Theoretical analysis of high efficient multi-effect distillation processes and their integration into concentrating solar power plants

RESEARCH WORK SUBMITTED TO THE DEPARTMENT OF ENERGY ENGINEERING, HIGHER TECHNICAL SCHOOL OF ENGINEERING, IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Theoretical Analysis of High Efficient Multi-Effect Distillation Processes and their Integration into Concentrating Solar Power Plants by Bartolomé Ortega Delgado ALMERÍA September 2016 UNIVERSITY OF SEVILLE Department of Energy Engineering In collaboration with SOLAR DESALINATION UNIT CIEMAT - PLATAFORMA SOLAR DE ALMERÍA II III Theoretical Analysis of High Efficient Multi-Effect Distillation Processes and their Integration into Concentrating Solar Power Plants Author: Bartolomé Ortega Delgado Supervisors: DR. DIEGO-CÉSAR ALARCÓN PADILLA SENIOR RESEARCHER AT CIEMAT DR. PATRICIA PALENZUELA ARDILA RESEARCH SCIENTIST AT CIEMAT DR. LOURDES GARCÍA RODRÍGUEZ PROFESSOR AT UNIVERSITY OF SEVILLE UNIVERSITY OF SEVILLE Department of Energy Engineering In collaboration with SOLAR DESALINATION UNIT CIEMAT - PLATAFORMA SOLAR DE ALMERÍA IV V El Dr. Diego-César Alarcón Padilla, Científico Titular de OPIS CIEMAT, la Dra. Patricia Palenzuela Ardila, Investigador Titulado Superior CIEMAT y la Dra. Lourdes García Rodríguez, Catedrática de la Universidad de Sevilla INFORMAN: Que el trabajo descrito en la presente memoria, titulado: “Theoretical Analysis of High Efficient Multi-Effect Distillation Processes and their Integration into Concentrating Solar Power Plants” ha sido realizado por D. Bartolomé Ortega Delgado bajo su dirección. Dicha tesis reúne los requisitos necesarios para su presentación y defensa. Y para que conste y surta los efectos oportunos, firma el presente en Sevilla, a ____ de ________________ de ________. El Director (1) La Directora (2) La Directora (3) y tutora Fdo.:Diego-César Alarcón Padilla Fdo.:Patricia Palenzuela Ardila Fdo.:Lourdes García Rodríguez VI VII Agradecimientos / acknowledgements Me gustaría en primer lugar expresar mi agradecimiento a Julián Blanco por la oportunidad que me ofreció para comenzar este período como estudiante predoctoral en el CIEMAT, a finales del 2012. Quiero agradecer muy especialmente a mis directores de tesis, Lourdes García, Diego Alarcón y Patricia Palenzuela, la completa dedicación, inestimable ayuda y amable generosidad que me han prestado y con la que me han tratado a lo largo de estos inolvidables años. Sus siempre acertados comentarios y oportuna guía han sido fundamentales para la realización de esta memoria. También quiero acordarme de mi excompañera de despacho Alba, por su apoyo y comprensión, con la que tan buenos momentos he compartido; y del resto de integrantes de la Unidad de Desalación Solar, Guillermo Zaragoza, Juan Antonio y Rafa, que siempre me han brindado su valiosa ayuda. Y, cómo no, quiero agradecer a las chicas de Tratamiento Solares del Agua, María, Isa, Margarita, Melina, Samira, Mercedes, Inma, Ana, Irene, Laura, Eli, Sara, Estefanía … por su amistad y por permitirme compartir con ellas tantos buenos ratos. También al resto de compañeros y personal de la PSA, Isa Oller, Pilar, Nacho, Carmen, Merche, Víctor, Luis, Laura, Elena, Asma, Aicha, David, Juanjo, Álex, Diego, Rafa, Lucía, Arantxa, Lidia, Javi… Y aunque no quiero olvidar a nadie, la lista de personas que he conocido durante mi estancia en la PSA es demasiado grande como para plasmarla aquí, por ello quiero mostrar mi mayor agradecimiento hacia todas y cada una de ellas. Y lo más importante, mi familia, por su gran ayuda durante esta etapa. VIII IX A mis padres, José y Carmen, y hermanos, José y María del Carmen “Consider your origin: you were not made to live like brutes, but to follow virtue and knowledge.” Dante Alighieri, Inferno Canto XXVI:85-142 Ulysses’s last voyage Abstract XVI technologies, Reverse Osmosis (RO) and multi-effect distillation. Results obtained show that the bet coupling option, which produce the lower levelised cost of water, is the RO process. Also, it is recommended its indirect integration with the CSP plant, connected directly to the local grid. It is concluded that, in view of the lack of agreement among scientific community about the most suitable technology for integrating with a CSP plant and due to the potential of the combined freshwater and power production with MED and CSP, further investigation with higher efficient MED plants is needed. In this regard, two methods to improve the efficiency of MED processes have been investigated: the increase of the number of effects, which leads to an increase in the Top Brine Temperature (TBT), and its coupling with thermocompressors. The first case has been assessed by using seawater pretreatments that permit to elevate the temperature of the MED process without scale formation, like the nanofiltration membranes. For this purpose, a detailed mathematical model has been developed for a MED plant with forward-feed configuration and the model was implemented within Engineering Equation Solver (EES) software environment. Such feed arrangement has been selected in order to minimize the scale risk on the tubes of the heat exchangers. Results show that the Gain Output Ratio (𝐺𝑂𝑅) is greatly improved (up to a 70%), while the specific heat transfer area and specific energy consumption are significantly reduced (11 and 45%, respectively). Despite of the great potential of this improvement to the MED process, the analysis of its integration into a CSP plant has been not pursued because there are not commercial MED plants using the forward feed scheme. Moreover, the increase of auxiliary consumption attributable to nanofiltration pretreatment may be not suitable for solar applications. Most commercial MED plants present a parallel/cross feed arrangement with thermal vapour compression, which presents several advantages for its coupling with power plants. Therefore, this technology has been selected for the analysis of its coupling with CSP plants. A preliminary evaluation has been performed for a particular case study, simulating the power and water productions of a parabolic trough CSP plant of 50 MWe, with features similar to commercial Andasol-I CSP plant, and a MED-TVC unit of 10,000 m3/d, based on commercial Trapani plant (Italy), during three representative days in winter and summer periods. Two different steam extractions have been considered to feed the MED-TVC unit, one from the high pressure turbine, and other from the low pressure turbine. The CSP model has been taken from the literature and implemented in MATLAB software environment, and the power block model, implemented in EES, has been developed to simulate part load conditions. From this assessment it is concluded that different integration schemes are needed to accomplish for the different profile demands of power and water during the year, in order to promote the power generation or the water production. Later, a parametric study of the integration of a parallel/cross MED-TVC, based on Trapani commercial plant, with a Rankine cycle power block similar to that one of Andasol-1 has been Abstract XVII carried out in order to identify the best coupling arrangement, in terms of efficiency and minimum specific heat transfer area. To that end, a detailed design mathematical model of the MED-TVC unit has been developed and validated against actual data. It is found that the maximum 𝐺𝑂𝑅 and minimum specific area are reached for a particular thermocompressor location, depending on the motive steam pressure fed into the thermocompressor. Also, an operation MED-TVC model has been developed, based on the design model, and used to determine the operational limits of the integration with a Rankine cycle power block that would allow the MED unit to work in nominal conditions (which has been possible by considering variable nozzle thermocompressors), for four different steam extractions of the turbines. For this purpose, the power block has been simulated at different loads, and a control algorithm has been also introduced in order to maintain the maximum brine salinity under 70,000 ppm and the end condenser temperature around its design value (37 °C). Finally, annual simulations of the coupling between a parallel/cross MED-TVC unit and a parabolic trough CSP have been performed, considering Almería (Spain) as the geographical location of the cogeneration plant. The models previously presented for the solar field, power block and desalination unit (nominal and off-design models) have been used. As a particular case study, the daily, monthly and yearly power and water productions have been estimated, using two different steam extractions, equal to those ones identified in the previous analyses: one from the high pressure turbine, at 45.4 bar, and other from the low pressure turbine, at 3.63 bar, which have been used alternatively to feed the MED-TVC unit depending on the monthly power demand for that location. Abstract XVIII XIX Table of contents Chapter 1. Objectives and Justification .................................................................................................... 1 List of figures ....................................................................................................................................... 2 Nomenclature ....................................................................................................................................... 3 1.1 Presentation .............................................................................................................................. 5 1.2 About the author ..................................................................................................................... 12 1.3 Objectives and scope .............................................................................................................. 13 1.4 Main contributions of this work ............................................................................................. 15 1.5 Methodology .......................................................................................................................... 16 1.6 Publishable results .................................................................................................................. 18 References.......................................................................................................................................... 21 Chapter 2. Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors ................................................................................................................................................ 25 List of figures ..................................................................................................................................... 27 List of tables ...................................................................................................................................... 28 Nomenclature ..................................................................................................................................... 29 2.1 Literature review on CSP+D .................................................................................................. 35 2.2 Thermoeconomic comparison of integrating seawater desalination processes in a concentrating solar power plant of 5 MWe ........................................................................................ 46 2.2.1 Concentrating Solar Power and Seawater Reverse Osmosis plants ................................ 48 2.2.2 Integration of a MED plant into the power production ................................................... 56 2.2.3 Comparative analysis of efficiency and production ........................................................ 60 2.2.4 Cost analysis ................................................................................................................... 61 2.2.5 Comparative results ........................................................................................................ 66 2.2.6 Conclusions of the thermoeconomic analysis ................................................................. 70 2.3 Conclusions ............................................................................................................................ 71 Appendix 2-A .................................................................................................................................... 72 Appendix 2-B ..................................................................................................................................... 74 References.......................................................................................................................................... 78 Table of contents XX Chapter 3. Opportunities of improvement of the MED seawater desalination process by pretreatments allowing high temperature operation ...................................................................................................... 83 List of figures ..................................................................................................................................... 84 List of tables ....................................................................................................................................... 86 Nomenclature ..................................................................................................................................... 87 3.1 Introduction ............................................................................................................................ 89 3.2 Forward feed MED model ...................................................................................................... 92 3.2.1 Process description ......................................................................................................... 92 3.2.2 Mathematical model ....................................................................................................... 93 3.2.3 Plant performance ......................................................................................................... 104 3.3 Validation of the FF-MED model and sensitivity analysis ................................................... 105 3.3.1 Validation of the model ................................................................................................ 105 3.3.2 Sensitivity analysis ....................................................................................................... 108 3.4 Analysis of the MED process with high heating steam temperature ..................................... 116 3.5 Conclusions .......................................................................................................................... 118 References ........................................................................................................................................ 120 Chapter 4. Preliminary model of TVC-MED plants coupled to parabolic trough concentrating solar power plants ......................................................................................................................................... 123 List of figures ................................................................................................................................... 126 List of tables ..................................................................................................................................... 127 Nomenclature ................................................................................................................................... 128 4.1 Introduction .......................................................................................................................... 131 4.2 Methodology ........................................................................................................................ 132 4.2.1 Solar field ..................................................................................................................... 132 4.2.2 Multi-effect distillation plant with thermal vapour compression .................................. 135 4.2.3 Power block .................................................................................................................. 137 4.3 Results .................................................................................................................................. 140 4.4 Conclusions .......................................................................................................................... 148 Appendix 4-A................................................................................................................................... 149 Appendix 4-B ................................................................................................................................... 151 References ........................................................................................................................................ 154 Table of contents XXI Chapter 5. Modelling of MED-TVC plants: parametric analysis ......................................................... 155 List of figures ................................................................................................................................... 157 List of tables .................................................................................................................................... 158 Nomenclature ................................................................................................................................... 159 5.1 Thermocompressor models ................................................................................................... 163 5.1.1 Literature review .......................................................................................................... 163 5.1.2 Models comparison....................................................................................................... 170 5.2 Introduction to the parametric analysis ................................................................................. 174 5.3 Process description ............................................................................................................... 176 5.4 Methodology ........................................................................................................................ 179 5.4.1 Mathematical model ..................................................................................................... 179 5.4.2 Plant performance ......................................................................................................... 189 5.4.3 Validation of the mathematical model .......................................................................... 190 5.5 Parametric study ................................................................................................................... 192 5.6 Results and discussion .......................................................................................................... 194 5.6.1 Base case ...................................................................................................................... 194 5.6.2 Parametric results ......................................................................................................... 195 5.7 Conclusions .......................................................................................................................... 202 Appendix 5-A .................................................................................................................................. 