Lam J.C., Thermal comfort and building energy consumption implications A review, Applied Energy, 115, (2014). 363

378 Control System for a Novel Photobioreactor in the Building Envelope Gulden Kokturk 1, Ayca Tokuc 2*, Anil Unal 1 1 Dokuz Eylul University, Faculty of Engineering, Department of Electrics and Electronics Engineering, Tinaztepe Yerleskesi, İzmir, 35160, Turkey 2 Dokuz Eylul University, Faculty of Architecture, Department of Architecture, Tinaztepe Yerleskesi, İzmir, 35160, Turkey * ayca.tokuc@deu.edu.tr Abstract Energy production and carbon neutrality are key concepts in designing for the future. The global warming debate and carbon emissions associated with the buildings make the necessity for utilizing green technologies in buildings evident. Photobioreactors (PBRs) are closed loop systems that create a medium for microalgal growth, which are photosynthetic organisms in other words they make use of carbon to generate oxygen. In addition, the algal biomass can be utilized in various energy systems. This project proposes the design of a building envelope element that is also a PBR. While there are many studies that deal with various renewable energy systems in the building envelope, there is little literature on the use of PBRs. For such an application there are many questions to answer including; which PBR system is both efficient and can be integrated into the building, which algae strain is more efficient for producing energy while staying alive in the provided medium, how can the cultivated algae be harvested, how can electrical energy be produced from the output of the system, and which component can be used to control the system parameters and outputs. The aim of this paper is the design of a control system to regulate various parameters inside the living medium of a PBR by using a programmable logic controller (PLC). In this paper, a flat panel PBR that can be integrated into the building envelope and with a tank volume surface area ratio is introduced for cultivation of Spirulina platensis. Within this scope, the control system deals with environmental conditions such as temperature, ph, carbondioxide (CO2), bicarbonate, nitrogen, and quantity of some ions inside the PBR. The system components are sensors, relays, external containers and a controller. Sensors will be used to measure the system parameters. External containers are going to be used to drain CO2, oxygen, biomass, other gasses and materials, which need to be taken from the microalgae culture, thus from the system. Relays are used for the drainage process. Keywords: Photobioreactor (PBR), control, building envelope. I. Introduction Energy production and carbon neutrality are key concepts both in the global and the local scales. While most of the energy produced today is produced from fossil fuels, the alternative energy sources are increasingly gaining attention. Photobioreactors (PBRs) and their implementation into daily life is a novel system with great potential. PBRs are used to cultivate microalgae, which are the source of nearly 50% of all O2 production (Chapman, 2013). Traditionally microalgae were cultivated in lakes, later they began to be cultivated in more efficient and controlled ponds, or under laboratory conditions. PBRs are closed loop systems that utilize a light source to grow phototrophic microorganism, in this case microalgae, cultures. Microalgae can be used in many ways to create a more sustainable earth and there are many studies that concentrate on their many application areas. They can be used a biomass source (Jorquera, 2010), a hydrogen source (Rashid et al., 2013), a carbon sink (Gonçalves et al., 2016), a valuable organic compound source (Harun et al., 2010), a biodiesel source (Chen et al., 2011) or a food source (Chacón-Lee and González-Mariño, 2010). Productivity is the most important indicator for PBR technology, yet it is very difficult to compare productivity of bioreactors due to various strains and scale of microalgae. PBRs are classified according to their shape and working principles. This paper proposes the growth of algae on the façade of a building. While there is a growing body of literature for various PBR systems, there is little literature on their use on buildings and there is only one live building application that uses a flat plate PBR. The system proposed in this paper is also a flat plate PBR. The culture conditions need to be carefully controlled and optimized to cultivate microalgae therefore modeling of microalgae is crucial to provide the required control strategy. There are several studies that make use of numerical models to estimate the behavior of microalgae under varying light intensities (Kommareddy & Anderson, 2003; Li et al., 2014), carbondioxide (CO2) (Widjaja, 2009), temperature (Sheng et al., 2011) and ph (Kumar et al., 2010). The proposed PBR can be integrated with the façade of a building. In the design phase of such a novel PBR, some of the main questions include; which PBR system is more efficient, which algae is more efficient 364

379 for producing energy and how can electrical energy be produced from the output of the system. The answers include a flat plate PBR will be used to cultivate microalgae for biomass production. Yet there are other questions to answer and the aim of this paper is to seek answers to the questions of which parameters are going to be controlled, and which components can be used to control the system parameters and outputs. After discussing the main parameters of a PBR system to be controlled, the design of a control system for temperature, liquid levels, etc. inside the PBR will be discussed. II. Design Procedure The designed PBR is used to cultivate photosynthetic microalgae. The photosynthesis diagram of microalgae is given in Fig.1. While designing a PBR system, the basic problems of mixing, light penetration, gas injection, ph and temperature control need to be considered in detail. However, all algae cannot tolerate agitation since they are sensitive to hydrodynamic stress. High mixing rate can cause damaging of the cells. Mixing in bubble column and air lift reactors can characterize with axial dispersion coefficient, mixing time, circulation time and Bodenstein number (Miron et al, 2004). Analysis of mixing in bubble column shows it has shorter time than airlift reactors. Bubbles beyond the suction pipe provide less blurry area and causes better exposure to the light. In addition, existence of suction pipe in airlift reactors causes more effective mixing because internal loop provides a circulation. Airlift reactor has higher fluid flow and gas-liquid mass transfer rate. Bubble column can cause unbalance in cell density and thus death of algae (Bitoga et al., 2011; Fan et al., 2007). II.2. Light penetration Microalgae need light for their photosynthesis therefore light penetration is another key parameter for a PBR especially for successful scale up. The microalgae growing photosynthetically needs light and the light intensity is the most significant limiting factor (Fernandes et al., 2010). Light intensity in a PBR is related to light wavelength, cell concentration, photobioreactor geometry and distance the transmitted light travels (Bitoga et al., 2011). Fig. 1: Photosynthesis diagram of microalgae ( ml) II.1. Mixing process Mixing provides approximately homogeneously distribution of light and nutrients to all cells and increases the gas transfer between culture medium and air. Mixing also helps sufficient CO2 transfer and maintains uniform ph inside the PBR In addition, mixing is necessary for preventing algae sedimentation and avoiding cell attachment to the reactor wall. Therefore, mixing increases the biomass productivity in PBR. Measurement of carbon supply for the photosynthesis process is of priority in a PBR. In very dense cultures, CO2 from air (includes % of CO2) and bubbles during the culture can limit algal growth. CO2 addition also creates a buffer against ph change in the water (Bitoga et al., 2011). If the mixing is poor, the cells clump together like different sizes of aggregates; therefore it leads 3 phase (solid-liquid-gas) system in the reactor. This situation tends to reduce the mass transfer. 365 The only light source for open ponds is the Sun. That is why alteration of raceway ponds is not possible. The depth of the pond is the only thing that can be changed. Thus mostly researches are going on closed systems to optimize light penetration. Most photobioreactors in lab scale are lightened by fluorescence lights from both inside and outside (Bitoga, 2011). Photosynthetic active radiation wavelength changes around nm, which equals the visible light (Kommareddy & Anderson, 2003). In intense cultures, light gradient changes over the photobioreactor radius due to the weakening of the light. Algae culture systems mostly use both sun and lamp light. Mostly lamp-lightened algae culture systems uses wider screens to be able to absorb more light from the system. For photosynthetic production, at least 50 % of the volume of PBR has to get enough light (Ogbonna et al., 1995). The light wavelength should be between nm to maximize photosynthesis. Light intensity depends on microalgae density. Higher algae density requires higher light intensity. Each type of microalgae has its own optimal light absorbing point. After the optimum point is exceeded, microalgae light absorption ratio decreases. After a specific point, light decreases the biomass production and this is called photoinhibition. Photoinhibition depend on time and after a few minutes of light stress, biomass loss starts, and more than 50 % damage can be seen after minutes. High light intensity limits algal growth, but has the benefit of higher lipid content and yield. It can be

380 seen in Ruangsomboon s study whose cultures exposed to low light intensity showed a higher biomass compared to others (Ruangsomboon, 2012). To increase the microalgae production, photoinhibition should be cut off or exceed to high light intense. In addition, photorespiration decreases the photosynthetic efficiency. Therefore the process has to avoid photorespiration. Photorespiration occurs when the oxygen concentration increases depending on the amount of CO2 (Bitoga, 2011). II.3. Gas injection CO2 is the natural carbon source for a photosynthetic microalgae culture. Oxygen release depends on the amount of carbon that is delivered to the medium. When the amount of carbon gets low, oxygen is produced by photolysis of water and is released to medium, thus CO2 input to the PBR is important. Sonnekus reported that the CO2 should make up % of the total gas flow and being careful about the CO2 input does not lower the ph of the culture (Sonnekus, 2010). Supplement of CO2 by bubbles is an important factor to be considered in designs. Injection of CO2 bases on giving CO2 to photobioreactor artificially. Researches show that rich ventilation of CO2 provides CO2 to algae, supports deooxygenation of suspension, to improve cycling provides mixing and limits the light inhibition (Zhang et al., 2002). But high ventilation rate leads to higher cost that is why in large scale of microalgae production it is not recommended. These researches results for microalgae production necessary optimum aeration rate of CO2 gas. Includes about 5% or 10% of CO2 (v/v), rate of vvm (Zhang et al., 2002). Volume of air/medium/ time is found cost effective for air mass culture (Bitoga, 2011). The amount of CO2 required for the growth relates to type of microalgae and PBR. Some types of algae strains are able to keep growing in high CO2 conditions; however for faster growth lower CO2 concentration is required. Virthie Bhola et al. reported in their studies that at 15% CO2 concentration there is a 3-fold decline in biomass yield when compared to the yield produced at a 4% CO2 concentration. This suggests that the strain under study could not endure CO2 concentrations greater than 4% (Bhola et al., 2011). Also Ebrahimzadeh et al. reported that increasing CO2 injection had a significant effect on microalgae growth (Ebrahimzadeh, 2011). Widjaja studied the effect of CO2 on growth and it was seen that this effect correlates directly to the lipid productivity since growth was enhanced tremendously by increasing the CO2 concentration (Widjaja, 2009). Algae can live in high CO2 concentration and other greenhouse gases such as nitrogen dioxide can become a food for algae. The exhaust gases from fossil fuels can not only feed algae production 366 facilities but also increase their efficiency. Research on the use of stack gases as a carbon source was done but the toxicity of the stack gas components could not be well documented (Bitoga et al., 2011). II.4. Temperature Algal growth also depends on temperature. For maximum growth there is a need to know the optimal temperature. Optimal temperature for microalgae cultures is usually between C. This can be different according to medium composition, type of culture and strain. Generally, the most cultured microalgae can tolerate the temperature between 16- and 27 C. The temperatures lower than 16 C will increase the duplication time and higher than 35 C will have a fatal effect on algae (Bitoga et al., 2011). However, these ranges can be changed by environmental factors such as salinity, ph, carbon dioxide etc. Temperature also changes the lipid production and composition (Sheng et al., 2011). The degree of unsaturation of algal membrane lipids increases if the cultures are maintained below their optimum temperatures (Sivakumara et al., 2012). Besides, temperature is significant for the solubility of carbon particles, which helps carbon to be used for photosynthesis. Temperature also effects respiration and photorespiration more than photosynthesis. However, if CO2 and light penetration are the limiting factors, the effect of temperature is insignificant. II.5. ph Microalgae medium require different ph values according to the culture and strain. The most common ph range for algal growth is around 7-9. The optimal ph for algae is between But it can change with different strains. There is a complex relationship between CO2 concentration and ph in microalgal bioreactor systems, owing to the underlying chemical equilibrium among such chemical species as CO2, H2CO3, HCO3 and CO3. Increasing CO2 concentrations can increase biomass productivity, but will also decrease ph and this causes important effect upon microalgal physiology (Kumar et al., 2010). In high ph concentration, the CO2 might be the limiting factor for growth and photosynthesis. Water contaminated with a high ph has negative effects on algal abundance (Bergstrom et al., 2007). If there is not enough CO2 gas supply, algae will utilize carbonate to maintain its growth (Widjaja et al., 2009). Although high concentration of carbon dioxide provides high biomass efficiency, on the other side higher contamination risk and effect of low ph on microalgae physiology occurs (Bitoga, 2011). Aside from the parameters mentioned above; there are also some other parameters, which affects algal growth and lipid accumulation. Nitrogen, phosphorus

381 and salinity are examples for these parameters (Moisander et al., 2002). Widjaja et al. studied the effect of nitrogen starvation on lipid accumulation. They reported that longer time of nitrogen starvation obviously results in higher accumulation of lipid inside the cells. Under all CO2 concentrations, the lipid content tends to increase, when the algae is exposed to nitrogen starvation and the total lipid content is higher than lipid obtained during normal nutrition conditions (Widjaja et al., 2009). Ruangsomboon found the highest biomass concentration under the highest phosphorus concentration (Ruangsomboon, 2012). Xin et al. have reported that lipid productivity was not at its highest when the lipid content was highest under nitrogen or phosphorus limitation (Xin et al., 2010). Yeesang and Cheirsilp also studied about nitrogen and salinity effect. They found an increase in algal biomass under nitrogen-rich condition for all strains and in the absence of a nitrogen source, no growth was observed. They reported that although some loss in algal biomass was found, the lipid contents of four strains increased. They also noticed that growth and lipid accumulation by these microalgae could be affected by salinity. Under nitrogen-rich condition, all strains survived at high salinity but growth of some strains decreased (Lin & Lin, 2011; Yeesang & Cheirsilp, 2011). III. Design The PBR proposed in this paper is flat plate, which offers advantages of high cell density, low energy consumption, high mass transfer capacity, reduction of oxygen increases, high photosynthetic efficiency, less dark volumes compared to other PBRs, lower dissolved oxygen concentration compared to the horizontal tubular PBRs, easy cleaning. In a suitably designed system, maximum cell mass and high photosynthetic activity can be achieved, but it also has some limitations (Borowitzka, 1999; Ugwu et al., 2008; Hu et al., 1996; Ramos de Ortega & Roux, and Khan, The selection of the algae species for this PBR was another challenge. We decided to use Spirulina platensis both for its high fertility rate in flat plate systems and its durability against environmental conditions such as temperature, humidity, ph quantity, etc. The materials that the Spirulina platensis culture requires are mainly; CO2, Bicarbonate, Nitrogen source (Ammonium chloride, urea), Phospate ions, Sodium-chloride ions, Magnesium ions, Copper ions, Chlorine ions, Sulfate ions. Volume-surface area ratio is a significant parameter for the algae culture in a flat panel PBR, however there are various parameters that the algae require to live. Some of the parameters that need to be controlled to satisfy the best medium for Spirulina platensis include: ph value of the gas inside the panel, ph value of the culture, Mass of the culture, Gas pressure in the panel, Mineral quantities in the panel, Amount of CO2 in the panel, Amount of oxygen in the panel, Panel temperature. Therefore a controller panel was required to regulate various elements inside the growing medium. In this study, the parameters of flat panel PBR temperature and liquid level in the tank are simulated using Matlab Simulink. The block diagram of the control system is given in Fig. 3. The reference value for temperature is determined as C, and the reference value of liquid level used in the system is 1 m. As seen in the block diagram, data is taken from the PBR by sensor, later it is reorganized in the programmable logic controller (PLC) and then directed to the output. When temperature increases inside the tank medium, the controller will set the unit that will decrease the medium temperature; likewise when temperature decreases inside the tank medium, the system will set the unit that will increase the medium temperature. The liquid level inside the PBR is also regulated by the PLC unit with the same logic. Fig. 2: Flat panel PBR schema (Bahadar and Khan, 2013) 1986; Iqbal et al., 1993). A schema of the designed flat plate PBR is given in Fig.2. We decided to create the PBR of glass with 90 cm lenghth, 2.6 cm internal width and 70 cm height as investigated by Bahadar 367 The system components are sensors, relays, external containers and a controller. Sensors are used to measure the system parameters. External containers are used to drain CO2, oxygen, biomass, other gasses and materials which need to be taken from the microalgae culture, thus from the system. Relays are use to do the drainage process, and a PLC system is used as the controller.

382 area of interest for many researchers. Photobioreactors (PBRs) are not yet popularized as a clean energy source yet they show promise as indicated by the growing body of literature. The use of PBRs in a building envelope is a new research area. This paper gives a design process for the use of PBR in the building envelope. Two of the main parameters for the living medium of algae in a flat panel PBR are temperature and liquid level. In the proposed design, these parameters are simulated and controlled with programmable logic controller (PLC). The simulation results show us that these parameters can be well controlled with PLC in a flat panel PBR. In future work, the desiged system will be set up inside laboratory conditions and all of the relevant parameters- detailed inside this paper- will be measured, analysed and controlled by the PLC. References Bahadar A., Khan M.B., Progress in energy from microalgae: A review, Renewable and Sustainable Energy Reviews, 27, (2013). Fig. 3: Control diagram of the flat panel PBR system The temperature data of the controller is given in Fig.4. The temperature is regulated inside the PBR according to the desired regime in a short time as can be seen from the figure. The liquid level control output is given in Fig.5, showing that the liquid level stabilised in a very short time. Fig. 4 : Output of the temperature control system Bergstrom C., McKeel C., Patel S., Effects of ph on algal abundance: a model of Bay Harbor, Michigan, Biological Station of University of Michigan (2007). Bhola V., Desikan R., Santosh S.K., Subburamu K., Sanniyasi E., Bux F., Effects of parameters affecting biomass yield and thermal behaviour of Chlorella vulgaris, Journal of Bioscience and Bioengineering, 111 (3): (2011). Bitoga J.P., Lee I.-B., Lee C.-G., Kim K.-S., Hwang H.-S., Hong S.-W., Seo I.-H., Kwon K.-S., Mostafa E., Application of computational fluid dynamics for modeling and designing photobioreactors for microalgae production, Computers and Electronics in Agriculture, 76: (2011). Borowitzka M.A., Commercial production of microalgae: ponds, tanks, and fermenters, Progress in Industrial Microbiology, 35 (C): (1999). Chapman R.L., Algae: the world s most important plants an introduction, Mitigation and Adaptation Strategies for Global Change, Volume 18, Issue 1, pp 5-12 (2013). Chen C.-Y., Yeh K.-L., Aisyah R., Lee D.-J., Chang J.-S., Cultivation, photobioreactor design and harvesting of microalgae for biodiesel production: A critical review, Bioresource Technology, Volume 102, Issue 1, January 2011, Special Issue: Biofuels - II: Algal Biofuels and Microbial Fuel Cells, (2011). Fig. 5 : Output of the liquid level controller IV. Conclusion The utilization of cleaner energy resources against the fossil fuels, which damage the environment, is an 368 Ebrahimzadeh M., Torkian Boldaji M., Hosseini Boldaji S. A., Ettehad Sh., Effects of varying sodium nitrate and carbon dioxide quantities on microalgae growth performance of a designed tubular photobioreactor and a channel photobioreactor, Iranian Journal of Plant Physiology 1(4): (2011).

383 Fan L., Zhang Y., Cheng L., Zhang L., Tang D., Chen H., Optimization of carbon dioxide fixation by Chlorella vulgaris cultivated in a membrane photobioreactor, Chemical Engineering & Technology 30 (8), (2007). Fernandes B.D., Dragoner G.M., Teixiera J.A., Vicente A.A., Light regime characterization in an airlift photobioreactor for production of microalgae with high starch content, Applied Biochemical Biotechnology, 161, (2010). Gonçalves A.L., Rodrigues C.M., Pires J.C.M., Simoes M., The effect of increasing CO2 concentrations on its capture, biomass production and wastewater bioremediation by microalgae and cyanobacteria, Algal Research, 14, (2016). Harun R., Singh M., Forde G.M., Danquah M.K., Bioprocess engineering of microalgae to produce a variety of consumer products, Renewable and Sustainable Energy Reviews, 14, (2010). Jorquera O., Kiperstok A., Sales E.A., Embiruçu M., Ghirardi M.L., Comparative energy life-cycle analyses of microalgal biomass production in open ponds and photobioreactors, Bioresource Technology, 101, (2010). Hu Q., Guterman H., Richmond A., A flat inclined modular photobioreactor for outdoor mass cultivation of photoautotrophs, Biotechnology and Bioengineering 51, (1996). Iqbal M., Grey D., Stepan-Sarkissian F., Fowler M.W., A flat-sided photobioreactor for culturing microalgae, Aquacultural Engineering, 12, (1993). Kommareddy A.R., Anderson G.A., Study of Light as a Parameter in the Growth of Algae in a Photo-Bio Reactor (PBR), ASAE Paper No ASAE, St. Joseph, Michigan (2003). Moisander P.H., McClinton E., Paerl H.W., Salinity Effects on Growth, Photosynthetic Parameters, and Nitrogenase Activity in Estuarine Planktonic Cyanobacteria, Microbial Ecology, 43(4), (2002). Ogbonna J.C., Yada H., Tanaka H., Light supply coefficient: A new engineering parameter for photobioreactor design, Journal of Fermentation and Bioengineering, 80(4), (1995). Ramos de Ortega A., Roux J.C., Production of Chlorella biomass in different types of flat bioreactors in temperate zones, Biomass, 10, (1986). Rashid N., Rehman M.S.U., Memon S., Ur Rahman Z., Lee K., Han J.-I., Current status, barriers and developments in biohydrogen production by microalgae, Renewable and Sustainable Energy Reviews, 22, (2013). Ruangsomboon S., Effect of light, nutrient, cultivation time and salinity on lipid production of newly isolated strain of the green microalga, Botryococcus braunii KMITL 2, Bioresource Technology 109: (2012). Sheng J., Kim H.W., Badalamenti J.P., Zhou C., Sridharakrishnan S., Krajmalnik-Brown R., Rittmann B.E., Vannela R., Effects of temperature shifts on growth rate and lipid characteristics of Synechocystis sp. PCC6803 in a bench-top photobioreactor, Bioresource Technology, 102(24), (2011). Sivakumara G., Xua J., Thompsonb R.W., Yangc Y., Randol-Smithd P., Weathersc P.J., Integrated green algal technology for bioremediation and biofuel, Bioresource Technology, 107, 1-9 (2012). Sonnekus M.J., Effects of Salinity on the Growth and Lipid Production of Ten Species of Microalgae from the Swartkops Saltworks: A Biodiesel Perspective, Nelson Mandela Metropolitan University (2010). Kumar A., Ergas S., Yuan X., Sahu A., Zhang Q., Dewulf J., Malcata F.X, Langenhove H., Enhanced CO2 fixation and biofuel production via microalgae: Recent developments and future directions, Trends in Biotechnology, 28(7), (2010). Li M., Hu D., Liu H., Photobioreactor with ideal light dark cycle designed and built from mathematical modeling and CFD simulation, Ecological Engineering, 73, (2014). Lin Q., Lin J., Effects of nitrogen source and concentration on biomass and oil production of a Scenedesmus rubescens like microalga, Bioresource Technology, 102 (2), (2011). Miron S.A., Garcia M.C., Camacho G.F., Molina G.E.Y., Mixing in bubble column and airlift reactors, Chemical Engineering Research and Design, 82 (A10): (2004). 369 Ugwu C.U., Aoyagi H., Uchiyama H., Photobioreactors for mass cultivation of algae, Bioresource Technology, 99, (2008). Widjaja A., Lipid production from microalgae as a promising candidate for biodiesel production, Makara Teknologi, 13(1), (2009). Widjaja A., Chien C.C., Ju Y.H., Study of increasing lipid production from fresh water microalgae Chlorella vulgaris, Journal of the Taiwan Institute of Chemical Engineers, 40(1),13-20 (2009). Xin L., Hong-ying H., Ke G., Ying-Xue S., Effects of different nitrogen and phosphorus concentrations on the growth, nutrient uptake, and lipid accumulation of a freshwater microalga Scenedesmus sp., Bioresource Technology, 101(14), (2010). Yeesang C., Cheirsilp B., Effect of nitrogen, salt, and

384 iron content in the growth medium and light intensity on lipid production by microalgae isolated from freshwater sources in Thailand, Bioresource Technology, 102(3), (2011). Zhang K., Kurano N., Miyachi S., Optimized aeration by carbon dioxide gas for microalgal production and mass transfer characterization in a vertical flat-plate photobioreactor, Bioprocess Biosystems Bioengineering, 25, (2002). 370

385 Energy and Exergy Analysis of a Solar Air Heater Having Transverse Wedge Shaped Rib Roughness Cihan Yildirim 1*, Ismail Solmus 2 1 Adana Science and Technology University, Faculty of Engineering and Natural Sciences, Mechanical Engineering Department, Yeşiloba Yerleşkesi Yeşiloba Mah. Öğretmenler Bulvarı Sokak No:3 Seyhan, Adana, 01180, Turkey 2 Atatürk University, Department of Mechanical Engineering, Erzurum, Turkey * dr.cihan.yildirim@gmail.com Abstract In this study, energy and exergy efficiencies of a solar air heater having transverse wedge shaped rib roughness elements are investigated and compared with the flat plate solar air collectors. The system is theoretically analyzed and simulated for different geometrical roughness parameters by help of energy balance between the elements of the collector. Geometrical roughness parameters such as ratio of relative roughness height, relative roughness pitch and rib wedge angle are examined for different Reynolds numbers. The effects of parameters on the energy and exergy efficiencies are investigated. It is observed that artificial roughness help us to enhance energy and exergy efficiency. Keywords: Solar air heater, energy, exergy, artificial roughness. I. Introduction Solar air heaters (collectors) become more popular in the recent years due to its simplicity and low cost. However the poor thermal properties of air prevent widespread using of the solar air heaters. Resercher aims to overcome this disadvantage and increase the efficiency of the solar air heaters. Artifical rougness usage on the absorber plate attracts great attention. There are many studies in the literature for solar air heaters. Tchinda (2009) summarized mathematical models for different types of solar air heaters. Chamoli et al. (2012) presented literature survey about double pass solar air heaters. Oztop et al. (2013) reviewed recent works on solar air heater based on energy and exergy analysis. Ong (1995a) proposed a mathematical model and investigated four different types of solar air heater. He (1995b) also compared theoretical results with experimental ones. In his study, he obtained satisfactory qualitative and quantitative agreement between theoretical and experimental works. Yıldırım and Solmuş (2014) investigated the thermal and thermohydraulic efficiency of double pass solar air heater theoretically. Their investigation was based on transient mathematical model. Their study explored the dynamics of double pass solar air heater during whole day. El-Sebaii (2011) theoretically and experimentally investigated double pass flat and v- corrugated plate solar air heaters. They observed that v-corrugated plate increase the thermal efficiency up to %. Njomo (2006) used sensitivity analysis method for optimization of four different types of solar air heaters. Hedayatizadeh et al. (2012) conducted an exergetic and parametric study for v-corrugated solar air heater. They highly recommended the v- corrugated solar air heater for high aspect ratio, low altitude of triangular duct, low mass flow rate and low inlet temperature applications. Gupta and Kaushik (2008) investigated optimum aspect ratio, optimum duct depth of flat plate solar air heater for maximum exergy delivery. Alta et al. (2010) conducted an experimental study for three different types of solar air heater and calculated energy and exergy output rates for different mass flow rates and different tilt angles. Akpınar and Koçyiğit (2010) investigated energy and exergy efficiencies of solar air heater with/without obstacles. They compared four different types of solar air heater configuration experimentally. Kurtbaş and Durmuş (2004) investigated five different types of solar air heaters with different absorber geometries. Esen (2008) investigated flat plate solar air heaters with/without obstacles. He compared four different types of solar air heaters by means of energy and exergy efficiencies. Altfeld et. al (1988a, 1988b) conducted an exergetic optimization study for solar air heaters. Another recent exergetic optimization study on solar air heater was conducted by Ajam et al. [17] and Farahat et al. (2009). Bhushan and Singh investigated the protruded absorber plate experimentally (2011) and theoretically (2012). Bhagoria et al. (2002) conducted an experimental study about transverse wedge shaped rib roughness element. They calculated the coralation of Nusselt number and friction factor with respect to rougness parameters. Yıldırım and Solmuş (2015) studied the parametric behaviour of a solar air heater having artificially protruded absorber plate. They examine the energy and exergy efficiency of the artificially roughned solar air heater. In this study, energy and exergy efficiencies of a solar air collector having transverse wedge shaped rib roughness elements have been investigated and compared with the flat plate solar air heaters. The system has been theoretically investigated for 371