204 References........................................................................................................................................ 208 Chapter 6. Operational analysis of the coupling between a MED-TVC unit and a Rankine cycle power block using variable nozzle thermocompressors .................................................................................. 211 List of figures ................................................................................................................................... 212 List of tables .................................................................................................................................... 213 Nomenclature ................................................................................................................................... 214 6.1 Introduction .......................................................................................................................... 217 6.2 Modelling of the system ....................................................................................................... 218 6.2.1 Rankine cycle power block ........................................................................................... 218 6.2.2 Multi-effect distillation with thermal vapour compression unit .................................... 221 6.3 Results .................................................................................................................................. 230 6.4 Conclusions .......................................................................................................................... 236 References........................................................................................................................................ 238 Table of contents XXII Chapter 7. Yearly simulations of the water and power productions in CSP+D plants .......................... 241 List of figures ................................................................................................................................... 242 List of tables ..................................................................................................................................... 245 Nomenclature ................................................................................................................................... 246 7.1 Introduction .......................................................................................................................... 247 7.2 Solar Field ............................................................................................................................ 247 7.2.1 Characteristics of the solar field.................................................................................... 247 7.2.2 Operation strategy ......................................................................................................... 250 7.3 Power block .......................................................................................................................... 252 7.4 Desalination unit ................................................................................................................... 252 7.5 Yearly simulations ................................................................................................................ 257 7.5.1 Methodology ................................................................................................................. 257 7.5.2 Solar energy resource quantification ............................................................................. 258 7.5.3 Yearly estimation of the power generation and fresh water production of the CSP+D plant ......................................................................................................................... 262 7.5.4 Daily simulations for representative months on summer and winter............................. 264 7.6 Conclusions .......................................................................................................................... 290 References ........................................................................................................................................ 291 Chapter 8. Conclusions and future works ............................................................................................. 293 Nomenclature ................................................................................................................................... 294 8.1 Conclusions .......................................................................................................................... 295 8.2 Future works ......................................................................................................................... 298 Chapter 1. Objectives and Justification Contents Chapter 1. Objectives and Justification ........................................................................................ 1 List of figures ........................................................................................................................... 2 Nomenclature ........................................................................................................................... 3 1.1 Presentation .................................................................................................................... 5 1.2 About the author .......................................................................................................... 12 1.3 Objectives and scope ................................................................................................... 13 1.4 Main contributions of this work ................................................................................... 15 1.5 Methodology ................................................................................................................ 16 1.6 Publishable results ....................................................................................................... 18 References .............................................................................................................................. 21 Chapter 1 Introduction Page 2 List of figures Figure 1.1. Areas with physical or economic water scarcity (IWMI, 2007). .............................. 5 Figure 1.2. Global water demand by utilization, 2000 and 2050 (OECD, 2012). ...................... 6 Figure 1.3. The Aral Sea in 1989 (left), and in 2014 (Lindsey, 2014). ....................................... 7 Figure 1.4. Yearly sum of direct normal irradiation global map (Meteonorm, 2015). ............... 9 Figure 1.5. Annual additional and cumulative desalination capacity, 1970 – 2014 (Global Water Intelligence, 2016)................................................................................................... 10 Figure 1.6. Main desalination technologies (adapted from (Li et al., 2013)). .......................... 11 Figure 1.7. Total worldwide installed desalination capacity by technology (Global Water Intelligence, 2016). ............................................................................................................ 11 Doctoral dissertation Bartolomé Ortega Delgado Page 3 Nomenclature Acronyms and abbreviations ABHP Absorption Heat Pump ADHP Adsorption Heat Pump BRIICS Brazil, Russia, India, Indonesia, China and South Africa CD Capacitive Deionization CR Central Receiver CSP Concentrating Solar Power ED Electrodialysis EDS European Desalination Society EES Engineering Equation Solver FF Forward Feed FM Freezing-Melting FO Forward Osmosis GOR Gain Output Ratio HDH Humidification-Dehumidification HTF Heat Transfer Fluid IE Ion Exchange IWMI International Water Management Institute LT Low Temperature MD Membrane Distillation MED Multi-Effect Distillation MIGD Million Imperial Gallons per Day MSF Multi-Stage Flash MVC Mechanical Vapour Compression OECD Organization for Economic Co-operation and Development PF Parallel Feed PTC Parabolic Trough Collectors PVD Passive Vacuum Desalination RO Reverse Osmosis SEGS Solar Electric Generation Station ST Solar Still STE+D Solar Thermal Electricity and Desalination TVC Thermal Vapour Compression Chapter 1 Introduction Page 10 been widely used in the food and chemical industry as a method for producing sugar and other products, since 19th century (El-Dessouky and Ettouney, 2002). Later, during 20th century this process was first applied for seawater desalination. One of the first MED desalination plants was built in Jeddah (Saudi Arabia) in 1907, comprised of two distillation units, and using submerged tubes technology, which led to significant problems related with the appearance of scaling in the tubes of the heat exchangers. Shortly afterwards they were replaced for two new units, using the same submerged tubes technology, with a total capacity of 135 m3/d. Other plants were installed but, the low productivity, maintenance stops, and low thermal efficiency led to look for other desalination alternatives, particularly Multi-Stage Flash (MSF) evaporation. That was one of the main reasons of the MSF desalination process growth, particularly in the Middle East countries. It was not until the 80s when new MED designs using high efficient falling-film evaporators and low operation temperatures (< 70 °C) re-activated the technology for desalination purposes, once the scale formation problems were controlled (Buros et al., 2000). The desalination market has experienced a big expansion in the last years, as it can be seen in Figure 1.5, where it has been represented the additional and cumulative annual desalination capacity worldwide, from 1970 to 2014. The global economic crisis surged in EEUU during 2008 affected to the growing tendency for new desalination plants, mainly due to financing and funds raising difficulties. Figure 1.5. Annual additional and cumulative desalination capacity, 1970 – 2014 (Global Water Intelligence, 2016). Currently, there are several desalination technologies and different classifications. Taking into account the element extracted from the process, water or salt, one categorization possible is the one illustrated in Figure 1.6. Within the processes based on water removal, two new categories can be proposed, with or without phase change. In the former group are Multi-Stage Flash, Multi-Effect Distillation, Solar Still (ST), Humidification-Dehumidification (HDH), Passive Doctoral dissertation Bartolomé Ortega Delgado Page 11 Vacuum Desalination (PVD), Membrane Distillation (MD), Freezing-Melting (FM) and the techniques based on heat pumps: Mechanical Vapour Compression (MVC), Thermal Vapour Compression (TVC), Absorption Heat Pump (ABHP) and Adsorption Heat Pump (ADHP). In the second group, without phase change, are Reverse Osmosis (RO) and Forward Osmosis (FO). In the salt-collecting processes are found Electrodyalisis (ED), Ion Exchange (IE) and Capacitive Deionization (CD), which are habitually used with brackish water. Figure 1.6. Main desalination technologies (adapted from (Li et al., 2013)). Despite of the significant number of desalination processes available, not all are suitable for handling seawater desalination and large capacities. What is more, some are in research and experimental stages. The most common commercial desalination techniques, with the largest installed capacities, are, in this order (see Figure 1.7): reverse osmosis (63%), multi-stage flash (23%), multi-effect distillation (8%) and electrodialysis (3%) (Global Water Intelligence, 2016). Figure 1.7. Total worldwide installed desalination capacity by technology (Global Water Intelligence, 2016). Desalination processes Collect water Collect salt Phase change Single change MED,MSF,MVC, TVC,ST,MD, PVD,HDH,FM, ABHP,ADHP RO, FO ED,CDI,IE Chapter 1 Introduction Page 12 At big scale, multi-effect distillation for seawater desalination is the most efficient technology for producing fresh water, among the thermal methods (Darwish and El-Dessouky, 1996), and RO the most used among the mechanical methods. Moreover, it is especially suitable for cogeneration purposes (combined electricity and freshwater production) as only need steam at 70 ºC to operate, which could be obtained from process steam of any industrial application or from any steam extraction of the turbines of a power plant. The commercially available MED plants use the thermal vapour compression to increase their efficiency, by means of steam ejectors (also known as thermocompressors). These devices are very simple but at the same time are cheap, robust and easy to operate. They are based on the Venturi effect and basically compress low pressure vapour extracted from one effect of the MED unit, using high pressure vapour (motive steam) from a boiler or other external source, up to an intermediate pressure between the two inlets (compressed vapour). Single MED-TVC units have capacities up to 36,000 m3/d (8 MIGD), like the plant at Layyah Power Station in United Arab Emirates, although the total daily capacity can be increased by the simultaneous operation of several units. For instance, Taweelah Power and Desalination Plant (Abu Dhabi - UAE) has 14 units of 17,143 m3/d each, reaching a total capacity of 240,000 m3/d (Veolia, 2016). The largest SWRO plant can handle up to 540,000 m3/d (Sorek plant in Israel), due to the modularity of this kind of technology. The specific heat and electric consumption of the MED-TVC process are between 50100 kWhth/m3 and 12 kWhe/m3, respectively, while the SWRO process have an electric consumption of about 3.95.6 kWhe/m3 (Khayet, 2013; Moser et al., 2011; Zak et al., 2013). 1.2 About the author The author of this research work, with an Industrial Engineering degree and Energy specialization obtained at the University of Seville (Spain), was awarded with a scholarship by CIEMAT research centre to complete a PhD program within the Solar Desalination Unit at Plataforma Solar of Almería, from 1st December 2012 to 30th November 2016, in the framework of the National Plan for Scientific Research, Development and Technological Innovation (“National R+D+I Plan” 2011), dependent on the Ministry of Science and Innovation of Spain. The topic of doctorate program was the integration between concentrating solar power plants and desalination plants, particularly using parabolic trough solar thermal technology and multi-effect distillation with thermal vapour compression process, for the production of electricity and water, respectively. The first year of the scholarship was dedicated in part to complete a Master Thesis in Thermal Energy Systems at University of Seville, as a previous requirement for joining the Doctorate program in Energy Engineering, Chemical and Environmental, which was done on 7th March Doctoral dissertation Bartolomé Ortega Delgado Page 13 2013. The Final Project of the Master consisted in developing a steady-state mathematical model for a forward feed multi-effect distillation plant with a sensitivity analysis of different designs and performance variables. Several papers resulting from his research activity at the Plataforma Solar de Almería have been submitted to conferences and journals, which are described in detail in Section 1.6. Also, the author has participated with oral presentations in two international conferences organised by the European Desalination Society (EDS) in Palermo and Rome (Italy) during 2015 and 2016. Oral presentations at SOLLAB in Germany, France and Switzerland during 2013, 2014 and 2015, respectively, have been done as well by the author. This work has been developed under the framework of a European project, STAGE-STE, within Work Package (WP) 10, focused on the investigation of sustainable methods and technologies for the combined power and water production using concentrating solar energy (EERA, 2015). In addition, the author has collaborated within the Solar Desalination Unit for different works related to the STAGE-STE project. Particularly, in this WP, studies focused in the modelling and evaluation of different systems or processes have been carried out: multieffect distillation plants, Rankine cycle power plants, parabolic trough solar fields, and steam jet ejectors. 1.3 Objectives and scope The main objective of this work is the theoretical evaluation of high efficient multi-effect distillation processes and their combination with parabolic trough CSP plants. In particular, this research work investigates the best coupling arrangement between the desalination unit and the CSP plant, which minimizes the energy consumption of the MED-TVC process and meet the water and power demand, both daily and seasonally, for the location considered. This study makes sense only in arid or semi-arid regions of the world that present high irradiation levels and have access to the sea. This research work is motivated as a continuation of the Thesis written by Patricia Palenzuela (2012), focusing in the MED-TVC technology and the use of thermocompressors for the coupling with Rankine cycle power blocks, as suggested in Palenzuela et al. (2011). Particularly, the study of the variable nozzle thermocompressors was analysed, due to their ability to maintain the MED efficiency near to the nominal even when the power plant is working at part load operation and the pressure level of the feeding vapour is reduced. The operational limits of the coupling between the two subsystems, solar thermal power plant and MED-TVC unit, working at part load operation, were thoroughly assessed and discussed in the present research work. Chapter 1 Introduction Page 14 The specific objectives and goals set out in the present research work are: 1. The development of steady-state mathematical models of high efficient MED processes. Different feed arrangements were considered: forward feed and parallel/cross. Moreover, possible enhancements for the increase of the thermal efficiency of the MED process have been analysed, by raising the number of effects and top brine temperature using pretreatments such as nanofiltration membranes, and introducing thermocompression. 2. To implement a detailed model of a parabolic trough solar field based on a commercial PT-CSP plant (Andasol-2, Spain) configuration in order to simulate the yearly thermal energy production with the highest accuracy possible. 3. To develop a mathematical model for a Rankine cycle power block, similar to those existing in parabolic trough solar thermal power commercial plants, working at nominal and part load operation. 4. To determine the position of the thermocompressor in a MED-TVC unit that produces the highest thermal efficiency of the desalination plant with the minimum specific heat transfer area. 5. To investigate the different coupling arrangements between a multi-effect distillation unit with thermal vapour compression and a Rankine cycle power block similar to that of a commercial PT-CSP plant (Andasol-1, Spain), in terms of freshwater and power production. 6. To identify the operational limits of the coupling between a Rankine cycle power block similar to that of a commercial PT-CSP plant (Andasol-1, Spain) that allow to keep the motive steam mass flow rate constant using variable nozzle thermocompressors. 7. To develop a partial-load model for the MED-TVC process to be integrated into a Rankine power block that considers the control of the operation by the adjustment of key variables such as the maximum brine salinity and the condensation temperature. 8. To develop a mathematical model for integrated CSP+MED-TVC plant as a tool to simulate throughout the year the power and fresh water production under different freshwater and power profile demands. Doctoral dissertation Bartolomé Ortega Delgado Page 15 1.4 Main contributions of this work The number of works dedicated to perform quasi-dynamic analyses of the integration of multieffect distillation processes for seawater desalination within concentrating solar power plants is scarce. The majority of the studies published in the literature either have been developed in steady-state conditions for a particular moment of the year, or assuming a 1-hour time step to simulate the yearly production of power and water (Casimiro et al., 2014; Ghobeity et al., 2011; Olwig et al., 2012), which makes the annual simulations less accurate because of the large difference between the time step and the solar irradiation variability. Other works use commercial simulation software with no access to governing physical equations (Iaquaniello et al., 2014; Wellmann et al., 2015). In the present research work, 10-s time step has been selected for the solar field performance simulation, which uses an accurate mathematical model developed by Llorente García et al. (2011) and validated against actual data from a commercial CSP plant. The simulations of the yearly power and water production have been done for 10-min time steps in order to save computational time, preserving a reasonable time difference to account for the instant variation of the solar irradiance. Similarly, there is a lack of MED-TVC models in the literature able to simulate part load conditions, which is essential to perform time-dependent analyses of the coupling with CSP plants. The time-dependent MED models published in the literature (Aly and Marwan, 1997; Dardour et al., 2005; de la Calle et al., 2015; El-Nashar and Qamhiyeh, 1990) have as the main objective to develop control schemes that regulate the response of the main operational variables against certain disturbances, which are not suitable for the yearly simulation of complex systems such as CSP+MED plants. On the other hand, most of steady-state MED models found are used for design purposes, where the geometry of the plant is determined by the model (Al-Mutaz and Wazeer, 2014; Bin Amer, 2009; El-Dessouky et al., 2000; Kamali et al., 2009), and they are not either suitable to simulate the partial load operation. In this research work operation models are developed in order to perform the simulations of the integration of a fixed MED-TVC unit with a CSP plant under part load operation. In addition, the developed MED-TVC models try to investigate more in detail some phenomena occurring inside the multi-effect distillation process, such as the saturation temperature losses of the vapour from its generation in one effect up to its condensation inside the evaporator of the next effect, in contrast with most of the models reported where this effect is either neglected or assumed constant. Also, other significant achievements of this work are:  Identification of misunderstandings in some published models of MED units. Chapter 1 Introduction Page 16  Thorough parametric study on thermal efficiency of MED-TVC plants with determination of the optimal reduction of heat transfer areas of the evaporators after the thermocompressor location that minimizes the specific heat transfer area.  Exhaustive parametric study on the operation of MED-TVC with the development of a specific control algorithm for the regulation of the maximum brine salinity produced and the condensation temperature. 1.5 Methodology In order to meet the mentioned objectives, this research work has been structured as follows: In Chapter 1, the justification and motivation of this research work is presented. Moreover, the main objectives and goals are described and the major contributions of this work in the scientific community are discussed. Finally, the publications derived from this work are shown. In the second chapter, a literature review on concentrating solar power and desalination is firstly performed in order to set the state of the technology. Later, a thermoeconomic comparison of the integration of a desalination process into a concentrating solar power plant is shown, considering the leading seawater desalination technology in terms of capacity, reverse osmosis, and the most efficient one among the thermal methods, multi-effect distillation. As a result of this study and from the extensive literature review performed, it is concluded that there is not a clear agreement among scientific community about the suitability of the RO process compared with MED technology for its coupling with a CSP plant. Also, it is suggested that further investigation should be carried out regarding the coupling of MED-TVC units with CSP plants, due to the fact that this configuration permits to de-couple the water and power productions (in contrast to LT-MED+CSP scheme) and meet the profile demands during the year. In Chapter 3, a highly efficient MED process is evaluated in order to investigate new methods for reducing the energy consumption and for improving the competitiveness of this technology against RO. The enhancement in the efficiency consists on the rise in the number of effects by increasing the top brine temperature, using specific seawater pretreatments such as nanofiltration membranes. In this case, forward feed configuration has been selected because, unlike the parallel feed, the maximum salinity takes place in the last effect, which has the lower temperature and therefore minimizes the scaling risks. The chapter also presents a detailed mathematical model for the FF-MED and a complete parametric study on the different design and operational parameters. Despite of the increase of the 𝐺𝑂𝑅 and reduction of the energy consumption observed by elevating the heating steam temperature and number of effects, the Doctoral dissertation Bartolomé Ortega Delgado Page 17 forward feed scheme is not commercially available yet and the model cannot be validated against actual data. Most of commercial plants are MED with thermal vapour compression and they are based on parallel feed configuration because their easier construction and operation. Therefore, next chapters are focused in developing and further improving the MED-TVC process in parallel feed configuration for its integration in CSP plants. A first approach to the investigation on the coupling between CSP plants and MED-TVC units is performed in Chapter 4. Particularly, a parabolic trough solar thermal power plant of 50 MWe with features similar to Andasol-1 is integrated with a MED-TVC unit of 10,000 m3/d based on the commercial Trapani plant (Italy). An off-design model of the power block is implemented in EES, while the solar field model, taken from the literature (Llorente García et al., 2011), is implemented in MATLAB. A design MED-TVC model, which is explained in detail in Chapter 5, is used to identify the optimum thermocompressor position. Two coupling arrangements are considered, using a high/medium pressure steam extraction (20.6 bar) and a low pressure steam extraction (1.224 bar) to feed the MED-TVC unit. Simulations during three representative days in winter and three days in summer are performed in order to compare the water and power productions and identify the best coupling scheme depending on the demands curves. For these simulations, a black-box model of the MED-TVC unit is used. From this study it is found that different coupling schemes should be considered depending on the period of the year and environmental conditions. Chapter 5 presents a parametric study of a Parallel Feed (PF) MED-TVC unit integrated with a Rankine cycle power block similar to that one in a commercial PT-CSP plant (Andasol-1, Spain), so as to identify the best coupling arrangement in terms of maximum thermal efficiency and minimum specific heat transfer area. A comprehensive design model for the PF-MEDTVC unit is developed and validated against actual data from the commercial plant located in Trapani (Italy). Four different steam extractions from the power block are selected in order to feed the MED-TVC unit, and the thermocompressor location is varied to investigate the position such as the minimum specific heat transfer area and maximum 𝐺𝑂𝑅 are obtained. As a result, four different and optimized coupling arrangements between the MED-TVC unit and the CSP plant are achieved. After the identification of the best coupling schemes between the CSP plant and the MED-TVC unit in nominal conditions, the operational limits of both the power block and the desalination plant are studied in Chapter 6. To that end, the power block is simulated at different loads to find the limits until which the MED-TVC can operate at nominal conditions. Also, an operation model for the MED-TVC unit is developed based on the former design model, and the simulation of the integrated plant (desalination unit into the Rankine power block) is performed for different coupling arrangements and loads of the power block. Chapter 1 Introduction Page 18 In Chapter 7, as an application of the simulation tools developed beforehand in the previous chapters, annual simulations of the water and power productions for the integrated PTCSP+MED-TVC plant using the off-design models are performed, taking Almería (Spain) as the location of the plant. Two of the proposed coupling arrangements are considered: one from the high pressure turbine at 45.4 bar and another from the low pressure turbine at 3.627 bar, which are used alternatively depending on the power demand profile for Andalusia (Spain) in 2015. Finally, the main conclusions reached during the research work are shown along with future works. 1.6 Publishable results The author has published and/or submitted for publication the following articles as a result of the work developed in the present research work: 1. Ortega-Delgado B., Palenzuela, P., Alarcón-Padilla, D.C., and García-Rodríguez, L. “Quasi-steady state simulations of thermal vapour compression multi-effect distillation plants coupled to parabolic trough solar thermal power plants”, Desalin. Water Treat. (2016) 1–12. doi:10.1080/19443994.2016.1173377 [In press]. Presented at EuroMed 2015: Desalination for Clean Water and Energy, Palermo, Italy, 10–14 May 2015. Organized by the European Desalination Society. 2. Ortega-Delgado B., García-Rodríguez, L., and Alarcón-Padilla, D.C., “Thermoeconomic comparison of integrating seawater desalination processes in a concentrating solar power plant of 5 MWe”, Desalination 392 (2016) 102–117. doi:10.1016/j.desal.2016.03.016. 3. Ortega-Delgado, B., Palenzuela, P., Alarcón Padilla, D.-C., “Parametric study of a multi-effect distillation plant with thermal vapor compression for its integration into a Rankine cycle power block”, Desalination. 394 (2016) 18–29. doi:10.1016/j.desal.2016.04.020. 4. Ortega-Delgado B., Cornali, M., Palenzuela, P., and Alarcón-Padilla, D.C., “Operational analysis of the coupling between a MED-TVC unit and a Rankine cycle power block using variable nozzle thermocompressors”. (To be submitted) (Presented at the EDS Conference, Rome (Italy), May 22-26, 2016.). Doctoral dissertation Bartolomé Ortega Delgado Page 19 5. Ortega-Delgado B., García-Rodríguez, L., and Alarcón-Padilla, D.C., “Opportunities of improvement of the MED seawater desalination process by pretreatments allowing high temperature operation”, (To be submitted). The author has also collaborated in the following works: 1. Palenzuela, P., Ortega-Delgado, B., Alarcón-Padilla, D.C., and Zaragoza, G., "Regeneration of concentrated solutions by multi-effect distillation in a RED close loop process". Presented at EuroMed 2016: Desalination for Clean Water and Energy, Rome, Italy, 22–26 May 2016. Organized by the European Desalination Society. 2. “Project deliverable 3.1: Low temperature regeneration assessment. Performance potentials of MED and MD processes using artificial aqueous solutions and effect of the addition of non-aqueous solvents”. Project: “Conversion of Low Grade Heat to Power through closed-loop Reverse Electro-Dialysis”. European Union’s Horizon 2020 research and innovation programme. 3. “Project deliverable 10.1: State of the art and model development of multi-effect distillation with thermocompression” STAGE-STE Task 10. Other studies carried out and presented by the doctorate at international conferences are listed below: 1. Ortega-Delgado B., Alarcón Padilla, D.C., and Blanco, J., “Modelling and optimization of integrated parabolic trough solar power and multi-effect desalination plants”, presented at 9th SOLLAB Conference in Hornberg Castle, Black Forest, Germany (France), May 13-14, 2013. 2. Ortega-Delgado B., Alarcón Padilla, D.C., García-Rodríguez, L., Zaragoza, G., and Blanco, J., “Analysis of the step time influence in the yearly simulation of dual purpose solar thermal concentrating plants”, Submitted to EuroMed 2014: Desalination for Clean Water and Energy, Limassol, Cyprus, 11–15 May 2014. Organized by the European Desalination Society. 3. Ortega-Delgado B., Palenzuela, P., Alarcón-Padilla, D.C., and García-Rodríguez, L., “Modelling and parametric study of thermal vapor compression multi-effect distillation plants”, presented at 10th SOLLAB Conference in Odeillo (France), June 23-25, 2014. 4. Ortega-Delgado B., Palenzuela, P., Alarcón-Padilla, D.C., and García-Rodríguez, L., “Analysis of the coupling between multi-effect distillation plants with thermal vapour Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 26 Contents Chapter 2. Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors ......................................................................................................................... 25 List of figures ......................................................................................................................... 27 List of tables ........................................................................................................................... 28 Nomenclature ......................................................................................................................... 29 Literature review on CSP+D ....................................................................................... 35 2.1 Thermoeconomic comparison of integrating seawater desalination processes in a 2.2 concentrating solar power plant of 5 MWe ............................................................................ 46 2.2.1 Concentrating Solar Power and Seawater Reverse Osmosis plants ..................... 48 2.2.2 Integration of a MED plant into the power production ........................................ 56 2.2.3 Comparative analysis of efficiency and production ............................................. 60 2.2.4 Cost analysis ........................................................................................................ 61 2.2.5 Comparative results ............................................................................................. 66 2.2.6 Conclusions of the thermoeconomic analysis ...................................................... 70 Conclusions ................................................................................................................. 71 2.3 Appendix 2-A ......................................................................................................................... 72 Appendix 2-B ......................................................................................................................... 74 References .............................................................................................................................. 78 Doctoral dissertation Bartolomé Ortega Delgado Page 27 List of figures Figure 2.1. Layout of the power cycle and thermodynamic properties of mass flows. ............. 51 Figure 2.2. Basic scheme of a SWRO plant. ............................................................................. 53 Figure 2.3. Basic layout of the productive core of a SWRO plant. ........................................... 53 Figure 2.4. Layout of the MED1 configuration: a MED plant replacing the CSP plant cooling system (only functionally). ......................................................................................................... 57 Figure 2.5. Basic layout of the MED2 configuration: a MED plant integrated within the CSP plant fed by an extraction. .......................................................................................................... 58 Figure 2.6. Analysis of the condensing thermal power of the Rankine cycle (𝑄𝑐𝑜𝑛𝑑), thermal power consumption in the MED (𝑄𝑀𝐸𝐷) and mass flow rate of exhaust steam (𝑞28) as function of the fresh water production (𝑞33) for MED2 configuration. .................................... 58 Figure 2.7. Effect of the fresh water production (𝑞33) on the thermal performance of the power cycle (𝜂𝑡ℎ), specific thermal energy added by the solar field (𝑄𝑎), specific work extracted in the low and high pressure turbines, (𝑊𝐿𝑃 and 𝑊𝐻𝑃, respectively) and specific work consumed by the condensing and feeding pumps (𝑊𝐶𝑃 and 𝑊𝐹𝑃, respectively) for MED2 configuration. .................................................................................................................. 59 Figure 2.8. Analysis of the extractions from the low pressure steam turbine (𝛼𝐴,𝛼𝐵,𝛼𝐶,𝛼𝐷 and total 𝛼𝑠𝑢𝑚) for different fresh water productions (𝑞33) in MED2 configuration. ............. 59 Figure 2.9. Block diagram of the material, thermal and mechanical interactions between the different subsystems in the MED1 and MED2 cases. ................................................................ 65 Figure 2.10. Block diagram of the material, thermal and mechanical interactions between the different subsystems in the RO1 and RO2 cases. ....................................................................... 65 Figure 2.11. Levelised electricity and water costs in the four cases considered. ...................... 67 Figure 2.12. Levelized water cost as function of the water flow rate. ....................................... 68 Figure 2.13. Sensitivity analysis of the LWC and LEC with the specific costs of solar field (a,b), MED (c), RO (d), the discount rate (e,f) and the capacity factor (g,h). ............................ 70 Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 28 List of tables Table 2.1. Some representative PT-CSP plants (NREL, 2016). ............................................... 48 Table 2.2. Design parameters (Zarza et al., 2006). ................................................................... 50 Table 2.3. Performance of individual items in which is based the configuration given in Figure 2.1 (Blanco-Marigorta et al., 2011). .......................................................................................... 50 Table 2.4. Main parameters of configuration given in Figure 2.1. ........................................... 52 Table 2.5. Comparison of selected configuration (regenerative Rankine cycle with reheating) with more simple Rankine cycles with the same maximum and minimum temperature and pressure values. .......................................................................................................................... 52 Table 2.6. Thermodynamic properties of the streams and design data for SWRO plant (Peñate and García-Rodríguez, 2012). .................................................................................................... 54 Table 2.7. Exergetic power and mass flow rates corresponding to different SWRO plant productions in the RO1 case. ..................................................................................................... 55 Table 2.8. Power and mass flow rates corresponding to different SWRO plant productions in the RO2 case. ............................................................................................................................. 