386 different geometrical roughness parameters by help of energy balance approach. Geometrical roughness parameters such as relative roughness height (RRH), relative roughness pitch (RRP) and rib wedge angle (φ) are examined for different Reynolds numbers. II. Method II.1. Mathematical Models Theoretical analysis of this study is based on thermal network of the solar air heater. Energy balance equations are constructed for each part of the collector and solved simultaneously. Roughness geometries are shown in Fig.1. The energy balance equations for the various parts of solar air heaters have been formulated under the following simplifying assumptions: (i) The system is steady state. (ii) There is no temperature gradient across the thickness of the system parts, so temperature of upper and lower sides are the same. (iii) There is no air leakage, so mass balance is kept. (iv) The air temperature varies linearly throughout section length. The mean air temperature is then equal to the arithmetic mean of inlet and outlet temperature of section. is the mean temperature matrix and [B] is the source matrix. This equation can be solved by simple matrix inversion. Calculation of heat transfer equations are discussed in the previous work of Yıldırım and Solmuş (2014). Bhagoria s (2002) correlation of Nusselt number and friction factor are used for roughened absorber plate and Dittus Boelter and Modified Blasius correlations are used for flat (smooth) absorber plate. For flat plate absorber Nu = 0.023Re 0.8 PR 0.4 (5) f = 0.085Re 0.25 (6) For transverse wedge shaped rib Nu = (Re) 1.21 RRH RRP 2.94 exp ( 0.71(ln(RRP 2 ))) ( φ 10 ) [exp ( 1.5 (ln ( φ 10 ) 2 ))] (7) f = 12.44Re 0.18 RRH 0.99 RRP 0.52 ( φ 10 ) 0.49 (8) First law efficiency (energy efficiency) of the collector is evaluated as; η Energy = m c p (T out T in ) I.A c (9) Second law efficiency (exergy efficiency) of the collector is evaluated as (Yıldırım and Solmuş 2014, Yıldırım and Solmuş 2015); Fig. 1: Schematic representation of solar air heater and absorber geometry. II.2. Energy Balance of the system Glass cover: 0 = Iα g A c q c,g amb q r,g sky + q r,p g + q c,a g (1) Air channel: 0 = q c,p a q c,a g m c p (T a out T a in ) (2) Absorber plate: 0 = Iα p τ g A c q c,p a q r,p g q loss,p amb (3) In general, energy balance of system can be displayed as matrix equation; [A][T] = [B] (4) where [A] is the heat transfer coefficient matrix, [T] 372 η Exergy = m (Δh T eδs) Ex Pump (1 T e Ts )IA c where (10) Δh = h out h in = c p (T out T in ) (11) s = s out s in = c p ln T out T in (12) Ex Pump = T amb m Δp T in η p ρ II.3. Solution Procedure (13) Considered mathematical model is valid for short collector so that long collector can be assumed to be divided into a number of sections (Ong 1995a). Temperature values of first section are initially guessed and heat transfer coefficients are calculated based on guessed values. Then new temperature values are evaluated by using heat transfer coefficients. This iterative process is repeated until all consecutive temperature values less than 0.01 o C. Consecutive section is evaluated as the same procedure. The outlet air temperature of the previous

387 section is set to inlet air temperature of the next one. The ambient temperature (10 o C) is used as initial values. Solar insolation is set to 500 W/m 2. Collector width is set to 1 m, length is set to 2 m and depth is set to 0.05 m. Tested parameters in this study are represented in Tab 1. roughness parameter. Increasing RRP increases energy (Fig 4) and exergy (Fig 5) efficiency up to a certain point (RRP=7.57) and then slightly decreases. This phenomenon may be explained by complex nature of flow separation and reattachment (Bhagoria 2002). Tab 1: Tested parameter values of the study RRH RRP φ 8 o 10 o 12 o 15 o III. Results and discussions RRP=3.730 RRP=4.700 RRP=5.670 RRP=7.110 RRP=7.570 RRP=10.00 RRP=12.12 A simulation program based on the mathematical model presented in the previous section has been written by MATLAB software. Variations of the energy and exergy efficiency with respect to Reynolds number are investigated for different roughness parameters. Relative roughness height (RRH) is the one of the roughness parameter. Energy efficiency is positively affected by increasing RRH and increasing Reynolds number (Fig 2). Exergy efficiency also increases by increasing RRH (Fig 3) but decreases by Reynolds number after some specific point due to increasing friction factor. I RHH=0.015 RHH=0.020 RHH=0.025 RHH=0.028 RHH= Re x 10 4 Fig. 2: Effect of the Relative Roghness Height (RRH) on the energy efficiency. II RHH= RHH=0.020 RHH= RHH=0.028 RHH= Re x 10 4 Fig. 3: Effect of the Relative Roghness Height (RRH) on the exergy (*100) efficiency. Relative roughness pitch (RRP) is another important 373 I I Re Re x 10 4 Fig. 4: Effect of the Relative Roughness Pitch (RRP) on the energy efficiency. II RRP=3.730 RRP=4.700 RRP=5.670 RRP=7.110 RRP=7.570 RRP=10.00 RRP=12.12 Fig. 5: Effect of the Relative Roughness Pitch (RRP) on the exergy (*100) efficiency. Energy and exergy efficiencies are also affected by rib wedge angle. Both values (Fig 6. and Fig 7.) increases with increasing rib wedge angel up to a certain point and then decreases. I II Fig. 6: Effect of the rib wedge angle on the energy efficiency Re = Re Re x =8 =10 =12 =15 I Re =19472 Re = Re Re x

388 =8 =10 =12 =15 IV. Conclusions In this study, energy and exergy efficiencies of a solar air collector having transverse wedge shaped rib roughness elements have been investigated. Main findings can be summarized as follows: II Re = Re Re x 10 4 Fig. 7: Effect of the rib wedge angle on the exergy (*100) efficiency. Benefit of the absorber plate having transverse wedge shaped rib roughness is seen from Fig 8. and Fig 9. Energy efficiency can be increased up to 38% times and exergy efficiency can be increased up to 89% times by using artificial roughness. I Fig. 8: Comparison of energy efficiency of smooth absorber and roughned with wedge shaped rib absorber II II 0.95 Fig. 9: Comparison of exergy (*100) efficiency of smooth absorber and roughned with wedge shaped rib absorber Re x Wedge shaped rib Smooth Wedge shaped rib Smooth Re x 10 4 Energy efficiency increases by increasing Reynolds number for all roughness parameters. Exergy efficiency increases by increasing Reynolds number up to its maximum value and then slightly decreases due to increasing friction of the air flow. Relative roughnes height (RRH) increases the energy and exergy efficiencies. Relative roughness pitch (RRP) increases the energy and exergy efficiencies up to a certain point and then decreases. Rib wedge angle increases the energy and exergy efficiencies up to a certain point and then decreases. Artificialy roughned absorber plate with transverse wedge shaped rib has superiority on smooth absorber plate. Nomenclature f : Friction factor I : Solar radiation (W/m 2 ) Nu : Nusselt number Re : Reynolds number RRH : Relative roughness height RRP : Relative roughness pitch Pr : Prandtl number Greek letters φ : Wedge edge angle ( o ) η : Efficiency Subscripts a-in : Air at the inlet a-out : Air at the outlet c : Collector c,a-g : Convection from air to glazing c,g-amb : Convection from glazing to ambient c,p-a : Convection from plate to air r,g-sky : Radiation from glazing to sky r,p-g : Radiation from plate to glass loss,p-amb: Loss from plate to amb References Ajam H., Farahat S. and Sarhaddi F., Exergetic Optimization of Solar Air Heaters and Comparison with Energy Analysis, International Journal of Thermodynamics, Vol. 8 No. 4, pp , Akpınar E. K. and Koçyiğit F., Energy and Exergy Analysis of a New Flat-Plate Solar Air Heater Having Different Obstacles on Absorber Plates, Applied Energy, Vol. 87, pp , Altfeld K., Leiner W. and Fiebig M., Second Law 374

389 Optimization of Flat-Plate Solar Air Heaters. Part 1: The Concept of Net Exergy Flow and the Modeling of Solar Air Heaters, Solar Energy, Vol. 41 No. 2, pp , Altfeld K., Leiner W. and Fiebig M., Second Law Optimization of Flat-Plate Solar Air Heaters. Part 2: Results of Optimization and Analysis of Sensibility to Variations of Operating Conditions, Solar Energy, Vol. 41 No. 4, pp , Alta D., Bilgili E., Ertekin C. and Yaldız O., Experimental Investigation of Three Different Solar Air Heaters: Energy and Exergy Analyses, Applied Energy, Vol. 87, pp , Bhagoria J.L., Saini J.S., Solanki S.C., Heat transfer coefficient and friction factor correlations for rectangular solar air heater duct having transverse wedge shaped rib roughness on the absorber plate, Renewable Energy, Vol.25, , Bhushan B., Singh R., Nusselt number and friction factor correlations for solar air heater duct having artificially roughned absorber plate, Solar Energy, Vol. 85, , Bhushan B., Singh R., Thermal and thermohydraulic performance of roughened solar air heater having protruded absorber plate, Solar Energy, Vol. 86, , Chamoli S., Chauhan R., Thakur N.S. and Saini J.S., A Review of the Performance of Double Pass Solar Air Heater, Renewable and Sustainable Energy Reviews, Vol. 16, pp , El-Sebaii A.A., Aboul-Enein S., Ramadan M.R.I., Shalaby S.M. and Moharram B.M., Investigation of Thermal Performance of Double Pass-Flat and V- Corrugated Plate Solar Air Heaters, Energy, Vol. 36, pp , , pp , Njomo D. and Daguenet M., Sensitivity Analysis of Thermal Performance of Flat Plate Solar Air Heaters, Heat Mass Transfer, Vol. 42, pp , Ong K.S., Thermal Performance of Solar Air Heaters- Experimental Correlation, Solar Energy, Vol. 55 No.3, pp , Ong K.S., Thermal Performance of Solar Air Heaters: Mathematical Model and Solution Procedure, Solar Energy, Vol. 55 No.2, pp , Oztop H. F., Bayrak F. and Hepbaşlı A., Energetic and Exergetic Aspect of Solar Air Heating (Solar Collector) Systems, Renewable and Sustainable Energy Reviews, Vol. 21, pp 59-83, Tchinda R., A Review of the Mathematical Models for Predicting Solar Air Heater Systems, Renewable and Sustainable Energy Reviews, Vol. 13, pp , Yıldırım C. and Solmuş İ., Çift Geçişli Hava Akışkanlı Güneş Toplacı Kanal Yüksekliğinin Termohidrolik Verime Etkisinin İncelenmesi, Journal of Thermal Science and Technology, Vol. 34(1), , Yıldırım C., Solmuş İ., First and second law analysis of various types of solar air heaters, 8th Conference on Sustainable Development of Energy, Water and Environment Systems, Dubrovnik-Croatia, September Yıldırım C., Solmuş İ., Energy and Exergy Analysis of a Solar Air Collector Having a Roughened Absorber with Circular Protrusions, International Conference on Environment and Renewable Energy, Vienna, Austria, May Esen H., Experimental Energy and Exergy Analysis of a Double-Flow Solar Air Heater Having Different Obstacles on Absorber Plates, Building and Environment, Vol. 43, pp , Farahat S., Sarhaddi F. and Ajam H., Exergetic Optimization of Flat Plate Solar Collectors, Renewable Energy, Vol. 34, pp , Gupta M.K. and Kaushik S.C., Exergetic Performance Evaluation and Parametric Studies of Solar Air Heater, Energy, Vol. 33, pp , Hedayatizadeh M., Ajabshirci Y., Sarhaddi F., Farahat S., Safavinejad A. and Chaji H., Analysis of Exergy and Parametric Study of a V-Corrugated Solar Air Heater, Heat Mass Transfer, Vol. 48, pp , Kurtbaş İ. and Durmuş A., Efficiency and Exergy Analysis of a Solar Air Heater, Renewable Energy, Vol. 375

390 Performance Analysis of Three Soft Computing Methods for Predicting the Heat Load of Buildings Cihan Turhan 1, Tugce Kazanasmaz 2, Gulden Gokcen Akkurt 1* 1 Energy Engineering Program, Izmir Institute of Technology, Gulbahce Campus, Urla, Izmir, Turkey 2 Faculty of Architecture, Izmir Institute of Technology, Gulbahce Campus, Urla, Izmir, Turkey * guldengokcen@iyte.edu.tr Abstract The performance of three soft computing (SC) methods; Artificial Neural Networks (ANNs), Fuzzy Logic (FL) and Adaptive Neuro-based Fuzzy Inference System (ANFIS) are comparatively evaluated by their prediction indices of the heat load of buildings in Turkey. Each method generated a soft model using heat transmission coefficient (U-value), total external surface area per volume of the building (A/V), total external surface area (TESA), and total window area per total external surface area (TWA/TESA). The outputs become the heat load values which are obtained from preceding simulation analyses. First, ANN model as a weight matrix is trained to learn how the heat load of residential buildings is estimated as a nonlinear complex problem. Second, FL model is constructed as a simple structure which is easier to understand than an ANN model. Finally, ANFIS model is derived from the combination of two previous SC techniques. The performance evaluation is based on prediction indices. The main conclusions are; i) the SC methods are powerful tools to model the heat load of buildings, ii) ANFIS model is able to obtain better prediction accuracy both in training and testing phases, iii) advantages of the SC methods compared to the energy simulation software are their simplicity, speed of calculation and effective learning from the less homogeneous data sets. Keywords: Heat load, existing buildings, ANN, fuzzy logic, ANFIS, soft computing. I. Introduction Building sector is the largest energy consuming sector in both Europe and Turkey (Republic of Turkey ministry of energy and natural resources (2014)). Since the energy consumption of the buildings is highly related with the heat load, its estimation in energy analysis is crucial in both new and existing buildings. Environmental and climatic conditions, building envelope materials and insulation are the driving forces increasing/or decreasing the heat load of a building significantly. Their interaction and relation are complicated and nonlinear. Reliable and robust tools are needed to obtain information about the heat load values to carry out advanced energy performance diagnostics. The soft computing (SC) methods are novel techniques which have come out owing to advances in the field of artificial intelligence (AI) in the last two decades (Islam, 2011). Many articles have been published on the use of SC methods which are Support Vector Machines (SVM), Genetic algorithms (GAs), Artificial Neural Networks (ANNs), Fuzzy Logic (FL) and Adaptive Neuro-based Fuzzy Inference System (ANFIS) into the prediction of the energy consumption and the heat load of buildings. Specifically, ANNs have been extensively employed for prediction of building energy consumption (Melo et al. 2014, Ekici and Aksoy, 2011). Turhan et al. (2014) succeeded to predict the heat load of existing buildings implementing a back propagation ANN model with a of prediction index. Outputs of the model were compatible with the ones obtained from a building energy simulation tool, KEP-IYTE-ESS (KEP-SDM, 2008). Fuzzy Logic (FL) models, on the other hand, provide a usual way to deal with real case problems for complicated non-linear systems (Ciabottoni et al.2014). BEP-TR (2010), which is the national building energy simulation tool of Turkey, was analysed by employing a FL model in a study (Kabak et al, 2014) Geometric, climatic and mechanical parameters were chosen to be the inputs. As a conclusion of this study, buildings were classified based on their overall energy performance. As it was proposed to combine the benefits of ANN and FL to come up with a more successful prediction, ANFIS is such an approach which has advantages of self-learning procedures like ANNs and easy to understand because of the structure of the FL rules. Ekici and Aksoy (2011) proposed an ANFIS model to forecast the building energy consumption in cold regions. Transparency ratios, azimuth angles, building form factors and insulation thicknesses were used as the model inputs. The ANFIS was effective with a successful prediction rate of In another study by Ayata et al. (2007), the demonstration of ANFIS provided information about the potential use of natural ventilation as a passive cooling system in new building designs in Turkey estimating the indoor air velocity values successfully. The paper aimed to assess the performance of three SC models to find out how they are adaptable and 376

391 effective in the heat load estimation of buildings and which one performs better. A building simulation tool was run to contribute its outputs in SC models and to calculate the prediction indices. The simulation tool is called The Standard Assessment Procedure for Energy Rating of Dwellings software (KEP-IYTE-ESS) (KEP-SDM, 2008). The models involved data from existing residential buildings in Izmir-Turkey. The MATLAB (2012) environment generated all SC models. In the simulations, two performance indices Mean Absolute Percentage Error (MAPE) and the Multiple Correlation Coefficient (R 2 ) were used to compare the results. II. Overview of SC techniques This section gives brief information of SC models which are used for this study. II.1. ANNs ANNs are self-adaptive mathematical models which are based on inspirations of our body s nervous system (McCulloch and Walter, 1943). The architecture of an ANN model corresponds to a network which is composed of a high number of interconnected- and weighted- neurons transmitting the signals in the entire structure to produce the output. Inputs contribute the model through the input layer. After introducing the data to the network, the hidden-layer receiving the weighted data from the preceding layer, associates them with bias and transfers them into a nonlinear transfer function. The process ends in the output layer where neurons perform a similar operation to generate the output (Tayfur, 2012). The schematic representation of a typical ANN is shown in Fig.1. (net j)=x 1w 1j+x 2w 2j+.+x nw nj+b i (1) The sigmoid function is usually employed as an activation function in the training step of the network. The sigmoid function is expressed as Eq. (2): y = 1/ (1+ e net j) (2) The objective of the model is to decrease error to an acceptable value that is called epoch or training cycle. The error is expressed by the root-mean-squared error value (RMSE), which can be calculated with following Eq. (3): E=½[ pσ iσ t ip-o ip ] ½ (3) where (E) is the RMSE, t the network output, and o the desired output vectors over all the pattern (p). II.2. FL Algorithm FL algorithms have the ability to describe knowledge like a human in the form of simple rules using linguistic variables (Ross, 2010).The general structure of the FL modelling is presented in Fig.2. Fig. 2: Schematic representation of a fuzzy system (Tayfur,2012) Fig. 1: Schematic representation of a three-layer ANN (Antognetti,1991) ANNs learning process is called the training phase during which a neuron has got the inputs from the layers, weights each input with a pre-arranged value, and combines these weighted inputs. The target output at each output neuron is minimized by adjusting the weights and biases through some training algorithm. Scalar input x 1,x 2,..,x n are multiplied by weights w 1j,w 2j,.w nj and the weighted values are fed to the summing confluence. The neuron has a bias bi that is summed with the weighted inputs in order to form the net input netj given in Eq. (1). 377 The FL algorithm consists of four processes basically. First, fuzzification transforms the numerical input values to the degree of memberships defined by appropriate verbal fuzzy information. Second, rule base contains the rules where the knowledge is stored for predicting output values. The rules are in the form of IF-THEN format instead of being mathematical equations. An example of a fuzzy rule can be constructed in IF-THEN format as below; IF the wall overall heat transfer coefficient is HIGH, building area/ volume ratio is LOW, total external surface area is MEDIUM total window area/total external surface area ratio is MEDIUM, THEN the heat load of building is VERY HIGH Fuzzy Inference Engine (FIS) takes into account all fuzzy rules in the fuzzy rule base to simulate non-linear behaviours and learns to transform inputs to the corresponding outputs. Defuzzification is responsible to transform the resulting outputs from the fuzzy engine to numbers. The most famous defuzzifiction method is centroid method expressed as the below Eq. (4);

392 K * x = [ i µ (K xi) K xi] / [ i µ (K xi)] (4) where K * x is the defuzzified output value, K ix is the output value in the ith subset, and μ(k ix) is the membership value of the output value in the ith subset. II.3. ANFIS ANNs are able to learn a high variety of nonlinear complex problems. Being black-box model makes this soft computing method difficult for researchers to understand. Limited or noisy training data can result in an illogical and meaningless output. Unlike ANNs, FL cannot learn by itself. In contrast, FL models are easy to manage because of the structure of IF-THEN rules. ANFIS which is claimed to be a universal predictor for nonlinear problems is the composition of both ANNs and FL approaches (Fuller, 2010). Fig.3 illustrates the structure of the ANFIS model. Fig. 3: Architecture of ANFIS ANFIS is a multilayer feed forward network where each node performs a particular function on incoming signals. It implements a first-order Sugeno fuzzy inference system which works well with optimization and adaptive techniques. The model uses a hybrid learning algorithm with a set of IF-THEN fuzzy rules to generate the previously stipulated input-output pairs (Köksal and Uğursal, 2008). In the first nodes, input variables are introduced to the system. The membership functions are used as node functions. The rules are constructed with the strength of corresponding layer and outputs are generated in the second nodes. Third nodes are an average layer that optimizes firing strength. Forth nodes are Consequent nodes that act as a defuzzifier. Finally, fifth nodes generate an output from the sum of each rule. III. Case study ANNs, FL and ANFIS were employed in a heat-load estimation based case study for apartment buildings in İzmir, Turkey. This case study aims to show the accuracy of the chosen SC techniques on prediction of heat load; and which one would be head of the others in terms of its prediction indices during this process. Thus, a reliable approach for making decisions about heat load would be attained for further research. are located in Izmir at geographical coordinates East, North. They involve approximately 5-11 storeys. The 3D City Guide by the Department of Geographical Information Systems (Fig.4) was used to choose the buildings. The case buildings correspond to the building stock in İzmir based on their architectural and constructional characteristics (Kazanasmaz et al., 2014). Fig. 4: Building selection by using 3D City Guide of İzmir ( Utilizing their architectural drawings, the values of area based input parameters were recorded. The heat load of buildings was calculated with the help of KEP-IYTE-ESS. The performance of ANNs, FL and ANFIS depends on the comparison of soft computing findings with the ones obtained from the software KEP-IYTE-ESS. KEP-IYTE-ESS is the software which is derived from KEP-SDM methodology (KEP-SDM, 2008). The software produces two outputs; annual energy consumption per unit floor area (kwh/m 2 year) and annual greenhouse gas emissions per unit floor area (kg CO2/m 2 year). The software was tested and validated using a well-known validation and diagnostic procedure, Building Energy Simulation Test (BESTEST) (Neymark et al., 2012) procedure and the outputs of the software were in the range of acceptable values of BESTEST (Akkurt et al.,2013). For further detail of the software KEP-IYTE-EES, please see (Turhan et al.2014 and Kazanasmaz et al.2014). IV. Application of SC modelling for the heat load estimation The area-based input parameters which contributed in the models were wall overall heat transfer coefficient (U), area/volume ratio (A/V), total external surface area (TESA), total window area/total external surface area ratio (TWA/TESA). Despite of having plenty of parameters for calculating the heat load, only four of them were involved in the models. Since three of them (x2 - x4) are architecturally-driven variables, they can be obtained by field surveys in existing buildings. On the other hand, the U-value (x1) can be calculated easily. The maximum and minimum values of input parameters are listed in Table 1. A number of 2136 apartments in residential buildings 378

393 Tab. 1: Input and output parameters used in three SC models Code Input parameters Data used in three SC models Minimum Maximum x 1 U (W/m 2 K) x 2 A/V (1/m) x 3 TESA (m) x 4 TWA/TESA (-) y 1 HL (kw/m 2 ) Initially, the data were normalized according to formula below, x i= (x i x min i)/(x max i x min i) (5) where xmax i and xmax i are the minimum and maximum values of i th node in the input layer for all feed data vectors, respectively. Intelligent systems learn how to analyse the inputs on examples. Before the application of the models, the data was divided into a training data set (80% of the data) carried out for learning and a test data set (20% of the data) implemented for the evaluation of the models. The multiple correlation coefficient (R 2 ) and mean absolute percentage error (MAPE) were used as the prediction indices at each SC process to compare the model performances (Eq. 6 and 7). The R 2 is expected to be close to 1, while the MAPE should be as close as to zero for the best performance. Trial-and-error iterations with a considerable number of neurons and layers and error procedures determine the optimal number of neurons and hidden layers. The iteration number was set to for the training of the model. No bias term was used. Learning rate was constant and equal to the 0.02 for the ANN model to predict the heat load of buildings. For further detail please see (Turhan et al., 2014). IV. 2. FL model construction Though ANN models are reliable, they are still black-box models. The user cannot interrupt and change the model easily during the operations. The model offers a weight matrix which defines the weights of interlayer connections, being optimized after thousands of iterations. To generate a simpler model, FL model was established including four inputs (U, A/V, TESA and TWA/TESA) and an output (HL). The fuzzy subsets of the variables were considered to have triangular and trapezoid membership functions. Three subdivisions of inputs and parameters namely were set as low (L), medium (M) and high (H) as represented in Fig.6 R 2 = 1 (Σ j t j-o j 2/ Σ j (o j) 2 ) (6) MAPE=1/pΣ j[ (t j-o j)/t j ]*100 (7) t is the target value, o is the output value and p is the number of input-output pairs (Antognetti,1991). IV. 1. ANN model construction A feed forward type of ANN model developed for the case buildings and published earlier (Fig.5). Fig. 5: Feed forward ANN model (Turhan et al.2014) The primary aim was to minimize the error at the output layer by changing a set of connection strengths. Based on this objective, 103 of the total data set were chosen for the training scope, while the rest for the validation of the ANN model. 379 Fig. 6: Fuzzy subsets for input and output parameters The min and centroid methods have been defined as the inference operator and defuzzification, respectively. Mamdani fuzzy inference systems (FIS)

394 determined the flow of the process to produce the rules that affect the output parameter. The fuzzy rules were expressed in the IF-THEN format. Table 2 lists an example of 81 fuzzy logic rules. Tab. 2: An example of 20 fuzzy rules set randomly from the total of 81 sets U value of the wall (W/m 2 K) A/V ratio (m 2 /m 3 ) TESA (m 2 ) TWA/TESA (-) HL (kw/m 2 ) L L L L VL L L L H L M L L L L H L L L M L L M M M L M H L M L L H H H L L H L M L M L L VL L L H L L L L H M M H H M H VH H L H M VH H L L H H M L L H H M L M H H H L M M VH H M M H VH H H L H H H H H H VH As an example, let us assume that the U value of the wall is 0.8 W/m 2 K, A/V ratio is 0.30 m 2 /m 3, TESA is 250 m 2 and TWA/TESA is 0.2. We want to find out the fuzzy output of the heat load of the building under these variables would be. As seen in Fig.6a, 0.8 W/m 2 K is a part of low and medium subsets of the U value of the wall with µ(u)=0.95 and µ(u)=0.05 membership degrees, respectively. Similarly, as seen in Fig.6b, 0.3 of A/V ratio is a part of low and medium subsets with membership degrees of µ(a/v)=0.05 and µ(a/v)=0.95., respectively. The fuzzy inference engine would consider the following rules from the fuzzy rule base related to the above example and find the degrees of membership of the heat load of building output by min operation. IF U value of the wall is low (µ(u)=0.90), A/V ratio is medium (µ(a/v)=0.83), TESA is low (µ(tesa)=0.98) and TWA/TESA is low (µ(twa/tesa)=0.32) THEN the heat load of the building is very low (µ(hl)= min (0.90, 0.83, 0.98,0.32)= When one employs Eq. (4) for the above example, the following output value would be obtained by weighted-average defuzzification; heat load*= [0.32*(0.0084) *( )/2]/ ( ) heat load*= kw/m 2 IV. 3. ANFIS model construction The ANFIS applies the hybrid learning algorithm which combines ANN and FL. A Sugeno-type ANFIS model with four inputs and an output was developed using MATLAB (MATLAB, 2012). To estimate the heat load of buildings, the formed ANFIS model was trained for 300 epochs with a certain number of parameters as summarized in Table 3. Tab. 3: Training parameters of the ANFIS for the heat load Parameters Heat load of buildings Number of nodes 193 Number of linear parameters 405 Number of nonlinear parameters 24 Number of fuzzy rules 81 Membership function gaussmf Epoch 300 Output MF type Linear Number of MF (input) Similar to ANN model, 80% of the total data set was chosen for the training scope, while the rest for the validation of the ANFIS model. U, A/V, TESA and TWA/TESA were used as input whilst the output was HL. Gaussian curve membership function (Gaussmf) was chosen to be the membership function to show the applicability of ANFIS model for predicting the heat load of buildings. V. Results and discussions By following the findings of calculations by three SC models (ANNs, FL and ANFIS), this study arrived at the possible best solutions with refined R 2 and MAPE values for the output. The interpretations derived from these prediction indices lead us to comparatively understand the SC models accuracy and usefulness in the prediction process. The coefficient of determination and the regression line indicate how the results of ANN model can fit well the outputs from the software as shown in Fig.7. IF U value of the wall is low (µ(u)=0.90), A/V ratio is low (µ(a/v)=0.17), TESA is medium (µ(tesa)=0.02) and TWA/TESA is medium (µ(twa/tesa)=0.68) THEN the heat load of the building is medium (µ(hl)= min (0.90, 0.17, 0.02,0.68)= Fig. 6e shows the output value of kw/m 2 corresponding to 0.32 degree of membership in the very low subset of the heat load of building and the output values of and kw/m 2 corresponding to 0.02 degree of membership in the medium subset of the heat load of building. Fig. 7: The statistical evaluation results of ANN and simulation results 380

395 The highest R 2 =0.977 and the MAPE=5.06 were obtained with a structure of LM learning algorithm and number of neurons. The comparison of KEP-IYTE-ESS and FL model results of building heat load set is given in Fig. 8. The figure indicates that the model is able to give a successful prediction of 98.6% with Mamdani FIS. observed that ANFIS model outperforms the other models. The ANFIS model improves significantly the prediction accuracy from MAPE of 5.06 to 2.43 for the testing sets. Fig. 8: The statistical evaluation results of FL and simulation results Fig. 9 shows the comparison of ANFIS model and simulation results. The figure indicates that the predicted values of the model had close match with the simulation software outputs (R 2 of 99%). Fig. 10: Forecasted and simulated heat load of buildings by ANN, FL and ANFIS Some final comments can be drawn in this section. ANN model is capable of capturing non-linear relationships among the parameters with 4 of MAPE. However, it is difficult to understand inside the model because of its black box structure. FL model can estimate the heat load of buildings with a good accuracy. Nevertheless, the model needs time and experience to write the verbal rules. ANFIS model combines ANN and FL models. This approach has self-organization without requirements of programming. Another benefit of the ANFIS model is giving immediate responses when you change or delete one of the heat load parameters. VI. Conclusions Fig. 9: The statistical evaluation results of ANFIS and simulation results Table 4 shows the comparison of the SC model results for training and testing phase, respectively. For the training phase of the heat load of buildings, the ANN, FL and the ANFIS models obtained the successful prediction rates (R 2 ) of 98.5%, 99.1% and 99.4%, respectively. Similarly, the testing results of the models are able to produce a good prediction with a R 2 of 97.7%, 98.6% and 99.0%, respectively. Tab. 4: Forecasting performance analysis of SC models for the heat load of buildings Training Testing R 2 MAPE R 2 MAPE ANN Fuzzy ANFIS The performances of all SC methods developed in this paper are shown in Fig. 10. Analysing the results during both training and testing phases, it can be 381 This study included the comparison of the application of three SC methods, ANN, FL and ANFIS in estimating the heat load of residential buildings. The purpose was to test their accuracy and prediction power by matching their findings with the ones obtained from the calculations by an energy simulation tool. Input parameters fed the models were U value of the wall, A/V ratio, TWA/TESA ratio and TESA. Two common statistical performance criteria (R 2 and MAPE) were adopted to evaluate the performance of the models. The results indicated that the SC methods were powerful tools to model the heat load of buildings and simpler than traditional forecasting approaches like physical and regression models. Indeed, whatever the proposed approach, the results of the SC methods were satisfactory. In terms of performance evaluation criteria, ANFIS model was able to obtain better prediction accuracy both in training and testing phases. The process of calculating the heat load which determines the required energy to heat the buildings contains several steps and a high number of input