56 Table 2.9. Main operational parameters in MED1 and MED2 configurations respect to the base case. .................................................................................................................................... 60 Table 2.10. Summary of results related to the four configurations analysed. .......................... 61 Table 2.11. Economic balances. ................................................................................................ 66 Appendix 2-B Table 2-B.1. Input data to the economic analysis. .................................................................... 75 Table 2-B.2. Summary of results related to the MED1 and MED2 configurations analysed in section 2.2.2. .............................................................................................................................. 76 Table 2-B.3. Summary of results related to the RO1 and RO2 configurations analysed in section 2.2.1. .............................................................................................................................. 77 Table 2-B.4. LEC and LWC for the four cases considered. ...................................................... 77 Doctoral dissertation Bartolomé Ortega Delgado Page 29 Nomenclature Variables 𝐴𝑆𝐹 Solar field aperture area, m2 𝐴𝑐 Solar collector aperture area, m2 𝐴 Levelised value of a quantity of money 𝑏 Molality, mol/kg 𝑐 Exergetic unit cost, €/kWh 𝐶󰇗 Cost rate, €/h 𝐶𝑅𝐹 Capital recovery factor 𝑒𝑥 Specific exergy, kJ/kg 𝐸1 Annual energy or output type 1 produced, kWh 𝐸𝑤 Yearly production of fresh water, m3/y 𝐸󰇗 Exergy rate, kW 𝐸󰇗𝑟𝑎𝑑 Exergy rate associated with the direct normal irradiance, kW 𝐸2𝑉𝑒2 Discounted sum of other revenues, € 𝑓 Conversion factor ℎ Specific enthalpy, kJ/kg 𝐻𝑅 Heat rate of the power block, kJ/kWh 𝐼𝑑 Direct normal irradiance at the design day, kw/m2 𝑘𝑛 Nominal discount rate or nominal cost of capital 𝐿𝑒 Discounted sum of input energy expenses, € 𝑀 Discounted sum of operating expenses, € 𝑚𝑑 Daily production of freshwater, m3/d 𝑁 Lifetime project, y 𝑁𝑐 Number of collectors 𝑛ℎ Hours of daily operation, h/d 𝑝 Pressure, bar 𝑃𝐼 Discounted investment costs, € 𝑃𝑂𝑀 Discounted O&M expenses, € 𝑃𝐹 Discounted fuel expenses, € 𝑃0 Value of an expenditure at the beginning of the first year, € 𝑄𝑎 Heat per unit of mass added to the power cycle, kJ/kg 𝑄 Heat per unit of mass, kJ/kg 𝑞 Mass flow rate, kg/s 𝑄󰇗𝑀𝐸𝐷 Heat rate added to the MED, MW 𝑄󰇗𝐶𝑂𝑁𝐷 Heat rate consumption in the condenser, MW Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 30 𝑄󰇗𝑠𝑢𝑛 Heat rate of the direct normal irradiance in the solar field, kW 𝑅 Discounted sum of replacement costs, € 𝑟𝑖 Annual inflation rate 𝑟𝑛 Nominal escalation rate 𝑠 Specific entropy, kJ/kg 𝑆𝑐 Collector aperture area, m2 𝑇 Temperature, °C 𝑇𝑜𝑢𝑡 Temperature of the heat transfer fluid in the outlet of the solar field, °C 𝑉𝑒2 Value of energy type 2 produced in year zero 𝑊𝑇 Specific total work extracted in the turbines, kJ/kg 𝑊𝑃 Specific total work consumed by the pumps in the power cycle, kJ/kg 𝑊󰇗𝑒 Net electric power, MW 𝑊𝑒,𝑛 Net electric energy, kwh/y 𝑊 Specific work, kJ/kg 𝑊𝑚𝑎𝑖𝑛 𝑅𝑂 Specific energy consumption of the main pumps, kwh/m3 𝑊𝑎𝑢𝑥 𝑅𝑂 Specific energy consumption of the booster pumps, kwh/m3 𝑥 Title of steam 𝑌 Specific cost 𝑍󰇗 Cost rate associated to capital investment and O&M, €/h Acronyms and abbreviations CFD Computational Fluid Dynamics COND Condenser CP Condensate Pump CPV Concentrator Photovoltaic CSP Concentrating Solar Power CSPonD Concentrating Solar Power on Demand DCF Discounted Cash Flow ED Electrodialysis EERA European Energy Research Alliance EV Evaporator FF Forward Feed FP Feeding Pump G Electric Generator GOR Gain Output Ratio HP High Pressure Turbine HR Heat Rate Doctoral dissertation Bartolomé Ortega Delgado Page 31 HTF Heat Transfer Fluid I+P Intake plus Pretreatment ISCC Integrated Solar Combined Cycle LEC Levelised Electricity Cost LWC Levelised Water Cost LP Low Pressure Turbine LT-MED Low Temperature Multi-Effect Distillation LWC Levelised Water Cost MD Membrane Distillation MED Multi-Effect Distillation MENA Middle East and North Africa MEH Multi-Effect Humidification MIT Massachusetts Institute of Technology MSF Multi-Stage Flash MVC Mechanical Vapour Compressor O&M Operation and Maintenance ORC Organic Rankine Cycle OT Once-Through PB Power Block PCM Phase Change Material PF Parallel Feed PH Preheater POWERSOL Mechanical POWER generation based on SOLar heat engines PT Parabolic Trough PV Photovoltaic RH Reheater RO Reverse Osmosis SEC Specific Energy Consumption SF Solar Field SH Superheater STPP Solar Thermal Power Plant SWRO Seawater Reverse Osmosis TBT Top Brine Temperature TDS Total Dissolved Solids TES Thermal Energy Storage TVC Thermal Vapour Compression Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 32 Subscripts a Heat added to the power block amb Ambient ap Apparent aux Auxiliary c Condensation d Daily production of fresh water e Electric F Fuel g Global I Investment n Net p Pump P Product out Outlet t Turbine th Thermal u Useful v Vapour w Water 0 First year 1 Energy type one 2 Energy type two Superscripts CI Capital Investment OM Operation and Maintenance RK Rankine RO Reverse Osmosis Greek 𝛼𝑖 Mass fraction of steam extracted from the bleed i 𝜂𝑔 Global efficiency 𝜂𝑏 𝑅𝑂 Internal efficiency of the booster pumps in the RO plant 𝜂𝑚 Mechanical efficiency 𝜂𝐻𝑃 𝑅𝑂 Internal efficiency of the high pressure pumps in the RO plant 𝜂𝑡ℎ Thermal efficiency Doctoral dissertation Bartolomé Ortega Delgado Page 33 𝜂𝑠ℎ Shaft efficiency 𝜂𝑋 Exergetic efficiency 𝜃 Incidence angle of the solar rays 𝜌 Density, kg/m3 𝜏 Annual average availability of the plant, h/y Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 34 Doctoral dissertation Bartolomé Ortega Delgado Page 35 Literature review on CSP+D 2.1 The literature regarding desalination processes powered by concentrating solar power (Concentrated Solar Power and Desalination, CSP+D) is very limited, despite the great interest aroused in the last years. The different approaches and studies that have been presented regarding this concept during the last decade are briefly reviewed hereafter. It is well known that seawater desalination is an energy-intensive consumer process, especially in the case of thermal methods. In this respect, solar irradiation may provide the energy necessary to drive such processes, particularly in arid regions of the world with high levels of annual solar irradiation, which normally lack from fresh water sources. Solar irradiation may be used in low temperature desalination processes without concentration, by means of flat plate collectors or evacuated tube collectors, up to 100  200 °C (Mekhilef et al., 2011). Also, solar energy can be concentrated in order to reach higher temperature levels, with single axis tracking collectors, such as linear Fresnel and Parabolic Trough (PT), obtaining temperatures of 300 °C and 400 °C, respectively (Suman et al., 2015). Further temperature levels may be attained with two axes tracking collectors, using central tower receivers or parabolic dish reflectors technologies, which achieves temperatures of 1500 and 2000 °C, respectively (Kalogirou, 2004). Within non-concentrating solar thermal technologies, some efficient methods for solar-driven desalination processes have been investigated. Garaway & Grossman (2006) developed a Multi-Effect Humidification (MEH) system for desalination purposes, using solar irradiation as the heat source of the system. They employed heat recovery from the condensed steam to preheat the seawater and used Computational Fluid Dynamics (CFD) to optimize the process and geometry of the facility. Results showed a low value for the distillate salinity (17 ppm) using saline feed water (23,000 ppm) and the operating efficiency (Gained Output Ratio, 𝐺𝑂𝑅) was comparable to that of ideal case in a MEH desalination device. Also, the possibilities of low temperature solar collectors for powering desalination systems were investigated by García-Rodríguez and Delgado-Torres (2007). They evaluated solar powered Rankine cycles providing shaft power to a Reverse Osmosis (RO) unit, for seawater or brackish water desalination. They concluded that desalination systems combined with solarpowered organic Rankine cycles had lower specific solar energy consumption compared to solar distillation and solar photovoltaic RO systems. This concept was later further exploited within the POWERSOL (mechanical POWER generation based on SOLar heat engines) Project (García-Rodríguez and Blanco-Gálvez, 2007), aimed to study and optimise solar thermal-driven mechanical power generation systems based on a solar-heated thermodynamic cycle, from low to medium temperature levels (80  250 °C). Different solar collectors were assessed: flat plate collectors, compound parabolic concentrators, linear Fresnel concentrators and parabolic trough collectors. Preliminary results showed that this technology could be more efficient and cost effective than photovoltaic systems. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 42 unit was considered, being the thermocompressor fed with motive steam from the high pressure turbine and entrainment steam from an intermediate effect of the MED unit. In the two latter configurations and in the case of RO+CSP, the three cooling methods mentioned above were considered for the sensitivity analysis. They concluded that, under certain conditions, that is, high temperature of the exhaust steam and high electricity consumption of the RO unit, the integration of a LT-MED process was more efficient than the coupling of a RO unit to the CSP plant, for the three cooling systems considered. The MED-TVC integration case resulted worse than RO option in all cases, although the LT-MED+TVC scheme was more efficient than RO for conditions similar to those of the Arabian Gulf region: high ambient temperature and high RO SEC (above 4.5 kWh/m3), using dry cooling or once-through cooling systems. The same authors carried out a techno-economic analysis of the configurations mentioned above for two study cases: a representative location in the Mediterranean Sea and a representative location in the Arabian Gulf (Palenzuela et al., 2015a). It resulted that the recommended desalination technology for integrating with a CSP plant in the Arabian region was the LT-MED+TVC arrangement, which was not better in terms of efficiency (LT-MED resulted the optimal option) but had other advantages such as the possibility of de-coupling the water production from the power generation and also being able to regulate the amount of fresh water produced, meeting the daily and seasonal demand of the location. If a Mediterranean area was considered for the CSP plant, the LT-MED scheme had better efficiency than RO case but not with evaporative cooling as the refrigeration method of the power block. The LTMED+TVC scheme was recommended only if dry-cooling was selected for the power cycle, although the differences with the RO option were not very large. Blanco et al. (2013) carried out a techno-economic evaluation of the coupling between desalination plants, based on MED and RO processes and a PT-CSP plant, taking Port Safaga (Egypt) as location. Three different coupling configurations were considered: LT-MED, LTMED-TVC, and RO, which were previously described. It was concluded that in cases in which the only feasible cooling option for the power block is done by means of dry condensers, which is the typical case in arid regions, the integration of a MED unit substituting the condenser of the cycle provided better thermal efficiency than the RO option, and production costs were similar. For such analysis, a high specific power consumption for the RO was selected (5.5 kWh/m3), which is habitual in arid regions dealing with harsh seawater. The LT-MEDTVC option resulted had the lowest efficiency, although in this case the water production is decoupled from the power generation, which is a major advantage. A techno-economic study of a cogeneration CSP plant for the combined production of power and water in Cyprus was performed by Fylaktos et al. (2014). They considered three different arrangements: only power production, power and water using RO and power and water using MED process. The capacity of the power plant and desalination unit were 4 MWe and Doctoral dissertation Bartolomé Ortega Delgado Page 43 5035 m3/d, respectively. A Discounted Cash Flow (DCF) method was selected for the financial analysis in order to estimate the annual electricity and water yields. The uncertainty of the input economic parameters was accounted using Monte Carlo simulations and pre-assigning them distribution functions. Results showed that the only-electricity production was the best option from an economic point of view, followed by the coupling of the MED, which provided slightly better LCOE and LCW. Also, it was suggested that the integration of desalination processes in CSP plants could be economically feasible in all cases and may be competitive with stand-alone CSP power production if water desalination by means of renewable energy were financially supported in the same manner than the electricity generation. Moreover, it was revealed that the financial performance of the plant is highly dependent on the electricity tariff in the three arrangements proposed. The accurate determination of the annual yield of power and water in CSP+D plants is important for performing thermo-economic and financial analyses, which are useful to predict the production costs and the feasibility of the project. In order to obtain the annual yield in CSP+D plants, time dependent analyses are required, due to the inherent variability and intermittence of the solar resource. Moreover, part load operation issues surged in MED units coupled to CSP plants is one of the reasons for the lack of real large-scale solar desalination plants in the market (Hassabou et al., 2013). Therefore, there is a necessity of further investigation in the part load operation in CSP+D plants and to develop mathematical models for predicting the performance of the system in such conditions. In this respect, Casimiro et al. (2014) presented a model for simulating the operation of a CSP plant coupled with a Forward Feed (FF) MED desalination unit acting as the condenser of the Rankine cycle, and taking as location San Diego (EEUU). In addition, four different cooling options were analysed: both the MED unit and a Once-Through (OT) condenser acting together, dry cooling, evaporative wet cooling and once-through cooling alone. Different control strategies were implemented, taking into account the minimum load for the operation of the power block and the MED unit. From the annual simulations performed it was obtained that the use of MED+OT cooling in a CSP was feasible and produced a penalty in the performance of the system below 5%, compared to the use of wet cooling method. Also, with the MED-OT option, fresh water is produced along with power. Also, Hassabou et al. (2013) analysed the transient operation of a 5000 m3/d MSF seawater desalination plant conceptually coupled to a CSP plant, located within the MENA region. Particularly, they selected a parabolic trough solar field with latent heat thermal energy storage (Phase Change Material, PCM) and a natural gas boiler as back-up system. The use of thermal energy storage extends the daily operation of the plant and permits to continue the operation when there is not available solar irradiation. Simulations with variable operation and changing weather conditions were performed over a year. Results obtained showed that the cost of the Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 44 energy source greatly depends on the selected solar fraction, plant location and energy prices. In addition, a solar fraction of 96% may be attained with average solar collector area and thermal storage size. The fresh water production costs of the MSF unit were almost three times higher than conventional technologies based on fossil fuels and RO system operating in the Arabian Gulf region. However, other features of the solar energy such as being a free and sustainable energy source or environmentally friendly technology should be taken into account when comparing with traditional power and water production plants. The water cost using CSP+RO was found to be competitive with conventional systems, although this technology has to face other problems in the Gulf region associated with high salinity and turbidity of the seawater, among red tides phenomena or high marine life. Finally, other desalination methods as MED process were suggested to be coupled with the CSP plant as they are more efficient thermodynamically than MSF. There are also studies on stand-alone solar desalination systems working out of design conditions. Fernández-Izquierdo et al. (2012) experimentally assessed the operation out of nominal conditions of the SOL-14 MED pilot plant located at Plataforma Solar de Almería solar research centre, in Almería (Spain). This vertically-stacked pilot plant consists of 14 effects with preheaters and a nominal capacity of 3 m3/h. After the experimental campaign conducted, it was concluded that the effects of the part load operation of the plant were higher on the distillate production (which decreased a 37% for a 19% heating steam temperature reduction, from 70 to 57 °C) than on the performance ratio (only suffered 11% of reduction, from 10 to 8.9, for the same case). Also, it was observed that the operation over the nominal values did not provide higher distillate production. Apart from parabolic trough, other CSP technologies could be considered for the solar thermoelectric production. Kalogirou (2013) investigated the optimal concentrating solar power technology for its implementation in Cyprus, among parabolic trough, central receiver and parabolic dishes. He obtained that the best option, in terms of the electricity production costs, was the central receiver. However, parabolic trough technology was preferred because of its maturity and proven operation in real plants. Moreover, PT plants have high global solar-toelectric efficiency and low area requirements per MWh. In opinion of the author, this plant should be located near to the sea, in order to use seawater as the cooling source of the condenser in the power block. Most conventional power stations are located in coastal areas for the same reason. In that case, it was recommended to combine the CSP plant with a desalination unit and produce fresh water as a second product. Also, the Fresnel solar collector technology has become commercial recently with a 30 MWe plant built in Puerto Herrado (Spain) in 2012 (Novatecsolar, 2016). In this regard, Hamed et al. (2016) studied the feasibility of the coupling between a Fresnel CSP plant and a MED-TVC desalination unit. A fossil back-up system was taken into account for supplying energy to the Doctoral dissertation Bartolomé Ortega Delgado Page 45 desalination plant when solar irradiation was not available. The proposed scheme was compared with a MED-TVC unit using only fossil fuel. Results showed that both cases provided similar Levelized Water Cost (LWC) for a fuel cost of $92/bbl, with the climatic conditions considered of Saudi Arabia (yearly DNI of 1132 kWh/m2). That means that up to this value, the scheme using only fossil fuel would provide lower LWC. If a location with higher annual DNI is selected (1937 kWh/m2), such as Almería (Spain), the breakeven fuel cost would decrease up to $52/bbl. From this value above, the Fresnel CSP would be more convenient. Moreover, with the specified climatic conditions, the combination of a Fresnel CSP plant and a MED-TVC unit without thermal energy storage resulted more economic than including a TES system. Kouta et al. (2016) investigated the coupling between a solar tower plant (central receiver) and a MED-TVC desalination unit for the co-production of power and water. In this case the power cycle selected was a supercritical CO2 Brayton cycle, and a thermal energy storage comprised of two tanks of molten salts was included. The entropy analysis revealed that the maximum entropy generation was produced by the solar tower (80%), followed by the MED-TVC unit and the sCO2 power cycle. The steam ejector was the component that had more entropy contribution in the desalination plant, followed by the end condenser. It was also shown that the specific entropy generation in the MED-TVC unit decreased when increasing the number of effects. Finally, a techno-economic analysis was carried out for Saudi Arabia from which it was drawn that the lowest LEC and LWC were obtained in regions with the highest solar irradiation levels during the year. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 46 Thermoeconomic comparison of integrating seawater desalination 2.2 processes in a concentrating solar power plant of 5 MWe This section deals with the case study of integrating seawater desalination processes in a specific CSP plant of 5 MWe, based on direct steam generation within absorber tubes of parabolic trough collectors. Zarza et al. (2006) designed and thoroughly described in this reference a CSP plant with these features. Based on this plant, the authors evaluate the power cycle and propose the best design in order to compare four coupling arrangements. Firstly, the coupling of a multi-effect distillation plant is analysed, thus integrating the fresh water and electricity production. Two different configurations are studied: the maximum water production by means of replacing the condenser by the distillation plant and the use of an intermediate steam extraction from the steam turbine to feed the MED unit. Secondly, the reverse osmosis desalination process is considered. Two options are examined: the integrated production of water and electricity, in which the desalination plant only produce water when the CSP plant operates, and the independent production of water and electricity. In this case two different plants can be installed in near but different locations. Since the use of distillation processes only makes sense for seawater desalination, the comparison is restricted to Sea Water Reverse Osmosis (SWRO). Moreover, a MED plant requires steam below 70 °C (to avoid the appearance of scaling in the tubes), hence the influence of this requirement on the power cycle efficiency is thoroughly calculated in this section, in order to compare distillation and membrane desalination processes. The cost of the fresh water production is also comparatively analysed in the four coupling arrangements considered. Although RO is currently the dominant technology in large-scale seawater desalination, some authors consider distillation as the only appropriate desalination process due to different reasons: - The requirement of extremely high quality water - concentration around a few ppb - to compensate the power cycle leakages, makes the RO process useless. Nevertheless, this is not a barrier for the RO process since further treatment of its product can be performed by means of other processes as electro-deionization. Moreover, the human consumption does not require any additional treatment, and then only a part of the water production would need it in that case. - The cooling requirement of the condenser is argued by some authors to be substituted by the distillation plant. Nevertheless, the seawater flow required by an MED plant is similar to the flow required by the corresponding once-trough cooled condenser. This is due to the need of cooling the end condenser in the distillation process. In addition, the hot brine disposal is an environmental problem similar to the once-through cooling of a Doctoral dissertation Bartolomé Ortega Delgado Page 47 CSP plant. Moreover, if a SWRO is installed, the condenser of the power cycle could be cooled by the feed, product or blowdown of the SWRO plant. - The self-consumption attributable to the electricity demand of the RO process, if it is integrated into the power production, is another drawback of RO process from the point of view of some authors. This is the case in which the electricity production has a direct subsidy. However, the coupling of a distillation process causes a decrease on the power cycle efficiency and needs a significant amount of energy for pumping, especially in the end condenser where the seawater is used as cooling stream. Furthermore, the temperature of the power cycle condenser in dry or evaporative cooling is higher than in once-trough cooling, in which an open circuit of water flows at ambient temperature and pass through the condenser. If a desalination process is considered together with the power generation, and a water stream is available at the CSP plant, then it can be used to partially or totally cool the power cycle condenser. Therefore, a realistic comparison should not consider the dry cooling as the only process to condense the power cycle steam. Due to above-mentioned reasons, the design of both, the desalination process and the condenser cooling should be carried out as a whole, and based on a realistic assessment of the power cycle efficiency for a given fresh water demand. Other key issue is related to the existence of conventional power backup for the electricity production, since in case of discontinuous operation distillation plants exhibit meaningful additional risk of scaling. Finally, in case of discontinuous operation, the nominal capacity of the desalination plant to supply a given fresh water demand should be significant higher. This does not make sense in case the water production is not only attributable to the self-consumption of the CSP plant. In addition, the interest in desalination coupled to solar power plants has been pointed out within the framework of the European Energy Research Alliance (EERA, 2015), by establishing a sub-programme on “Concentrated Solar Power plus Desalination” within the Joint Programme on CSP. Both, distillation and reverse osmosis, may be considered as the desalination process to be coupled to a solar power plant. This section is focused on the thermoeconomic comparison between distillation and reverse osmosis desalination technologies integrated in a parabolic trough solar thermal power plant, based on direct steam generation within the solar field. Since the comparison between heat or electricity consumptions is complex, the well-known thermoeconomic methodology is selected in order to assess the actual cost of the steam consumption of the distillation process. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 48 2.2.1 Concentrating Solar Power and Seawater Reverse Osmosis plants 2.2.1.1 Concentrating Solar Power (CSP) plant The first step of considering the coupling between CSP and desalination is to estimate realistic performances of both processes. Table 2.1 shows a record of representative PT-CSP plants, specifically those installed in the Mojave Desert, with the thermal efficiencies of the Rankine cycle. Table 2.1. Some representative PT-CSP plants (NREL, 2016). Plant Location Year 𝑾󰇗𝒆 (MWe) 𝑻𝒐𝒖𝒕 (°C) 𝑨𝑺𝑭 (m2) 𝜼𝒕𝒉 % Power block Back-up Solana Arizona, US 2013 250 390 2,200,000 - 100 bar, reheat Molten salts Andasol I Granada, Spain 2008 49.9 393 510,120 37.5 100 bar, reheat Molten salts Nevada Solar I Boulder City, NV 2007 64 390 357,200 37.6 100 bar, reheat None APS Saguaro Tucson, AZ 2006 1 300 10,340 20.7 ORC None SEGS IX Harper Lake, CA 1991 80 390 483,960 37.6 100 bar, reheat HTF heater SEGS VIII Harper Lake, CA 1990 80 390 464,340 37.6 100 bar, reheat HTF heater SEGS VI Kramer Junction, CA 1989 30 390 188,000 37.5 100 bar, reheat Gas boiler SEGS VII Kramer Junction, CA 1989 30 390 194,280 37.5 100 bar, reheat Gas boiler SEGS V Kramer Junction, CA 1988 30 349 250,500 30.6 40 bar, steam Gas boiler SEGS III Kramer Junction, CA 1987 30 349 230,300 30.6 40 bar, steam Gas boiler SEGS IV Kramer Junction, CA 1987 30 349 230,300 30.6 40 bar, steam Gas boiler SEGS II Daggett, CA 1986 30 316 190,338 29.4 40 bar, steam Gas boiler SEGS I Daggett, CA 1985 13.8 307 82,960 31.5 40 bar, steam 3-hrs TES 𝑊󰇗𝑒 is the nominal electric power generated, 𝑇𝑜𝑢𝑡 is the heat transfer fluid outlet temperature in the solar field, 𝐴𝑆𝐹 is the solar field aperture area and 𝜂𝑡ℎ is the thermal efficiency of the power block. The power block is analysed considering limits in temperature and pressure of the solar field described in Zarza et al. (2006). The sizing of the solar field is consistent with this reference. It has been assumed the same heatarea ratio than that one of the mentioned work. Therefore, the area of the solar field is function of the power generated in the plant, the thermal efficiency of the Rankine cycle and the solar field area and heat rate provided by the collectors in Zarza et al. (2006). Doctoral dissertation Bartolomé Ortega Delgado Page 49 𝑄󰇗𝑎 ∗ 𝐴𝑆𝐹 ∗=𝑄󰇗𝑎 𝐴𝑆𝐹 ⇒𝐴𝑆𝐹 =𝑄󰇗𝑎 𝑄󰇗𝑎 ∗𝐴𝑆𝐹 ∗=𝑊󰇗𝑒/𝜂𝑡ℎ 𝑊󰇗𝑒,𝑛 ∗𝐻𝑅𝑛 ∗𝐴𝑆𝐹 ∗ (Eq. 2.1) where 𝑄󰇗𝑎 is the heat rate added by the solar field, 𝐴𝑆𝐹 is the solar field area, 𝑊󰇗𝑒 is the electric power produced in the generator, 𝜂𝑡ℎ is the thermal efficiency of the cycle, 𝐻𝑅 is the heat rate of the cycle, the subscript 𝑛 represents the net value and the superscript * is related to that variable in Zarza et al. (2006). The thermal and economic analysis have been carried out using Engineering Equation Solver (EES) (Klein, 2013) as the main software environment for the calculations. This software is intended for solving non-linear equation systems simultaneously by the Newton-Raphson method. Moreover, it contains libraries to obtain the thermophysical properties of the pure water/steam, using the “IAPWS Formulation 1995 for the Thermodynamic Properties of Ordinary Water Substance for General and Scientific Use” (Wagner and Pruß, 2002), and seawater, using the formulation by Sharqawy et al. (2010). Table 2.2 shows the main parameters used in the design of the solar field and power block, which has been carried out at solar noon on 21st June. The solar field comprises 70 parabolic trough collectors, which generate steam at 400 °C and 60 bar for the turbines of the power block. The condensation of the cycle by means of evaporative cooling occurs at 30 °C with a temperature increase of 10 °C. Pressure loses in the solar field and main valves at the inlet of the high pressure turbine are considered to be of 5%. Besides that, the selected configuration of the power cycle, in which a deaerator and three closed preheaters are included, is shown in Figure 2.1. Each point of the cycle has been thermodynamically characterized and the equations used for solving the model are presented in Appendix 2-A. Performances of main items (turbines, pumps, etc.) are given in Table 2.3. Notice that the solar field efficiency, which varies with the change in the operation conditions of the cycle, has been calculated using the Eq. (2-A.8) in Appendix 2-A. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 50 Table 2.2. Design parameters (Zarza et al., 2006). Parameter Value Solar field Parabolic trough collector model ET-100 Number of parallel rows 7 Number of collectors per row 10 Overall length of a single collector, m 98.5 Parabola width, m 5.76 Net collector aperture per collector, m2 548.35 Design point: solar noon on 21st June Direct solar irradiance, W/m2 875 Longitude of the location 5°58′ W Latitude of the location 37°24′ N Air temperature, °C 20 Incident angle 13.7° Power block Live steam temperature, °C 400 Live steam pressure, bar 60 Condensation temperature, °C 30 Net power, kWe 5175 Net heat rate, kJ/kWh 14,460 Table 2.3. Performance of individual items in which is based the configuration given in Figure 2.1 (Blanco-Marigorta et al., 2011). Parameter 𝜼 𝜼𝒎 High pressure turbine 0.85 0.98 Low pressure turbine 0.85 0.98 Condensate pump 0.8 0.85 Feeding pump 0.8 0.85 Generator 0.97 - Turbine shaft - 0.98 Solar field 0.443* - *Calculated in the base case Doctoral dissertation Bartolomé Ortega Delgado Page 51 Figure 2.1. Layout of the power cycle and thermodynamic properties of mass flows. HP=high pressure turbine; LP=low pressure turbine; FP=feeding pump; CP=condensate pump; G=electric generator; PH=preheater; EV=evaporator; SH=super-heater; RH=reheater The key parameters of the power cycle are summarized in Table 2.4, showing the specific work extracted by the turbines, the specific work consumed by the pumps, the specific heat provided by the solar field and the thermal efficiency of the cycle, among others. In addition, in order to facilitate the comparison and permit the adaptation of results from this section to other configurations, Table 2.5 compares the principal parameters of the selected cycle with a simple Rankine cycle and the cycle with reheating. It is also worth mentioning that the reheating pressure has been optimized in order to achieve the maximum thermal efficiency of the power block. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 58 Figure 2.5. Basic layout of the MED2 configuration: a MED plant integrated within the CSP plant fed by an extraction. Figure 2.6. Analysis of the condensing thermal power of the Rankine cycle (𝑄󰇗𝑐𝑜𝑛𝑑), thermal power consumption in the MED (𝑄󰇗𝑀𝐸𝐷) and mass flow rate of exhaust steam (𝑞28) as function of the fresh water production (𝑞33) for MED2 configuration. 020 40 60 80 100 120 140 160 180 200 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 0 2000 4000 6000 8000 10000 12000 q33 [m3/h] q28 [kg/s] q28 q28 Q [kW] QMED QMED QCOND QCOND Doctoral dissertation Bartolomé Ortega Delgado Page 59 Figure 2.7. Effect of the fresh water production (𝑞33) on the thermal performance of the power cycle (𝜂𝑡ℎ), specific thermal energy added by the solar field (𝑄𝑎), specific work extracted in the low and high pressure turbines, (𝑊𝐿𝑃 and 𝑊𝐻𝑃, respectively) and specific work consumed by the condensing and feeding pumps (𝑊𝐶𝑃 and 𝑊𝐹𝑃, respectively) for MED2 configuration. Figure 2.8. Analysis of the extractions from the low pressure steam turbine (𝛼𝐴,𝛼𝐵,𝛼𝐶,𝛼𝐷 and total 𝛼𝑠𝑢𝑚) for different fresh water productions (𝑞33) in MED2 configuration. 020 40 60 80 100 120 140 160 180 200 0.31 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0 450 900 1350 1800 2250 2700 3150 q33 [m3/h] hth hth W, Q [kJ/kg] WLP WLP WHP WHP WCP WCP WFP WFP Qa Qa 020 40 60 80 100 120 140 160 180 200 0 0.2 0.4 0.6 0.8 1 1.2 q33 [m3/h] a aA aB aB aC aC aD aD asum asum Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 60 Table 2.9. Main operational parameters in MED1 and MED2 configurations respect to the base case. Parameter Unit Base case MED1 MED2 𝑞𝑤 m3/h - 186.4 158.6 𝑇𝑐 ºC 30 70 30 𝑊𝑇 kJ/kg 1090.7 908.2 941.6 𝑊𝑃 kJ/kg 8.35 8.35 8.35 𝑄𝑎 kJ/kg 2908.5 2891.1 2908.5 𝜂𝑡ℎ - 0.372 0.311 0.321 𝜂𝑋 - 0.668 0.566 0.576 𝐴𝑆𝐹 m2 248,110 296,640 287,750 𝑥 - 0.927 0.99 0.927 𝑄󰇗𝑀𝐸𝐷 kW 0 11,865 10,094 2.2.3 Comparative analysis of efficiency and production Table 2.10 summarizes the four cases considered for water and electricity production based on different configurations. In the MED1 arrangement the distillate produced cannot be regulated because it depends on the power block condensation. The quantity of water obtained is then fixed by the exhaust steam coming from the low pressure turbine. Besides, the daily fresh water production is limited to an average of 5.73 hours in the MED1, MED2 and RO1 cases due to the dependence on the solar resource, while the RO2 case may produce 24 hours per day, which is a major advantage. Notice how, in the RO1 configuration, the fresh water production can be chosen based on the quantity of electric energy consumed from the generator. As an example, there has been taken two values, corresponding to the fixed production in the MED1 case and the maximum production (when all the electric power generated has been consumed), respectively. Doctoral dissertation Bartolomé Ortega Delgado Page 61 Table 2.10. Summary of results related to the four configurations analysed. Parameter Unit MED1 MED2 RO1 RO2 Product flow regulation No Yes Yes Yes Average daily operation h/d 5.73 5.73 5.73 24 Fresh water production m3/h 186.4 158.6 186.4 1,658 186.4 Plant capacity m3/d 1067.3 908.8 1067.3 9500.3 4473.6 Net electric power kW 4577.5 4635 4401.3 0 4958.7 Thermal power consumption kW 11,865 10,094 0 0 0 Electric power consumption kW 372.8 317.2 557.3 4958.7 557.3 Pump power of the cycle kW 49.7 47.9 41.3 41.3 41.3 Specific consumption of auxiliaries kWh/m3 2♠ 2 0.85♣ 0.85 0.85 Specific consumption of the process Thermal energy kWh/m3 63.6 62.7 - - - Electric energy kWh/m3 - - 2.14♣ 2.14 2.14 ♠ Palenzuela et al. (2015b) ♣ Peñate and García-Rodríguez (2012) 2.2.4 Cost analysis 2.2.4.1 Thermoeconomic analysis Thermoeconomy combines the exergetic analysis alongside the economic principles to provide the designer of thermal systems information not available by means of classic thermal and economic analysis (Bejan et al., 1996). This helps them to identify possible improvements to the system and production costs. The exergetic balance gives the inefficiency associated to each element of the system considered: exergy destructions, which are internal and inherent to the component due to irreversibilities, and exergy losses, which are related to material and energy streams exiting the system without any further technical utilization. In systems with two products, like in this case, electricity and water, it is of great interest to know the cost of production for each asset and the services used to generate them so that these costs can be properly charged. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 62 2.2.4.2 Assumptions The well-known thermoeconomic analysis (Bejan et al., 1996) is applied to compare the four configurations described in sections 2-3. Main hypothesis applied are the following:  Steady-state operation.  The useful life of the CSP and desalination plants is 20 years.  Input data are referred to their values in the base year and updated for the following years according to the annual inflation rate 𝑟𝑖 (which is assumed constant and equal to 2%) and the nominal discount rate or nominal cost of capital 𝑘𝑛, established to 6% (real discount rates for technology comparison analysis are in the 5-10 percent range (International Energy Agency, 1991)).  The economic analysis is performed in current value because it includes the effect of inflation in the general interest definition.  The total capital investment is done by means of a bank loan at a given interest.  The depreciation and production taxes are not considered.  