396 parameters. Formulas and algorithms define the correlations between each step and parameter. Climatic conditions and building parameters are integrated into this process. Simulation tools provide energy analyses satisfactorily, however, require experienced users. The traditional prediction models like regression and SVR models require significant time. The SC methods are alternative techniques which have the capability to result in successful predictions in the field of building energy performance. Utilizing a very few number of input variables when compared to a simulation tool, SC models suggest a simple and accurate solutions and provide feedback information, especially, about existing buildings. Acknowledgements This research was supported by The Scientific and Technological Research Council of Turkey (TÜBİTAK) (Project Number: 109M450). References Akkurt GG., Sahin CD., Takan S., Arslan ZD., Testing a simplified building energy simulation program via building energy simulation test (BESTTEST), CLIMAMED 7th Mediterranean Congress of Climatization October (2013): Istanbul, Turkey; Antognetti VMP., Neural Networks, Concepts Applications and Implementations, New Jersey, USA, Prentice Hall,(1991). Ayata T., Çam E., Yıldız O., Adaptive neuro-fuzzy inference systems (ANFIS) application to investigate potential use of natural ventilation in new building design in Turkey, Energy Conversion and Management, 48 (5), , (2007). BEP-TR, National Building Energy Performance Calculation Methodology of Turkey. (No: YİG/ )), Turkish Official Journal, (2010). Ciabottoni L.,Grisostomi M.,Ippoliti G., Longhi S., Fuzzy logic home energy consumption modelling for residential photovoltaic plant sizing in the new Italian scenario, Energy, 74, , (2014). Ekici BB., Aksoy UT., Prediction of building energy consumption by using artificial neural networks, Advanced Engineering Software, 40, , (2011). Fuller R., Introduction to Neuro-Fuzzy Systems, ISBN , US, Springer, (2010). Kabak M., Köse E., Kırılmaz O., Burmaoğlu S., A fuzzy multi-criteria decision making approach to access building energy performance, Energy and Buildings, 72, , (2014). Kazanasmaz T.,Erlalelitepe Uygun İ., Akkurt GG.,Turhan C., Ekmen KE., On the relation between architectural considerations and heating energy performance of Turkish residential buildings in Izmir, Energy and Buildings, 72,38-50, (2014). KEP-SDM, Dwelling Energy Performance-Standart Assessment Procedure, Chambers of Mechanical Engineers, Izmir,Turkey, (2008) Koksal MA., Ugursal IV., Comparison of neural network, conditional demand analysis, and engineering approaches for modeling end-use energy consumption in the residential sector, Applied Energy, 85, , (2008). MATLAB, Version 2012b ed., The Mathworks, (2012). McCulloch W., Walter P., A Logical Calculus of Ideas Immanent in Nervous Activity, Bulletin of Mathematical Biophysics, 5 (4), , (1943). Melo AP., Costola D., Lamberts R., Hensen JLM., Development of surrogate models using artificial neural network for building shell energy labelling, Energy Policy, 69, ,(2014). Neymark J., Judkoff R., Knabe G., Le HT., Durig M., Glass A., et al., Applying the building energy simulation test (BESTEST) diagnostic method to verification of space conditioning equipment models used in whole-building energy simulation programs, Energy and Buildings,34, ,(2012). Republic of Turkey ministry of energy and natural resources, Retrieved 05st March 2014, Fromhttp:// TKB_2010_2014_Stratejik_Plani_EN.p df. Ross TJ., Fuzzy Logic with Engineering Applications. Third Edition, ISBN: , United Kingdom,Wiley,(2010). Tayfur G., Soft computing methods in water research engineering, Southampton, UK, WIT Press, (2012). Turhan C., Kazanasmaz T., Erlalelitepe Uygun İ., Ekmen KE.,Akkurt GG., Comparative study of a building energy performance software (KEP-IYTE-ESS) and ANN-based building heat load estimation, Energy and Buildings,85, ,(2014). Islam UIB., Comparison of Conventional and Modern Load Forecasting Techniques Based on Artificial Intelligence and Expert Systems, IJCSI International Journal of Computer Science Issues, 8 (5), , (2011). 382

397 Ventilation Strategies for the Preventive Conservation of Manuscripts in Necip Paşa Library, İzmir-Turkey Turgay Coskun 1, Cem Dogan Sahin 2, Ozcan Gulhan 1, Zeynep Durmus Arsan 3*, Gulden Gokcen Akkurt 1 1 Izmir Institute of Technology, Energy Engineering Program, Gülbahçe, Urla, Izmir, 35430, Turkey 2 Izmir Institute of Technology, Engineering Faculty, Mechanical Engineering Department, Gülbahçe, Urla, Izmir, 35430, Turkey 3 Izmir Institute of Technology, Faculty of Architecture, Department of Architecture, Gülbahçe, Urla, Izmir, 35430, Turkey * zeynepdurmus@iyte.edu.tr Abstract Libraries are specific spaces, of which indoor microclimate should meet rigorous requirements such as thermal comfort of humans and conservation of books, manuscripts and cultural property. Inadequate indoor microclimate (mainly temperature, relative humidity and of fluctuations) in libraries may possibly cause chemical, biological and mechanical degradations on paper-based collections. In this paper, indoor microclimate of Necip Paşa Library, the historical library located in Tire-Izmir, Turkey, was discussed from the perspective of preventive conservation of manuscripts. The library, which has no active heating, cooling and ventilation system, was modeled with the help of the building energy simulation tool, DesignBuilder. The indoor temperature and relative humidity was monitored throughout one year period and the model was calibrated with respect to the measurements. In order to reduce the degradation risks upon the manuscripts, ventilation srategies including natural and mechanical control were proposed. The results show that chemical degradation risks can be diminished to some extent. Keywords: Indoor microclimate, ventilation strategy, preventive conservation of manuscripts, dynamic simulation. I. Introduction Historical buildings are the significiant part of world cultural heritage. Historical library buildings contain manuscripts, the preservation of which is vital for human culture. If sufficient indoor microclimate is supplied in libraries, they may survive for centuries (Margus and Targo, 2015). Insufficient indoor microclimate conditions in a historical building may possible cause mechanical, biological and chemical degradation on cultural property. Therefore, conservation approaches have been improved to avoid degradation risks on cultural properties in the historical buildings. Two types of conservation approaches, namely direct and indirect physical interventions, can be implemented. The former means physical reactions on the property like stabilization, consolidation and disinfestations. Indirect intervation, contrarily, is defined as the preventive conservation, and mainly deals with environmental monitoring and control of storage area, good housekeeping, pest management and education of staff (Krüger and Diniz, 2011). The main objective of this study is to examine extant conservation conditions of manuscripts in the historic library, and improve it from the preventive point of view. In the literature, mechanical, chemical and biological degradation risks on the paper-based collections were investigated from the side of preventive conservation approach (Krüger and Diniz, 2011), (Bülow, 2002), (Fabbri and Pretelli, 2014). Fluctuations in temperature (T) and relative humidity (RH) are the main reason of mechanical degradation causing dimensional alterations, shrinking and swelling of manuscripts. The main reason behind chemical degradation is the amount of moisture content in hygroscopic materials, which may result in deterioration in text and discoloration in papers. Biological degradation is related to T, RH and substrate. As a result of biological degradation mould growth may be seen on the surface of manuscripts. Finally, the indoor microclimate must be controlled in order to avoid from these degradations with respect to defined instructions by (Bülow, 2002), (Silva and Henrique, 2015) and (Martens, 2012). A historical church was monitored a long-term period and a statistical risk-based analysis was created by Silva and Henrique. The indoor climate was evaluated according to standards defined by the EN 15757, PAS 198 and the historic set-point of 20⁰C and 50 %. Martens studied climate risk assessment in museums in his PhD thesis. He investigate the influence of set points for T and RH on the indoor microclimate by changing the values in the model. Two museums was observed during a long-term period and micro-climate of them was analysed by Bülow. According to results, he proposed the importance of ventilation and humidification for the indoor environment of the museums to preserve collections. Novel climate control systems are generally adapted to historical buildings to adjust the indoor environment in preventive conservation manner. 383

398 The motivation of this paper is to fulfill suitable indoor microclimate to preserve the manuscripts in Necip Paşa Library, Tire-İzmir, Turkey. The Library has 1156 manuscripts, 1312 books printed in the era of Ottoman Empire and more than 9000 books in Latin letters (Yıldırım, 2004). There is no HVAC system in the main hall, while doors and windows are used for ventilation of the building (Fig 1). A wooden octagonal shaped cage-like-structure where the manuscripts are separately preserved was located in the middle of the main hall (Fig 1). Preliminary results of the ongoing study in this library depict that indoor microclimate might cause the chemical degradation risk for manuscripts, while there is no indication about the mechanical and biological risks. III.1. Measurements Indoor and outdoor microclimate of Necip Paşa Library was monitored during one year period via automatic sensors. Five data loggers, which record data with ten minutes intervals, were used to measure T and RH from September 1, 2014 to August 31, Only one of them was placed to outside while the others were placed to inside. Locations of the data loggers are given in Fig 1. In this study, two control strategies which are mechanical and natural ventilation are proposed to reduce the chemical degradation risk of manuscripts in Necip Paşa Library with the help of building energy performance (BEP) tool, DesignBuilder (Version ). II. Necip Paşa Library Necip Paşa Library, which was built in 1827, located in the west of Turkey, Tire-Izmir and lies on north-south direction. The library consists of three zones: main hall, entrance and manuscript zone located inside of the main hall (Fig 1). The thickness of external walls is 1.08, 1.16 and 1.25 m for east, west and north (and south) walls, respectively. The lead covered brickwork roof within the shape of dome was used on the top of the main hall. All windows were single glazed with wooden frames and iron shuttered from inside. A split air-conditioner is placed into entrance for cooling and heating purposes, while there is no any HVAC system in the main hall and manuscript zone. A small number of specialist visit the Library on weekdays. Fig. 1: Schematic view of Necip Paşa Library (Source: UMART Architects and Engineering) III. Methodology Fig. 2: Methodology III.2. Modelling and calibration A BEP tool, namely DesignBuilder (v ), was used to model the Library. The model was divided into three thermal zones; main hall, manuscript zone and entrance zone, Fig 1. The Library was built 1.5 m height from the ground. Blower door test was carried out to measure airtightness value of the library and found as 0.5. h 1. The overall heat transfer coefficients (U) of external walls made of limestone and brickwork are 1.64 W/m 2 K, 1.76 W/m 2 K and 1.89 W/m 2 K for the south/north, west and east wall, respectively. The roof had a shape of dome, with a U value of 1.51 W/m 2 K. All windows of the Library are modelled as wooden frame with single glazing (U=5.89 W/m 2 K). The top surface material is the only difference between the sections of ground floor of the manuscript zone (U=1.24 W/m 2 K) and the main hall (U=1.38 W/m 2 K). The main scope of this study is to decrease chemical degradation risk on the manuscripts which kept in the Necip Paşa Library by using natural and mechanical ventilation systems. The steps of the study are given in Fig

399 Equation (3) and (4). LMx = ( 50% ) 1.3 Ea e R ( 1 Tx ) (3) RH x elm = 1 Ea 1 1 (50% n RHx )1.3 ( e R Tx n ) x=1 (4) Fig. 3: South facade (left) and BEP model (right) of Necip Paşa Library The weather data (based on T and RH), integrated into the model, was obtained from the one-year-outdoor measurements. The calibration of model was carried out with respect to the comparison of measurements (T and RH) and simulation results according to ASHRAE Guideline 14 (ASHRAE Guideline 14, 2002). Two dimensionless error indicators, mean bias error (MBE) and coefficient of variation of root-mean-square error (CV(RMSE)), were used as the criteria for calibration process and calculated by the formulas (1) and (2), respectively (Kandil and Love, 2014): MBE = N i (M i S i ) i=1 N i M i=1 i CV(RMSE) = 1 [[ N 2 i [(M i S i )2 ] i=1 ]] N i 1 N i M N i i i=1 (1) (2) In equations (1) and (2), measured and simulated data at instance i is shown by M i and S i, respectively. The number of used data is shown by N i. The upper limit for CV(RMSE) and MBE values were defined as 30% and ±10% for hourly measurements according to ASHRAE Guideline 14. If the calculated values lower then the upper limits, the model can be assumed as calibrated. III.3. Simulation for mechanical and natural ventilation In Equation (3) and (4), Ea is activation energy (100 kj/mol for degradation of cellulose), R is gas constant and the indice x indicates the value at instance in equation (3) and (4). The risk levels for the interpretation of elm values are given in Table 1. It is stated that the lifetime for the most objects duplicate every 5K decrement in T that is around 20 C (Michalski, 2003). Thus, mechanical and natural ventilation systems were integrated to model in order to provide elm values within acceptable limits and reduce the chemical degradation risk. Tab. 1: Interpretation of the elm values Ideal Good Some risk Potenti al risk High risk elm >2.2 [ [ [1-1.7[ [0.75-1[ <0.75 III.4. Comparison LM values of simulations and measurements were compared with each other and general risk assessments for the indoor microclimate of the Library were done with respect to the climate classes specified in ASHRAE Chapter 21 (Martens, 2012). IV. Results and discussion IV.1. Measurements In Fig 4, the manuscript zone and outdoor temperatures were depicted yearly. Equation (3) was used to calculate LM values for manuscript zone. The horizontal blue dashed lines in Fig. 4 indicate the LM risk levels. The restoratation of the Library was started at the beginning of July and the date was indicated with a vertical dashed line in Fig.4. Mechanical and natural ventilation systems were designed and introduced to the model to figure out whether the manuscripts are under chemical degradation risk or not between the dates where the preliminary results exhibited chemical degredation risk. The parameter called as Lifetime Multiplier (LM) which corresponds to the number of time spans an object remains unstable when compared to indoor climate of 20 C and 50%RH (Silva and Henrique, 2015, Martens, 2012), was used to investigate chemical degradation risk of the manuscripts. Equivelant lifetime multiplier (elm) measure the annual response of the objects. The elm values below than 0.75 and greater than 1 is an indicator of high risk and low risk, respectively. In between is considered as medium risk level (Martens, 2012). LMx and elm values are calculated by using 385 Fig. 4: Results of the preliminary study According to the results, the medium risk of chemical degradation on the manuscripts was seen in the months of April and November. As seen in Fig.4, the manuscripts are under high risk from May to October

400 while there is no risk between December and March. The elm value for the manuscripts is calculated as 0.89 and they are under potential risk according to Tab.1. As can be seen in Fig.4, T and LM are inversely proportional. Fig. 5: Change of LM values with respect to T and RH The cooling potential of both system were assessed in this study. IV.4. Comparison The LM values for mechanical and natural ventilation cases were calculated and the chemical degradation risk levels were shown in Fig.6 and Fig.7, respectively. The restoration of the library started after the vertical black dashed line in both figures. Introducing to mechanical and natural ventilation systems to the model reduce the chemical degradation risk on manuscripts for critical months which are from May to November. The results point out that the chemical degradation risks on manuscripts are moving from high risk level to medium risk level after implementing both cases in May and October. The highest value of LM was observed in January where the T is the lowest. RH values are changed between the 40% and 70% from September to July. As can be seen in Fig.5, the manuscripts is under chemical degredation risk when T exceeds 20⁰C. IV.2. Model calibration The calibration was made with respect to hourly T and RH measurements. All error values given in Table 2 were below the threshold values and satisfy the ASHRAE Guideline 14 based on MBE and CV(RMSE) calculations. Tab. 2: Discrepancies between the measurements and simulation results Temperature Relative Humidity Spaces MBE (%) CV(RMSE) (%) MBE (%) CV(RMSE) (%) Entrance Zone Main Hall Manuscri pt Zone ASHRAE Limits ±10 30 ±10 30 IV.3. Simulation for mechanical and natural ventilation Case 1: Mechanical Ventilation A mechanical ventilation system was introduced to the Main Hall and its T was set with respect to outdoor T. If the outdoor T below than 20⁰C, the system becomes active. The maximum airflow rate is selected as 3 ac/h for the Case 1. Case 2: Natural Ventilation Fig. 6: Comparison of the measurements with mechanical ventilation system Fig. 7: Comparison of the measurements with natural ventilation system The elm values of measurements were calculated as 0.89 from the Equation (4). After implementation of mechanical and natural ventilation to the model, the elm values were calculated as 0.97 and 0.92, respectively. The cases expand the life-span of the manuscripts and it can be said that mechanical ventilation is more effective solution for the Library. In natural ventilation system, a daily schedule were introduced to the BEP model to reduce T in the manuscript zone. In the schedule, all windows and door in the Main hall is opened from 22:00 to 08:

401 Tab. 3: Comparison of elm values Natural Natural Behavior Ventilaiton Mechanical Ventilation elm Improvement - % 3.4 % 9 Region Potential risk Potential risk Potential risk As it can be seen in Fig.5, when T decreases, LM value increases. The affect of RH on LM values can be seen in Fig.8 and 9. The sharp rise and drop in LM values are observed with respect to RH values. RH and LM is inversely correlated with each other. values. The cases increase the life span of the manuscripts. Finally, it can be concluded that chemical degredation risk on manuscripts can be diminished totally by controlling the T and RH together in the Library. To achieve this goal, T and RH in the Library need to be controlled by an HVAC system. Acknowledgements The authors wish to thanks Ali İhsan Yıldırım, who is the administrator of the library, for his valuable help and kind patience throughout the study. Besides, the authors would also like to thank the Prime Ministry Directorate General of Foundations, Republic of Turkey for the permission to take measurements in the library and their co-operation. References Fig. 8: Change of LM values with respect to T and RH (Natural ventilation) ASHRAE, Guideline , Measurement of energy and demand savings, American Society of Heating, Ventilating, and Air Conditioning Engineers, Atlanta, GA, (2002). Bülow A.E., Preventive conservation for paper-based collections in historic buildings, PhD thesis, De Montfort University, Leicester (2002). Fig. 9: Change of LM values with respect to T and RH (Mechanical ventilation) V. Conclusions Indoor microclimate of a historical Library is studied in this paper. According to preliminary results, manuscripts in the Library is under chemical degradation risks for some critical months while there is no biological and mechanical degedation risks. In this study, mechanical and natural ventilation systems are integrated to the historic library via BEP model to reduce the chemical degradation risks on the manuscripts. Before doing that the BEP model was calibrated hourly by tuning the parameters which affect the hygro-thermal conditions. The chemical degradation risks on the manuscripts are diminished in some extent after the integration of the ventilation strategies. This improvement is not elimate the risk on the manuscripts completely. The lifetime of the manuscripts is shown by elm 387 Design Builder, Version Available: m_docman/task,doc_details/gid,53/itemid,30/ (accessed ). Fabbri K., Pretelli M., Heritage buildings and historic microclimate without HVAC technology: Malatestiana Library in Cesena, Italy, UNESCO Memory of the World, Energy and Buildings, 76, (2014). Kandil A., Love J.A., Signature analysis calibration of a school energy modelusing hourly data, J. Building Perform. Simul. 7 (5) , (2014). Krüger E.L., Diniz W., Relationship between indoor thermal comfort conditions and the time weighted preservation Index (TWPI) in three Brazilian archives, Applied Energy, 88(3), pp (2011). Martens M.H.J., Climate risk assessment in museums, degradation risks determined from temperature and relative humidity data. PhD thesis, Eindhoven University of Technology (2012). Michalski S., Double the life for each five-degree drop, more than double the life for each halving of relative humidity, ICOM committee for conservation, 13th triennial meeting Rio de Janeiro preprints vol. 1, pp 66-72, (2003). Napp M., Kalamees T., Energy use and indoor climate of conservation heating, dehumidification and adaptive ventilation for the climate control of a mediaeval church in a cold climate, Energy and

402 Buildings, 108, (2015). Silva H.E., Henriques F.M.A., Preventive conservation of historic buildings in temperate climates. The importance of a risk-based analysis on the decision-making process, Energy and Buildings, 107, (2015). Şahin C.D., Arsan Z.D., Akkurt G.G., Preventive conservation of manuscripts in historic libraries via thermo-hygrometric approach: A case study of Necip Paşa Library, Tire-İzmir, Turkey, Building and Environment (under review). Yıldırım A.İ., Tire Necip Paşa Library catalogue of illuminated manuscripts, Tire (in Turkish), (2004). 388

403 Investigation of Thermodynamic and Environmental Performance Based on Subcooling Of Refrigerants in Direct Expansion System for Supermarket Applications Onder Altuntas 1*, M. Ziya Sogut 2, Enver Yalcin 3, T. Hikmet Karakoc 4 1 Faculty of Aeronautics and Astronautics, Anadolu University, TR26470, Eskisehir, Turkey 2 Department of Mechanical Engineering, Engineering Faculty, Orhangazi University, Bursa 3 Department of Mechanical Engineering, Engineering Faculty, Balikesir University, Balikesir 4 Department of Airframe and Powerplant Maintenance, Faculty of Aeronautics and Astronautics, Anadolu University * oaltuntas@anadolu.edu.tr Abstract This study focuses on effect of subcooled refrigerants on coefficient of performance (COP), exergy efficiency and environmental performances based on Total Equivalent Warming Impact (TEWI) and Human Health Damage (HHD) by considering different refrigerants. In the analyses, direct expansion (DX) system used commonly in supermarkets and the evaporating temperatures at -35 C and 0 C are taken as reference. The existing refrigerants such as R-22, R-404A and R-507 and the alternative refrigerant R-407C are examined separately. The results of analysis show that energy saving is at the range of 4% and 8% via the subcooling of 5 K in these systems while provided increase in the exergy efficiency is at the range of 3% and 25% depending on the refrigerant kinds. Besides, HHD and TEWI values also decrease with increasing of subcooling temperature. At the end of the study, it was emphasised that why the subcooling applications are important and why exergy efficiency is needed to these types of thermal applications especially DX systems. Keywords: DX systems; COP, refrigerants; exergy analyses; CO2 emission. I. Introduction Supermarkets use high amount of refrigerants and consume high-energy. Therefore, they should be addressed because of these reasons. Supermarkets with a sales area changing at the range of sqm are structures in where heavy energy consumption in commercial applications with an annual average energy consumption of 2 3 million kwh (Saba et al., 2009). Energy consumptions of these systems used for the protection of food in supermarkets can reach approximately 50% of the total energy consumption of a market. This ratio goes up to 65% in low capacity market applications (Baxter and David, 2003; Arthur, 2002). Around 60% 70% of the energy consumption in these systems is consumed in compressors and condensers (IEA, 2003). These systems have gained importance because of their negative impacts on total consumption costs as well as the environmental threats caused by refrigerants and energy consumption. This study held for developing of refrigeration systems is based on the reduction of energy requirements in the systems and their negative effects and on the choice of green refrigerants with low GWP (Global Warming Potential) and ozone depletion potential among existing refrigerants. With respect to the storage areas in market applications, it was observed that many refrigeration systems were used on basis of commercial requirements and that the Direct Expansion (DX) system was remarkable among them with a ratio of up to 60%. This system is followed by secondary loop systems, distributed systems, low-charge multiplex systems and advanced self-contained systems. Many studies about cooling applications in supermarkets have been made in the literature (Yang and Zhang, 2011;Kwon et al., 2012; Waqas and Kumar, 2011; Sogut et al., 2012). These studies can be classified as the studies based on energy analysis, the studies of mechanical cooling design and analyses based on refrigerants, and exergetic and environmental analyses of cooling systems. However, according to the literature review, the studies containing effects of subcooling on energetic, exergetic and environmental parameters have been rarely presented. Although subcooling applications are increase COP and exergy efficiency, investigation of these effects based on refrigerants and connected with a load has not encountered in the literature. In this study, first, the energy and exergy analyses, which were based on different refrigerants, were carried out for the DX systems. In addition, energetic, exergetic and environmental analysis of the subcooling applications between 0 K and 5 K were examined based on different refrigerants for DX systems. After all, environmental performances considering Human Health Damage (HHD) and Total Equivalent Warming Impact (TEWI) values were calculated for different refrigerants and its subcooling application at the range of 0 K and 5 K. 389

404 II. Direct Expansion System (DX) and Subcooling in Supermarket Applications II.1. Direct Expansion System (DX) Supermarkets with increasing importance in meeting daily food requirements are commercial structures of heavy energy consumption with an annual average of 1000 kwh/m 2. The cooling applications used in this type of structure are responsible about 50% of the total energy consumption changing with the cooling and air conditioning systems of different properties. In small capacities, cooling systems consume approximately 65% of total energy consumption. In these structures, where cooling applications of various types with large volumes are used, the charge amount of the refrigerants used in the system exceeded up to 2500 kg. In the supermarket applications of cooling systems, different cooling temperatures are needed to keep fresh different products. DX systems have four main components: evaporator, compressor, condenser and expansion valve, and the main work principle is based on a vapour compression cooling cycle. The flow schema of a DX system is given in Fig.1. of the application depends mainly on the cooling fluid. Subcooling caused increasing in the cooling capacity of the system needs an extra heat transfer surface area. Subcooling is used in old or new type direct expansion vapour compression system as the most usable system. The subcooling system can be either a dedicated mechanical sub-cooling system or an integrated mechanical sub-cooling system. Tab. 1: Thermodynamics features of refrigerants Thermodynamics features Refrigerants R-22 R-404A R-507 Chemical Formula CHC1F2 CHF2CF3/ CH3CF3/ CH2FCF3 CHF2CF3/ CH3CF3 R-407C (Alternative refrigerant) CH2CF3/ CH2F2/ CH2FCF3 Molecular weight (kg/kmol) Boiling point at (1.013 bar) Enthalpy of evaporation (kj/kg) Critical temp. (ºC) Critical Pressure (bar) Cplikid (kj/kgk) Cpvapor (kj/kgk) Environmental features ODP GWP Atmospheric life (year) Condenser Desuperheating Condensing Desuperheating Condenser III. Exergy Concept in Cooling Systems Power Compressor Subcooling Expansion valf Evaporating Superheating Evaporator Power Compressor Subcooling Superheating Evaporating Evaporator Condensing Heat exchanger As a measure of the machine s working activities in cooling systems, the thermal efficiency and the coefficient of performance (COP) have been defined. The coefficient of performance (COP) for a cooling machine is; a) Normal system b. System with subcooler Fig.1: Direct expansion system Q W L COP (1) net Direct expansion systems are the dominant technology for supermarkets world-wide. Choice of refrigerant in current applications includes R-22 gas from HCFC s and R-507 and R-404A from HFCs and for medium temperature applications is a smaller extent R-134a gas. Furthermore, in this study, refrigerant R-407C, which is a proposed as alternative gas, is examined and compared with current refrigerants. The thermodynamic and environmental parameters of these refrigerants are given in Table 1. II.2 Subcooling of refrigerants Why subcooling? The first aim is to cover pressure losses in liquid line due to friction and local losses. If it has not been applied, the evaporation of refrigerant begins before reaching to TXV. It is named as flashing. As a result of this, cooling capacity decreases lower values. Subcooling of the refrigerants in cooling systems provides an indirect saving by increasing the cooling capacity of the system. It works best at high and constant outdoor temperatures throughout the year. The effectiveness 390 where QL indicates the temperature moving away from the cooled environment and Wnet indicates the net effort input into the compressor the compressor (Sogut, 2011). Cooling machines are systems that work based on the cycle principle and the energy saving for the cycle may be expressed as follows: W net Q Q (2) H L where QH is the heat output from the machine. In cooling systems, COP is expressed as efficiency according to the first law of thermodynamics. In these systems, the working ability of a machine is explained by the second law of thermodynamics and this is defined as exergy. In the cooling system, exergy efficiency is compared to the possible highest COPtr of the real COP in the same work parameters. In a cooling machine, COPtr can be expressed as; COP tr TH TL 1 1 (3) where TH is the outside temperature and TL is the environmental temperature removed (Akpinar and

405 Hepbasli, 2007). In this case, the exergy efficiency of the cooling machine can be expressed as; COP II (4) COP tr If equation (1) and equation (3) are united ; T Q H L Q TL L II (5) W net The improvement of the system or process based on exergy efficiencies depends on reduction of the exergetic losses or destruction of the system. The exergy loss or irreversibility is based on the difference between the exergy input and output (Van Gool, 1997). The improvement potential (IP) is developed for irreversibility processes used in thermal applications (Hammond, 2007). IP 1 ).( Ex E x ) (6) ( II in out refrigeration systems, which are directly affect climate change, are expressed by direct and indirect emissions. The direct impact of emissions for refrigerants contains effects based on leakages, maintenance and losses in the systems. Electrical energy consumption of the system relates to indirect effects in terms of CO2 emission potentials. Besides, this effect includes indirect parameters of systems like material processing, maintenance, and waste disposal (Eurammon, 1996). The CO2 emission for the production of electricity is dependent upon the mixture of energy carriers used in power plants. The greenhouse effect directly is a global problem caused by Earth s climate change. The study is based on published data on CO2 emissions from power plants and utilizes the average value for the regions of North America, Japan and Europe. In this study, the average value is 0.47 kg CO2/kWh (Hellmann and Barthélemy, 1997). IV.2 HHD concept IV. Environmental Parameters for Cooling Systems IV.1 TEWI concept A concept as called Total Equivalent Warming Impact (TEWI) was proposed by the US Department of Energy and Oak Ridge National Laboratories for evaluating global warming effect in the Copenhagen meeting. According this, emissions potential effects of the greenhouse gas such refrigerants used in air conditioning and cooling systems, which caused global warming and climate change, can easily be calculated. The concept was firstly developed and defined as standard of EN 378:2000. The concept can directly be used to compare relative impact of the refrigerant systems. This concept is basically based on environmental effects, which relates global warming potential (GWP) of the refrigerants, and energy consumption of the system dependent on its life time. Besides, changes of the TEWI value shows overall effects of the system as a measure of the poorer on environmental (Horst, 2000). TEWI can be expressed as; TEWI [ GWP. L. n ] [ GWP. m.(1 recov ery)] [ n. Eannual. ] (7) Leakagelosses Direct GWP Recov ery losses Energyconsumptio n Indirect GWP where E annual is the annual energy consumption, β is the CO 2 emission produced by power generation, n is the system operating time, m is the refrigerant charge, α recovery is the recovery or recycling factor, L is the amount of leakage, GWP is the global warming potential, TEWI (kgco2/year) is the total equivalent warming impact (Moore, 2005). Evaluation of the GWP effect for the substances or gasses is made by comparison of CO2, which is used an index. For this, GWP value of the substances or gasses are found by divided GWP impact of the unit CO2 covered 100, 500 year time span (DUPONT, 2005). Emission potentials of the refrigerants in the 391 Climate change, caused by ozone which is formed by the formation of Global Warming Gases (especially CO2), primary effects biological diversity, and creates a significant health threat on human life. Emission of refrigerants, used in refrigerant systems, and CO2 emission, produced by power plants due to consumption of more energy, because of air-conditioning systems, used in providing comfort conditions. This indirectly leads to serious effects on human health (Medindia, 2012). In this instance, the prevention of human health damages are so critical. HHD, one of damage categories, is expressed as disability-adjusted life years (DALY). DALY is expressed as the number of years lost due to problems with health, disability, or early death. One DALY essentially can be thought of as one lost year of healthy life. HHD consists of respiratory and carcinogenic damages, global exchange, and increasing radiation rains, all causes of ozone layer depletion (Goedkoop and Spriensma, 2001). In this study, HHDs are evaluated through the Eco-indicator-99 software, an impact assessment method for the LCA (Life Cycle Assessment). HHDs of TEWI were calculated using the SimaPro software. Also, Eco-indicator 99 has a damage assessment step: This means that the impact category indicator results that are calculated in the characterisation step are added to form damage categories. Addition without weighting is justified here because all impact categories that refer to the same damage type have the same unit. This procedure can also be interpreted as grouping. The damage categories (and not the impact categories) are normalised on a European level (damage caused by one European per year), mostly based on 1993 as a base year, with some updates for the most important emissions (PRé Consultants, 2006). V. Results and Discussion

406 DX Systems with varying cooling capacity from 20 kw to 1000 kw are generally operated at low conditions (-35 C<T<0 C) for the refrigerants of R-22, R-404A and R-507. In addition to these refrigerants, R-407C gas was examined as an alternative refrigerant. According to these references, some parameters are taken as limit values for the analyses. The properties and cycle parameters of the DX systems used in the study are given in Table 2. Tab. 2: The parameters of DX systems Features of DX systems Features Min Max Cooling capacity (kw) Refrigerant charge (kg) Leakage rate % 3 33 Energy consumption (kwh) Approximate percentage of sector refrigerant emissions in subsector (%) 47 Cycle parameters for analyses Cooling capacity (kw) 1 Evaporator temperature C -35/0 Condenser temperature C 40 Evaparator Superheating ºC 7 Condenser Super-cooling ºC 0-5 Fig.3: Exergy efficiency distrubitions of refrigerants based on sub-cooling temperatures The average change in the exergy efficiency of the R-404A gas, which is most commonly used in DX systems, was found to be 28.91% for the sub-cooling range of 0K to 5K. In the R-22 gas with restricted use, this value had a change rate of 3.71%. For the R-407C gas, which was assessed as an alternative for the DX systems, the change ratio in the exergy efficiency based on the sub-cooling change was found to be 8.31%. These changes in the system performance will directly affect energy consumption. The power consumption changes of the system depending on this data and refrigerant fluids are given in Fig.4. In the study, the COP and exergy efficiencies were examined separately for four refrigerants by taking into consideration each sub-cooling temperature and -35 C/0 C evaporator temperatures. COP changes of each refrigerant depending on the sub-cooling temperatures are given in Fig.2. Fig.4: Power consumptions of refrigerants based on sub-cooling temperatures Fig.2: COP distributions of refrigerants based on sub-cooling temperatures. According to COP analyses, excessive cooling affects efficiency in the range of 3.6% and 8.71%, depending on the choice of refrigerant fluid. The lowest COP change was 2.71 in 0 K sub-cooling for the R-22 gas while this value is 2.81 for 5 K sub-cooling. For the R-507 gas, the highest COP change was found to be 2.34 for 0 K and 2.53 for 5 K. The average change ratio for the R-404A gas, which was most commonly used in the system, was found to be 7.71%. In the alternative refrigerant fluid R-407C gas, the COP value increases from 2.57 at 0 K to 2.70 at 5 K sub-cooling. The changes in the exergy efficiencies, depending on sub-cooling temperatures, were examined and results are given in Fig.3. Power consumption and changes of each refrigerant for the cooling capacity of 1 kw were examined depending on the sub-cooling application in DX systems. The power consumption of the R-22 gas with restricted use was found to be kw for 0 K and kw for 5 K while the sub-cooling based saving ratio was found to be 4.53%. The commonly used R-404A gas was found to be In addition, this fluid was found to have the highest energy consumption value among all refrigerants. The energy save potential based on the subcooling temperature change was found to be 5.14% for the R-407 gas, which was examined as an alternative fluid. This change was low compared to the R-404A gas while the energy consumption of the R-407C gas was found to be 8.38% lower for 0K. Moreover, when the energy consumption depending on sub-cooling changes was compared, the save potential average of the R407C gas was found to be 16.61% lower than the R-404A gas. A similar study was conducted for the R-507 gas used in the DX systems. The energy save potential of the R-507 gas was found to be 8.01% on average. When the save potential of this gas was compared to the R-407C gas, the R-407C 392

407 gas was found to have 17.15% lower energy consumption. The energy consumption of the R-404A gas was found to be 0.6% lower than the R-507 gas. These changes in the energy consumptions of the refrigerant fluid based DX systems for the subcooling applications will affect the environmental performances of the systems, as well. In environmental performances, CO2 equivalent TEWI values were examined separately by taking into consideration the minimum and maximum leakage ratios in DX systems and the temperatures at 0K and 5K. Based on the minimum leakage ratios, the changes in TEWI values of the DX systems with a cooling load of 1 kw are given in Fig.5. Fig.5 TEWI values of refrigerants based on sub-cooling temperatures for 3% leakage rate Power consumption and changes of each fluid for the cooling capacity of 1 kw were examined depending on the sub-cooling application in DX systems. The power consumption of the R-22 gas with restricted use was found to be kw for 0 K and kw for 5 K while the sub-cooling based saving ratio was found to be 4.53%. The commonly used R-404A gas was found to be In addition, this fluid was found to have the highest energy consumption value among all refrigerants. In DX systems, the saving values in the CO2 equivalent TEWI values of the R-22, R-404A, R-507 and R-407C refrigerants for the 0 K and 5K sub-cooling based on 3% leakage rate were 4.40%, 6.98%, 7.62% and 5.01%. In the systems, the gas with the highest CO2 equivalent emission effect is the R-507 gas with kg/y at low leakage rate. The TEWI value of the R-404A gas, which is commonly used in DX systems, was found to be kg/y. The emission values of the R-404A and R-507 gases were found to be 17.67% and 19.2% higher than the R-22 gas for low leakage rate and 0 K sub-cooling value. Emission values of the R-407C gas examined for DX systems were found to be kg/y and kg/y based on the 0 K and 5 K sub-cooling values respectively. The emission value changes of this gas were 5.01% higher than the R-22 gas and 10.53% and 11.68% lower than the R-404A and R-507 gases for 0 K. The said changes for 5 K were respectively 4.40% higher and 8.67% and 9.19% were lower. In this change, the effect of the sub-cooling applications was observed. A similar evaluation was made for the maximum leakage rate of 33% for the DX systems and the TEWI values based on the sub cooling temperatures are given in Fig.6. Fig.6: TEWI values of refrigerants based on sub-cooling temperatures for 33% leakage rate In the DX systems, the reductions of the TEWI values of the R-22, R-404A, R-507 and R-407C between the sub-cooling temperatures of 0 K and 5 K were found to be 3.38%, 4.69%, 5.13% and 3.99% respectively. The highest TEWI value for the current gases was for the R-507 gas with kg/y for 0 K sub-cooling and kg/y for 5 K sub-cooling. The TEWI values of the R-404A gas were found to be very close to the R-507 gas for both sub-cooling temperatures. The TEWI values of the alternative refrigerant R-407C gas were found to be kg/y for 0 K and kg/y for 5 K for a leakage ratio of 33%. In HHD of TEWI, the DALY per kwh of generated energy is taken into consideration for the calculation of indirect TEWI contribution. DALY, one of the damage categories unit of Eco-indicator 99, means different disability caused by diseases are weighted. In this study, it is accepted that HHD of 1 CO2e is about DALY. HHD of TEWI based on sub-cooling temperature for 3% and 33% leakage rate are shown in Figs. 7 and 8, respectively. Fig. 7: HHD of TEWI based on sub-cooling temperature for 3% leakage rate Fig. 8: HHD of TEWI based on sub-cooling temperature for 33% leakage rate 393

408 As shown in these figures, while the maximum HHD values was calculated as DALY/kWh at 0 C in 33% leakage rate, respectively, the minimum HHD values was found to be DALY/kWh at 5 C in 3% leakage rate. For a low human health impact on demand, higher temperature and lower lekage values should be selected. HDD of TEWI decreased with increasing of sub-cooling temperature. While maximum decrease ratio of HDD was found to be 1.60% and 1.06% for 3%, and 33% leakage rates, respectively, at R507, minimum ratio was calculated as 0.90% and 0.69% for 3% and 33% leakage rates, respectively, at R-22. VI. Conclusions In this study, COP, exergy and environmental performances of the DX systems considering sub-cooling applications in the range of 0 K and 5 K were examined by taking into consideration the minimum and maximum leakage ratios of the DX systems. The results according to the analyses are given below: a) When the system is sub-cooling, the COP performance of R-404A gas, which is extensively used in refrigerant, increases from 2.36 to 2.55, and the COP change is 7.71%. The COP value of alternative refirgerant R-407C increases from 2.57 to 2.70 and the COP change is 5.1%. b) Among current refrigerants, exergy efficiency of R-407C gas was found 22.67% and it has better performance 10.07% than R-404A gas. Compared to exergy efficiencies of R-22 (HCFC gas) with other gases the average exergy efficiencies of the R-404A, R-507 and R-407C gases were found, respectively, 14.33%, 13.58% and 4.82% lower than the R-22 gas. c) When the compressor energy consumption based on per unit cooling capacity for the 0K and 5K subcooling applications was examined, it was found that can be obtained saving respectively 4.53% for R-22 gas, 7.33% for R-404A gas, 8.01% for R-507 gas and 5.14% for R-407 gas. temperature. If there is some reduction with the HHD, subcooling temperature should be increased. The study shows that exergy analyses and CO2 equivalent TEWI parameter depending on refrigerant choice taken as a reference will be a significant guide for the evaluation of environmental effects and thermal efficiency in the system design especially DX systems. Moreover, in supermarket applications, a significant saving is observed when the sub-cooling application is considered in DX systems with high energy consumption and when this efficiency is achieved in condencer surface areas. The using of subcooling applications in old DX system applications with low efficiency will be provided a significant saving. References Akpinar E.K., Hepbasli A., A comparative study on exergetic assessment of two ground- source (geothermal) heat pump systems for residential applications, Building and Environment, 42 (2007) Arthur D. Little, Inc. (ADL)., Global comparative analysis of HFC and alternative technologies for refrigeration, air conditioning, foam, solvent, aerosol propellant, and fire protection applications, Final Report to the Alliance for Responsible Atmospheric Policy (2002). Available at Baxter, Van D. and David H. Walker., Analysis of advanced low-charge refrigeration for supermarkets, ASHRAE Transactions (2003). DUPONT Dupont refrigerants the science of cool, Du Pont de Nemours (Deutschland) GmbH, Germany (2005). Eurammon, Evaluation of the environmentally friendly refrigerant ammonia according to the TEWI Concept, NH3 for ecologically friendly future, Frankfurt, Germany (1996). d) According to the CO2 equivalent TEWI values, R-22 gas, among the current gases, has 17.67% and 19.20% lower emission than the R-404a and R-507 gases respectively. For the range of 0K and 5K and the leakage rate 3%, the average CO2 equivalent emission saving can be provided 6.98% for R-404A gas and 5.01% for R-407C gas. Besides the average CO2 equivalent emission savings for the leakage rate 33%, are found 4.69% for the R-404A gas and 3.99% for the R-407C gas. e) For a low HHD on demand, higher temperature and lower lekage values should be selected. According to HHD calculations, HDD of TEWI decreased with increasing of subcooling 394 Goedkoop, M. and Spriensma, R. The Eco-Indicator 99. A Damage Oriented Method for Life Cycle Impact Assessment, Methodology Report, Pré Consultants [online] (2001) (accessed 01 April 2012). Hammond A.J. Stapleton, Exergy analysis of the United Kingdom energy system. Proceedings of the Institute of Mechanical Engineers (2001);215 (2): Hellmann J., and Barthélemy P. AFEAS-TEWI III study: results and evaluation of alternative refrigerants, Solvay Fluor und Derivate GmbH Technical Service-product Refrigerants, Bulletin no. C/11.97/06/E, page 5, (1997).

409 Horst K. Refrigerant use in Europe, ASHRAE journal September (2000), IEA International Energy Agency, IEA Annex 26: Advanced süpermarket refrigeration/heat recovery systems, Final Report Volume 1 Executive Summary,Compiled by Van D. Baxter, Oak Ridge National Laboratory, April (2003). Kwon L., Hwanga Y., Radermachera R., Kim B., Field performance measurements of a VRF system with sub-cooler in educational offices for the cooling season Volume 49, Pages , June (2012). Medindia, Health Effects of Global Warming, (2012) /health-effects-of-global-warming.htm) Moore D. A comparative method for evaluating industrial refrigerant systems, Sabroe Ltd. (reva). Nov. (2005), Saba S., Slim R., Palandre L., Clodic C. D., Inventory of direct and indirect ghg emissions from stationary air conditioning and refrigeration sources, with Special Emphasis on Retail Food Refrigeration and Unitary Air Conditioning, Final Report, State of California Air Resources Board Research Division, PO Box 2815 Sacramento, CA 95812, Page 60, March (2009). PRé Consultants SimaPro End User License Agreement (EULA) and Service Level Agreement (SLA), (2006) [online] (accessed 1 March 2012). Sogut Z. A study on the exergetic and environmental effects of commercial cooling systems, Int. J. Exergy, Vol. 9, No. 4. (2011). Sogut M.Z., Yalcin E., Karakoc H., Refrigeration inventory based on CO2 emissions and exergetic performance for supermarket applications, Energy and Buildings (2012) Yang L., Zhang C.L. On subcooler design for integrated two-temperature süpermarket refrigeration system, Energy and Buildings (2011) Van Gool W. Energy policy: fairly tales and factualities, in: O.D.D. Soares, A. Martins da Cruz, G. Costa Pereira, I.M.R.T. Soares, A.J.P.S. Reis (Eds.), Innovation and Technology Strategies and Policies, Kluwer, Dordrecht, (1997). Waqas A., Kumar S. Thermal performance of latent heat storage for free cooling of buildings in a dry and hot climate: An experimental study, Energy and Buildings 43, page (2011). 395

410 Techno-Economic Assessment of Solar-Geothermal Based Multigeneration System for a Community Farrukh Khalid 1 *, Ibrahim Dincer 1, Marc A. Rosen 1 1 Faculty of Engineering and Applied Science, University of Ontario Institute of Technology, 2000 Simcoe Street North, Oshawa, Ontario, L1H 7K4, Canada farrukh.khalid@uoit.ca Abstract A techno-economic assessment is conducted for a multigeneration system comprising two renewable energy subsystems geothermal and solar to supply electrical power, cooling, heating, and hot water for a community. The proposed system is evaluated in terms of energy and exergy efficiencies, and the values are determined to be 20.2% and 19.2%, respectively. A parametric study is carried out to assess the effect of various parameters on the system energy and exergy efficiencies. In order to perform the economic assessment, the hybrid optimization of multiple energy resources (HOMER) software is used. HOMER analysis shows that the net present cost of the optimized electrical power system is $472,443 and levelized cost of electricity of $0.088/kWh. Keywords: Energy, exergy, efficiency, geothermal, solar, cost assessment. I. Introduction The increase in the demand of energy for buildings in many countries and the extensive use of fossil fuels to meet this demand creates many challenges. Efforts are being made as a consequence to find more sustainable options for building energy systems. The use of renewable energy and multigeneration systems can assist these efforts. Solar energy is one of the most abundant form of renewable energy available in the nature. The main drawback associated with the solar energy is its fluctuating nature. This can in some cases be overcome by integrating it with other renewable energy resources such as geothermal (Bicer and Dincer, 2016; Khalid et al., 2015; Dincer and Zamfirescu, 2012; Dagdougui et al., 2008). Recently, much research has been reported on the integration of different renewable energy sources. Calise et al. (2016) investigate a geothermal and solar based system for producing energy and water. They found that the share for geothermal energy is more prominent than that for solar for their plant supply, as the geothermal source is mostly constant throughout the year for the location considered in their study. Chauhan and Saini (2016) examine an integrated renewable energy system for a community, considering nine combinations of renewable energy sources. Ayub et al. (2015) demonstrate that a system comprising of solar and geothermal is more economic than solar or geothermal energy alone. Ghasemi et al. (2014) show that a hybrid system consisting of solar and geothermal energy is more efficient than a solar or geothermal system alone. Suleman et al. (2015) investigate a multigeneration system based on solar and geothermal sources comprised of two organic 396 Rankine cycles for electricity and an absorption chiller for cooling. Through energy and exergy efficiency analyses of a mulitgeneration system based on solar and geothermal resources, Ali and Dincer (2014) find that the maximum exergy destruction take place in the solar collector system. Zhou et al. (2013) examine a hybrid system that uses geothermal and solar energy. They find that the hybrid system can be more efficient on both energy and exergy bases, and report that there can be a 20% reduction in the cost of electricity production when the hybrid system is used instead of a standalone geothermal system. Tempesti et al. (2012) study two micro combined heat and power systems using organic Rankine cycles based on geothermal and solar energy, and find that highest exergy destruction takes place in the parabolic trough collector. Bakos and Tsagas (2003) study a system comprised of various renewable energy sources for a house in Greece. The above studies shows that combining solar with other renewable such as geothermal can be advantageous. In the present study a new hybrid system based on geothermal and solar energy is proposed, developed and assessed. The specific objectives of this paper are summarized as follows: To develop a new hybrid system based on solar and geothermal energy for a community, in order to provide multiple useful outputs such as cooling, electric power, heating and hot water. To thermodynamically analysis the developed system using energy and exergy analyses. To carry out a parametric study to determine the effects of various parameters like ambient temperature on the overall performance of the

411 system in terms of energy and exergy efficiencies of the overall system. To do a cost assessment of the developed system in terms of levelized cost of electricity and net present cost. II. System Description A hybrid multigeneration system based on solar and geothermal energy for a community is developed. The system is designed to provide electricity, hot water, heating, and cooling. The developed system consists of a concentrated solar power (CSP) collector, a geothermal source, three Rankine cycles, and a vapor absorption chiller. Fig. 1 shows a schematic diagram of a multigeneration system... Duratherm oil at state 13 enters the solar collector and, after being heated, exits at state 14. The oil then enters heat exchanger 2, which also acts as a storage tank and is utilised in heating the water that enters the storage tank at state 20 to provide the energy requirement of Rankine cycle Condenser 1 7 Generator 16 CSP HEX 1 14 Chilled water from building 11 EV 1 9 Evaporator 1 12 Chilled water to building 10 2 Pump 1 1 Absorber 5 EV 2 6 HEX 2 HEX Steam generator RC 1 35 Pump 2 23 HPST HPST Condenser 2 Air Hot air 30 HPST 3 31 RC 3 Separator LPST MC 33 Pump 3 24 RC 2 Condenser Water Hot water FC 1 36 Separator 2 40 Condenser 4 34 FC Geothermal production well Reinjection well Fig. 1 Schematic illustration of integrated solar-geothermal multigeneration system. The oil is then passes to heat exchanger 3 which is used to heat stream 35 (water) to stream 23 to drive high pressure steam turbine 2 (HPST2) in Rankine cycle 2. Heat is transferred from the hot oil at state 16 to the generator of the vapor absorption chiller and the oil leaves the generator at state 13 and is pumped back again to the solar collector. Stream 17 (water) enters high pressure turbine 1 (HPST1) to 397 generate electricity. After leaving HPST1, stream 18 is sent to condenser 2. Stream 19 passes through a pump and is returned to heat exchanger 2 in order to be heated by the oil. The stream leaving heat exchanger 3 (23) is sent to HPST2 to produce electricity. After HPST2, stream 24 is sent to condenser 3, which provides hot water for the community. The stream leaving condenser is

412 directed to the reinjection well. The stream from the geothermal production well (29) is sent to a flash chamber 1 (FC1) which is then passed to separator 1 where it is separated into two streams (31 and 36). Stream 31 (vapor) enters high pressure steam turbine 3 (HPST3) and then the mixing chamber. Stream 36 (liquid) is further flashed in flash chamber 2 (FC2) and then separated in separator 2and the liquid portion (stream 39) is directly sent to the reinjection well. The vapor from the second separator (stream 38) is sent to a low pressure steam turbine (LPST), from which the exit stream is conveyed to the mixing chamber along with stream 32. The outlet of the mixing chamber is condensed and pumped to heat exchanger 3 in order to be heated and used in high pressure turbine 2. III. Analysis III.1. Balance Equations General mass, energy, entropy, and exergy balance equations for the system shown in Fig. 1 can be written as follows: i m i Q W + e m e gz e ) = de cv dt i m i = dm cv dt m i(h i + V i + gz i ) m e(h e + V e 2 s i + Q k k T k Ex Q + m i(ex i ) i i 2 2 III.2. Energy Efficiencies i m e s e + S gen = ds cv dt e + 2 e m e (ex e ) Ex WEx d= dex cv dt (1) (2) (3) (4) The energy efficiencies can be defined for the systems considered here as the ratio of useful energy output to the total energy input. In this study the energy efficiencies of Rankine Cycle 1, the absorption chiller and the overall system are presented below. Rankine Cycle 1 The energy efficiency of the Rankine Cycle 1 can be expressed as η en,rc1 = W HPST1 W P1 m 14h 14 m 15h 15 (5) Absorption Chiller The energetic COP of the absorption chiller can be written as COP en,ac = Q e1 Q generator (6) Overall System Here, the energy efficiency of the system is defined as the ratio of total useful outputs (hot water, cooling, electric power and heating) to the total energy inputs (solar and geothermal). The energy efficiency of the overall system can be written as follows: η en,ov = (Q e1+m 28h 28 m 27h 27 +m 22h 22 m 21h 21 +W HPST1+W HPST2+W HPST3+W LPST W P2 W P3) m 29h 29 m 26h 26 +Q sol (7) where Q sol and m 29h 29 m 26h 26 are the solar and geothermal energy inputs to the overall system, respectively. Also, Q e1, m 28h 28 m 27h 27, m 22h 22 m 21h 21, and W HPST1 + W HPST2 + W HPST3 + W LPST W P2 W P3 represent the cooling, heating, hot water, and net electric outputs of the overall system, respectively. III.3. Exergy Efficiencies The exergy efficiencies for Rankine Cycle 1, the absorption chiller and the overall system are written as follows: Rankine Cycle 1 The exergy efficiency of the Rankine Cycle 1 can be expressed as η ex,rc1 = W HPST1 W P1 m 14ex 14 m 15ex 15 (8) Vapor Absorption Chiller The exergetic coefficient of performance of the vapor absorption chiller can be expressed as COP ex,ac = Q e1( T 0 Te1 1) Q generator(1 T 0 Ts ) (9) Overall System The exergy efficiency of the overall system can be written as follows: η en,ov = Q (Eẋ cooling +m 28ex 28 m 27ex 27 +m 22ex 22 m 21ex 21 +W HPST1+W HPST2+W HPST3+W LPST W P2 W P3) Q m 29ex 29 m 26ex 26 +Eẋ sol IV. Results and Discussion (10) The results obtained from the analysis show that the energy and exergy efficiencies of the overall system are 20.2% and 19.2%, respectively. It is also found that the energy efficiency of Rankine Cycle 1 is 20.2% (see Table 1). The developed system can satisfy a cooling load of 140 kw. Fig. 2 shows the exergy destruction rate of the main components of the system. The greatest exergy destruction rate occurs in the CSP, the next highest in HEX 2 and the third highest in condenser 3. The variation of the energetic and exergetic COPs of the vapor absorption cycle with ambient temperature is shown in Fig. 3. As the ambient temperature increases from 15 to 35, there is no change in 398

413 the energetic COP while the exergetic COP increases from 0.11 to This is due to the fact that as the ambient temperature increases, the exergetic output of the evaporator increases resulting in the increase in exergetic COP. The effect of ambient temperature on the energy and exergy efficiencies of Rankine Cycle 1 is shown in Fig. 4. As the ambient temperature changes, the energy efficiency remains the same while the exergy efficiency increases from 47.54% to 52.45%. This trend is due to the fact that, as the ambient temperature increases, the exergetic input of Rankine Cycle 1 decreases. chiller increases, resulting in the increase in exergy efficiency of the system. Energy Efficiency h en,rc1 h ex,rc T 0 ( o C) Fig. 4 Effect of ambient temperature (T0) on the energy and exergy efficiencies of RC Exergy Efficiency Fig. 2 Exergy destruction rates of selected units of system. Table 1 Parameter values from modeling and energy and exergy analyses of system. Parameter Value Output of LPST (kw) Output of HPST 1 (kw) Cooling load (kw) 140 Energy efficiency of RC 1 (%) 20.2 Exergy efficiency of RC 1 (%) 49.8 Energy efficiency of the overall system (%) 20.2 Exergy efficiency of the overall system (%) Energy Efficiency Fig. 5 Effect of ambient temperature (T0) on the energy and exergy efficiencies of the overall system. Fig. 6 shows the effect of oil outlet temperature on the energy and exergy efficiencies of the overall system. By increasing the oil outlet temperature both energy and exergy efficiencies of the overall system increases h en,ov h ex,ov T 0 ( o C) Exergy Efficiency COPen COP en,ac COP ex,ac T 0 ( o C) Fig. 3 Effect of ambient temperature (T0) on the energetic and exergetic COPs of the vapor absorption chiller. The effect of ambient temperature on the energy and exergy efficiencies of the overall system is shown in Fig. 5. As the ambient temperature changes from 15 to 35, there is no change in the energy efficiency while the exergy efficiency of the overall system changes from 18.8% to 19.7%. The reason for this trend is that, on increasing the ambient temperature, the exergetic input of the sun decreases and the exergetic output of the absorption COPex 399 Energy Efficiency Fig. 6 Effect of oil outlet temperature (T14) on the energy and exergy efficiencies of the overall system. Fig. 7 shows a schematic diagram for the system design in HOMER. The total connected load is 1,151 kwh daily. Table 2 shows that the requirements are 120 kw HPST1, 50 kw HPST2, 10 kw and 5 kw LPST for the optimized electric power system to meet the electricity demand of the community. Table 3 shows the cost summary of the optimized electric power system. In the table, the cost of electricity is h en,ov h ex,ov T 14 ( o C) Exergy Efficiency

414 seen to be $0.088/kWh and the total cost of the system as $472,443. while HPST1 and HPST2 are comparable in terms of operation cost. The monthly production of the electricity is shown in Fig. 9. It can be noted from the figure that the electricity produced by HPST1 is the greatest followed by HPST2. Table 2 List of components and their capacities for optimized electric power system. Component Capacity (kw) HPST1 120 HPST2 50 HPST3 10 LPST 5 Fig. 7 Schematic of electric power system design in HOMER. Fig. 8 shows the cash flow summary of the optimized electric power system. The capital cost of HPST1 is around 60% of the total capital cost of the system, Table 3 Cost summary of the optimized electric power system. Cost parameter Value Total net present cost $472,443 Levelized cost of energy $0.088/kWh Operating cost $5,667/y Fig. 8 Cash flow summary of the optimized electric power system. V. Conclusions Fig. 9 Monthly production of electricity for optimized electric power system. In this study, a new solar and geothermal based hybrid system is proposed to meet the electric, cooling, heating, and hot water demands of a community. The performance of the proposed 400 system is evaluated in terms of the energy and exergy efficiencies. The results show that the overall system has energy and exergy efficiencies of 20.2%, and 19.2%, respectively. Furthermore, it is found that the maximum exergy destruction occur in the concentrated solar panel. The cost assessment