All the costs are levelized, i.e., converted to a series of different money quantities paid or collected to a constant annuity affected by the effective annual discount rate and the nominal escalation rate. 2.2.4.3 Exergy costing The exergy costing study allows quantifying the energy quality of the different flows considered entering and exiting the system, as well as the degree of inefficiency of the process or component. In this method, a unitary cost is associated to every exergetic power stream. The thermoeconomic analysis is done by means of the economic balance: ∑𝐶𝑒 󰇗 ∀𝑒 =∑𝐶𝑖󰇗 ∀𝑖 +𝑍󰇗 (Eq. 2.2) where 𝐶𝑒 󰇗 and 𝐶𝑖󰇗 are the cost rate associated with outlet and inlet streams respectively, both energetic and matter, in €/h, while 𝑍󰇗 is the cost rate of the capital investment and the operation and maintenance of each component, in €/h, also referred to as carrying charges. The cost rate is defined by: Doctoral dissertation Bartolomé Ortega Delgado Page 63 𝐶𝑗󰇗=𝑐𝑗𝐸𝑗󰇗 (Eq. 2.3) being 𝑐𝑗 the exergetic unit cost, in €/kWh, and 𝐸𝑗󰇗 is the exergetic power of the stream 𝑗, in kW. Assuming known the exergetic unit cost of all inlet flows, the exergetic power of the matter and energy streams, the cost rates associated with the capital investment and O&M, 𝑛𝑒−1 additional equations are needed to close the problem involving the levelized costs per unit of exergy 𝑐𝑖, where 𝑛𝑒 is the the number of outlet streams. The investment and operation costs of a system or component are written as follows: 𝑍󰇗=𝑍󰇗𝐶𝐼 +𝑍󰇗𝑂𝑀 (Eq. 2.4) where 𝑍󰇗𝐶𝐼 is the cost rate associated with capital investment and 𝑍󰇗𝑂𝑀 the cost rate related to the operation and maintenance, in €/h. It accounts for the expenditures associated with the purchase and operation and maintenance of the component. It is calculated as the levelized cost divided per the number of hours of annual operation. 2.2.4.4 Economic balances It is useful to make a diagram numbering all the matter and energy streams in the system. In this way the data is ordered and the streams can be defined as fuels or products. Fuel (F) is the resource used to generate the desired utility, whether it is one or several simple or compound streams, of matter or energy. Product (P) is defined as the flow or flows of interest, which can be simple or compound as well, of matter or energy. It is the purpose for which the component or system is planned and designed. The definition of fuel and product for every component of the system in study is presented in Table 2.11. The following considerations have been taken into account when performing the cost assignments to the streams of exergy:  It is assumed that the unitary exergy costs of inlet streams are known.  The direct normal irradiance and the seawater intake have null costs due to they are a free source of energy and matter, respectively.  The unitary exergy costs related to the waste streams are assumed to be zero as its dispersion cost to the environment is considered to be null. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 64  The cost per unit of exergy related with the electricity production does not vary both in condensing and cogeneration mode due to the penalty caused by the water production is charged to the solar field, which will have to be bigger enough in order to maintain the 5 MWe of electric power in the generator.  The electricity for pumping is taken from the self-production in the MED1, MED2 and RO1 cases, thus the unitary exergy cost associated is the same that the one from the electricity generated. In the fourth case, RO2, with the desalination plant connected to the grid, the electricity is purchased according to the high-voltage power access tariffs established by the liberalized energy market. The exergy rates of the pumping, extracted steam, condensate water and electricity were determined in Section 3. The exergy rate associated with the direct normal irradiance is obtained using the following equation: 𝐸󰇗𝑟𝑎𝑑(𝑄𝑠𝑢𝑛) =𝑄󰇗𝑠𝑢𝑛(1− 𝑇𝑎𝑚𝑏 𝑇𝑎𝑝,𝑠𝑢𝑛) (Eq. 2.5) being 𝑄󰇗𝑠𝑢𝑛 the heat rate of the direct normal irradiance in the solar field, in kW, 𝑇𝑎𝑚𝑏 the temperature of the environment, in K, and 𝑇𝑎𝑝,𝑠𝑢𝑛 the apparent temperature of the Sun, taken as 5770 K. The term 𝑄󰇗𝑠𝑢𝑛 is determined with: 𝑄󰇗𝑠𝑢𝑛 =𝑆𝑐⋅𝑁𝑐⋅𝐼𝑑⋅𝑐𝑜𝑠(𝜃) (Eq. 2.6) where 𝑆𝑐 is the collector aperture area, in m2, 𝑁𝑐 is the number of collectors, 𝐼𝑑 the direct normal irradiance at the design day (June 21st at solar noon) and 𝜃 the incidence angle of the solar rays. The cost rate of capital investment and O&M are determined using the methodology described in Appendix 2-B. The levelised cost of energy, 𝐿𝐶, for both electricity and water, is defined by the International Energy Agency (1991) as: 𝐿𝐶=𝑃𝐼+𝐿𝑒+𝑀+𝑅−𝐸2𝑉𝑒2 𝐸1∑1 (1+𝑘𝑛)𝑡 𝑛 𝑡=1 (Eq. 2.7) where 𝑃𝐼 is the discounted investment cost, 𝐿𝑒 the discounted sum of input energy expenses, 𝑀 is the discounted sum of operating expenses, 𝑅 is the discounted sum of replacement costs, 𝐸1 Doctoral dissertation Bartolomé Ortega Delgado Page 65 is the annual energy or output produced and 𝐸2𝑉𝑒2 is the discounted sum of other revenues. This factor is equal to the unitary exergy costs of the electricity and freshwater streams in the thermoeconomic diagram. Figs. 2.9 and 2.10 along with Table 2.11 present the diagrams of main components and input parameters. Figure 2.9. Block diagram of the material, thermal and mechanical interactions between the different subsystems in the MED1 and MED2 cases. Figure 2.10. Block diagram of the material, thermal and mechanical interactions between the different subsystems in the RO1 and RO2 cases. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 66 Table 2.11. Economic balances. STPP=Solar Thermal Power Plant; I=Intake; P=pretreatment; A.e.=Auxiliary equations Subsystem Fuel Product Economic balance A. e. MED1 STPP (1-2)+6 5+(3-4) 𝑐2𝐸󰇗2+𝑐3𝐸󰇗3+𝑐5𝐸󰇗5= 𝑐1𝐸󰇗1+𝑐4𝐸󰇗4+𝑐6𝐸󰇗6+𝑍󰇗𝑃𝑃 𝑐3=𝑐4 𝑐2=0 I+P 8 (9+9’)-7 𝑐9𝐸󰇗9+𝑐9′𝐸󰇗9′ =𝑐7𝐸󰇗7+𝑐8𝐸󰇗8 +𝑍󰇗𝑃𝑅𝐸 𝑐9=𝑐9′ MED (3-4)+10 +(9’-13) +(9-11) 12 𝑐11𝐸󰇗11+𝑐12𝐸󰇗12+𝑐13𝐸󰇗13+𝑐4𝐸󰇗4 =𝑐3𝐸󰇗3+𝑐9𝐸󰇗9+𝑐9′𝐸󰇗9′+𝑐10𝐸󰇗10 +𝑍󰇗𝑀𝐸𝐷 𝑐13 =0 𝑐11 =0 MED2 STPP (1-2)+6-14 5+(3-4) 𝑐2𝐸󰇗2+𝑐3𝐸󰇗3+𝑐5𝐸󰇗5+𝑐14𝐸󰇗14 =𝑐1𝐸󰇗1+𝑐4𝐸󰇗4+𝑐6𝐸󰇗6+𝑍󰇗𝑃𝑃 𝑐3=𝑐4 𝑐2=𝑐14 =0 I+P 8 (9+9’)-7 𝑐9𝐸󰇗9+𝑐9′𝐸󰇗9′ =𝑐7𝐸󰇗7+𝑐8𝐸󰇗8 +𝑍󰇗𝑃𝑅𝐸 𝑐9=𝑐9′ MED (3-4)+10 +(9’-13) +(9-11) 12 𝑐11𝐸󰇗11+𝑐12𝐸󰇗12+𝑐13𝐸󰇗13+𝑐4𝐸󰇗4 =𝑐3𝐸󰇗3+𝑐9𝐸󰇗9+𝑐9′𝐸󰇗9′+𝑐10𝐸󰇗10 +𝑍󰇗𝑀𝐸𝐷 𝑐11 =0 𝑐13 =0 RO1 & RO2 STPP (1-2)-3+5 4 𝑐2𝐸󰇗2+𝑐3𝐸󰇗3+𝑐4𝐸󰇗4= 𝑐1𝐸󰇗1+𝑐5𝐸󰇗5+𝑍󰇗𝑃𝑃 𝑐2=0 𝑐3=0 I+P 7 8-6 𝑐8𝐸󰇗8=𝑐6𝐸󰇗6+𝑐7𝐸󰇗7+𝑍󰇗𝑃𝑅𝐸 - RO 9-10 11-8 𝑐10𝐸󰇗10+𝑐11𝐸󰇗11 = 𝑐8𝐸󰇗8+𝑐9𝐸󰇗9+𝑍󰇗𝑅𝑂 𝑐10 =0 2.2.5 Comparative results The levelised electricity and water costs for the four cases considered, MED1, MED2, RO1 and RO2 are depicted in Figure 2.11. It is observed that the lowest water cost (0.76 €/kWh) is achieved in the RO2 arrangement, where the desalination plant is connected to the grid operating 24 hours per day. The main reason is because of the larger value corresponding to the net electricity produced in the power block, compared with the other cases analysed, as the only internal electricity consumption is due to the pumping in the cycle. Besides that, the total cost (capital investment and O&M) is lower in the RO case. Doctoral dissertation Bartolomé Ortega Delgado Page 67 The LWC obtained in the multi-effect distillation cases is higher than those achieved with the reverse osmosis technology, both directly coupled (RO1) and with the indirect integration (RO2, connected to the grid). The values obtained are 1.239 and 1.265 €/kWh for the MED1 and MED2 configurations, respectively. The LEC is constant and equal to 0.1308 €/kWh as it has been assumed that this cost, calculated in the only-electricity mode, does not change when operating in cogeneration mode. Comparing the two cases proposed in the distillation technology, there are no significant differences in the water production cost when choosing to substitute the condenser of the cycle by the MED plant or to feed the MED in parallel to the first feedwater heater at 63 °C. A more detailed description of the different terms used to calculate the LEC and LWC is showed in Appendix 2-B. Similar values are obtained using conventional power generation plants (AlHengari et al., 2014; Mahbub et al., 2012) and solar thermal power plants (Palenzuela et al., 2013). Additionally, it has been analysed the case of a feed-in tariff on the electricity production, with a constant value of 28.5 c€/kWh during the project lifetime. Results indicate that if grants are offered for this kind of plants, there are no costs associated with the water production. Figure 2.11. Levelised electricity and water costs in the four cases considered. Actual prices of water in Almería for industrial use are 0.463 €/m3 for a consumption lower than 50 m3 and 1.163 €/m3 for more than 50 m3 (“Aqualia - Almería,” 2014). The average electricity prices for industrial use, according to the annual statistics report of the year 2012 from the Spain’s Ministry of Industry, Energy and Tourism, are 0.13 and 0.097 €/kWh for high tension tariffs 3.1A (<450 kW) and 6.1 (>450 kW), respectively (MINECO, 2014). Comparing with the LWC and LEC obtained, slight difference is found between production and purchasing costs. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 74 Appendix 2-B The cost rate associated with the capital investment and the operation and maintenance of a component or system is defined as follows: 𝑍󰇗=𝑍󰇗𝐶𝐼 +𝑍󰇗𝑂𝑀 =𝑃𝐼+𝑃𝑂𝑀 𝜏⋅∑1 (1+𝑘𝑛)𝑡 𝑁 𝑡=1 (2-B.1) where 𝑃𝐼 are the discounted investment costs, in €, 𝑃𝑂𝑀 are the discounted O&M expenses, in €, 𝜏 is the annual average availability of the plant, in h/y, 𝑁 is the project lifetime, in years and 𝑘𝑛 is the discount factor or cost of capital. All the costs considered are levelized, that is the conversion of different payments or charges to a financially equivalent constant quantity over a particular time period (usually a year), affected by the inflation rate and the discount rate. The levelization of a quantity of money 𝑃0 paid in the base year is done as stated below: 𝐴=𝑃0⋅∑(1+𝑟𝑛 1+𝑘𝑛)𝑡 𝑁 𝑡=1 ∑1 (1+𝑘𝑛)𝑡 𝑁 𝑡=1 =𝑃0⋅𝑘⋅(1−𝑘𝑁) 1−𝑘 ⋅𝑘𝑛⋅(1+𝑘𝑛)𝑁 (1+𝑘𝑛)𝑁−1 =𝑃0⋅𝑘⋅(1−𝑘𝑁) 1−𝑘 ⋅𝐶𝑅𝐹 (2-B.2) 𝑘= 1+𝑟𝑛 1+𝑘𝑛 (2-B.3) 𝐶𝑅𝐹=𝑘𝑛⋅(1+𝑘𝑛)𝑁 (1+𝑘𝑛)𝑁−1 (2-B.4) with 𝑟𝑛 the nominal escalation rate, 𝑘𝑛 the nominal discount rate and 𝐶𝑅𝐹 the capital recovery factor. Besides that, Table 2-B.1 shows input data required for the economic analysis. Regarding the capital costs of desalination technologies, Palenzuela et al. (2015) estimated a specific investment cost of 1207 $/(m3/d) for a SWRO plant. Also, capital costs of 1054 €/(m3/d) and 875 €/(m3/d) were reported for the “Alicante 2” and Barcelona-Llobregat SWRO plants (Pankratz, 2011), while 1562-1181 $/(m3/d) were presented for distillation plants in the same reference. Previous values of capital costs reported (Wangnick, 2002) for distillation plants with capacities below 10,000 m3/d are above 2000 $/(m3/d). RO plants are modular, unlike Doctoral dissertation Bartolomé Ortega Delgado Page 75 distillation plants, therefore the capital cost increases significantly in distillation plants as plant capacity decreases. Table 2-B.1. Input data to the economic analysis. Concept Value Ref Solar thermal power plant Project lifetime, years 20 Assumption Full operation hours, h/y 2090 (Zarza et al., 2006) Average daily operation hours, h 5.726 Calculated Plant construction period, years 2 Assumption Electric power in the generator, kW 5000 (Zarza et al., 2006) Specific cost of the power block, €/kW 1500 (Balsa Escalante, 1999) Water consumption in the STPP, m3/h 30 Assumption Solar field Specific cost, €/m2 300 Assumption Heat rate, kJ/kWh 14,460 (Zarza et al., 2006) Unitary heat rate 4.02 (Zarza et al., 2006) Net electricity power, kW 5175 (Zarza et al., 2006) Heat rate provided by the solar field, kW 20,786.3 (Zarza et al., 2006) Collector area of the solar field, m2 38,385 (Zarza et al., 2006) MED plant Specific cost, €/(m3/d) 1400 Assumption Average daily operation hours, h 5.726 Calculated O&M man power, €/m3 0.04 (Perera, 1999) O&M pretreatment, €/m3 0.04 (Perera, 1999) Pretreatment cost, % of total cost 10 Assumption RO plant Specific cost, €/(m3/d) 1200 (Palenzuela et al., 2015a) Average daily operation hours, h 5.726/24 Calculated O&M man power, €/m3 0.086 (Medina, 2004) O&M membrane replacement, €/m3 0.036 (Medina, 2004) O&M pretreatment, €/m3 0.054 (Medina, 2004) Pretreatment cost, % of total cost 0.17 (Sanchez, 2008) Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 76 Table 2-B.2 presents the exergy rate and exergetic unit cost associated with the streams of the MED1 and MED2 cases, and Table 2-B.3 shows the same for the RO1 and RO2 configurations. Notice how the negative exergy rate of the brine reject and condensing thermal power means that these streams would produce work if there were a machine able of taking them to the zero state. A detailed description of the different terms used to calculate the LEC and LWC is showed in Table 2-B.4. Table 2-B.2. Summary of results related to the MED1 and MED2 configurations analysed in section 2.2.2. MED1 MED2 Stream Description 𝐸󰇗 (kW) 𝑐 (€/kWh) 𝐸󰇗 (kW) 𝑐 (€/kWh) 1 Direct normal irradiance 30,945 0 30,945 0 2 Global thermal losses in solar field 0 0 0 0 3 Heating steam from PB to MED 1636 0.041 1171 0.046 4 Condensed heating steam from MED 65.96 0.041 39.7 0.046 5 Electric power in generator 5000 0.131 5000 0.131 6 Pump power 49.69 0.131 47.9 0.131 7 Seawater intake 0 0 0 0 8 Aux. electric power intake+pretreat. 372.8 0.131 317.2 0.131 9 Feed seawater to MED 0 0 0 0 9’ Seawater entering the condenser 0 0 0 0 10 Auxiliary electric power for cooling 0 0 0 0 11 Brine reject -124.7 0 -106.1 0 12 Distillate water 297.8 1.239* 244.8 1.265* 13 Cooling seawater rejected to the sea 126.2 0 107.3 0 14 Condensing thermal power of the PB - - 0 0 * In €/m3 Doctoral dissertation Bartolomé Ortega Delgado Page 77 Table 2-B.3. Summary of results related to the RO1 and RO2 configurations analysed in section 2.2.1. RO1 RO2 Stream Description 𝐸󰇗 (kW) 𝑐 (€/kWh) 𝐸󰇗 (kW) 𝑐 (€/kWh) 1 Direct normal irradiance 30,945 0 30,945 0 2 Global thermal losses in solar field 0 0 0 0 3 Condensing thermal power of the PB -310.7 0 -310.7 0 4 Electric power in generator 5000 0.131 5000 0.131 5 Pump power in the power block 41.32 0.131 41.32 0.131 6 Seawater intake 0.323 0 0 0 7 Aux. electric power intake+pretreat. 158.4 0.131 37.8 0.048 8 Seawater feed to RO process 0 0 0 0 9 Main electric power RO process 398.9 0.131 95.17 0.048 10 Brine reject 185.2 0 44.17 0 11 Permeate 292 1.054* 69.67 0.76* * In €/m3 Table 2-B.4. LEC and LWC for the four cases considered. 𝑞𝑤 (m3/h) 𝑛ℎ (h/d) 𝑞𝑑 (m3/d) 𝑃𝐼 (€) 𝑃𝑂𝑀 (€) 𝑃𝐹 (€) 𝑊𝑒,𝑛 (kWh/y) 𝐸𝑤,𝑛 (m3/y) 𝐿𝐸𝐶 (€/kWh) 𝐿𝑊𝐶 (€/m3) 𝐿𝑊𝐶∗ (€/m3) STPP 0 5.726 0 15,236,313 306,753 0 10,363,641 0 0.131 - - MED1 186.4 5.726 1067.3 18,229,070 763,153 0 9,566,912 326,876 = 1.239 -4.099 MED2 158.6 5.726 908.1 17,732,279 694,070 0 9,686,878 268,774 = 1.265 -6.278 RO1 186.4 5.726 1067.3 16,504,665 1,245,088 0 9,198,809 326,876 = 1.055 -4.835 RO2 44.5 24 1067.3 16,504,604 1,245,043 642,991 10,363,641 326,876 = 0.76 -5.875 𝐿𝑊𝐶∗= levelized water cost with feed-in tariff for the electricity production Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 78 References Al-Hengari, S., ElMoudir, W., El-Bousiffi, M.A., 2014. Economic assessment of thermal desalination processes. Desalin. Water Treat. 1–14. doi:10.1080/19443994.2014.957982 Al-Karaghouli, A., Kazmerski, L.L., 2013. Energy consumption and water production cost of conventional and renewable-energy-powered desalination processes. Renew. Sustain. Energy Rev. 24, 343–356. doi:10.1016/j.rser.2012.12.064 Ali, M.T., Fath, H.E.S., Armstrong, P.R., 2011. A comprehensive techno-economical review of indirect solar desalination. Renew. Sustain. Energy Rev. 15, 4187–4199. doi:10.1016/j.rser.2011.05.012 Alrobaei, H., 2008. Novel integrated gas turbine solar cogeneration power plant. Desalination 220, 574–587. doi:10.1016/j.desal.2007.01.058 Aqualia - Almería [WWW Document], 2014. URL http://test.aqualia.es/almeria/gestion_clientes/t_vigente.asp (accessed 12.3.14). Balsa Escalante, P., 1999. Operation of Parabolic Trough Solar Collectors. Sol. Therm. Electr. Gener. Train. Mobil. Res. Program. CIEMAT. Bejan, A., Tsatsaronis, G., Moran, M.J., 1996. Thermal Design and Optimization. John Wiley and Sons, New York etc. Blanco-Marigorta, A.M., Victoria Sanchez-Henríquez, M., Peña-Quintana, J.A., 2011. Exergetic comparison of two different cooling technologies for the power cycle of a thermal power plant. 5th Dubrovnik Conf. Sustain. Dev. Energy, Water Environ. Syst. 36, 1966–1972. doi:10.1016/j.energy.2010.09.033 Blanco, J., Palenzuela, P., Alarcón-Padilla, D., Zaragoza, G., Ibarra, M., 2013. Preliminary thermoeconomic analysis of combined parabolic trough solar power and desalination plant in Port Safaga (Egypt). Desalin. Water Treat. 51, 1887–1899. Casimiro, S., Cardoso, J., Alarcón-Padilla, D.-C., Turchi, C., Ioakimidis, C., Mendes, J.F., 2014. Modeling Multi Effect Distillation Powered by CSP in TRNSYS. Energy Procedia 49, 2241–2250. doi:10.1016/j.egypro.2014.03.237 Damerau, K., Williges, K., Patt, A.G., Gauché, P., 2011. Costs of reducing water use of concentrating solar power to sustainable levels: Scenarios for North Africa. Energy Policy 39, 4391–4398. doi:10.1016/j.enpol.2011.04.059 De Munari, A., Capão, D.P.S., Richards, B.S., Schäfer, A.I., 2009. Application of solarpowered desalination in a remote town in South Australia 248, 72–82. doi:10.1016/j.desal.2008.05.040 EERA, 2015. Concentrated Solar Power (CSP) | EERA [WWW Document]. URL http://www.eera-set.eu/eera-joint-programmes-jps/concentrated-solar-power-csp/ (accessed 2.8.16). El-Sayed, Y.M., 2007. The rising potential of competitive solar desalination. Desalination 216, Doctoral dissertation Bartolomé Ortega Delgado Page 79 314–324. doi:10.1016/j.desal.2007.01.009 Fernández-Izquierdo, P., García-Rodríguez, L., Alarcón-Padilla, D.C., Palenzuela, P., MartínMateos, I., 2012. Experimental analysis of a multi-effect distillation unit operated out of nominal conditions. Desalination 284, 233–237. doi:10.1016/j.desal.2011.09.004 Fylaktos, N., Mitra, I., Tzamtzis, G., Papanicolas, C.N., 2014. Economic analysis of an electricity and desalinated water cogeneration plant in Cyprus. Desalin. Water Treat. 1– 18. doi:10.1080/19443994.2014.940219 Garaway, I., Grossman, G., 2006. Investigation of a solar-powered desalination system employing regeneration. Desalination 197, 63–74. doi:http://dx.doi.org/10.1016/j.desal.2005.12.016 García-Rodríguez, L., Blanco-Gálvez, J., 2007. Solar-heated Rankine cycles for water and electricity production: POWERSOL project. Desalination 212, 311–318. doi:10.1016/j.desal.2006.08.015 García-Rodríguez, L., Delgado-Torres, A.M., 2007. Solar-powered Rankine cycles for fresh water production. Desalination 212, 319–327. doi:10.1016/j.desal.2006.10.016 Gastli, A., Charabi, Y., Zekri, S., 2010. GIS-based assessment of combined CSP electric power and seawater desalination plant for Duqum-Oman. doi:10.1016/j.rser.2009.08.020 Ghobeity, A., Noone, C.J., Papanicolas, C.N., Mitsos, A., 2011. Optimal time-invariant operation of a power and water cogeneration solar-thermal plant. Sol. Energy 85, 2295– 2320. doi:10.1016/j.solener.2011.06.023 Hamed, O.A., Kosaka, H., Bamardouf, K.H., Al-Shail, K., Al-Ghamdi, A.S., 2016. Concentrating solar power for seawater thermal desalination. Desalination 396, 70–78. doi:10.1016/j.desal.2016.06.008 Hassabou, A.H., Spinnler, M., Polifke, W., 2013. Tecnoeconomic Analysis of Medium and Large-sacle Desalination Plants Driven by Concentrated Solar Systems in the Mena Region. Energy Procedia 42, 735–744. doi:10.1016/j.egypro.2013.11.076 Iaquaniello, G., Salladini, A., Mari, A., Mabrouk, A.A., Fath, H.E.S., 2014. Concentrating solar power (CSP) system integrated with MED–RO hybrid desalination. Desalination 336, 121–128. doi:10.1016/j.desal.2013.12.030 International Energy Agency, 1991. Guidelines for the economic analysis of renewable energy technology applications. International Energy Agency, Chateau Montebello, Quebec, Canada. Kalogirou, S.A., 2013. Solar thermoelectric power generation in Cyprus: Selection of the best system. Renew. Energy 49, 278–281. doi:10.1016/j.renene.2012.01.014 Kalogirou, S.A., 2004. Solar thermal collectors and applications. Prog. Energy Combust. Sci. 30, 231–295. doi:10.1016/j.pecs.2004.02.001 Klein, S.A., 2013. Engineering Equation Solver Software (EES). Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 80 Kouta, A., Al-Sulaiman, F., Atif, M., Marshad, S. Bin, 2016. Entropy, exergy, and cost analyses of solar driven cogeneration systems using supercritical CO2 Brayton cycles and MEE-TVC desalination system. Energy Convers. Manag. 115, 253–264. doi:10.1016/j.enconman.2016.02.021 Mahbub, F., Hawlader, M.N.A., Mujumdar, A.S., 2012. Combined water and power plant (CWPP) — a novel desalination technology. Desalin. Water Treat. 5, 172–177. doi:10.5004/dwt.2009.581 Medina, J.A., 2004. 20 years evolution of desalination costs in Spain, in: Proceedings of International Conference on Desalination Costing. Limassol, Cyprus. Mekhilef, S., Saidur, R., Safari, A., 2011. A review on solar energy use in industries. Renew. Sustain. Energy Rev. 15, 1777–1790. doi:10.1016/j.rser.2010.12.018 MINECO, 2014. Estadísticas eléctricas anuales - Estadísticas y Balances Energéticos - Energía - Ministry of Industry, Energy and Tourism [WWW Document]. URL http://www.minetur.gob.es/energia/balances/Publicaciones/ElectricasAnuales/Paginas/Ele ctricasAnuales.aspx (accessed 12.3.14). Novatecsolar, 2016. NOVATEC SOLAR · Concentrating Solar Power [WWW Document]. URL http://www.novatecsolar.com/56-1-PE-2.html (accessed 9.1.16). NREL, 2016. NREL: Concentrating Solar Power Projects - Concentrating Solar Power Projects with Operational Plants [WWW Document]. URL http://www.nrel.gov/csp/solarpaces/projects_by_status.cfm?status=Operational (accessed 4.25.16). Palenzuela, P., Alarcón-Padilla, D.-C., Zaragoza, G., Blanco, J., 2015a. Comparison between CSP+MED and CSP+RO in Mediterranean Area and MENA Region: Techno-economic Analysis. Energy Procedia 69, 1938–1947. doi:10.1016/j.egypro.2015.03.192 Palenzuela, P., Zaragoza, G., Alarcón-Padilla, D.C., 2015b. Characterisation of the coupling of multi-effect distillation plants toconcentrating solar power plants. Energy 82, 986–995. doi:10.1016/j.energy.2015.01.109 Palenzuela, P., Zaragoza, G., Alarcón-Padilla, D.C., Blanco, J., 2013. Evaluation of cooling technologies of concentrated solar power plants and their combination with desalination in the mediterranean area. Appl. Therm. Eng. 50, 1514–1521. doi:10.1016/j.applthermaleng.2011.11.005 Palenzuela, P., Zaragoza, G., Alarcón-Padilla, D.C., Guillén, E., Ibarra, M., Blanco, J., 2011a. Assessment of different configurations for combined parabolic-trough (PT) solar power and desalination plants in arid regions. Energy 36, 4950–4958. doi:10.1016/j.energy.2011.05.039 Palenzuela, P., Zaragoza, G., Alarcón, D., Blanco, J., 2011b. Simulation and evaluation of the coupling of desalination units to parabolic-trough solar power plants in the Mediterranean region. Desalination 281, 379–387. doi:10.1016/j.desal.2011.08.014 Pankratz, T., 2011. IDA Desalination Yearbook (2009-2010). Media Analytics Ltd, UK. Doctoral dissertation Bartolomé Ortega Delgado Page 81 Peñate, B., García-Rodríguez, L., 2012. Seawater reverse osmosis desalination driven by a solar Organic Rankine Cycle: Design and technology assessment for medium capacity range. Desalination 284, 86–91. doi:10.1016/j.desal.2011.08.040 Perera, J.C.I., 1999. Desalación de aguas. Colegio de Ingenieros de Caminos Canales y Puertos, Madrid. Philipps, S., Bett, A.W., Horowitz, K., Kurtz, S., 2015. Current Status of Concentrator Photovoltaic (CPV) Technology. United States. Dept. of Energy. Office of Solar Electric Technology, Washington, D.C. Sanchez, J., 2008. Estado del arte en el diseo del proceso de plantas desaladoras de agua de mar de gran capacidad. Tecnol. del Agua 28, 64–69. Sharaf, M.A., Nafey, A.S., García-Rodríguez, L., 2011. Thermo-economic analysis of solar thermal power cycles assisted MED-VC (multi effect distillation-vapor compression) desalination processes. Energy 36, 2753–2764. Sharqawy, M.H., Lienhard V, J.H., Zubair, S.M., 2010. Thermophysical properties of seawater: A review of existing correlations and data. Desalin. Water Treat. 16, 354–380. doi:http://dx.doi.org/10.5004/dwt.2010.1079 Slocum, A.H., Codd, D.S., Buongiorno, J., Forsberg, C., McKrell, T., Nave, J.-C., Papanicolas, C.N., Ghobeity, A., Noone, C.J., Passerini, S., Rojas, F., Mitsos, A., 2011. Concentrated solar power on demand. Sol. Energy 85, 1519–1529. doi:10.1016/j.solener.2011.04.010 Suman, S., Khan, M.K., Pathak, M., 2015. Performance enhancement of solar collectors—A review. Renew. Sustain. Energy Rev. 49, 192–210. doi:10.1016/j.rser.2015.04.087 Trieb, F., Müller-Steinhagen, H., 2008. Concentrating solar power for seawater desalination in the Middle East and North Africa 220, 165–183. doi:10.1016/j.desal.2007.01.030 Trieb, F., Müller-Steinhagen, H., Kern, J., Scharfe, J., Kabariti, M., Al Taher, A., 2009. Technologies for large scale seawater desalination using concentrated solar radiation. Desalination 235, 33–43. doi:10.1016/j.desal.2007.04.098 U.S. Department of Energy (DOE), 2009. Concentrating Solar Power Commercial Application Study. United States. doi:10.2172/1218186 Wagner, W., Pruß, A., 2002. The IAPWS formulation 1995 for the thermodynamic properties of ordinary water substance for general and scientific use. J. Phys. Chem. Ref. Data 31, 387–535. Wangnick, K., 2002. 2002 IDA Worldwide Desalting Plants Inventory. Report No. 17. IDA (International Desalination Association). Werner, M., Schäfer, A.I., 2007. Social aspects of a solar-powered desalination unit for remote Australian communities. Desalination 203, 375–393. doi:10.1016/j.desal.2006.05.008 Wilf, M., Awerbuch, L., 2007. The guidebook to membrane desalination technology : reverse osmosis, nanofiltration and hybrid systems : process, design, applications and economics. Balaban Desalination Publications, L’Aquila, Italy. Chapter 2 Seawater desalination integrated in solar thermal power plants based on parabolic trough collectors Page 82 Zarza, E., Rojas, M.E., González, L., Caballero, J.M., Rueda, F., 2006. INDITEP: The first precommercial DSG solar power plant. Sol. Power Chem. Energy Syst. 80, 1270–1276. doi:10.1016/j.solener.2005.04.019 Chapter 3. Opportunities of improvement of the MED seawater desalination process by pretreatments allowing high temperature operation Contents Chapter 3. Opportunities of improvement of the MED seawater desalination process by pretreatments allowing high temperature operation ................................................................... 83 List of figures.......................................................................................................................... 84 List of tables ........................................................................................................................... 86 Nomenclature.......................................................................................................................... 87 3.1 Introduction .................................................................................................................. 89 3.2 Forward feed MED model ........................................................................................... 92 3.2.1 Process description ............................................................................................... 92 3.2.2 Mathematical model ............................................................................................. 93 3.2.3 Plant performance ............................................................................................... 104 3.3 Validation of the FF-MED model and sensitivity analysis ........................................ 105 3.3.1 Validation of the model ...................................................................................... 105 3.3.2 Sensitivity analysis ............................................................................................. 108 3.4 Analysis of the MED process with high heating steam temperature ......................... 116 3.5 Conclusions ................................................................................................................ 118 References ............................................................................................................................ 120 Chapter 3 Opportunities of improving MED seawater desalination process by … Page 90 membrane and thermal processes coupled to power plants. He highlighted the promising concept of NF pretreatment for the removal of the ions responsible for alkaline and nonalkaline scale on the tubes of the evaporators. Tests carried out on a MSF pilot plant operating at a TBT of 130 °C resulted successful, increasing the product recovery up to 70% with respect 35% of conventional MSF plants. Also, Zhou et al. (2015) pointed out the benefits of introducing nanofiltration pretreatment into diverse seawater desalination technologies in order to reduce the costs associated with the scale formation. Particularly, if MED process is considered, the top brine temperature could be raised up to 125 °C without the risk of scaling. Moreover, the recovery ratio and water production could be improved. Accordingly, the introduction of pretreatment processes for eliminating the risk of scaling in MED desalination technology and increase the TBT would result in a significant improvement of the thermal efficiency and reduction of the overall energy consumption. Particularly, the detailed analysis of the increment of the TBT in MED units and the influence on the main design and performance parameters is of great interest and suggested by the above-mentioned authors. Notice that, among the different MED configurations, the forward feed is the most suitable one to investigate the augmentation of the maximum temperature of operation due to the lower risk of scaling (the maximum brine concentration is reached in the effect of lowest temperature) (El-Dessouky and Ettouney, 2002). One of the first steady-state mathematical models found for MED process with Forward Feed (MED-FF) arrangement was developed by El-Sayed and Silver (1980), relying on simplified assumptions such as constant thermophysical properties of the seawater or equal heat rates on the evaporators. The effect of the vapour pressure losses caused by friction was accounted by augmenting the 𝐵𝑃𝐸. They obtained useful analytical expressions to calculate the performance ratio, the thermal loads and heat transfer surface areas. In addition, they considered preheaters and flash boxes in the system analysis. El-Dessouky et al. (1998) developed a detailed mathematical model for MED-FF units, including preheaters and flash boxes. They assumed constant heat transfer areas in evaporators and preheaters, the influence of vapour leaks on the venting system and the effect of thermodynamic losses due to the 𝐵𝑃𝐸, Non-Equilibrium Allowance (𝑁𝐸𝐴) and vapour pressure drops through the demisters, connecting lines and during the condensation inside the tubes of the evaporators. Moreover, the thermophysical properties of the seawater were calculated as function of the temperature and salinity, and the effect of the non-condensable gases on the condensation heat transfer coefficients was accounted. They concluded that the thermal performance of the unit is nearly independent of the top brine temperature (for a fixed number of effects) and significantly affected by the number of effects. Also, the overall heat transfer coefficients in evaporators and preheaters increased with the temperature, being higher for evaporators than for the preheaters (2.3‒2.7 and Doctoral dissertation Bartolomé Ortega Delgado Page 91 1.9‒2.1 kW/(m2-°C), respectively). Other steady-state mathematical model for MED-FF plants, based on mass and energy balances applied on the different components of the system, was presented by Mistry et al. (2013). They used a simultaneous equation solver, which provides more flexibility to the model as it does not require developing any algorithm to reach the convergence and at the same time reduces the number of assumptions. The model was compared with others found in the literature by simulating the Gain Output Ratio (𝐺𝑂𝑅) and specific heat transfer area (𝑠𝐴) as function of different operational variables. The results showed that the model agreed quite well with the one from El-Sayed and Silver (1980), providing more details about the temperature profiles in the MED plant. Moreover, it had much simpler implementation and less assumptions that the model from El-Dessouky et al. (1998). However, only the 𝐵𝑃𝐸 was considered in the calculation of the thermodynamic losses, ignoring the saturation temperature losses due to the pressure drops in the demisters, connecting lines and during the condensation inside the evaporators. Due to the fact that the specific heat transfer area is greatly affected by the saturation temperature losses of the generated vapour, from its generation in one effect to the condensation inside the evaporator of the following effect, they should be properly accounted while modelling MED plants. In addition, some approximations were done, such as neglecting the effect of the non-equilibrium allowance as a result of the flashing processes of the brine in the effects and the distillate in the flash boxes. The present model for MED-FF units takes advantage of the flexibility of a simultaneous equation solver while calculating in detail the thermodynamic losses, including boiling point elevation, non-equilibrium allowance and saturation temperature reduction of the vapour as consequence of the pressure drop in the demister, connecting lines and during the condensation inside the tubes of the evaporator. The objective of this chapter is to investigate the effect of augmenting the top brine temperature and number of effects in MED-FF units on the main design and operation parameters (𝐺𝑂𝑅, specific heat transfer area and specific energy consumption) and further improve the existing FF-MED models in the literature. Seawater pretreatment like nanofiltration would permit to increase the TBT without the risk of scaling and fouling, by retaining the bivalent ions and rejecting the microorganisms. For that purpose, a detailed mathematical steady-state model of a MED with Forward Feed (MED-FF) arrangement has been developed and validated against data found in the literature. This model includes some improvements with respect others previously published, like the detailed calculation of the thermodynamic losses. Also, a sensibility analysis regarding different design and operational parameters of the MED process (number of effects, terminal temperature difference in the first effect, brine salinity of the intake seawater, etc.) has been carried out. Chapter 3 Opportunities of improving MED seawater desalination process by … Page 92 3.2 Forward feed MED model 3.2.1 Process description A multi-effect distillation unit for seawater desalination consists basically on a sequence of evaporation and condensation processes taking place inside a train of connected vessels, called effects, each one at lower pressure and temperature than the previous. In essence, this system takes advantage of the enthalpy of condensation of the generated vapour in one effect for promoting a new evaporation process in the following effect, repeating the sequence up to the last effect. Among the different feed arrangements of the MED system, the forward feed configuration is characterized by the equal direction of the vapour and feed flows in the system. Furthermore, the lowest salinity of the brine is reached in the first effect and is progressively increasing up to the last effect. Hence, this configuration is specially indicated for higher temperatures of the external heating steam as the risk of scale formation is minimized. In this system, each effect is comprised of a horizontal tube falling film evaporator, a demister and a preheater, except the last one which does not have a preheater but a condenser, called end condenser (see Figure 3.1). The external thermal energy, usually saturated steam (heating steam) below 70 °C to avoid the appearance of scaling in the tubes, is introduced exclusively in the evaporator of the first effect and represents the primary energy source which drives the entire distillation process. Firstly, the seawater enters the system from intake beach wells or submarine pipelines. It is directed to the end condenser where is used to condensate the vapour generated in the last effect. At the outlet of the end condenser, the seawater is divided in two streams: the feed seawater, which goes to the first effect passing through the preheaters of each effect, and the cooling seawater, which rejects the waste heat back to the sea. The preheating of the seawater permits to reduce the energy requirements of the process thanks to the condensation of a fraction of the total vapour generated in each effect. The feed seawater (feedwater), after being preheated, is sprayed over the tube bundle of the first evaporator where is partially evaporated due to the heat released by the condensation of the external vapour, which returns as saturated liquid to the steam generation source. From one side, vapour is produced, while the unevaporated brine remains at the bottom of the effect and constitutes the feedwater for the next effect. The vapour produced, considered free of salts, passes through a demister in order to retain the brine droplets, and is directed to a preheater where part of it condenses. The rest of the vapour is brought to the evaporator of the second effect and constitutes the driven force of the new evaporation process, at lower pressure and temperature. In this effect, the feedwater is the brine generated in the previous, which undergoes a flash process and produces additional vapour. Both the distillate produced in the preheater and inside Doctoral dissertation Bartolomé Ortega Delgado Page 93 the evaporator are collected in a flash box, producing additional flash vapour which is introduced in the vapour space of the effect. This process is repeated sequentially up to the last effect. Figure 3.1. Scheme of the FF-MED plant. 