415 shows that net present cost of the optimised electric power system is $472,443 and that the levelized cost of energy is $0.088/kWh. Acknowledgement The authors acknowledge the support provided by the Natural Sciences and Engineering Research Council of Canada. Nomenclature COP : Coefficient of performance ex : Specific exergy (kj/kg) Eẋ : Exergy rate (kw) g : Acceleration due to gravity (m/s 2 ) h : Specific enthalpy (kj/kg) m : Mass flow rate (kg/s) P : Pressure (kpa) q : Solar radiation received per unit area (W/m 2 ) Q : Heat transfer rate (kw) s : Specific entropy (kj/kg K) T : Temperature ( 0 C) W : Work rate or power (kw) Greek Letters η : Efficiency Subscripts ac : Absorption chiller cv : Control volume d : Destruction e : Exit en : Energy ex : Exergy gen : Generation hex : Heat Exchanger i : Inlet ov : Overall sol : Solar Acronyms COE : Cost of Electricity COP : Coefficient of Performance CSP : Concentrated Solar Power EV : Expansion Valve HEX : Heat Exchanger HOMER : Hybrid Optimization of Multiple Energy Resources HPST : High Pressure Steam Turbine LPST : Low Pressure Steam Turbine MC : Mixing Chamber NPC : Net Present Cost RC : Rankine Cycle Ayub M., Mitson A., Ghasemi H., Thermoeconomic Analysis of a Hybrid Geothermal Power Plant, Energy , (2015). Bakos GC.,Tsagas NF., Techno-economic assessment of a hybrid solar/wind installation for electrical energy saving, Energy and Buildings 35, (2003). Bicer Y., Dincer I., Development of a new solar and geothermal based combined system for hydrogen production, Solar Energy 127, , (2016). Calise F., Accadia DM., Macaluso A., Piacentino A., Vanoli L., Exergetic and exergoeconomic analysis of a novel hybrid solar-geothermal polygeneration system producing energy and water, Energy Conversion and Management 115, , (2016). Chauhan A., Saini RP., Techno-economic feasibility study on integrated renewable energy system for an isolated community of India, Renewable and Sustainable Energy Reviews 59, , (2016). Dagdougui H., Minciardi R., Ouammi A., Robba M., Sacile R., Modeling and optimization of a hybrid system for the energy supply of a green building, Energy Conversion and Management 64, , (2012). Dincer I., Zamfirescu C., Renewable-energy-based multigeneration systems, International Journal of Energy Research 36, , (2012). Ghasemi H., Sheu E., Tizzanini A., Paci M., Mitsos A., Hybrid solar geothermal power generation: optimal retrofitting, Applied Energy 131, , (2014). Khalid F., Dincer I., Rosen MA., Energy and exergy analyses of a solar-biomass integrated cycle for multigeneration, Solar Energy 112, , (2015). Suleman F., Dincer I., Chaab MA., Development of an integrated renewable energy system for multigeneration, Energy 78, , (2014). Tempesti D., Manfrida G., Fiaschi D., Thermodynamic analysis of two micro CHP systems operating with geothermal and solar energy, Applied Energy 97, , (2012). Zhou C., Doroodochi E., Moghtaderi B., An in-depth assessment of hybrid solar geothermal power generation, Energy Conversion and Management 74, , (2013). References Ali M., Dincer I., Energetic and exergetic studies of a multigenerational solar-geothermal system, Applied Thermal Engineering 71, 16-23, (2014). 401

416 PV Array Based Smart Home Automation System Ahmet Senpinar College of Technical Sciences, Department of Electronics Technology, Firat University 23100, Elazig, Turkey Abstract The utilization of renewable energy sources have been increasing day by day. Solar, wind, biomass, wave, etc. are some of the renewable energy sources. Solar, wind and hydro are more commonly used than others. Solar energy has many advantages such as abundance, renewability, continuation, and pollution free etc. Advancing technology increases people s life standards. Over time, high life standards have stopped being a luxury and entered people s homes with smart home automation systems. With the advances in Internet technologies, people now expect to be able to control their homes and workplaces over the Internet too. This study controls six different Computer/ Internet-based systems of smart home automation in a house which is included solar arrays. These systems were illumination of three different rooms, temperature control of a room, garden irrigation system control, and home security system control. With the software written, all these systems were monitored online on a website. These controls were successfully experimented on the model house with the smart home automation system designed. Keywords: Internet, PV Cells, Renewable Energy, Smart Homes. I. Introduction Human needs are changing increasingly by the day along with the advancing technology. Millions of engineers and scientists around the world are working towards achieving these advances. Recent human needs entail the facilitation of home and work life. Many previously manual operations can now be performed automatically with the automation systems developed with the help of advancing technology. Such efforts are known by the name Smart Home Automation Systems. The main goal of this concept is to enable a safer, easier, more comfortable and economic life. The energy consumption has been also increasing since these systems are increasing. So, it is utilized from renewable energy sources to meet the energy. Commonly used methods in smart home automation systems include: Phone control, Internet based control, PLC control, Computer control, Even though smart buildings were defined for the first time at the beginning of the 1980s in the USA, it only reached Turkey in At first, applications in the country were based on observation. Many applications around this time and thereafter targeted the comfort of healthy and ordinary people. Yumurtacı et al. studied smart home systems and the technologies usually used in these systems (Arkin and Paciuk, 1997). Yet another study designed a smart home automation system over GSM technology. This technology works over a phone line by processing digitized tones, and has the advantage of being used independently of time and place. Işık et al. used a different method to design another mobile phone based smart home automation system (Arslanoglu, 2009). Bekcibası and Tenruh (2011), studied control over telephone and concluded that it may be preferred owing to its ease and reasonable maintenance costs (Aydogan et al. (2011)). There are many other studies on GSM technology (Bayram, 2006, Birgül and Cansever, 2008, Bushby,1997, Can and Ipek, 2010). For Internet control, the design is made with a server card embedded into the web interface and accessing the system over the Internet (Chen et al. 2006, Cincirop, 2009, Cayıroglu and Gorgunoglu, 2010 and Gencoglu, 2008). Other researchers studied computer controlled smart home automation systems. In this method, control is achieved via the interface program on a desktop or laptop computer by accessing micro-controlled circuits over ports (Guillemin and Morel, 2001) and Isik and Altun, 2005). Smart home systems designed for Plc control operate by accessing the Plc over a written Scada (Inan and Akcayol 2009). There are also other studies on smart home automation systems, most of which used micro controller design (Kahraman and Boz, 2009, Kaklauskas and Zavadskas, 2006). Other researchers have investigated the general structure of smart home automation systems (Kaklauskas and Zavadskas, 2002, Korkmaz 2007, Kua and Lee, 2002, Naimavičienė et al., 2008, NanoLOC TRX, 2009, Neelamkavil, 2009, Ochoa and Capeluto, 2008, Omer, 2008, Park et al. 2005, Ríos-Moreno et al. 2007, Schacht, 2004, Seward and Quayle,1996, Soungho and Naruo, 2007). Staedter, (2003) studied the control parameters and benefits of smart home automation. Some researchers have investigated the renewable energy and energy systems of smart home automation systems (Sahin and Hazer, 2010, 402

417 Tosunoglu and Gokturk, 2008). Smart home automation systems may be used to control electronic tools and systems such as security (Uzun,2009, Wong et al.2008), temperature (Wong et al. 2005, Wood and Newborough, 2007), RF command (Woodman and Harle, 2009, Yabanova et al. 2010), illumination (Yumurtaci and Kecebas, 2009), TV, sound systems, irrigation systems, curtain control, and garden/ garage gate control. These systems may particularly benefit disabled and old people s lives. When people are away from their homes, they wish to control what is happening over smart home automation systems. For instance, having left for their summer vacation, a family may want to check their home security system or irrigation for the garden from afar. Similarly, a family may want to check the temperature at home on the internet and adjust it before they arrive home after an evening out. This study aims to enable the control of different home electronic systems by means of renewable energy sources away from home. The energy which is required for this purpose is provided by pv arrays. These systems controlled via a computer were the illumination of three rooms, temperature control of a room, home security system control, and the control of garden irrigation system. II. System Design The designed smart home automation system was accessed over the internet. The system basically works over a computer program. System information is entered on a database through a website. This database is kept on a rental server. The computer program reads and transmits data regularly from the database to the control circuit. There are six active outputs from the computer in the system. These outputs are used to illuminate three rooms, start and stop air conditioning for temperature in a room, start and stop the garden irrigation system, activate and stop the home security system. Optionally, the number of outputs can be increased; any device on a power line, for instance an iron, tv, or electric oven may be controlled. The designed system has 8 outputs and six of them are ready for active use. Further, the system may also be used to control the temperature of three rooms. Another feature on the system is that it allows the eight input unit to be checked over a website. The system is not affected by power outages because it is supported by pv arrays. Fig.1 shows the block diagram of the designed system. Fig. 2 shows the model house where the designed smart home automation system could be established. The designed system has five main components: pv array, internet access, computer software program, control circuit and driver circuit. Fig. 1: Block diagram of the designed system II.1. Pv Array Fig. 2: Model house A PV cell is a specialized semiconductor material with a p n junction. It converts sunlight into electricity through a basic process called photovoltaic effect. The energy generated by the cell is in direct proportion with the visible light it has been exposed to. Additionally, conversion efficiency also depends on extending the plane. The amount of the current and the voltage changes depending on the amount of sunlight shining on the cell. Then, the I V equation is: I= Il I0 ( e (qv)/(kt) -1) (1) where Il is the component of the PV cell current due to photons. II.2. Internet Access Internet access enables remote access to the system. It allows people to control their smart home automation system by using a predefined website. As the website supports mobile devices, the system may also be accessed over smart phones. Website design has two components. The first one entails the inclusion or exclusion of the electronic devices or systems to be controlled. The buttons here are used to switch on/switch off. The first component shows the current temperature in a pre-specified room on the screen. Based on this temperature, the user can control the air conditioning system for heating in winter or cooling in sun. The second component includes the activation/ deactivation of the home security system. When an alarm goes off while the security system is in use, this component notifies us and provides video over to website. Fig. 3 shows the system website. 403

418 Fig.3: Web page II.3. Computer Software Program The computer software program is a desktop application that controls the designed system. Fig.4 shows the flow chart of it. This application provides the connection between the Internet site and the control circuit. Owing to this program, the devices on the smart home automation system may be controlled over the home computer when there is no Internet. Below is a screenshot from when the application runs for the first time (Fig.5): can be checked. It has been designed to check 6 devices and includes 12 buttons for the switch on/switch off operations. The left side of the dual buttons is used for energizing, and the right side for de-energizing. All buttons have active and passive positions. These positions provide information about whether that device is working online or not. This enables us to view the device that we wish to control and change its position accordingly. In addition, the temperature values of the rooms can also see in this section if want. The system has 6 active outputs and has been designed to control up to 8 outputs with the same design. All controls made over the computer application can be viewed simultaneously on the website designed. In other words, the designed pv array based smart home automation system may be controlled both from a distance over the website and from home computer. II.4.1. Control Circuit The control circuit is the unit that enables communication between the computer and the devices to be controlled. This circuit is connected to the computer output and controls the driver circuit according to the commands received. At the same time, it sends the information from the security system and the value from the temperature sensor to the computer. The microprocessor used was Microchip s 16F877A from the PIC family. This microprocessor has enough input/output ports for the controls. It also enables the RS-232 series communication protocol that the system requires to communicate with the computer. The control circuit diagram and card are shown in Fig.6. Fig. 4: The flow chart of software program Fig. 6: Control Circuit Diagram and Card II.4.2. Electronics Driver Circuit Fig. 5: Screenshot of the computer application This application has two sections. The first one includes the settings for the connection between the application and the control circuit. The COM port to connect to and the speed for the connection are selected here. When the COM port is clicked, currently usable ports are displayed. If the port that our device has been connected to is not shown among the ports, the connection between the port and the electronic circuit needs to be checked. The second section is where the devices on the system 404 The electronics driver circuit is used to switch on and off the loaded devices depending on logic-1 (+5V) and logic-0 (GND) from the control circuit. It has a relay and a transistor. The transistor provides the energy needed for the relay by elevating according to the logic information received. With the led connected to the circuit, it may be understood whether the driver is active or passive. This circuit was designed as a model to control only one output. For every other output, the circuits can be multiplied. Below is a diagram and complete state of a driver circuit for controlling a device (Fig.7).

419 Fig. 7: Driver Circuit Diagram and Card II.5.The Input Units of System used to put this system into use. First, pv array and charge regulator are connected to system. Then, the energy needed for the system is met by them. To begin with, the desktop application was downloaded on the laptop. Following this, connection was established between the control and driver circuits and the computer. Finally, the loads we planned to control were connected to the driver circuit and the system was ready to be used. The overview of the experimental system can be seen in Fig.8. The input units of the system are those that transmit any alarm incidents to the control card when the alarm output is active. The designed circuit supports eight alarm incidents. During the experiments, 4 active alarm inputs were used. These were as follows: II.5.1. Humidity Sensor The sensor used in the system measures humidity and yields an appropriate analog output. The output signal is between 0-5V. This sensor is used with a comparator circuit. With the help of a potentiometer on the circuit, the desired humidity level is set. If the humidity level surpasses this pre-specified level, the circuit yields logic-1 information. Afterwards read with a microprocessor, this information is shown in the computer software and the Internet site. II.5.2. Fire Sensor The fire sensor in the system is one that yields analog output by perceiving ultraviolet (UV) light. It is used with a comparator circuit. With the help of a potentiometer on the circuit, the desired fire level is set. If the fire level in the environment surpasses this at any given time, we are informed over the computer and the website via the microprocessor. Fig. 8: Overview of the experimental system In order to run the smart home automation system, the desktop application was operated to observe the temperature data and alarm status from the control circuit. Then, a personal website was reached from the computer to observe the change in the alarm status connected to the security system in the circuit. The alarm indicator that is normally is green turns red if the alarm is activated. The changes in room temperature at different times can be observed from the internet page (Fig.9). II.5.3. Motion (PIR) Sensor Known as the PIR sensor, this motion sensor yields 3V digital output based on the motion level. For as long as motion exists and depending on the time set, this 3V output continues. When there is no motion, the output is 0V. As the microprocessor does not perceive 3V as logic, an additional logic inverter has been used in the circuit between output and the microprocessor. II.5.4. Flooding Sensor The sensor functions on dry contact logic. It has 2 ends. One is connected to the feeding voltage. The other is connected to the microprocessor. If the water level rises, the sensor closes its contacts and thus logic1 is fed into the microprocessor. III. Experimental Application The designed smart home automation system may be controlled over a website or the computer depending on the user s preference. A laptop was 405 Fig. 9: Changes in Temperature When all outputs are active, the 6 led lights on the driver card give off light. The small computer fan representing the air-conditioning system keeps working, and the water pump representing the irrigation system moves the water in the glass (Fig.10). When a load at an output is active, the led connected to that output is on (Fig.10). This shows whether an output is active or passive. When will outputs are active as above, all of the 6 different controls on the smart home automation system may be made: illumination of 3 rooms, temperature (air-conditioning) control in one room, garden irrigation system control, and home security system control.

420 Fig.10: Observation of Led and Loads with all Active Output As input to the system, four active inputs were tried. The changes in the website and desktop application were observed online by boiling water in a kettle and applying vapor on the humidity sensor, by bringing a lighter close to the fire sensor, by moving a hand in front of the motion sensor, and by pouring water into a glass with the flood sensor in it (Fig.11). consumption and harmful emissions to the environment. Solar energy used for industrial processes is an inevitable trend in future. Solar energy is clean and safe in many industrial sectors. This study focused on using the Internet to control the illumination system, temperature system, irrigation system, and security system for pv array based smart home. The energy required for the system has been provided by a 80W pv array. It is an advantage of the designed smart home automation system that it may be controlled both over a website and with a computer. The reasonable costs of the automation system and its ease of application are the other advantages. This system facilitates human life. Future studies may regularly record home controls and temperature information in graphics. This would enable an observation of how long a device works daily and how much energy they consume on average. The system may also be made more useful by add on such as mobile device software and RF command. V. Acknowledgement Fig.11: Observation of Input Units In addition, the temperature change that occurred when one of the room temperature sensors was heated with a lighter was observed instantly on the website and the desktop application (Fig.12). Author received his Ph.D. in Solar Energy Systems in 2005 from Firat University, Elazig, Turkey. Currently, he is assistant professor at College of Technical Sciences, Department of Electronics Technology, Firat University and is also chairman of Department of Electronics and Automation. His research interests are in the areas of solar energy, solar angles, tracking systems and PV systems. Nomenclature q k T : Electrical load ( C), : Boltzman constant ( j/k) : The cell temperature in Kelvin. References Fig.12: Observation of temperature changing Arkin H., Paciuk M., Evaluating intelligent building according to level of service system integration, Automation in Construction 6, pp , (1997). The designed smart home automation system has been also supported by renewable energy sources, such as pv array. That is why the energy consumption needed for the control and driver cards in the system is approximately 5-10W. The energy consumption of a laptop computer controlling these is approximately between 60-70W. The modem for the internet connection also consumes an average of 5W, thus necessitating a total of approximately W energy consumption. As can be seen, the energy required for the system may be provided by a 100W solar panel. In this experiment, the pv array for 80 W was been used. IV. Conclusion The use of solar energy in building is one important contribution for the reduction of fossil fuel 406 Arslanoğlu G., Rf Ev Otomasyonu, Gazi Üniversitesi Elektrik-Elektronik Mühendisliği Yüksek Lisans Tezi, (2009). Aydoğan T., Çakır A., Akça M.A., ve Polat Y.E., Cep Telefonu İle Acil Çağrı Otomasyonu, Elektrik-Elektronik ve Bilgisayar Sempozyumu FEEB, (2011). Bayram U., Akıllı Ev Otomasyonu, Çanakkale Onsekiz Mart Üniversitesi Bilgisayar Bölümü Yüksek Lisans Tezi, (2006). Bekçibaşı U., ve Tenruh M., Telefon Şebekesi Üzerinden Şifre Güvenlikli Akıllı Ev Kontrol Sistemi, 13.(Akademik Bilişim 11), 2-4, Malatya,(2011). Birgül L., ve Cansever G., Mikrokontrollör ile Akıllı

421 Ev Otomasyonu ve Bilgisayar Arayüzü, ELECO 2008, Elektrik Elektronik - Bilgisayar Mühendisliği Sempozyumu ve Fuarı (2008). Bushby ST., BACnet: a standard communication infrastructure for intelligent buildings, Automation in Construction 6, pp (1997). Can C., İpek A., Azapoğlu E., ve Türker R.O., Lazer Güvenlik Sistemi ve Gsm Haberleşme, MYO-OS Ulusal Meslek Yuksekokulları Oğrenci Sempozyumu, (2010). Chen Z, Clements-Croome D., Hong J., Li H., Xu Q., A multicriteria lifespan energy efficiency approach to intelligent building assessment, Energy and Buildings 38 (5),pp , (2006). Cincirop B., Gsm Kontrollü Akıllı Ev Otomasyonu, Sakarya Üniversitesi Elektronik ve Bilgisayar Eğitimi Yüksek Lisans Tezi, (2009). Çayıroğlu İ., ve Görgünoğlu S., Mobil Telefon ve Pic Mikrodenetleyici Kullanarak Uzaktan Esnek Kontrol Sağlanması, Int. J. Eng. Research & Development, Vol.2,No.1, (January 2010). Gençoğlu M.T., Akıllı evler, I.Mühendislik ve Teknoloji Sempozyumu, (2008). Guillemin A., Morel N., An innovative lighting controller integrated in a selfadaptive building control system, Energy and Buildings 33 (5), pp , (2001). Işık H., ve Altun A.A., "Mikrodenetleyici Kullanılarak Cep Telefonu Kontrollü Akıllı Ev Uygulaması", Selçuk Üniversitesi, Teknik-online Dergi, (2005). İnan K., ve Akçayol M. Ali, GSM Tabanlı Akıllı Ev Uygulaması, Bilişim Teknolojileri Dergisi, Cilt: 2, Sayı: 2, (2009). Kahraman F., ve Boz A.F., Web Tabanlı Otomasyon Sistemi Tasarımı ve Yapımı, 5. (IATS 09), 13-15, Karabük, Türkiye, (2009). Kaklauskas A., Zavadskas E., Ditkevičius R., An intelligent tutoring system for construction and real estate, Lecture Notes in Computer Science 4101, pp , (2006). Kaklauskas A., Zavadskas EK., Internetinė sprendimų parama (Web-based decision support), monograph, Technika, Vilnius, 2002, 216 pp. [in Lithuanian](2002). Korkmaz H., Akıllı Ev Otomasyonunun Mikrodenetleyici ile Gerçekleştirilmesi, Yıldız Teknik Üniversitesi Elektrik Mühendisliği Yüksek Lisans Tezi, (2007). Kua H.W., Lee SE., Demonstration intelligent building a methodology for the promotion of total sustainability in the built environment, Building and 407 Environment 37 (3), pp , (2002). Naimavičienė J., Kaklauskas A., and Burinskiene M., Intelligent knowledge and device based assisted residential environment: doctoral dissertation: technological sciences, civil engineering (02T), Vilnius Gediminas Technical University, Technika, Vilnius, 134 pp., (2008) NanoLOC TRX - High Performance RF Transceiver with Location Awareness and Robust Wireless Communication, Fact Sheet. In C-NA5TR1.pdf Nanotron Technologies, (2009). Neelamkavil J., Automation in the prefab andmodular construction industry, th International Symposium on Automation and Robotics in Construction (ISARC 2009) Austin, Texas, USA, June 24 27, pp , (2009). Ochoa CE., Capeluto IG., Strategic decision-making for intelligent buildings: comparative impact of passive design strategies and active features in a hot climate, Building and Environment 43 (11), pp , (2008). Omer A.M., Renewable building energy systems and passive human comfort solutions. Renewable and sustainable, Energy Reviews 12 (6),pp , (2008). Park JH., Park HS, Lee SJ., Choi J., Lee DG., Intelligent multimedia service system based on context awareness in smart home, Lecture Notes in Computer Science 3681, pp , (2005). Ríos-Moreno GJ., Trejo-Perea M., Castañeda-Miranda R., Hernández-Guzmán VM., Herrera-Ruiz G., Modelling temperature in intelligent buildings by means of autoregressive models, Automation in Construction 16 (5),pp , (2007). Schacht K., New developments: smart materials (cutting edge, shape memory properties, column), Appliance Manufacturer 52 (1), pp14 16, (2004). Seward D.W., Quayle S.D., System architectures and safety for mobile construction robots, 13th Int. Symp. on Robotics in Construction, Tokyo, Japan, pp , (June 1996). Soungho C., Naruo K., Application of location information by stereo camera images to project progress monitoring, ISARC 2007, September 2007, 24th International Symposium on Automation and Robotics in Construction, Indian Institute of Technology Madras, Kochi, India, pp , (September 2007). Staedter T., The smart antenna (Visualize, brief article), Technology Review Cambridge 106 (1), pp81 82, (2003).

422 Şahin H., ve Hazer O., Akıllı Konut Teknolojileri, Gazi Üniversitesi Endüstriyel Sanatlar Eğitim Fakültesi Dergisi, Sayı: 269, (Mayıs 2010). Tosunoğlu O., ve Göktürk M., Akıllı Ev Sistemlerinde Ethernet Üzerinden Cihaz Kontrolü Ve İzlenmesi, ELECO 2008, Elektrik- Elektronik- Bilgisayar Mühendisliği Sempozyumu ve Fuarı (2008). Uzun F.Cem., Akıllı Binalarda Otomasyon, Makine Mühendisliği Yüksek Lisans Tezi, (2009). Wong J, Li H., Lai J., Evaluating the system intelligence of the intelligent building systems: Part 1: development of key intelligent indicators and conceptual analytical framework, Automation in Construction 17 (3),pp , (2008). Wong JKW, Li H, Wang SW, Intelligent building research: a review, Automation in Construction 14 (1),pp , (2005). Wood G, Newborough M., Energy-use information transfer for intelligent homes: enabling energy conservation with central and local displays, Energy and Buildings 39, pp ,(2007). Woodman O., Harle R., RF-based initialisation for inertial pedestrian tracking, in: H. Tokuda, et al., (Eds.), Pervasive Computing, 7th International Conference, Pervasive 2009, Nara, Japan, May 11 14, 2009, Proceedings. Lecture Notes in Computer Science, 5538, Springer, pp , (2009). Yabanova İ., Taşkın S., Ekiz H., ve Çimen H., Denetleyici Alan Ağı Üzerinden Mekatronik Bir Sistemin Kontrolü, Makine Teknolojileri Elektronik Dergisi, Cilt: 7, No: 2, (2010). Yumurtacı M. ve Keçebaş A., 5.Uluslararasleri Teknolojiler Sempozyumu (IATS 09), Mayıs Karabük, Türkiye, (2009). 408

423 Energy and Exergy Analyses of a Solar Energy Driven Multigeneration System for Green Buildings Yunus Emre Yuksel 1*, Murat Ozturk 2, Ibrahim Dincer 3 1 Afyon Kocatepe University, Education Faculty, Department of Elementary Science Education, ANS Campus, Afyon, 03200, Turkey 2 Suleyman Demirel University, Faculty of Technology, Department of Mechatronics Engineering, Cunur, West Campus, Isparta, Turkey 3 University of Ontario, Faculty of Engineering and Applied Science, Institute of Technology (UOIT), 2000 Simcoe Street North, Oshawa, Ont., Canada L1H 7K4 * yeyuksel@aku.edu.tr Abstract In this paper, an integrated solar energy system for green buildings is proposed and investigated as it consists of the parabolic trough collector (PTC), two-stage organic Rankine cycle (DS-ORC), quadruple effect absorption cooling system (QEACS), storage system, hydrogen production and utilization system. This multigeneration system is designed to produce five useful outputs, such as electricity, heating, cooling, hot water and hydrogen, to meet the demands of the green buildings. A thermodynamic assessment of integrated system is carried out through energy and exergy analyses. Also, the parametric studies are conducted to investigate the influences of varying system design parameters on the integrated system performance. The thermodynamic analysis results illustrate that the exergy efficiency of integrated system increases from 53.75% to 56.18% with increasing solar intensity from 400 W/m 2 to 1000W/m 2. Keywords: Energy, exergy, efficiency, solar energy, green buildings, multigeneration I. Introduction The world s demand for energy has been increasing mainly due to the population growth and improved living standards. Oil, natural gas and coal are recognized as primary energy sources and meet the largest portion of the energy demand of world. In addition, primary energy combustion emissions cause major environmental damages, such as the global warming or global climate change, the stratospheric ozone depletion potential, the acid rain, etc. Such global damages require potential solutions, namely renewable energy technologies. Due to increased efficiency and reduced environmental impact and cost, the multigeneration systems driven by renewable energy sources have become increasingly important. Several studies have defined and investigated thermodynamically of the integrated systems for multigeneration aims. These integrated system studies primarily focus on the renewable energy based multigeneration process. Also, to develop the sustainability of societies, the implementation of a renewable energy based integrated system for the green buildings is very necessary. The green buildings are recognized as buildings that offset all of its energy consumption by using alternative energy technologies. The green energy approach for buildings should be achieved with two main indicators, such as (i) decreasing building s energy demands and (ii) producing power, heating, cooling and hot water using by the renewable energy technologies. In the first indicators, the passive actions play an important role in addressing green buildings. The second indicator affects energy production, storage and transmission requirements. Many researchers have investigated various applications of integrated systems for the building applications. Bakos et al. (2003) have presented a technoeconomic assessment of a building integrated PV system for energy savings. They have evaluated the technical and economic factors by using a computerized renewable energy technologies (RETs) assessment tool. Zhai et al. (2007) have investigated a solar integrated energy system which produces heating, cooling, natural ventilation and hot water. This integrated solar energy system supplies energy for 460 m 2 building area. After one-year operation, they have concluded that solar system contributed 70% of total energy need for the building. Besides green buildings, zero energy buildings which are producing the certain amount of energy that they need play key role in sustainability. Hamdy et al. (2013) have introduced efficient, transparent, and time-saving simulation-based optimization method for cost-optimal and nearly-zero-energy building solutions. According to their study results, minimizing the cost of heating applications play key role for achieving cost-optimal systems. Rosato et al. (2014) have compared a building integrated micro cogeneration system with the building using conventional system composed of a natural gas fired boiler and grid connection for electricity. Their analysis results demonstrated that in comparison to 409

424 the conventional building, the proposed system achieved the reduction of energy consumption. The main objective of this paper is to investigate the effects of varying system design parameters on the efficiencies of the proposed integrated system. II. System description The proposed integrated system for buildings is evaluated and presented in this section. Each sub-system investigated here is designed to meet the electricity, heat, cold and hot water needs of buildings. Power needs in the buildings are designed by taking into account the power requirements of systems or heat losses of buildings. Today, power in buildings is commented for sustainable and constant conditions in system design. In addition to that, continuously to meet power and other needs, the hydrogen production, storage and utilization sub-system and heat working fluid storage system are designed in the integrated system. Also, the process flow diagram for the integrated system includes the parabolic trough collector (PTC), double state organic Rankine cycle (DS-ORC), quadruple effect absorption cooling system (QEACS), storage system, hydrogen production and utilization system. The technology of multigeneration systems is improved to accomplish the challenges associated with producing reliable useful outputs using by the renewable energy technologies which characterized by its primary power such as solar radiation. In this respect, a parabolic trough collector sub-system is used to collect solar radiation and heat the working fluid to a high temperature which is utilized in the DS-ORC sub-system for electricity generation. A portion of the thermal energy as well as electrical energy generated by the DS-ORC are used to drive the PEM electrolyzer sub-system to produce hydrogen. The producing pure hydrogen is stored in a hydrogen tank for later usage. The PEM fuel cell is used to provide power when the solar energy is not available (e.g., during the night). Also, in order to maintain continuous operating a thermal energy is stored in the heat storage tank which utilize it for later use during cloudy days or night time. Li-Br absorption chiller cycle is selected, since this cooling system is one the most promising and applicable cooling systems. III. Energy and exergy analyses The present thermodynamic assessment study of the integrated system for buildings can be conducted by taking the mass, energy and exergy balance equations of different components of the solar based integrated system. The mass rate balance for all processes can be written as follows: m in = m out (1) The general energy rate balance for all processes can be written as follows: 410 m inh in + Q = m out h out + W (2) where Q and W are the heat transfer rate and work transfer rate, respectively. The general exergy rate balance equation for all processes can be given as follows; E x Q + in m inex in = E x W + m outex out out (3) Here, E x D is the exergy destruction rate, E x Q is the heat transfer exergy rate crossing the control volume boundary, and can be written as follows: E x Q = (1 T o T)Q (4) And E x W is the shaft work exergy rate, and can be given as follows: E x W = W (5) In this study, the kinetic and potential exergy are considered negligible. Based on this assumption, the specific exergy (ex) is written as: ex = ex ph + ex ch (6) where ex ph and ex ch are the physical and chemical exergy, respectively. ex ph = (h h o ) T o (s s o ) (7) ch n = [ x i ex ch n i=1 i + RT 0 i=1 x i lnx i ] (8) ex mix The flow exergy rate is defined as follows: E x = m ex (9) The energy efficiency equation for any processes can be generally written as follows: η = Useful output energy Energy input (10) The energy efficiency of the sub-systems and whole system can be written as follows: η DS ORC = W ORC+Q Heating+Q Hotwater η storage = η hydrogen = Q HST Q HEX I W PEM Q HEX II Q WPH+W PEM el η absorption = Q Cooling+Q Heating Q HEX III+W P V η system = W ORC+W PEM+Q Cooling+Q Heating+Q Hotwater Q PTC (11) (12) (13) (14) (15) The energetic COP of the QEACS can be calculated as follows:

425 COP en = Q Eva+Q Con W P V+Q HEX III (16) The exergy efficiency equation for any processes can be generally written as follows: ψ = Useful output exergy Exergy input (17) The exergy efficiency of the sub-systems and whole system can be written as follows: ψ DS ORC = W Q Q ORC+E x Heating +E x Hotwater Q (18) ψ storage = ψ hydrogen = E Q x HST E x HEX II Q (19) E x HEX I W PEM Q E x WPH +W PEM el Q Q +E x Heating Q E x HEX III +W P V Q Q Q +E x Heating +E x Hotwater ψ absorption = E x Cooling (20) (21) ψ system = W ORC+W PEM+E x Cooling Q (22) E x PTC The exergetic COP of the QEACS can be calculated as follows: COP ex = E Q Q x Eva+E x Con W P V+E x HEX III Q (23) To see the amount of energy losses exergy destruction rates of the integrated system components, the performance of each parts of process can be calculated by determining properly its input-output and available standard thermochemical definitions. The exergy efficiency equations of the integrated system components are given in Table 1. III.1. Solar concentrating collector The heat available on the parabolic trough collector surface can be calculated as given below: Q = F R A A [ρ R α A I t C U L (T A T o ) εσ(t A 4 T o 4 )] (24) where F R is the heat removal factor, A A is the area of absorber, ρ R is the reflectivity of reflector, α A is the absorptivity of absorber, I t is the global solar radiation, C is the concentration ratio of parabolic collector, U L is the heat transfer coefficient from absorber to surroundings, T A is the absorber/receiver temperature, ε is the emissivity of absorber and σ is the Stefan-Boltzmann constant. The useful heat available at the receiver surface of concentrating collector can be calculated as follows: Q u = Q (1 T o T ) (25) The energetic efficiency of the PTC should be given as the ratio of heat energy available at the receiver to the incident solar radiation on the collector, and also can be calculated as follows: Q η PTC = (26) I t A C where A C is the is the area of the solar collector. The exergetic efficiency of PTC should be given as the ratio of produced useful heat energy at the receiver to the incident solar exergy (E x solar ) on the collector, and also can be calculated as follows: ψ PTC = Q u E x solar A C (27) E x solar = I t [ (T o ) 4 4 ] (28) T s 3 T s where T s is the outer surface temperature of sun (Petela, 2005), where T s is taken as 5777K. III.2. PEM electrolyzer T o In this paper, the PEM electrolyzer, which is given on the left side of Fig. 1, is used to produce hydrogen. To drive the electrochemical reaction in the electrolyzer, the electricity and heat come from the system. The water at ambient temperature enters the water pre-heating subsystem that heats it to the electrolyzer working temperature before it goes to the PEM electrolyzer. The produced hydrogen at the cathode side of electrolyzer is cooled to the ambient condition, and should be stored in the steel hydrogen storage tank at a pressure of 150 to 200 bar and a reference temperature of approximately 25 C for later use. Also, the produced oxygen at the anode site of electrolyzer is separated from the water/oxygen mixture, and then dissipates heat to the surroundings. The unused water is sent to the water pre-heater for the next hydrogen generation cycle. The total energy needed for PEM electrolyzer can be written as follows: ΔH = ΔG + TΔS (29) where ΔG is the Gibb s free energy, TΔS represents the heat energy demand for PEM electrolyzer, ΔH and ΔS are the enthalpy and entropy change, respectively, and T is the reaction temperature. The mass flow rate of hydrogen gas can be obtained as follows (Ahmadi, Dincer, & Rosen, 2012): N H 2,out = J 2F = N H 2 O,reacted (30) where J is the current density and F is the Faraday constant. The electrolyzer voltage can be written as; V = V o + V act,a + V act,c + V ohm (31) where V o is the reversible potential, and can be determined using by the Nernst equation as given below: V o = x10 4 (T Pe 298) (32) where TPe is the PEM electrolyzer temperature. In Eq. (31), V act,a is the anode activation over-potential, V act,c is the cathode activation over-potential, and V ohm is the electrolyte ohmic over-potential. 411

426 Fig. 1. The schematic diagram of the integrated process based on the solar energy Table 1. Exergy efficiencies of the integrated system parts System parts Exergy efficiencies Trough collector ψ PTC = (m 1ex 1 m 8ex 8 ) Q solar(1 T o T PTC ) HEX-I ψ HEX I = (m 15ex 15 m 14ex 14 ) (m 2ex 2 m 3ex 3 ) Valve-I ψ Val I = (m 2ex 2 + m 4ex 4 ) m 1ex 1 Hot storage tank ψ HST = (m 9ex 9 Q L,HST(1 T o T HST )) m 15ex 15 Cold storage tank ψ CST = (m 12ex 12 Q L,CST(1 T o T CST )) m 13ex 13 Pump-I ψ Pum I = (m 10ex 10 m 9ex 9 ) W Pum I Expander-I ψ Exp I = W Exp I (m 16ex 16 m 17ex 17 ) Separator ψ Sep = (m 18ex 18 + m 21ex 21 ) m 17ex 17 Turbine ψ Tur = W Tur (m 18ex 18 m 19ex 19 ) Resorber-I ψ Res I = (m 27ex 27 m 26ex 26 ) (m 19ex 19 m 20ex 20 ) Mixing chamber ψ MC = m 23ex 23 (m 20ex 20 + m 22ex 22 ) Water preheater ψ WPre = (m 34ex 34 m 33ex 33 ) (m 31ex 31 m 32ex 32 ) PEM electrolyzer ψ PEM elec = (m 35ex 35 + m 36ex 36 + m 37ex 37 ) (m 34ex 34 + W PEM elec) Hydrogen ψ HCS = m 38ex 38 (m 37ex 37 + W HCS) system PEM fuel cell ψ PEMFC = W PEMFC m 38ex 38 VHTG ψ VHTG = (m 53ex 53 + m 54ex 54 m 52ex 52 ) (m 30ex 30 m 31ex 31 ) Expansion ψ ExV I = m 79ex 79 m 78ex 78 valve-i VHT-HEX ψ VHT HEX = (m 72ex 72 m 54ex 54 ) (m 51ex 51 m 52ex 52 ) Con-HEX ψ Con HEX = (m 65ex 65 m 63ex 63 ) (m 41ex 41 m 68ex 68 ) Condenser ψ Con = Q Con(1 T o T Con ) (m 64ex 64 + m 65ex 65 m 69ex 69 ) Evaporator ψ Eva = Q Eva(1 T o T Eva ) (m 71ex 71 m 70ex 70 ) Absorber ψ Ab = Q Ab(1 T o T Ab ) (m 71ex 71 + m 79ex 79 m 39ex 39 ) 412

427 The reversible potential is concerning with the difference in free energy between reactants and products. The local ionic conductivity of PEM can be given as follows (Ni, Leung, & Leung, 2008) : σ PEM [λ(x)] = [0.5139λ(x) 0.326]exp [1268 ( T )](33) where λ(x) is the water content at a location x in PEM, and can be calculated as follows: λ(x) = λ a λ c D x + λ c (34) where D is the membrane thickness, λ a and λ c are the water contents at the anode mebrane interface and the cathode mebrane interface, respectively. The overall ohmic resistance of PEM can be given as follows (Ni, Leung, & Leung, 2008): R PEM = D dx 0 σ PEM [λ(x)] (35) The ohmic overpotantial of PEM can be given using by the Ohm s law as; V ohm,pem = JR PEM (36) The activation overpotential (V act ) of PEM electrolyzer can be expressed based on the deflection of current from its equlibrium and electron transfer reaction. V act,i = RT F sinh 1 ( J 2J o,i ), i = a, c (37) where J o is the exchange current dencity, which is a significant indicator in determining the activation overpotential, and subscripts a and c are the anode and cathode, respectively. The exchange current density of electrolyzer can be given as follows (Ni, Leung, & Leung, 2008) calculated as: J 0,i = J ref i exp ( E act,i ), i = a, c (38) RT where J ref i is the pre-exponential factor, E act,i is the activation energy for the anode and cathode. III.3. PEM fuel cell The generated voltage for PEM fuel cell can be written as follows: V(I) = V r V ir (39) where V r is the PEM fuel cell reversible voltage, and can be given as V r = x10 3 (T Cell ) x10 5 T Cell ln [( C H2 ) ( C 1 O2 ) 2 ] (40) Here, V ir is the irreversible cell voltage, and can be written as follows: 413 V ir = η act + η ohmic + η con (41) where η act, η ohmic and η con are the activation, ohmic and concentration over potentials, respectively η act can be evaluated as follows: a η act = η act c + η act (42) a where η act is the activation over potentials for the c anode catalyst layer, and η act is the activation over potentials for the cathode catalyst layer in fuel cell stack. The ohmic over potential can be evaluated as; η ohmic = η a bp c + η bp + η a e + η c e + η m (43) a where η bp is the ohmic losses from the anode layer c of the fuel cell stack, η bp is the ohmic losses from a the cathode layer of the fuel cell stack, η e is the c ohmic losses from the anode backing side, η e is the ohmic losses from the cathode backing side, and η m is the over potential associated with the membrane of PEM fuel cell. η con is related with the critical mass transfer restriction at higher current density, and can be calculated as follows: a η con = η con c + η con (44) a where η con is the concentration over potentials of c the anode electrode, and η con is the concentration over potentials of the cathode electrode. The generated power from the one fuel cell stack can be calculated as follows: W cell = V(I)xIxA cell (45) where I is the current density and Acell is the PEM fuel cell area. The produced power from the PEM fuel cell sub-system can be calculated as follows: W PEM_fc = n fc xw cell (46) Here, n fc is the number of the cells in the fuel cell sub-system. III.4. Quadruple effect absorption cooling sub-system In this paper, the QEACS is used for cooling and heating applications of buildings, and the ammonia-water mixture is chosen as a working fluid in the cooling sub-system. The mass balance equations for the ammonia-water working fluid and oil can be given as follows: m 52x 52 = m 53x 53 + m 54x 54 (47) m 30 = m 31 = m 32 (48) The energy and exergy balance equations of the very high temperature generator (VHTG) can be written as follows:

428 m 52h 52 + m 30h 30 = m 31h 31 + m 53h 53 + m 54h 54 (49) m 30ex 30 + m 52ex 52 = m 31ex 31 + m 53ex 53 + m 54ex 54 + E x D,VHTG (50) The energy and exergy balance equations for the very high temperature-hex (VHT-HEX) are written as follows: m 51h 51 + m 54h 54 = m 52h 52 + m 72h 72 (51) m 51ex 51 + m 54ex 54 = m 52ex 52 + m 72ex 72 + E x D,VHT HEX (52) The energy and exergy balance equations for the condenser can be given as follows; m 69h 69 + Q con = m 64h 64 + m 65h 65 (53) m 69ex 69 + Q Con(1 T o T Con ) = m 64ex 64 + m 65ex 65 + E x D,Con (54) The following energy and exergy balance equations are used to evaluate the heat absorbed to the evaporator. m 70h 70 + Q eva = m 71h 71 (55) m 70ex 70 + Q Eva (1 T o T Eva ) = m 71ex 71 + E x D,Eva (56) The given below energy and exergy balance equations can be used to evaluate the heat rejected from the absorber. m 71h 71 + m 79h 79 = m 39h 39 + Q ab (57) m 71ex 71 + m 79ex 79 = m 39ex 39 + Q Ab(1 T o T Ab ) + E x D,Ab (58) III.5. Double-stage ORC The heated working fluids enter the expander-i at point 16 to generate power electricity. The energy and exergy balance equation of expander-i can be expressed as follows: m 16h 16 = m 17h 17 + W Exp I (59) Under steady-state and steady-flow conditions, the energy and exergy balance equations of the resorber-i can be written as follows: m 19h 19 + m 26h 26 = m 20h 20 + m 27h 27 (65) m 19ex 19 + m 26ex 26 = m 20ex 20 + m 27ex 27 + E x D,Res I (66) Under the steady-state and flow conditions, both energy and exergy balance equations of the mixing chamber can be given as follows: m 20h 20 + m 22h 22 = m 23h 23 (67) m 20ex 20 + m 22ex 22 = m 23ex 23 + E x D,MC (68) IV. Results and discussion In this study, energy and exergy analyses and parametric studies of the multigeneration for buildings is given detailed in this subsection. Therminol-59 is chosen as the working fluid for the concentrating collector and ammonia-water (NH3-H2O) mixture is chosen as the working fluid for the DS-ORC and QEACS. The reference temperature and pressure are given as 25 C and kpa, respectively. Also, the exergy destruction rate, exergy destruction ratios and exergy efficiency of the integrated system components are given in Table 2. The results illustrated that, the highest three exergy destruction rates are occurred in the PTC, PEM fuel cell and DS-ORC turbine with the values of 1955 kw, kw, and kw, respectively. In addition, exergy destruction ratios of these three system parts are calculated as 18.97%, 5.09% and 4.05%, respectively. COP en COP en COP ex COP ex m 16ex 16 = m 17ex 17 + W Exp I + E x D,Exp I (60) The energy and exergy balance equations of the separator unit can be written as follows: m 17h 17 = m 18h 18 + m 21h 21 (61) m 17ex 17 = m 18ex 18 + m 21ex 21 + E x D,Sep (62) The energy and exergy balance equations are written for the DS-ORC turbine as follows under steady-state and flow conditions: m 18h 18 = m 19h 19 + W Tur (63) m 18ex 18 = m 19ex 19 + W Tur + E x D,Tur (64) T 0 ( o C) Fig. 2. Effect of ambient temperature on COPen and COPex of QEAC subsystem As seen from Fig. 2, the reference environment temperature increases from 5 0 C to 40 0 C, while energetic coefficient performance remains the same, exergetic coefficient of performance increases from about 2 to nearly 5. The reason of energetic COP remaining the same is that energy analysis is independent from the reference temperature. However exergetic COP increases because the definition of exergy tells that exergy is related to the environment.

429 Table 2. Thermodynamic analysis results for the integrated system components System parts Exergy destruction (kw) Exergy destruction ratio (%) Exergy efficiency (%) Parabolic trough collector HEX-I Valve-I Valve-II HST CST Pump-I Expander-I Separator Turbine Resorber-I Mixing chamber Water preheater PEM electrolyzer Hydrogen compression/storage PEM fuel cell Very high temperature generator High temperature generator Medium temperature generator Low temperature generator Expansion valve-i Very high temperature HEX High temperature HEX Medium temperature HEX Low temperature HEX Condenser HEX Condenser Evaporator Absorber Another parametric study is performed to find out how PEM fuel cell temperature affects the power output and exergy efficiency. Fig. 3 shows that there is a direct proportion between PEM fuel cell temperature and exergy efficiency of PEM. Correspondingly, an increase in PEM fuel cell temperature causes the increase in the amount of work done by PEM. W PEM-FC (kw) T PEM-FC ( o C) W PEM-FC y PEM-FC Fig. 3. Effect of PEM fuel cell temperature on power outputs and exergy efficiency If solar energy is used in any conventional or multigenerational energy production system, solar intensity plays a significant role in producing power. As seen from Fig. 4, as solar intensity increases from 400 W/m 2 to 1000 W/m 2, the produced work increases nearly from 350 kw to nearly 470 kw. The reason of this increase is that higher solar intensity means much more heat energy available for the system. ypem-fc Q out (kw) Solar intensity (W/m 2 ) Q hot water Q heating Q cooling W total power Fig. 4. Effect of solar intensity on hot water, heating, cooling and power production rate Another important factor influencing the total power produced by the system is the ambient temperature as shown in Fig. 5. According to this figure, while ambient temperature increases, the amount of produced heating, hot water and total power increases, although cooling performance of the system decreases. The effect of DS-ORC turbine inlet pressure on exergy efficiencies of the power, the combined heat and power (CHP), the tri-generation and the multigeneration is illustrated in Fig. 6. Also, this figure illustrates that an increase in the DS-ORC pressure increases the exergy efficiencies for these generations. The energy balance equation for a W out (kw) 415

430 control volume of HEX-II illustrates that when the energy input from the PTC is constant, the decrease in DS-ORC turbine inlet enthalpy increases the turbine mass flow rate. Therefore, the power production rate from the DS-ORC system, and heating load and hot water production rate from the resorber-i and II, respectively, increase with increasing turbine inlet pressure. Q out (kw) Fig. 5. Effect of ambient temperature on hot water, heating, cooling and power production rate Exergy efficiency T 0 ( o C) DS-ORC inlet pressure (kpa) Fig. 6. The effect of DS-ORC inlet pressure to the exergy efficiency of different energy production options V. Conclusions Single generation Cogeneration Trigeneration Multigeneration Q hot water Q heating Q cooling W total power In this paper, the thermodynamic assessment of the solar energy based integrated system for buildings and their component is performed. Exergy destruction rates and exergy efficiencies of entire system and its components are presented. Some parametric analyses are performed in order to see how some variables affect the system outputs. The parametric variables investigated in this paper are ambient temperature, PEM fuel cell temperature, solar intensity, and DS-ORC inlet pressure. Some concluding remarks are extracted as given below; The parameters, namely intensity of solar radiation, reference temperature have great influence on the performance of integrated system. Multigeneration energy production systems have much higher exergy than the other production options such as trigeneration, cogeneration or single generation W out (kw) 416 The energy and exergy analyses of investigated integrated system for buildings can supply a beneficial base to design a more performance and sustainable integrated system to meet energy needs of buildings from renewable energy based processes. References Ahmadi, P., Dincer, I., & Rosen, M. (2012). Energy and exergy analyses of hydrogen production via solar-boosted ocean thermal energy conversion and PEM electrolysis. International Journal of Hydrogen Energy, 38(4), Ahmadi, P., Dincer, I., & Rosen, M. (2012). Energy and exergy analyses of hydrogen production via solar-boosted ocean thermal energy conversion and PEM electrolysis. International Journal of Hydrogen Energy, 38(4), Bakos, G. C., Soursosb, M., & Tsagas, N. F. (2003). Technoeconomic assessment of a building-integrated PV system for electrical energy saving in residential sector. Energy and Buildings, 35, Hamdy, M., Hasan, A., & Siren, K. (2013). A multi-stage optimization method for cost-optimal and nearly-zero-energy building solutions in line with the EPBD-recast Energy and Buildings, 56, J.J.Baschuk, & Li, X. (2003). Mathematical model of a PEM fuel cell incorporating CO poisoning and O2 (air) bleeding. International Journal of Global Energy Issues, 20(3), Ni, M., Leung, M., & D.Y.Leung. (2008). Energy and exergy analysis of hydrogen production by a proton exchange membrane (PEM) electrolyzer plant. Energy conversion and management, 49, Ni, M., Leung, M., & Leung, D. Y. (2008). Energy and exergy analysis of hydrogen production by a proton exchange membrane (PEM) electrolyzer plant. Energy conversion and management, 49, Petela, R. (2005). Exergy analysis of the solar cylindrical-parabolic cooker. Solar Energy, 79, Rosato, A., Sibilio, S., & Scorpio, M. (2014). Dynamic performance assessment of a residential building-integrated cogeneration system under different boundary conditions. Part I:Energy analysis. Energy Conversion and Management, 79, Zhai, X. Q., Wang, R. Z., Dai, Y. J., Wu, J. Y., Xu, Y. X., & Ma, Q. (2007). Solar integrated energy system for a green building. Energy and Buildings, 39,

431 Achieving Sustainable Buildings via Energy Efficiency Retrofit: A Case Study of an Industrial Building Abstract Muhsin Kilic*, Ayse Fidan Altun Uludag University, Engineering Faculty, Mechanical Engineering Dept., Gorukle Campus, Bursa, 16059, Turkey mkillic@uludag.edu.tr Among the buildings in any country, existing buildings are more energy inefficient than the new ones. As a result, building energy efficiency retrofit plays a crucial role in order to achieve sustainable building targets. Retrofitting, not only create possibilities to reduce energy consumption and greenhouse gas emissions of all buildings, but also improves occupants health and thermal comfort. The focus of this paper is to asses an industrial building which is energy efficiently retrofitted. Energy consumption of a cement factory during its operation, before and after the energy retrofit is examined. Results from the analysis prove that energy efficiency retrofit, helped to decrease in energy consumption of the building and eventually reduced operational expenditures and greenhouse gas emissions in a very short payback period. Reduction in energy consumption in the building, leaded to an energy cost savings of $ per year. Carbon dioxide emissions decreased by % 87. Key words: Green buildings, industrial facilities, sustainable design, energy retrofitting I. Introduction Global warming, environmental pollution, extinction of fossil fuels and the price of energy rises are the biggest problems of the humanity. Construction industry consumes of the 40% of global resources, 12% of potable water reserves, 55% of wood products, 45-65% produced waste, 40% raw materials, and the emission of 48% of hazardous greenhouse emissions, which results to environmental pollution and global warming Suzer (2014). Fig.1: Sectoral energy consumption in Turkey (Bulut H., 2009) Over the past years, as being the 17 th largest economy in the world, Turkey has experienced considerable growth and energy use has grown at an annual rate of about 4.5% from 2013 to 2015 (Ferdos, 2015). Turkey petroleum and other liquids consumption and production amount is shown in Figure 2 and sectoral energy consumption can be seen in Figure 1. As a result, government seeks to reduce the country s dependence on imported fuels, which is almost the whole amount. For this reason, energy efficiency is a critical issue for Turkey. Energy retrofit of industrial and residential buildings is an effective solution to decrease Turkey s high and rising energy consumption demands. The aim of this paper is energy efficiency retrofit of an existing industrial building with optimal HVAC system solutions which can meet the demands of the users. And eventually, prove that sustainably renovated buildings can generate numerous benefits in terms of energy and CO2 reduction, cost savings and healthier environments. 417 Fig.2: Turkey petroleum and other liquids consumption/production ( II. Green buildings Green building is a method which is environmentally responsible, supports resource efficiency throughout the buildings life cycle: during design, construction, maintenance, renovation and demolition Azouz and Kim (2015). In other words, a green building is one, whose construction and lifetime of operation assure the healthiest environment while minimizing resource utilization and greenhouse gas emissions ( There are numerous sustainable assessment schemes all around the world which aim to rate a building s environmental performance. Some of the widely used green building assessment schemes are GBTool (Canada),

432 CASBEE (Japan), BREEAM (Britain), BEAM (Hong Kong), Green Star (Australia), LEED (USA), DNGB (Germany), and TSE Green & Secure Buildings (Turkey). Generally, they aim to assess the impact of the building on its environment. Improving energy efficiency of existing buildings plays a vital role in order to achieve the goals of sustainable buildings. Energy efficiency retrofit, reduce greenhouse gas emissions, utility bills and maintenance costs and create jobs and career opportunities Pengpeng (2012). There are a number of approaches of building retrofitting such as installation of renewable energy systems, enhancing building envelope and upgrading HVAC systems Zhou et al. (2015). Heating, ventilation and air conditioning (HVAC) system is the highest energy consuming component in a building, therefore improving HVAC system contributes in greater energy savings within the building Ruparathna et al. (2015). In this study, significant attention has been paid to improve the efficiency of building s HVAC system. In order to enhance the HVAC system, high efficiency heat pumps, solar thermal system and waste heat recovery system are used, coal boiler is replaced with high efficiency condensing boilers. III. Case Study Bursa Cement Factory was selected in this study to serve the research purpose. It was established as an incorporated company on 14 July 1966 in Kestel/Bursa. Today, tons clinker production capacity and tons cement grinding capacity and tons cement stock capacity are available. The facility has 6 main areas which are; upper housing & warehouse, lower housing, manager housing, administrative building, production facility and social facility& dining hall. The coal-fired central heating system was used for the building. Since heating, ventilation and air conditioning (HVAC) system is the highest energy consuming component in a building, the methodology will focus on HVAC retrofitting of the building. As the first step of building energy retrofit projects, it is important to diagnose and analyze building energy consumption Ruparathna et al. (2015). As a result, energy requirement calculations is made for every single part of the building by using climate data of Bursa which is shown in Figure 3. According to the results, optimal HVAC upgrading solutions is given. III. 1. Lower Housing & Warehouse Lower housing has 1050 m 2 building area and consists of 8 flats. Warehouse has 300 m 2 building area. Heating was done by a central boiler located in mechanical room. Losses arose from the distance between the boiler room and the lower housing, and the deformation of the thermal insulation caused severe energy inefficiency. Fig.3: Bursa city climate data ( The heating energy needs in the lower housing calculated as 119 kwh. The heating energy needs in the warehouse calculated as 58 kwh. The water heating energy needs calculated as 77 kwh Recommended Action Lower housing and warehouse system selection; kw natural gas-fired, condensing boiler system kw, 55 C air source heat pump system 3. Thermal solar system liters double serpentine water heater III. 2. Upper Housing Upper housing has 2400 m 2 building area and consists of 16 flats. Heating energy was generated by the central boiler located in mechanical room. The heating energy needs in the upper housing calculated as 209 kwh. The water heating energy needs calculated as 154 kwh Recommended Action kw natural gas-fired, condensing boiler system kw, 55 C air source heat pump system 3. Thermal solar system liters double serpentine water heater III. 3. Manager Housing Manager housing has 250 m 2 building area. Heating energy was generated by the central boiler located in mechanical room. The heating energy needs in the upper housing calculated as 21 kwh. The water heating energy needs calculated as 10 kwh 418

433 Recommended Action kw natural gas-fired, condensing boiler system liters single serpentine water heater retrofit implementations. Energy efficiency retrofit has a significant impact on the reduction of energy consumption and polluting emissions. III. 4. Administrative Building Administrative building has 2000 m 2 building area. Heating energy was generated by the central boiler located in mechanical room The heating energy needs in the upper housing calculated as 218 kwh. Recommended Action kw natural gas-fired, condensing boiler system During the transitional seasons (spring, autumn), existing VRF system of the building will provide required energy for the heating. Fig.4: CO2 emissions comparison (kg/m 2 ) III. 5. Production facility Production facility has 2175 m 2 building area. Heating energy was generated by a waste-heat recovery system which uses compressor s waste heat. The heating energy needs in the production facility calculated as 159 kwh. Recommended Action kw natural gas-fired, condensing boiler system 2.Waste-Heat recovery system which uses compressor s heat (70 C, 10 m 3 /h) Fig 5: Tonne of oil equivalent (TOE) comparison III. 6. Social facility and dining hall Social facility and dining hall heating energy was generated by the central boiler located in the boiler room. The heating energy needs in the social facility calculated as 146 kwh, and in the dining hall calculated as 77 kwh. Recommended Action kw natural gas-fired, condensing boiler system kw 30 C/65 C water sourced heat pump (which uses the WHR cooling tower s waste heat) 3. Heat recovery system (70 C 10 m 3 /h) which uses the waste heat from compressor liters water heater ( lt+2x5000 lt) IV. Results and discussion IV.1. Environmental- benefit analysis As it can be seen from the Fig.4 and Fig.5, replacement of HVAC system is a beneficial action in terms of energy savings and environmental protection. CO2 emissions decreased 87% after the 419 IV.1. Cost-benefit analysis The economic problem of allocating limited resources to various needs often requires making cost- benefit analysis where costs and benefits over the lifetime of the project are evaluated and investments with positive net benefits are considered to be acceptable Bhattacharyya (2011). In this section, costs related to the investment is estimated and appropriateness of the investment is determined. In order to finance a project, organizations often borrow money from banks or other leading organizations and projects financed in this way cost more than similar projects financed organization s own funds because of the interest charges Bureau of Energy Efficiency (2007). It is estimated that, Cement factory will finance its investment by borrowing money from a bank. Monthly repayment value; C = A.k [1 1 (k+1)n ] (1)

434 where; C = Monthly repayment n= Repayment period k= Monthly interest rate A= capital that is borrowed If the interest rate is 1.15 % and the repayment period is 48 months, monthly repayment value is calculated with Eq.1. Total capital cost with interest rate can be found as $. The energy costs for electricity (0,06$/kWh), and natural gas (0,35$ /m 3 ) were obtained from data provided by Bursagaz and UEDAS. Net Annual Savings is calculated as $. n= investment economic lifetime R= annually energy savings Decision Rule; If PW 0, then the investment is attractive PW is calculated as $ with Eq. 2. Cost savings can be seen from the Table 3. and Fig.6. Project related costs and benefits are identified by comparing, after retrofit, before retrofit annual energy costs. As shown above, retrofit will provide, $ annual savings, which proves profitability of the investment. Simple payback period is calculated as 4.14 years. Tab. 1: The annual costs of the old system Item Cost Labor & Maintenance Cost $/year Coal Cost $/year Total Annual Cost (AC1) $/year Tab. 2: The annual costs of the retrofitted system Item Cost Electricity Cost $/year Natural Gas Cost $/year Total Annual Cost (AC 2) $/year Before making an investment, present worth analysis should take into consideration; Fig 6: Cost savings of the building (%) PW = R 1 + R 2 +. (1+i) 1 (1+i) 1 Areas of the factory R n (1+i) n A (2) Tab. 3. Cost savings table of the retrofit Operational cost of Financial the new system savings ($/year) ($/year) Operational cost of the old system ($/year) Implementation Cost Imp. Cost with interest rate Lower Housing & Warehouse $ $ $ $ $ Upper Housing $ $ $ $ $ Manager Housing $ $ $ $ $ Administrative Building $ $ $ $ $ Production facility $ $ $ $ $ Social facility and dining hall $ $ $ $ $ V. Conclusions Acknowledgements This paper provides a literature review about green buildings and building energy efficiency retrofits. The case study of an industrial building has emphasized that how HVAC retrofitting, can increase energy efficiency and decrease operational costs. In this study, as a result of the retrofitting, the annual energy cost decreased by %40. Coal boiler is replaced with air sourced, water sourced heat pump system, solar thermal system, condensing boiler system and waste heat recovery system. Energy efficiency in heating and cooling systems is improved through the use of automatic thermostats and the hybrid control ability of the system. For the financial analyses, cost/savings table is represented. Energy efficiency retrofit, undoubtedly play a vital role in achieving energy efficiency goals. As a result, comprehensive retrofit programs should be developed to achieve the Turkey s energy targets. 420 The successful completion of this study is due to the contributions of Biytaş Company. These contributions are gratefully acknowledged. References Suzer O., A comparative review of environmental concern prioritization: LEED vs other major certification systems, Journal of Environmental Management, 154, , (2015). Azouz M., Kim J., Examining Contemporary Issues for Green Buildings from Contractors Perspectives, Procedia Computer Science Journal, 118, , (2015).

435 Filippi M., Remarks on the green retrofitting of historic buildings in Italy, Journal of Energy and Buildings, 95, 15-22, (2015). Bureau of Energy Efficiency, Energy Economy Handbook, (2007), India Mehta P., Wiesehan M., Sustainable Energy in Building Systems, Procedia Computer Science Journal, 19, , (2013). Hewage K., Ruparathna R. and Sadiq R., Improving the energy efficiency of the existing building stock: A critical review of commercial and institutional buildings, Renewable and Sustainable Energy Reviews, 53, , (2015). Pengpeng X., A model for sustainable building energy efficiency retrofit using energy performance contracting mechanism for hotel buildings in China, Ph.D Thesis, Hong Kong PolyTechnic University, Hong Kong, (2012). Ferdos N., An approach for cost optimum energy efficient retrofit of primary school buildings in Turkey, M.Sc. Thesis, İstanbul Technical University, Turkey, (2015). Tahsildoost M., Zomorodian Z., Energy retrofit Techniques: An experimental study of two typical school buildings in Tehran, Energy and Buildings Journal, 104, 65-72, (2015) Audenaert A., Boeck L., De Mesmaeker L., Verbeke S., Improving the energy performance of residential buildings: A literature review, Renewable and Sustainable Energy Reviews, 52, , (2015). U.S. energy information administration website ( Investment Support and Promotion Agency of Turkey Turkish State Meteorological Service Antunes C. et al., A comparison between cost optimality and return on investment for energy retrofit in buildings, Sustainable Cities and Societies Journal, 21, 12-25, (2015). Bianco N. et al., A new methodology for investigating the cost-optimality of energy retrofitting a building category, Energy and Buildings Journal, 107, , (2015). Hogan M., A design approach to achieving the passive house standard in a home energy retrofit, MSc. Thesis, University of Oregon, USA, ( 2011). Touchie M., Improving the energy performance of multi-unit residential buildings using air-source heat pumps and enclosed balconies, PhD. Thesis, University of Toronto, Canada, ( 2014). Becchio et al., The cost optimal methodology for evaluating the energy retrofit of an ex-industrial building in Turin, Energy Procedia Journal, 78, , ( 2015). Gundogan H., Motivators and barriers for green building construction market in Turkey, MSc. Thesis, Middle East Technical University, Turkey, (2012). Dirksen T., McGowan M., Greening existing buildings with LEED-EB, MSc. Thesis, Massachusetts Institute of Technology, USA, (2008). Bhattacharyya S., Energy Economics, Springer Press, University of Dundee, UK, Bulut H., Energy consumption in Turkey, Harran University, Turkey, 2009 ( f) United Nations Environment Programme website, Turkish Standards Institution website, Turkish Green Building Association s website: 421

436 Passive Thermal Management of a Photovoltaic Panel: Influence of Fin Arrangements Ceren Yuksel, Cem Kalkan, Mustafa Aydin, Güven Nergiz, Mehmet Akif Ezan * Dokuz Eylul University, Department of Mechanical Engineering, Izmir, Turkey * mehmet.ezan@deu.edu.tr Abstract It is well known the fact that the efficiency of a photovoltaic panel (PVP) dramatically decreases with increasing the cell temperature. For a solar cell without any thermal control unit, the surface temperature can reach up to 80 C, which significantly reduces the power generation efficiency of the PVP. In the current paper, a phase change material (PCM) reservoir with and without a fin is numerically investigated to overcome the overheating problems in the PVPs. A numerical model is developed in ANSYS-FLUENT software in which the incident solar radiation and power generation of the PVP are implemented into the energy equation as source terms. Fins with different lengths and intervals are placed on the PV surface to improve the rate of heat transfer through the PCM. Furthermore, a parametric survey is conducted to reveal the influence of the thickness of the PCM on the system efficiency. Keywords: Phase change material, photovoltaic panel, thermal management I. Introduction The efficiency of a solar panel depends on three major factors: the amount of incident sunlight on the panel surface, the quality of the semiconductor panel material, and the operating temperature of the semiconductor cell. Among these factors, the operating temperature of the photovoltaic cell may be the most important factor (Du et al., 2013). The efficiency of the PV cells is well known to decrease with increasing cell temperature. Therefore, it is required to keep the panel temperature as low as possible to increase the efficiency of power generation (Browne et al., 2015). Upon ascending above the nominal operating temperature of 25 C, as defined by the industry standard STC (Standard Test Conditions), an increase in the panel temperature causes a decrease in conversion rate by 0.5%/ C (Emery et al., 1996). Recently, Yuksel et al. (2015), investigated the panel variation of temperature depending on the climatic conditions for the city of Izmir/Turkey. They have indicated that the highest insolation acting on building integrated photovoltaic panel (BIPV) is in October, and the PV temperature can reach to 57 C in this month. The increment in the panel temperature corresponds to a reduction of panel efficiency by 16%. Consequently, panel temperature is vital importance for the efficiency of PV cells. Only about 16% of the radiation incident on the photovoltaic cell is converted into electrical energy. Therefore more than 80% of the absorbed energy is waste heat (Ingersoll, 1986). Thermal control of the PV panel has become increasingly important in recent years and different passive and active thermal management techniques have been used to keep PV at lower temperatures (Hasan et al., 2010). Active control methods usually use a pump or fan to circulate the heat transfer fluid to reduce the PV panel temperature. Passive cooling does not require additional devices such as fan or pump. This technique enables to provide an acceptable temperature level of cooling for PV, by using a high heat capacity material such as PCM or the use of a duct for cooling via natural convection. Thus, it may be considered to be one of the most effective methods for thermal management of PVPs (Cuce et al., 2011). Phase change materials (PCMs) absorb a significant amount of energy as latent heat at phase transition temperature. Owing to the high storage capability of the PCMs, they are used as the thermal energy storage materials. By incorporating a PCM cavity adjacent to a PV panel the temperature of the PV cell could be maintained at low temperature. During the daytime, PCM can absorb the excess heat by melting and prevents the overheating. At nighttime, on the other hand, the absorbed heat inside the PCM is released through the low-temperature external environment, so that the material transforms into the solid phase again. Accordingly, phase change materials can repeatedly be used without deterioration, and because it is environmentally friendly, is a preferable material as a heat storage material. Huang et al. (2004) studied PCM melting in dimensions of 300 mm x 132 mm x 40 mm aluminum container, under the solar radiation values varying from the range of W/m 2. They used a finite volume model to analyze the heat transfer in a twodimensional container. Huang et al. (2006a, b) obtained temperature distributions on the front surface and inside PCM reservoir with and without fin at different insolation values. Temperature distribution was estimated numerically with 2D and 3D 422

437 computational models and the model was verified experimentally. It was found that the temperature rise in this system can be reduced by greater than 30 C for 130 min. Huang et al. (2011) dealt with the PV-PCM system with metallic fins, which is previously investigated by Huang et al. (2006a, b). Natural convection effects are also included of the PCM melting and solidification process in a rectangular container. In this way they have achieved improvement in temperature regulation. Cellura et al. (2008) analyzed the same geometry with partial differential equation (PDE) solver. In the model, they considered that the PCM is a pure material. They assumed that the PCM melts at a constant temperature which means that temperature of the material does not change during the melting process. However, this is not valid for many PCMs in building applications. Phase transition temperature varies within a certain range. They reported that a PCM with a melting temperature between 28 C and 32 C could increase the efficiency by about 20% in the summer period. Biwole et al. (2013) investigated the finned geometry that is examined by Huang et al. (2004) using finite volume method. Natural convection effect in PCM has also included the study. They reported that the PV with PCM took 80 min to reach 40 C while PV without PCM took only 5 min to arrive at the same temperature. Tonui and Tripanagnostopoulos (2007, 2008) have placed the fins into the air duct on the back of the panel for increasing heat transfer area. They found that heat removal from the PV module can be enhanced with fins for better electrical and thermal energy production. Cuce et al. (2011) used an aluminum heat sink for remove waste heat from the panel surface. The experiments were carried out at different ambient temperatures and various insolation values under a solar simulator. Results showed that about 20% enhancements in power output can be achieved with the aluminum rectangular fins at the 800 W/m 2 solar radiation value. In this paper, phase change material reservoir with and without a finned surface is numerically investigated. The fins with different lengths and spacing are placed into the PCM domain to increase the heat transfer surface area. Moreover, to study the effect of PCM on the system efficiency, a set of parametric analyses are conducted to by varying the thickness of the PCM. Numerical analyzes were performed on ANSYS-FLUENT software. To validate the current numerical model the results are compared with our previous study. II. Material and Method Integrated thermal system, which consists of PV and PCM domains, has been created on ANSYS Workbench. Schematic representation of the system is given in Figure 1. It is assumed that the PV consists of 3 mm glass layer and 1 mm plastic material. For the case of PCM embedded system, the PCM thickness is varied from 1 cm to 3 cm to examine the effect of PCM mass on the performance of a PV unit. Moreover, fins are placed in the PCM reservoir with different 423 lengths and also intervals. Half and full fin width fin cases are studied in which a total of 15, 25 and 40 fins are used to enhance the rate of heat transfer. The thermo-physical properties of the materials that are utilized in the current model are given in Table 1. Table 1. Thermo-physical properties of the materials Material c ρ k L T m (J/kgK) (kg/m 3 ) (W/mK) (J/kg) (K) Glass PV/Plastic PCM Fin/Al Figure 1. Schematic representation of the system Bottom (y = 0), top (y = H) and right (x = W) surfaces of the system is considered to be adiabatic ( T/ y y=0,h = 0 and T/ x x=w = 0). The surface on the left (x = 0), on the other hand, is exposed to the outside environment. Heat transfer between the surface and the external environment comprise the solar irradiation as well as the emissive power and mixed convection. A portion of the solar energy falling on the control surfaces transfers to the environment, and the rest passes through the PV by conduction. The energy balance on the external surface of the glass is written as dt k q q q dx where y x 0 solar x q radiation q convection I solar PVP solar convection radiation PCM FIN (1) q represents the insolation falling on the per unit area, q convection and q radiation, indicate the convective and radiative heat losses from the surface. The amount of solar radiation transmitted into photovoltaic panels depends on the absorption coefficient of the surface, and the incident radiation. Heat transfer by convection from the surface to the external environment is expressed as the sum of the natural and forced components, q convection htotal ( Tsurface T ) (2)

438 where T surface and T represent the outer surface temperature of the PV and the ambient temperature, respectively. h total is the combined convection heat transfer coefficient between the surface and ambient. It consists of natural and forced components (Hendricks and Sark, 2011). The heat transfer by radiation from the surface can be calculated with the well-known Stefan-Boltzmann equation. In the current model, it is assumed that the incident solar radiation is absorbed by the glass layer and PV cells. Depending on the efficiency of the PV only a small portion of the absorbed energy can be transformed into the electricity and the rest causes heating the cell. On the glass layer, on the other hand, since there is no way to generate electricity, the absorbed energy directly increases the temperature of the layer. A total of three source terms are developed in C++ language and interpreted into the ANSYS-FLUENT software to calculate the power output and the heat generation in the PV cell and also the glass layer. According to the literature survey of the authors, there is no attempt has been made to simulate the energy transformation in a PV cell in ANSYS-FLUENT software. In the previous models in ANSYS-FLUENT researchers considered the PV as a plate with no power output or absorbed heat. In the current, it is the first time that the panel is defined a power generator and the heat generations in both PV cells and the glass layer is simulated using an User- Defined-Function (UDF). lower isolation values, such as 200 and 300 W/m 2. For the higher insolations, on the other hand, a small discrepancy is observed, which is less than 1 K. The comparative results prove that the current model is consistent with the results in the literature. III.2. Parametric Results Analyses were performed under constant insolation values for 8 hours W/m 2 was chosen as the theoretical maximum value. Since a BIPV could not receive that rate of insolation for 8 hours, the analyses were conducted for different insolation values according to the annual average insolation values for Izmir City, Turkey (Photovoltaic Geographical Information System, Interactive Maps). In Table 2 the monthly average values of Izmir City are given. Three representative insolation values was chosen as 500, 300 and 200 W/m 2, regarding the monthly average values. III. Results and Discussions In this section, first the validity of the model is proven by comparing the Yuksel et al. (2015). Later on, the parametric results are discussed for the systems with and without fin configurations under different solar radiation conditions and also fin configurations. III.1. Validation A numerical model for the passive thermal control of a PV unit was developed by Yuksel et al.(2015). They developed the model in C++ language by discretizing the energy equation with using the control volume approach. They determined the system performance of a PV unit with three different PCM thicknesses. Numerical analyses were carried out for constant and variable solar radiation cases. The validity of the model was revealed by comparing the predicted numerical results with the exact solution for solidification problem in a half space which was given by Ozisik (1993). In this work, to prove the validity of the current model in ANSYS-FLUENT, the problem that is simulated by Yuksel et al. (2015) is reproduced. The results are compared with the reference work for four different insolation values, 200, 300, 500 and 1000 W/m 2. In Figure 2, the time-wise variation of the PV temperatures are compared with the ones in the reference work. Here the dashed lines represent the current results and the solid lines are for the reference work. It was found that the results are full compliance with the numerical code especially for Figure 2. Comparative PV temperature results Table 2. Annual average insolation values for Izmir City Months Irradiation(W/m 2 ) January February March April May June July August September October November December In the earlier study of the authors (Yuksel et al., 2015) it was found that a complete melting could not be obtained at lower insolations, when the PCM thickness was selected to be 3 cm. Even though the PCM domain has the highest thermal energy storage capacity when the total amount of PCM increases, since the thermal conductivity of the material is not 424

439 that high, the conductive thermal resistance inside the domain reduces the rate of heat transfer. That is, improving the speed of heat transfer within the PCM domain for higher PCM amounts may increase the effectiveness of the system. Since the incomplete melting problem was observed in the case in which 3 cm of PCM was implemented, in the current analyses the thickness of the PCM domain is fixed to be 3 cm. The influence of the extended surface configurations, namely the length and the spacing, are represented. which half and full fin configuration are simulated for 500 W/m 2. The results are given at t = 20,000 s. Since there is a cyclic nature along the y-direction, only a small representative section with a constant height is represented by each fin design. In Figure 3, the time-wise variation of the mean temperature of the PCM is given for 1000 W/m 2 of insolation. Here the solid black line represents the case in which the PV does not include fins. It is clear that the temperature slope of the curve after 6 hours of sunlight exposure. Fin insertion at the rear side of the PV enhances the rate of heat transfer and let the meltdown PCM more quickly. With increasing the fin length or number of fins, the rate of heat transfer increases, so that the higher amount of PCM could be melted. In other words, the required time to reach the steady-state condition reduces with fin insertion. Without the fin the PCM temperature reaches to the melting temperature in seconds. The required time to arrive at the same temperature is reduced to s with the addition of 40 full-length fins. Consequently, a higher amount of thermal energy could be stored in the domain, which corresponds improved energetic effectiveness, in the case the fins are implemented. It is also interesting to note that the difference between the half and full-width fins becomes more significant when the number of fins is increased. For the configuration in which forty-fins are used, the difference between the half and full-length scenarios is nearly 35%, regarding the time to reach 310 K. On the other hand, for five fins configuration there is no remarkable change between the full and half designs. (a) Five fin (b) Twenty-five fin Figure 3. PCM temperature with and without extended surface configurations for 1000 W/m 2 insolation In Figure 4, the temperature distributions inside the computational domains are compared for the cases in 425 (c) Forty fin Figure 4. Temperature contours at t = 20,000 s for I solar = 500 W/m 2 For five-fin design higher temperature gradients are observed on the PV/PCM interface, which

440 corresponds, non-uniform temperature distributions inside the domain. The highest PV temperature are observed in the case in which five-fins are used. Increasing the fin length from half to full, lead penetration of heat transfer into PCM and a local cold spot is observed on the PV layer. Increasing the number of total fins provides homogeneous temperature distributions inside the PCM domain and reduces the temperature of the PV. It is clear that the implementation of the finned surface the PCM domain is heated more effectively. PV temperatures are given in Figure 5 for four different insolation values as 1000, 500, 300 and 200 W/m 2. In Figure 5(a), the time-wise variations of the PV are given for the highest insolation value. It is found that the utilization of fins inside the PCM shift the curves through left, which means that the total time for melting reduces. The mass of PCM reduces with inserting fins but the thickness of the fins are 4 mm so that one can assume that the heat capacity of the PCMs is almost same for each design. Since the storage capacities are nearly identical, the only difference could be obtained in terms of the heat transfer speed. The usage of fins keeps PV at a lower temperature at the early periods of the process. With increasing the number of fins and/or length of the fins enhance the rate of heat transfer, so that reduces the mean PV temperature. Beyond the complete melting of the PCM, the PV temperature rapidly increases and reaches to the steady state temperature. The maximum temperature difference between the plain PV surface and the finned one reaches almost 7.5 K for the design in which forty full length fins are used. Figures 5(b, c & d) represent the PV temperature variations for 500, 300 and 200 W/m 2 of insolations, respectively. It is clear that decreasing the insolation value suppresses the influence of fin usage. At 500 W/m 2, the temperature difference between the configuration in which forty fins are used and the one in which there is no fin is nearly 4 K. However, the maximum temperature difference reduces through 1 K for the lower insolations, such as 300 and 200 W/m 2. At lower insolation values the usage of fins does not cause any significant change, so that it may not be feasible to use finned surface considering the cost of the fin material and the manufacturing complexities. On the other hand, for the higher insolation values, optimum fin configuration may be selected regarding the capital cost and the total volume that is occupied by the PCM. Increasing the total number of fins will reduce the mass of PCM so that the heat capacity of the passive thermal controller will decrease. Moreover, the natural convection phenomena inside the PCM cavity will be affected by the fin arrangement. Decreasing the gap between the fins restrict the convection currents and reduce the rate of heat transfer due to natural convection will be decreased. Using the full-length fin, on the other hand, will generate small confined zones. The buoyancy forces enhance as the characteristic height increases. The effect of the natural convection currents will be weaker as the domain is divided into small zones by using the 426 full-length fins. Consequently, more in-depth analyses should be conducted to end up with the optimum fin arrangements for the passive thermal controller unit of a PV unit. (a) I solar = 1000 W/m 2 (b) I solar = 500 W/m 2 (c) I solar = 300 W/m 2

441 (d) I solar = 200 W/m 2 Figure 5. Time-wise variations of the PV temperature for different insolation values and fin arrangements Power outputs for the selected insolation values are given in Figure 6 under various fin arrangements. For the highest insolation value, i.e W/m 2, there is an increment in terms of the power output at the early periods of the process. Increasing the number of fins and/or the length of the fins improves the power output at the beginning of the process. Nevertheless, for the configurations with fins, after the complete melting achieved, the PV temperature increases rapidly, that is, the power output drops down the reference case, i.e. w/o fin. At lower incident solar radiation values, the difference between with and without fin configurations reduces. The maximum difference between the configurations with forty-full fin and without fin are 5 W/m 2, 0.75 W/m 2, 0.75 W/m 2, 0.2 W/m 2 and 0.1 W/m 2 for 1000, 500, 300 and 200 W/m 2, respectively. (a) I solar = 1000 W/m 2 (b) I solar = 500 W/m 2 IV. Conclusions In the current study a numerical model for a BIPV panel is developed in a commercial CFD solver, ANSYS-FLUENT. The model includes power generation in the PV cell and also considers the absorbed heat inside glass and layers. According to the comparative results, one may infer that, the finned configurations are effective if only if the PV panel is exposed to higher incident solar radiation values, i.e. insolations greater than 500 W/m 2. Finned design improves the rate of heat transfer and prevents the temperature jump of the PV. Higher power outputs could be obtained at the early periods of the process by increasing the number of fins and/or length of the fin. Since the performance of the proposed design hardly depends on the incident solar radiation, the number of fins and also properties of the selected PCM, a more comprehensive work should be conducted to find the optimum configuration with considering the investment costs and payback time of the proposed designs. (c) I solar = 300 W/m 2 427

442 (d) I solar = 200 W/m 2 Figure 6. Power output for different insolation values and fin arrangements Nomenclature c : Specific heat (J/kgK) h total : Heat transfer coefficient (W/m 2 K) I solar : Incident solar radiation (W/m 2 ) k : Thermal conductivity (W/mK) P out : Power output (W) q solar : Insolation falling on the per unit area (W/m 2 ) q conv : Convective heat losses from the surface (W/m 2 ) q rad : Radiative heat losses from the surface (W/m 2 ) T m : Melting temperature (K) T PV : PV temperature (K) T : Ambient temperature (K) T sky : Sky temperature (K) : PV thickness (m) w PV References Biwole P. H., Pierre E., and Frédéric K., "Phasechange materials to improve solar panel's performance." Energy and Buildings 62, 59-67, (2013). Browne M. C., Norton B., and McCormack S. J., "Phase change materials for photovoltaic thermal management." Renewable and Sustainable Energy Reviews 47, , (2015). Cellura M., Valerio L. B., and Antonino M., "582. A Photovoltaic panel coupled with a phase changing material heat storage system in hot climates", (2008). Cuce E., Bali T., and Sekucoglu S. A., "Effects of passive cooling on performance of silicon photovoltaic cells." International Journal of Low-Carbon Technologies, ctr018, (2011). Du D., Darkwa J. and Kokogiannakis G., "Thermal management systems for Photovoltaics (PV) 428 installations: A critical review." Solar Energy 97, , (2013). Emery K., Burdick J., Caiyem Y., Dunlavy D., Field H., Kroposki B., Moriatry T., Temperature dependence of photovoltaic cells, modules and systems, in: Proceedings of the 25th IEEE PV Specialists Conference, Washington, DC, USA, May 13 19, pp , (1996). Hasan A., McCormack S. J., Huang M. J, Norton B., "Evaluation of phase change materials for thermal regulation enhancement of building integrated photovoltaics." Solar Energy 84.9, , (2010). Hendricks, J. H. C., and W. G. J. H. M. Sark, "Annual performance enhancement of building integrated photovoltaic modules by applying phase change materials." Progress in Photovoltaics: Research and Applications 21.4, , (2013). Huang, M. J., Eames P. C., and Norton B., "Thermal regulation of building-integrated photovoltaics using phase change materials." International Journal of Heat and Mass Transfer 47.12, , (2004). Huang, M. J., Eames P. C., and Norton B., "Phase change materials for limiting temperature rise in building integrated photovoltaics." Solar Energy 80.9, , (2006a). Huang, M. J., Eames P. C., and Norton B., "Comparison of a small-scale 3D PCM thermal control model with a validated 2D PCM thermal control model." Solar energy materials and solar cells 90.13, , (2006b). Huang, M. J., Eames P. C., and Norton B., "Natural convection in an internally finned phase change material heat sink for the thermal management of photovoltaics." Solar Energy Materials and Solar Cells 95.7, , (2011). Ingersoll J.G., Simplified calculation of solar cell temperatures in terrestrial photovoltaic arrays, ASME J. Solar Energy Eng. 108, , (1986). Photovoltaic Geographical Information System, Interactive Maps, Tonui J.K., Tripanagnostopoulos Y., Air-cooled PV/T solar collectors with low cost performance improvements. Solar Energy, 81: , (2007). Tonui J.K., Tripanagnostopoulos Y., Performance improvement of PV/T solar collectors with natural air flow operation. Solar Energy, 82:1 12, (2008). Yuksel C., Ezan M.A., Aydın M., Kalkan C., Nergiz G., A transient model for vertical PV module with PCM; Constant & Variable insolation, Journal of Thermal Engineering, (2015).

443 Multi-Criteria Selection Factors for Evaluation of Intelligent Buildings; A Novel Approach for Energy Management Elnaz Asadian*, Katayoun Taghizadeh Azari, Ali Vakili Ardebili, Samira Mahmoodkelayeh Tehran University, Faculty of fine art, 16th Azar St., Enghelab Sq., Tehran, , Iran * elnaz.asadian@yahoo.com Abstract Among various challenges that the new generation encounter, the environmental and energy concerns have been attracted a great deal of attention in recent years. The indiscriminate energy use not only have a direct influence on human s life quality, but also is associated with heavy environmental impacts such as energy resources depletion, global warming, climate change, carbon emission, acidic rain and waste accumulation. As a result, the efficient management of energy supplies through rational use of present resources has become a great concern during the past few decades. Taking into account that buildings are one of the largest energy consumers, improving energy conservation and sustainable developments play a critical role in the construction section. The development of sustainable construction has introduced a new concept entitled Intelligent Buildings (IBs). This concept is based on the use of technology and process to create a building that is safer and more productive for its occupants and more operationally efficient for its owners. Therefore, IBs are seeking a balance between all three environmental, social and economic sustainability features. Various definitions have been proposed for IBs in the literature. However, most existing definitions are either too vague to be a reliable guide for detailed designs or do not consider all the involved parameters. These definitions are mostly focused on the technological aspects or only consider financial features. Since there is no consensus agreement on the definition of intelligent buildings, a strong need for a framework which includes all the involved criteria as a decision-making tool is felt. Herein, a multi-criteria framework composed of 68 sub-factors is proposed as a comprehensive tool regarding selection of intelligent buildings. In this regard, 8 quality environment modules were considered as the main factors including environmental and energy indicators, space flexibility, cost effectiveness, user comfort, working efficiency, safety, culture and technological factors. The results provide a better insight towads choosing the priority levels of each factors in IBs. Keywords: Energy consumption, intelligent buildings, multi-criteria decision-making, sustainable construction. I. Introduction Among various challenges that the new generation encounter, the environmental and energy concerns have been attracted a great deal of attention in recent years (Perez-Lombard et al, 2008). In one hand, the rise in energy consumption reflects the economic growth of industry in developing countries. However, on the other hand this issue resulting in the difficulties with respect to energy supply. With the increasing energy demands due to the population growth and considering the limited renewable energy resources such as fossil fuels, finding the alternative energies is more important than ever. Building are one of the largest energy consumers in any developing countries (Junnila and Horvath, 2006). Hence, the study of energy consumption and the environmental impact of buildings during their life cycles play a critical role in this regard. Improving the energy efficiency in the new generation buildings is one of the easiest way to overcome this issue and reduce the energy consumption, operating costs and CO2 emission. The increasing need for sustainable or green design has led to the development of a new concept; Intelligent Buildings (IBs), which requires a continuous process of balancing between all three environmental, social and economic sustainability features (ALwaer and Clements-Croome, 2010). IBs not only try to bring flexibility and comfort to their inhabitants and occupants while maintaining the cost effectiveness, but also deals with the user s safety as well as attaining higher environmental performance standards (Wong et al, 2005). Based on the fact that a sustainable development should combine three basic issues including People (owners, users, occupants and inhabitants), Products (equipment, materials and facilities) and Processes (maintenance, facilities management and performance evaluation) as well as the relationships between them, a profound survey that integrates all these factors is critical. Therefore, in the present study we propose a multi-criteria decision-making framework (including 68 sub-factors) which provides a major insight into the selection of sustainable intelligent buildings indicators. The results can helps the stakeholders (clients, architects, engineers and construction managers) to obtain a better perception on the priority levels of each factor in IBs. I.1. Energy consumption As previously mentioned, the rapidly growing world energy consumption faces the governments with 429

444 some serious environmental and economic issues. The statistics gathered by the International Energy Agency on the energy consumption trends are so alerting; during the last two decades ( ), the primary energy has grown by 49% and CO2 emissions by 43%, with an average annual increase of 2% and 1.8%, respectively (Fig. 1). Fig. 1 The plot of world primary energy consumption, CO2 emission and the population growth between 1984 and 2004 (Perez-Lombard et al, 2008). The problems associated with the innumerable use of energy sources has become a substantial worldwide challenge from developing countries to more industrialized ones. According to a survey done by the Lawrence Livermore National Laboratory (LLNL), more than half (58%) of the total energy produced in US is wasted due to inefficiencies such as waste heat from power plants, vehicles and light bulbs. Moreover, the lack of energy quandary also involves the energy-rich countries. For instance, Iran holds 10% of the world s proven oil reserves and 15% of its gas and is the second largest OPEC s exporters. As one of the most energy-intensive countries of the world, its per capita energy consumption is 15 times that of Japan and 10 times that of European Unions (Iran energy balance sheet, 2012). Although Iran is among the countries with abundant natural energy resources such as oil and gas but it is not an exception in facing the energy issues and according to the International Energy Agency statistic the energy produced in the country is not accountable for its use (Fig. 2). Fig. 2 The plot of total energy production and consumption in Iran (IEA, 2015). In February 2010 and after the social problems associated with energy subsidies, the government reform the energy prices to manage the increasing trend of energy consumption. According to the so-called Targeted Subsidies Law which passed by the parliament, energy (petrol, oil, liquefied gas and kerosene) prices would increase up to 90 percent of the border prices in five years (at least 75 percent of the export prices for natural gas) as shown in Fig. 3 (Moshiri, 2013). productivity and efficiency in the insulation systems cause a lot of energy waste. As an example, 36% of the total energy consumption in Iran is consumed in building sectors as showed in Fig. 4 (Iran energy balance sheet, 2012). High energy prices may lead to reduced energy consumption to some extent. However, this also affects the wellbeing and comfort of users which in turns influences human s productivity, morale and satisfaction (Iran energy balance sheet, 2012). In the meantime, the construction sector is one of the largest energy consumers. In addition to the high energy consumption in this section, the improper design, non-standard materials and lack of 430 Fig. 3 The energy prices growth in Iran (Moshiri, 2013).

445 Fig. 4 Iran energy consumptions in different sectors (Iran energy balance sheet, 2012) It should be noted that these amount of consumption is not only unique to Iran and about 40% of world total energy consumption is in the building sector (Fig. 5) (Perez-Lombard et al, 2008). Fig. 5 Primary energy use in US commercial and residential buildings (IEA, 2015). Considering the high levels of energy consumption in this sector and the need for energy saving, developing strategies for optimal management of energy consumption is essential. The emergence of new technologies can be a precursor to developments in this regard. Hence, the next section is devoted to one of the most recent related technologies i.e. Intelligent Buildings (IBs). I.2. Intelligent buildings (IBs) The term intelligent building was first used by UTBS Corporation (United Technology Building Systems Corporation) in 1981 in the USA. About 2 years later, their efforts became a reality and the City Place Building in Hartford (Connecticut, USA) was named as the world s first intelligent building (So et al, 1999). Since then, various definitions have been proposed for IBs. The initial definitions only focused on the technological aspects without taking the user s requirements into account (Powell, 1990). However, the recent definitions also consider the occupant s interactions as well as their relation with the surrounding environment (Wong et al, 2005), (Yang and Peng, 2001), (Martins et al, 2012). According to the Intelligent Building Institute of the United States, an intelligent building can be defined as a building that provides a productive and 431 cost-effective environment through optimization of four basic elements including structures, systems, services and management and the interrelationships between them (Wong et al, 2005), (Wigginton and Harris, 2002). On the other hand, the European Intelligent Building Group described IB in a way that creates an environment which maximizes the effectiveness of the building s occupants, while at the same time enabling efficient management of resources with minimum life-time costs of hardware and facilities (Wong et al, 2005), (Wigginton and Harris, 2002).Besides these general definitions, there exist numerous descriptions for IBs in the literature (Wong et al, 2005). For instance, in a literature review which was done by Wigginton and Harris (Wigginton and Harris, 2002) 30 separate definitions were listed in relation to intelligence and building. In 1983, Cardin defined an intelligent building as one which has fully automated building service control systems (Wong et al, 2005). However, this definition was further developed as one which integrates various systems to effectively manage resources in a coordinated mode to maximize: technical performance, investment and operating cost savings, flexibility by the Intelligent Building Institution in Washington in 1988 (Arkin and Paciuk, 1997). More recently, Seo et al. have proposed that intelligent buildings are not intelligent by themselves, but they can furnish the occupants with more intelligence and enable them to work more efficiently (Wong et al, 2005). Meanwhile, it is also suggested that intelligent buildings represent a key benefit that can reduce the initial capital outlay, as well as enabling a higher potential return on investment (ROI). Some researchers (Wong et al, 2005) introduce the intelligent building as a multidisciplinary effort to integrate and optimize the building structures, systems, services and management in order to create a productive, cost effective and environmentally approved environment for the building occupants. The importance of IBs not only have been drawn a great deal of attention in the construction field in USA and Europe, but also has a significant impact in Asian countries. Based on the definition that published by Public Works Department of Singapore government, an intelligent building must accomplish three distinction conditions: (1) The building should have advanced automatic control systems to monitor various facilities, including air-conditioning, temperature, lighting, security, fire etc. to provide a comfortable working environment for the tenants. (2) The building should have good networking infrastructure to enable data flow between floors. (3) The building should provide adequate tele-communication facilities (Wong et al, 2005) As can be seen, there is not a general agreement about the definition of IBs. However, the main goal of

446 implementation of IBs is to reduce energy consumption through energy saving and conservation. The percentage of reduced energy consumption (%) is shown in Fig. 6. order to identify the perceived critical selection criteria and to gather all the evaluation factors that affect the performance of an IB. for this purpose, the main factors were collected from eight quality environment modules (QEMs) (from M1 to M8) including (So et al, 1999): M1; Environmental and energy M2; Space utilization and flexibility M3; Cost effectiveness M4; Human comfort M5; Working efficiency M6; Safety and security M7; Culture M8; Technological factors Fig. 6 The percentage of reduced energy consumption (%) (IEA, 2015). For sure, a comprehensive definition which considers all involved parameters is essential for a better decision-making since without a precise understanding, achieving a smart building that gather the best combination of social, environmental and economic values seems impossible. II. Methodology outline Since the IBs are considered as a complex system made up of various components, the best approach to analyze them is to divide the system to its elements. In this regard, a general survey was undertaken in Then in the next step, each of these eight key modules were assigned to a number of facilities which are marked as secondary and sub-factors (Fig. 7). In order to achieve a complete and comprehensive set of key elements, a statistical population of 76-members were chose from the experts who are mostly active in the field of Building Management Systems (BMSs). It is noteworthy to mention that since the IB industry is an emerging and developing industry in Iran, only a very limited number of professionals were identified for the survey. The evaluation criteria of IBs that influence the whole life cycle of building and derived from the previously mentioned 8 main factors (quality environment modules) are shown in Fig

447 Fig. 7. Intelligent buildings multi-criteria selection framework (ALwaer and Clements-Croome, 2010), (Wong et al, 2005), (So et al, 1999), (Wong and li, 2008) 433

448 III. Selection factors framework The main challenge usually facing the project design team is to choose the optimum configurations that meet the considerations of developers in one hand while strike a balance between these goals and the expectations of the users on the other hand (Wong and li, 2008). The wide distribution of these multi-dimensional perspectives augment the complexities involved in the evaluation and selection of the control systems for the intelligent buildings. Therefore, there is an essential need for the selection evaluation tools to be recognized in order to facilitate the achievement a logical decision. As a result, we propose a comprehensive list of evaluation criteria which helps the decision makers to select proper categories and reach the customer s satisfaction. IV. Conclusions In conclusion, efficient intelligent buildings has become a trend for the future of construction industry. The main challenge of designing such buildings is the multiplicity of the factors influencing the decision making process. In the present study, a comprehensive framework of factors affecting the development of intelligent buildings is gathered. Using 8 quality environment modules the main, secondary and sub-factors were driven including 68 key elements. The results of this study can provide a better insight for the design and development of a more sustainable intelligent building. A.T.P. Soalvin C.W. WongK-c. Wong, A new definition of intelligent buildings for Asia, Facilities, 17, , (1999). J.A. Powell, Intelligent design teams design intelligent building, Habitat International, 14, 83 94, (1990). J. Yang, H. Peng, Decision support to the application of intelligent building technologies, Renewable Energy, 22, 67 77, (2001). J. F. Martins et al., Smart Homes and Smart Buildings, 13th Biennial Baltic Electronics Conference (BEC2012), (2012). M. Wigginton, J. Harris, Intelligent Skin, Architectural Press, Oxford, UK, (2002). H. Arkin, M. Paciuk, Evaluating intelligent building according to level of service system integration, Automation in Construction, 6, , (1997). J.K.W. Wong, H. Li, Application of the analytic hierarchy process (AHP) in multi-criteria analysis of the selection of intelligent building systems, Building and Environment, 43, , (2008). References L. Perez-Lombard, J. Ortiz, C. Pout, A review on buildings energy consumption information, Energy and buildings, 40, , (2008). S. Junnila, A. Horvath, A. Acree Guggemos, Life-Cycle Assessment of Office Buildings in Europe and the United States, Journal of infrastructure systems, 12, 10 17, (2006). H. ALwaer, D.J. Clements-Croome, Key performance indicators (KPIs) and priority setting in using the multi-attribute approach for assessing sustainable intelligent buildings, Building and Environment,45, , (2010). J.K.W. Wong, H. Li, S.W. Wang, Intelligent building research: a review, Automation in Construction, 14, , (2005). Iran energy balance sheet, (2012). International Energy Agency, IEA, (2015). S. Moshiri, Energy Price Reform and Energy Efficiency in Iran, International Association for Energy Economics, second quarter, 33 36, (2013). 434

449 HYDROGEN GENERATION, STORAGE AND TECHNOLOGY 435

450 Effect of Nitrogen Doping on Hydrogen Storage of Graphene-TiO2 Nanocomposites Zahra Gohari Bajestani *1, Alp Yurum 2, Omid Akhlaghi 1, Yuda Yurum 1 1 Sabanci University, Faculty of Engineering and Natural Sciences, Tuzla, Istanbul, 34956, Turkey 2 Sabanci University, Nanotechnology Research and Application Center, Tuzla, Istanbul, 34956, Turkey * zgohari@sabanciuniv.edu Abstract TiO2 nanoparticles with exposed (001) facets were synthesized and in-situ decorated over nitrogen-doped reduced graphene oxide sheets (N-rGO) by a solvothermal method. Hydrogen storage capacity of nanocomposite was measured at room temperature and pressures relevant for practical on-board storage systems (9.5 bar). Nitrogendoping was found to induce homogeneous dispersion of TiO2 nanoparticles over graphene sheets that significantly improves the hydrogen storage capacity of the system (0.84 wt%). Comparing the adsorption curves of the samples, it is revealed that rgo has the lowest hydrogen uptake and the surface modification improved the hydrogen adsorption capacity of the substrate (TiO2-N-rGO > TiO2-rGO > N-rGO > rgo). Keywords: graphene, nitrogen-doping, hydrogen adsorption I. Introduction The realization of innovative hydrogen storage materials has worldwide strategic importance. Among the many candidates for hydrogen storage technologies, much effort has been devoted to carbon-based systems due to remarkable properties such as high specific surface area, chemical stability and low mass density (Ataca, Aktürk et al. 2008, Srinivas, Zhu et al. 2010, Chen, Chung et al. 2013, Spyrou, Gournis et al. 2013, Ma, Zhang et al. 2014). Recently, graphene-based carbon materials have gained significant attention as a promising hydrogen storage medium (Gadipelli and Guo 2015, Kostoglou, Tarat et al. 2016). Unfortunately, weak binding energy between carbon nanostructures in any form and H2 is a key factor that leads to low H2 adsorption capacity of these materials (Gadipelli and Guo 2015).Theoretical calculations predicted that doping or functionalization by alkali, alkali earth or transition metal (TM) atoms can significantly improve the hydrogen adsorption/desorption properties of carbonaceous materials (Ma, Zhang et al. 2015). Besides, after doping by TM, hydrogen binding ability can be enhanced by Kuba s type interaction due to presence of empty d-orbitals in TM (Nachimuthu, Lai et al. 2014). However, theoretical and experimental works examined that TMs have tendency of clustering on carbon surfaces as a result of high cohesive energies of TM atoms. This clustering affects the nature of hydrogen bonding and decreases the hydrogen uptake of system (Sun, Wang et al. 2005, Yang, Yoon et al. 2008, Ma, Zhang et al. 2015), thus chemically active sites are required to hold TM atoms and overcome the metal cohesion (Ma, Zhang et al. 2015).One strategy to eliminate this clustering tendency is to introduce different types of defects to the structure of carbonaceous substrate (e.g., substitutional doping atoms and vacancy). In this view, defects act as chemically active sites that can hold TM strongly and enhance the bonding between metal atoms and the substrate (Beheshti, Nojeh et al. 2011, Parambhath, Nagar et al. 2012, Kim, Lee et al. 2014, Seenithurai, Pandyan et al. 2014, Ma, Zhang et al. 2015). Recently, doping with substitutional atoms such as nitrogen has been reported to significantly affect the hydrogen adsorption/desorption process of carbon-based materials (Beheshti, Nojeh et al. 2011, Nachimuthu, Lai et al. 2014). Higher electronegativity of nitrogen compared to carbon gives rise to formation of considerably polarized C N bonding and thus positive charge on the carbon atom. Indeed, nitrogen doping improves the surface chemical activity of parent matrix in terms of polarity and basicity (Parambhath, Nagar et al. 2012). Consequently, these positively charged centers become the active sites for pinning the transition metal/metal oxide nanoparticles grown on the surface of nitrogen-doped (N-doped) carbonaceous supports (Wu, Santandreu et al., Chen, Liu et al. 2015). Vinayan et al. (Parambhath, Nagar et al. 2012) showed 212 % increase in hydrogen uptake of Pd-decorated N-doped graphene compared to Pddecorated graphene at ambient temperature and under pressure of 2MPa. This increase in hydrogen capacity was correlated with high dispersion of nanoparticles and strengthened interaction between N-doped graphene sheets and nanoparticles. TiO2 has been widely studied due to its photochemical, catalytic and dielectric characteristics. Theoretical studies showed the potential of TiO2-anchored carbonaceous materials as a promising hydrogen storage media for room-temperature applications (Wang, Lee et al. 2009, Liu, Gao et al. 2013). Recent experimental results on TiO2-decorated expanded graphite (Yu, Zhao et al. 2012) and TiO2-carbon nanotubes (CNT) composites (Rather, Mehraj-ud-din et al. 2009) revealed the enhanced electrochemical hydrogen uptake of graphene and higher H2 gas adsorption of CNT after impregnation with TiO2 436

451 nanoparticles. In order to achieve TiO2 nanoparticles with high reactivity and stability, controlling the particle shape in synthesis process is a critical factor. As surface-dependent properties such as reactivity are related to the surface chemistry of nanoparticles, using different adsorbate atoms to bind to different crystalline planes can change the occurrence of specific facets (Selloni 2008, Yang, Sun et al. 2008). In case of anatase TiO2, (001) facets are the highly reactive surfaces in the equilibrium state. Hence these faces are rapidly eliminated during crystal growth to minimize the total surface energy of crystal (Yang, Sun et al. 2008, Wu, Yang et al. 2013). According to Wu et al.(wu, Yang et al. 2013), hydrochloric acid (HCl) and hydrofluoric acid (HF) strongly bind to (101) and (001) facets, respectively. Therefore, use of HF as a capping agent would favor the occurrence of (001) rather than (101) planes. Here we report the preparation and characterization of TiO2-decorated N-doped graphene nanocomposite via solvothermal method. Morphological and structural properties of nanocomposites were evaluated by different characterization methods such scanning/transition electron microscopy, X-ray photoelectron spectroscopy and Raman spectroscopy. Results of TEM revealed the stability of high reactive (001) faces during the growth process. To investigate the role of nitrogen doping on hydrogen storage capacity of graphene, hydrogen adsorption measurements were carried out in room temperature and hydrogen pressures up to 9.5 bar. The homogenous dispersion of TiO2 nanoparticles with high reactive facets throughout the N-doped graphene sheets was found to enhance the hydrogen storage capacity of pristine graphene by ~394%. Higher hydrogen uptake of TiO2-N-doped nanocomposite compared to that of graphene sample was linked to strong attachment of fully dispersed and highly reactive nanoparticles to the underlying graphene sheets. II. Experimental Methods II.1. Materials Natural graphite flake (99%), sulfuric acid (H2SO4, 98%), sodium nitrate (NaNO3), potassium permanganate (KMnO4) and hydrochloric acid (HCl, 37%) were purchased from Sigma Aldrich. Hydrogen peroxide (H2O2, 30%), ethanol (C2H6O), hydrofluoric acid (HF), and titanium isopropoxide (C12H28O4Ti) were obtained from Merck. All reagents were analytical grade and used without further purification. 2.2 Preparation method Graphite oxide (GO) was prepared from natural graphite powder using Hummers method (Poh, Sanek et al. 2012). Nitrogen-doped graphene (NrGO) was obtained by treating the as-prepared graphite oxide in a flow of ammonia gas at 1000 C. TiO2 nanoparticles were decorated over N-rGO support by solvothermal method. In a typical synthesis, 100 mg of N-rGO was dispersed in 50 ml of a mixture of ethanol and NMP (25/1 vol/vol) by ultrasonication for 2 h. The required amount of 437 titanium isopropoxide and HF was added to the solution drop-wise under mechanically stirred conditions. Solvothermal treatment was performed at 150 C for 24h. The same solvothermal treatment was applied to GO (without ammonia treatment), named as TiO2-rGO nanocomposite. Suspensions were filtered, washed by ethanol and vacuum-dried at 60 C. As a reference, GO was thermally reduced at 1000 C to attain fully exfoliated graphene oxide (rgo). Preparation of TiO2-N-rGO as well as TiO2-rGO nanocomposite is schematically shown in Fig. 1. Fig. 1: Schematic illustration for the fabrication of TiO2-N-rGO nanocomposite II.2 Characterization methods X-Ray powder diffraction (XRD) patterns were performed using Bruker AXS diffractometer fitted with a Siemens X-ray gun using nm Cu Kα radiation. Raman spectroscopic analysis was recorded from samples by using Renishaw invia reflex Raman spectrometer with a 532 nm laser beam in the range of cm-1 while samples were loaded on silica wafer and focused with a 50 objectives. X-ray photoelectron spectroscopy (XPS) analyses were conducted on a Thermo K-alpha X-ray photoelectron spectrometer with a monochromated Al Kα supported by a low energy electron/ion flood gun for charge neutralization. High resolution transmission electron microscopy (HR-TEM) and scanning electron microscopy (SEM) equipped with energy dispersive spectroscopy (EDS) were carried out by JEM-ARM 200CF; JEOL, Tokyo, Japan with the accelerating voltage of 200 kev and Leo Supra 35VP field emission scanning electron microscope with an acceleration voltage of 2 20kV, respectively. Hydrogen storage capacity of samples was measured by using Intelligent Gravimetric Analyzer (IGA 001, Hiden Isochema). The samples were degassed at 100 C for 12h under high vacuum (~ 10-7 mbar) prior to measurements and then, the hydrogen adsorption isotherm were measured at room temperature and hydrogen pressures up to 9.5 bar. III. Results and Discussion Fig.2a shows the Raman spectra of rgo samples after N-doping and incorporation of TiO2 nanoparticles. The corresponding spectra of graphite and GO were also shown for comparison. Two typical bands of graphite appeared at Raman shifts of ~ 1348 cm -1 (Dband) and 1583 cm -1 (G-band). G-band corresponds to vibration of the sp 2 -bonded carbon atoms in a 2D hexagonal lattice while D-band represents in-plane stretching motion of symmetric sp 2 C C bonds

452 (Giovanni, Poh et al. 2012). Along the process of doping, the G band experiences a blue shift caused by the strain present in the specimen and the D band grows in intensity. Upon oxidation, the G band of GO was detected at higher frequencies of 1595 cm -1 compared to that of graphite (1583 cm -1 ).The most plausible explanation for observation of this blue shift is the presence of isolated C C double bonds in graphitic areas of GO. After thermal reduction, this band shifted back to 1589 cm -1 that is attributed to a graphitic self-healing phenomenon (Kudin, Ozbas et al. 2008). In case of N-rGO, the G-band displayed an upshift to 1596 cm -1 that evidences the incorporation of nitrogen atoms to the graphene sheets (Parambhath, Nagar et al. 2012).The ID/IG of 0.88 was calculated from Raman spectrum of GO sample while it changes to 1.01, 1.03 and 1.05 in N-rGO, TiO2-rGO, andtio2-n-rgo, respectively. The relatively higher intensity of the D compared to G band in doped samples suggests that graphene structure contains a series of defects such as vacancy and grain boundary (Shao, Zhang et al. 2010, Cheng, Yang et al. 2012). In addition, increasing ID/IG of N-rGO after deposition of TiO2 nanoparticles points out the presence of interaction between these heteroatoms and graphene sheets (Shao, Zhang et al. 2010, Shao, Tian et al. 2015). Besides the G and D bands, nanocomposite exhibited characteristic Raman peaks at 151 cm -1, 390 cm -1, 515 cm -1 and 629 cm -1 (Fig. 2b). These peaks are respectively assigned to Eg, B1g(1), A1g + B1g(2), and Eg(2) modes of TiO2 that arise from the external vibration of anatase phase (Choi, Jung et al. 2005, Zhang, Sun et al. 2012, Chang-Jun, Mao-Wen et al. 2013). However, a large frequency blue-shift was detected in of Eg mode of doped TiO2 compared to that of anatase single crystal (144 cm -1 ). A similar large frequency shift was also reported by Zheng et al. for TiO2 nanocrystals fabricated by solution chemical process. They showed when dimensions of TiO2 crystallites decrease to nanometer scale frequency shift occurs in Eg mode as a result of phonon confinement. Fig. 2: Raman spectra of (a) graphite, GO, rgo, N- rgo, TiO2-rGO and TiO2-N-rGO (b) TiO2-rGO and TiO2-N-rGO in the range cm 1 The XRD patterns of the as-prepared samples are shown in Fig. 3. In GO, the characteristic peak of (001) emerged at 2θ 10.5 with d-spacing of 8.4 A can be seen. This peak demonstrates the presence of oxygen containing functional groups on graphene sheets (Poh, Sanek et al. 2012). After thermal reduction, observation of a broad peak ranging from 2θ of 18 to 30, indexed as (002), in the absence of (001) peak in rgo and N-rGO implies the complete removal of functional groups from GO sheets and formation of poorly ordered graphene-like structure along the stacking direction (Xiang, Li et al. 2012). After addition of TiO2, intensity and full width at half maximum (FWHM) of the (002) peak were changed that suggests the restacking of graphene nanosheets was prevented during solvothermal process. This phenomenon illustrates that TiO2 nanoparticles that are attached to the surface of graphene sheets act like a barrier for agglomeration of graphene layers (Bai, Zhang et al. 2014, Shao, Tian et al. 2015). Moreover, we did not observe any peak related to TiO2 nanoparticles in diffraction pattern of nanocomposite. It is noticeable that the main peak of TiO2 at 25.3 and rgo at 24 are at the same region and thus the characteristic peak of anatase TiO2 may be screened by the main peak of rgo. The same observation has also been reported by Shah et al. (Sher Shah, Park et al. 2012), Zhou et al. (Zhou, Wang et al. 2015) and Dhanabalan et al. (Dhanabalan, Li et al. 2013) for TiO2-rGO, SnO2-N-doped graphene and SnO2- Graphene composite, respectively. 438

453 Fig. 3: XRD patterns of GO, N-rGO, TiO2-GO and TiO2-N-rGO samples The morphology and dispersion state of the nanoparticles were analyzed by SEM and HR-TEM. As seen in Fig 4a, TiO2 nanoparticles are poorly distributed on the surface of graphene sheet and agglomerated colonies are clearly detectable. On the other hand, high magnification SEM image of TiO2-NrGO nanocomposite (shown in Fig. 4b) displays almost complete coverage of TiO2 nanoparticles on graphene layers with no sign of agglomeration. Therefore, formation of homogenously distributed nanoparticles on N-doped graphene reveals that N- doping can create a defective substrate which acts as nucleation centers for TiO2 nanoparticles.this observation is in agreement with previous reports (Parambhath, Nagar et al. 2012, Jung, Park et al. 2016) that introducing the heteroatom to the graphene sheets is effective approach to obtain dispersed particles. It should be noted that formation of fine and distributed nanoparticles throughout the sample is highly preferred to increase the hydrogen adsorption capacity of graphene based nanocomposites (Tsao, Tzeng et al. 2010). For this aim, EDS analysis was carried out to get more insight about distribution of nanoparticles in thetio2-n-rgo nanocomposite. Corresponding elemental mapping images of Ti, N, O and C are shown in Fig. 5. Detection of distributed patterns of elements in the elemental mapping confirms the presence of Ti, N and O over the entire graphene sheets with good incorporation into graphene matrix. Fig. 4: SEM images of (a) TiO2-rGO, (b) TiO2-N-rGO Fig. 5: SEM elemental mapping analysis of TiO2-NrGO nanocomposite: (a) selected area and corresponding elemental mapping for (b) oxygen (c) titanium and (d) nitrogen HR-TEM was carried out to characterize the morphology and crystal structure of TiO2 nanoparticles. Fig. 6a shows TiO2 nanoparticles with average particle size of ~ nm. The square- 439

454 shape of nanoparticles is clearly detectable from welldefined edges and corners. Closer observation, presented in Fig. 6 b-c, show the lattice fringes of TiO2 with a spacing of and nm which corresponds to the d-spacing of the (101) and (004) plane of TiO2 anatase phase. The observed (004) planes which are parallel to the particle surface present direct evidence for the existence of {001} exposed nanofacets (Selloni 2008, Gu, Wang et al. 2012, Ding, Huang et al. 2015). Yang et al. (Yang, Sun et al. 2008) investigated the effect of various adsorbate atoms such as H, C, N, O, F and Cl on surface stability of anatase TiO2 by first-principle quantum chemical calculations. Calculated surface energy (γ) of clean and terminated (001) and (101) faces showed that termination with F atoms minimizes the value of γ for both the (001) and (101) surfaces. Low bonding energy of F-F bonding leads to achieve the {001} which is more stable than (101) facets. In another work, Yu et al. (Yu, Xiang et al. 2010) experimentally illustrated that accumulation of F on the surfaces reduces the γ of (001) to below that of (101), thus appearance of TiO2 with exposed (001) planes. Fig. 6d exhibits a schematic illustration of the truncated octahedral bipyramidal structure of TiO2 nanoparticles. The top and bottom surfaces represent (001) facets which are shown in Fig.6 b-c. Fig. 6: Representative TEM and HR-TEM images of TiO2-rGO (a) Low-magnification TEM image of TiO2- rgo (b) and (c) HR-TEM images of TiO2-rGO (d) 3D model of truncated TiO2 particles XPS measurement was performed to determine the elemental quantification in TiO2-N-rGO nanocomposite. Fig.7 shows the core level XPS spectra of C 1s, N 1s, Ti 3d, O 1s and F 1s orbitals of sample. C 1s XPS spectrum in Fig. 7a depicts four peaks centered at bonding energies of , , and ev that are assigned to graphitelike sp 2 C, N sp 2 C, N sp 3 C and satellite peak of π- π * interaction, respectively. The high intensity of the peak at indicates that most of the C atoms are 440 arranged in a conjugated honeycomb lattice. However, N sp 2 C and N sp 3 C bonds are originated from substitution of carbon atoms by nitrogen atoms (Wei, Liu et al. 2009, Wang, Zhang et al. 2011, Dave, Park et al. 2015). Various types of nitrogen functional groups such as pyridinic-n, pyrrolic-n and graphitic N are present in N-doped carbon system (Shao, Zhang et al. 2010, Parambhath, Nagar et al. 2012). These groups are schematically demonstrated in Fig. 7b. The pyridinic groups refer to the N atoms at the edge of graphene planes that make two chemical bonds with carbon atoms and donate one π-electron to the carbon network. The pyrrolic-n atoms are bonded to two carbon atoms and contribute to the π system with two π-electrons (Wu, Santandreu et al., Wei, Liu et al. 2009, Shao, Zhang et al. 2010). XPS results in Fig. 7c represent that our sample contains all these three functional groups; including pyridinic-n (397.9eV, 0.9 at.%), pyrrolic-n ( ev, 2.12 atom%), and graphitic N (401.14eV, 1.44 atom%) (Shao, Zhang et al. 2010, Parambhath, Nagar et al. 2012). In Fig. 7d, Ti2p doublet pair exhibits two peaks centered at ± 0.1 and 459 ± 0.2 ev assigned respectively to the Ti 2p1 and Ti 2p3 spin orbital splitting photoelectrons in the Ti4+ state. The splitting energy between two Tibands was 5.7 ev that is in agreement with the normal state of Ti4+ (Sher Shah, Park et al. 2012, Bai, Zhang et al. 2014). The presence of oxygen is also detected at 530 and 532 ev, as shown in Fig. 7e, which are attributed to bonding energy of O in Ti O C bond (Wang, Wang et al. 2012) and surface adsorbed oxygentio2 (Wang, Li et al. 2015), respectively. Nitrogen doping has been found to improve the surface chemical activity of parent matrix in terms of polarity and basicity (Parambhath, Nagar et al. 2012) Since nitrogen has higher electronegativity compare to carbon, C-N bonds are considerably polarized and create a positive charge on the carbon atom adjusted to nitrogen and it becomes active sites for oxygen adsorption (Wu, Santandreu et al.). As a result, bonding between Ti, O, and C confirms the integration of nanoparticles into graphene sheets. In order to understand the interaction between Ti and F, XPS spectrum of F 1s was presented in Fig. 7f. The measured bonding energy of ev originates from ligand exchange between F and surface hydroxyl groups. This observation is in agreement with the oxidation state of the Ti element (Fig.7c) which is identical to that of Ti in bulk TiO2. As no peak was observed at ev (F in solid solution of TiO2-xFx (Wang, Li et al. 2015)), it can be concluded that diffusion of F species to the interior lattice and subsequently lattice substitution of F - for O -2 was ruled out (Yang, Sun et al. 2008, Yu, Xiang et al. 2010, Wang, Li et al. 2015).