3.2.2 Mathematical model A steady-state mathematical model for a FF-MED process is developed by applying the mass and energy balance equations over the different components of the plant, along with the heat transfer equations associated with the heat exchangers (evaporators, preheaters and end condenser). The generic FF-MED system includes 𝑁 effects, 𝑁−1 preheaters and 𝑁−1 distillate flashing boxes (see Figure 3.1). By default, the input variables needed to solve the model are the following: the temperature of the heating steam, 𝑇𝑠, the temperature and salinity of the seawater intake, 𝑇𝑖𝑛 and 𝑋𝑖𝑛, respectively, the temperature and salinity of the brine in the last effect, 𝑇𝑁 and 𝑋𝑁, respectively, the temperature of the cooling seawater, 𝑇𝐹, the minimum temperature difference in the end condenser, 𝐷𝑇𝑇𝑐, and last preheater, 𝐷𝑇𝑇1, and the geometric and physical characteristics of the demister, connecting lines, and evaporator tubes. The model has been implemented in Engineering Equation Solver (EES) software environment (Klein, 2013), which solves simultaneously all the nonlinear equations set in the system (using the Newton-Raphson method) after a proper initialization of the variables. This software is useful for the characterization of thermal systems like the one studied here because it includes libraries for the thermophysical properties of numerous substances, particularly, pure water, using the IAPWS Formulation 1995 (Wagner and Pruß, 2002), and seawater (Sharqawy et al., 2010). Moreover, it does not need to create an algorithm to sequentially solve the equations, providing more degrees of freedom. On the other hand, the convergence is strongly subject to proper initial guesses and a reasonable range of variation of the variables. Cooling seawater Intake seawater Brine Distillate 12N P2 FB2FBN i Pi FBi P1 Feedwater End condenser Splitter Flash box Falling-film Evaporator Demister Preheater Brine Boiling vapor Heating steam Flash vapor Chapter 3 Opportunities of improving MED seawater desalination process by … Page 94 In the development of the model, the following assumptions and approximations have been taken into account: - Constant and equal heat transfer areas in evaporators and preheaters. This is a common practice in the real thermal desalination industry due to economic reasons. - The thermophysical properties of the seawater are function of the temperature and salinity. - The distillate produced is considered salt-free. - The thermal losses to the environment are neglected as the equipment is supposed to be well insulated and the operation temperatures are relatively low (40‒120°C). - The temperature of the vapour is considered equal to the brine temperature in each effect. That means that the vapour is slightly superheated by the boiling point elevation. - In each evaporator both the inlet steam and the exiting condensed liquid are supposed to be in saturation conditions. - The thermodynamic losses have been taking into account: the boiling point elevation, the non-equilibrium allowance in the effects and flashing boxes, and the saturation temperature decrease of the vapour due to the pressure losses in the demister, connecting lines and during the condensation inside the evaporators. - The vapour suffers an isenthalpic process while passing through the demister. - The temperature of the flashing vapour in the flash box is equal to the temperature in the vapour space of the effect. - Due to the utilization of a vacuum system, no vapour leaks have been considered. 3.2.2.1 Global mass and salt balances The global mass and salinity balance applied to the complete system leads to: 𝑞𝐹= 𝑞𝐷+𝑞𝐵 (Eq. 3.1) 𝑞𝐹𝑋𝐹=𝑞𝐵𝑋𝑁 (Eq. 3.2) where 𝑞𝐹 is the mass flow rate of feed seawater entering the first effect, 𝑞𝐷 is the total mass flow rate of distillate produced, 𝑞𝐵 is the mass flow rate of brine exiting the last effect, 𝑋𝐹 is the salinity of the feed seawater and 𝑋𝑁 the salinity of the brine in the last effect. Notice that as local mass balances have been applied in the effects, the global mass balance is only for Doctoral dissertation Bartolomé Ortega Delgado Page 95 verification purposes. Similarly, the global salinity balance is equivalent to the salinity balance in the last effect. The total mass flow rate of distillate generated in the plant is the sum of the mass flow rate of vapour produced by boiling in each effect 𝑞𝐷𝑖 plus the vapour produced by flashing 𝑞𝐹𝐸𝑖 (except in the first effect where there is not flashing phenomena): 𝑞𝐷=∑𝑞𝐷𝑖 𝑁 𝑖=1 +∑𝑞𝐹𝐸𝑖 𝑁 𝑖=2 (Eq. 3.3) 3.2.2.2 Temperature profiles The temperature of the brine in a generic effect 𝑖 is equal to the saturation temperature of the vapour formed by boiling plus the boiling point elevation, which takes into account the presence of salts in the water: 𝑇𝑖=𝑇𝑉𝑠𝑎𝑡,𝑖 +𝐵𝑃𝐸𝑖 (Eq. 3.4) The boiling point elevation is obtained with the correlation proposed by Sharqawy et al. (2010), which is function of the temperature and salinity of the brine. Also the temperature of the brine and the temperature of the vapour in each effect are considered equal: 𝑇𝑖= 𝑇𝑉𝑖 (Eq. 3.5) It is supposed that the vapour suffers an isenthalpic process through the demisters, reaching a temperature 𝑇𝑉𝑖 ′ in the vapour space, with the respective pressure drop. The saturation temperature of the generated vapour decreases in the path to the next evaporator, because of the pressure drop in the demister, the connecting lines and the condensation inside the tubes. This temperature drop causes a reduction in the temperature difference between effects, which is the driven force of the process. Therefore, the condensation temperature of the vapour generated in the effect 𝑖, which takes place inside the evaporator of the effect 𝑖+1, from 𝑖 =1..𝑁−1, is obtained with: 𝑇𝑐,𝑖 =𝑇𝑉𝑖 −𝐵𝑃𝐸𝑖−(Δ𝑇𝑚,𝑖 +Δ𝑇𝑙,𝑖 +Δ𝑇𝑐,𝑖) (Eq. 3.6) Chapter 3 Opportunities of improving MED seawater desalination process by … Page 96 where Δ𝑇𝑚,𝑖, Δ𝑇𝑙,𝑖 and Δ𝑇𝑐,𝑖 are the saturation temperature drops in the demister, connecting lines and condensation process, all referred to the effect 𝑖 and starting from the second effect. Notice that the condensation temperature in the first evaporator is the saturation temperature of the heating steam and the corresponding condensation temperature of the effect 𝑖+1 is 𝑇𝑐,𝑖, from 𝑖 = 1..𝑁−1. The decrease in the saturation temperature of the vapour after passing through the demister of a generic effect 𝑖 is the difference between the saturation temperature of the vapour generated before the demister (𝑇𝑉𝑠𝑎𝑡,𝑖) and that one after the demister (𝑇𝑉𝑠𝑎𝑡,𝑖 ′). Similarly, the saturation temperature drop occurring in the connecting lines between the effect 𝑖 and effect 𝑖+1 is equal to 𝑇𝑉𝑠𝑎𝑡,𝑖 ′ minus the saturation temperature of the vapour at the inlet of the following evaporator (𝑇𝑐,𝑖 ′). Finally, the saturation temperature losses of the vapour during the condensation in the tube bundle are defined as the difference between 𝑇𝑐,𝑖 ′ and the condensation temperature (𝑇𝑐,𝑖). Δ𝑇𝑚,𝑖 =𝑇𝑉𝑠𝑎𝑡,𝑖 −𝑇𝑉𝑠𝑎𝑡,𝑖 ′ (Eq. 3.7) Δ𝑇𝑙,𝑖 =𝑇𝑉𝑠𝑎𝑡,𝑖 ′−𝑇𝑐,𝑖 ′ (Eq. 3.8) Δ𝑇𝑐,𝑖 =𝑇𝑐,𝑖 ′−𝑇𝑐,𝑖 (Eq. 3.9) The above mentioned temperatures are calculated with the pressure drops of the formed vapour while flowing to the tubes of the next evaporator. Particularly, the pressure decrease in the demister is obtained with the correlation proposed by El-Dessouky and Ettouney (2002). For the pressure drop in the connecting lines, due to the friction with the walls, the Unwin’s formula (Nayyar, 2006) has been used. Finally, the pressure drop due to the condensation of the vapour inside the tubes of the evaporator has been determined by means of the methodology described in ESDU (1993), which relies on the Friedel’s correlation (Friedel, 1979). 3.2.2.3 First effect The first effect is different from the rest of the MED unit as is the place where the external energy is introduced (see Figure 3.2). The mass balance applied to a Control Volume (CV) containing the first effect establishes: 𝑞𝐹= 𝑞𝐵1 +𝑞𝑇1 (Eq. 3.10) Doctoral dissertation Bartolomé Ortega Delgado Page 97 where 𝑞𝐵1 is the mass flow rate of brine exiting the first effect and 𝑞𝑇1 the total mass flow rate of vapour generated within the first effect, which in this case is only the one produced by boiling (𝑞𝐷1). Notice that in this effect the seawater introduced does not suffer a flash process because its temperature is below the saturation temperature at the existing pressure inside the effect. The salinity balance applied to the same CV is shown below, where it has been assumed that the vapour generated is free of salts: 𝑞𝐹𝑋𝐹=𝑞𝐵1𝑋1 (Eq. 3.11) with 𝑞𝐵1 and 𝑋1 the mass flow rate and salinity of the brine in the effect 1, respectively. Finally, the energy balance in this first effect, considering all the streams entering and exiting the CV, is as follows: 𝑞𝑠𝜆𝑠+𝑞𝐹ℎ𝑝𝑟𝑒ℎ2 =(1−𝛼1)𝑞𝑇1ℎ𝑉1 ′+𝛼1𝑞𝑇1ℎ𝐶1 ′+𝑞𝐵1ℎ𝐵1 (Eq. 3.12) where 𝑞𝑠 is the mass flow rate of heating steam, 𝜆𝑠 is the specific enthalpy of condensation of the heating steam at 𝑇𝑠, ℎ𝑝𝑟𝑒ℎ2 is the specific enthalpy of the feed seawater before entering the first preheater, at 𝑡𝑝𝑟𝑒ℎ2, 𝛼1 is the fraction of the total steam that condenses in the first preheater, ℎ𝑉1 ′ is the specific enthalpy of the steam in the vapour space at 𝑇𝑉1 ′, after passing through the demister, ℎ𝐶1 ′ is the specific enthalpy of the condensate in the preheater at 𝑇𝑉1 ′, and ℎ𝐵1 is the specific enthalpy of the brine at the bottom of the effect at 𝑇1. Figure 3.2. Schematic diagram of the first effect. qs, Ts x=1 qB1,X1 1 T1 Ts (1-α1)qT1,T'v1 qs, Ts x=0 qT1, T'v1 P1 qF,tpreh1 α1qT1,T'v1 qD1,Tv1 qF,tpreh2 Chapter 3 Opportunities of improving MED seawater desalination process by … Page 98 The area of the evaporator is obtained by applying the heat transfer equation in this component. The rate of heat transfer (𝑄1) that takes place between the condensing steam and the sprayed seawater in this first evaporator accounts both for the sensible and latent heat added: 𝑄1=𝑞𝐹𝑐𝑝1(𝑇1−𝑡𝑝𝑟𝑒ℎ1)+𝑞𝐷1𝜆𝑉1 =𝐴1𝑈𝑒1(𝑇𝑠−𝑇1) (Eq. 3.13) where 𝑐𝑝1 is the specific heat at constant pressure of the feedwater between 𝑇1 and 𝑡𝑝𝑟𝑒ℎ1, 𝑡𝑝𝑟𝑒ℎ1 is the temperature of the feedwater after passing through the preheater associated with the first effect, 𝑞𝐷1 is the mass flow rate of vapour produced by boiling in the first effect, 𝜆𝑉1 is the specific enthalpy of evaporation of the water at 𝑇𝑉𝑠𝑎𝑡,1, 𝐴1 is the heat transfer area of the evaporator and 𝑈𝑒1 is the overall heat transfer coefficient of the evaporator, which is calculated using the correlation proposed by El-Dessouky and Ettouney (2002) as function of the temperature: 𝑈𝑒1 =1.9695+1.2057⋅10−2𝑇1−8.5989⋅10−5𝑇1 2+2.5651⋅10−7𝑇1 3 (Eq. 3.14) 3.2.2.4 Effects from 2 to N-1 The mass balance applied to the CV defined by the generic effect 𝑖 (see Figure 3.3) is as follows: 𝑞𝐵𝑖 =𝑞𝐵,𝑖−1 −𝑞𝐷𝑖 −𝑞𝐹𝐸𝑖 (Eq. 3.15) being 𝑞𝐵𝑖 and 𝑞𝐵,𝑖−1 the mass flow rates of the brine exiting and entering the effect 𝑖, repectively. Similarly, the salt balance in the CV establishes that: 𝑞𝐹𝑋𝐹=𝑞𝐵𝑖𝑋𝑖 (Eq. 3.16) where 𝑞𝐵𝑖 and 𝑋𝑖 are the mass flow rate and salinity of the brine in the effect 𝑖. The energy balance applied to the same CV gives: (1−𝛼𝑖−1)𝑞𝑇,𝑖−1𝜆𝑐,𝑖−1 +𝑞𝐹𝐵𝑖ℎ𝑉𝑖 ′′ +𝑞𝐵,𝑖−1ℎ𝐵,𝑖−1 =(1−𝛼𝑖)𝑞𝑇𝑖ℎ𝑉𝑖 ′+𝛼𝑖𝑞𝑇𝑖ℎ𝑐𝑖 ′+𝑞𝐹𝑐𝑝,𝑝𝑟𝑒ℎ,𝑖(𝑡𝑝𝑟𝑒ℎ,𝑖 −𝑡𝑝𝑟𝑒ℎ,𝑖+1)+𝑞𝐵𝑖ℎ𝐵𝑖 (Eq. 3.17) where 𝛼𝑖 is the fraction of vapour condensed in the preheater 𝑖, 𝜆𝑐,𝑖−1 is the specific enthalpy of condensation of the vapour inside the evaporator of the effect 𝑖, at 𝑇𝑐,𝑖−1, 𝑞𝐹𝐵𝑖 is the mass Doctoral dissertation Bartolomé Ortega Delgado Page 99 flow rate of vapour produced by flash in the flash box 𝑖, ℎ𝑉𝑖 ′′ is the specific enthalpy of the flashing vapour at 𝑇𝑉𝑖 ′′, 𝑐𝑝,𝑝𝑟𝑒ℎ,𝑖 is the specific heat of the feedwater at constant pressure and mean temperature between 𝑡𝑝𝑟𝑒ℎ,𝑖 and 𝑡𝑝𝑟𝑒ℎ,𝑖+1 and 𝑞𝑇𝑖 is the total mass flow rate of vapour produced by boiling within the effect (𝑞𝐷𝑖), by flashing of the brine (𝑞𝐹𝐸𝑖) and by flashing of the distillate (𝑞𝐹𝐵𝑖): 𝑞𝑇𝑖 =𝑞𝐷𝑖 +𝑞𝐹𝐸𝑖 +𝑞𝐹𝐵𝑖 (Eq. 3.18) Figure 3.3. Schematic diagram of the generic effect 𝑖. In the effects from 2 to 𝑁, the brine entering the effect suffers a flashing process because of being slightly superheated and discharged into a lower pressure effect. This process is described with the following balance: 𝑞𝐹𝐸𝑖𝜆𝐹𝐸𝑖 =𝑞𝐵,𝑖−1𝑐𝑝,𝐹𝐸𝑖(𝑇𝑖−1 −𝑇𝐵,𝐹𝐸𝑖) (Eq. 3.19) where 𝜆𝐹𝐸𝑖 is the specific enthalpy of evaporation of the seawater at 𝑇𝑖, 𝑐𝑝,𝐹𝐸𝑖 is the specific heat at constant pressure and mean temperature between 𝑇𝑖−1 and 𝑇𝐵,𝐹𝐸𝑖, with 𝑇𝐵,𝐹𝐸𝑖 the temperature of the unevaporated brine in the effect 𝑖 after the flashing process. The temperature of the brine is higher than the boiling temperature 𝑇𝑖 by the non-equilibrium allowance, which represents the deviation of the real process respect to the ideal one in equilibrium. This variation is mainly caused by the finite time period in which the flash process occurs (Fiorini et al., 2001). For that reason, the equilibrium conditions between the liquid and the vapour cannot (1-αi)qTi, T'vi i Tci-1 qDi,Tvi qFBi qTi,T'vi Ti Ti'' αiqTi, T'vi qFBi, Tvi'' qF , tpreh,i+1 Pi FBi qFEi,TVi Tvi'' qC,i-1,T''i-1 qCi,T''i qBi,Xi qB,i-1,Ti-1 qF ,tpreh,i (1-αi-1)qT,i-1 T'c,i-1 Chapter 3 Opportunities of improving MED seawater desalination process by … Page 106 For the validation of the MED-FF model, results from Mistry et al. (2013) have been taken. Particularly, the variation of the 𝐺𝑂𝑅 and 𝑠𝐴 with the number of effects and heating steam temperature was considered. To perform the comparison, the same specifications for the MED were selected and presented in Table 3.2. The model has been calibrated by minimizing the thermodynamic losses: only the demister of the first effect has been taken into account and the diameters of the connecting lines and the tubes of the evaporators has been selected large enough in order to decrease the friction losses. Table 3.2. Inputs taken for the validation of the model. Parameter Value Number of effects 319 Fresh water production, kg/s 1 Heating steam temperature, °C 70 Intake seawater temperature, °C 25 Intake seawater salinity, ppm 42,000 Brine blow down temperature, °C 40 Brine blow down salinity, ppm 70,000 Minimum 𝑇𝑇𝐷 in preheaters, °C 5 Temperature rise in the end condenser, °C 10 Figure 3.5 shows the 𝐺𝑂𝑅 and 𝑠𝐴 as function of the number of effects, along with the results obtained by Mistry et al. (2013). It can be seen how the relative error made is lower than 2% for the 𝐺𝑂𝑅 and 7% for the 𝑠𝐴, following the same trend in both cases. It is observed that when the number of effects is increased, the efficiency grows but progressively decreases because of the elevation of the thermodynamic losses and the increase of the specific heat of evaporation. Also, the overall heat transfer coefficient associated with the heat exchangers decreases along the effects of the MED plant, which degrades the heat transfer process and reduces the freshwater production. Regarding the 𝑠𝐴, a good agreement with Mistry et al. (2013) is also found. As shown, the 𝑠𝐴 increases considerably with the number of effects due to the difference of temperature between effects become smaller. Doctoral dissertation Bartolomé Ortega Delgado Page 107 Figure 3.5. Comparison of the 𝐺𝑂𝑅 and 𝑠𝐴 as function of the number of effects using the model from Mistry et al. (2013) and the present one. Other significant variable for the assessment of the MED design is the maximum temperature reached by the brine (top brine temperature), which in the case of FF arrangement takes place in the first effect. There is a practical limit for this parameter at nearly 70 °C due to the appearance of scaling (salts precipitation) on the tubes of the evaporators, which is favoured by the increase of the seawater temperature. Both models are also compared with respect the 𝐺𝑂𝑅 and 𝑠𝐴 by varying the heating steam temperature, with the number of effects fixed to 8 (see Figure 3.6). As it can be seen, the curves for both parameters present a good agreement with the results obtained by Mistry et al. (2013). The maximum relative errors found are lower than 2 and 8% for the 𝐺𝑂𝑅 and 𝑠𝐴, respectively. It is also observed that the thermal efficiency decreases only slightly (9.5%) when the heating steam temperature is elevated from 60 to 100 °C (40%). The decrease on the specific heat transfer area is more considerable, on the contrary. Chapter 3 Opportunities of improving MED seawater desalination process by … Page 108 Figure 3.6. Comparison of the 𝐺𝑂𝑅 and specific heat transfer area (𝑠𝐴) as function of the heating steam temperature (𝑇𝑠) using the model from Mistry et al. (2013) and the present one. 3.3.2 Sensitivity analysis In this section several key parameters for the design and operation of MED-FF systems, such as number of effects, temperature difference between effects, fraction of steam condensed in the preheaters, mass flow rate of distillate produced in each effect, etc. are analysed using the developed model. The base case selected for the sensitivity analysis is the one described in Mistry et al. (2013), previously defined in Section 3.3.1, but changing the last effect temperature for the 𝑇𝑇𝐷 in the end condenser, which is chosen as 5 °C. As the temperature increase of the seawater at the end condenser is fixed (10 °C), the condensation temperature is also specified to 40 °C. Also, the thermodynamic losses have been minimized by removing all the demisters except the first one and increasing the diameters of the connecting lines and evaporator tubes. 3.3.2.1 Temperature difference between effects and preheaters The difference of temperature between effects is an important parameter for the design of MED units and represents the driven force of the evaporation process in each effect. An increase of the number of effects, maintaining the total temperature difference constant, produces a decrease of the temperature drop between effects, as depicted in Figure 3.7, and consequently, a significant growth of the specific heat transfer area. Therefore, there is a practical limit in the maximum number of effects in MED plants related to an allowable temperature difference Doctoral dissertation Bartolomé Ortega Delgado Page 109 between effects, which usually lies between 2 and 3 °C (Cipollina et al., 2005). Noteworthy, the effective temperature difference should discount the thermodynamic losses which further limit the practical number of effects. Moreover, the temperature drop profile along effects show only a slightly variation for each case. Similar trends are observed for the temperature difference between preheaters, as shown in Figure 3.8. It can be seen how the temperature variation decreases with the elevation of the number of effects, ant their values are approximately equal to those of the temperature differences between effects. In the particular case simulated, if 3 °C is selected as a reasonable value for the temperature difference, the number of effects should be below 10. Figure 3.7. Difference of temperature between effects for different number of effects, from 𝑁 = 4 to 18. Figure 3.8. Difference of temperature between preheaters for different number of effects, from 𝑁 = 4 to 18. Chapter 3 Opportunities of improving MED seawater desalination process by … Page 110 3.3.2.2 Mass fraction of vapour condensed in the preheaters In this model it has been assumed that a fraction of the total vapour generated in each effect (by boiling in the evaporator, flash of the sprayed brine, and flash of the distillate collected in the flash box), denoted by 𝛼, condenses in the outer surface of the tubes in the preheater. The amount of condensed steam is directly associated with the boundary conditions imposed for solving the model, specifically the level of preheating of the feedwater. In this case the 𝑇𝑇𝐷 of the preheater associated with the first effect is fixed to 5 °C, which is the difference of temperature between the TBT and the seawater entering the first effect, while for the rest there are not restrictions imposed. Figure 3.9 shows how the mass of vapour condensed is higher within the first preheaters and decreases gradually up to the last, possibly due to the lower temperature level of the vapour from the last effects. Figure 3.9. Fraction of vapour condensed in each preheater as function of the number of effects, from 𝑁 = 4 to 18. 3.3.2.3 Terminal temperature difference of the preheaters The terminal temperature difference of the preheaters for different number of effects is depicted in Figure 3.10, tending to increase along the plant from the minimum value, 5 °C, which is reached in the preheater associated with the first effect. These parameters are important because they have a great influence on the temperature of the seawater entering the first effect. The higher this temperature is, the lower thermal consumption needed, due to the less thermal energy required to preheat the seawater up to the saturation temperature. However, the area needed in the heat exchangers increases significantly, so the selection of this parameter should account for this trade-off. Doctoral dissertation Bartolomé Ortega Delgado Page 111 Figure 3.10. Terminal temperature difference in the preheaters as function of the number of effects, from N=4 to 18. In the base case of study, the terminal temperature difference of the preheater associated with the first effect was set to 5 °C, as a conservative value (typically ranges from 3 ‒ 5 °C). The effect of the variation of this parameter on the 𝐺𝑂𝑅 and 𝑠𝐴 is presented in Figure 3.11. As mentioned above, a decrease in 𝑇𝑇𝐷𝑝𝑟𝑒ℎ1 improves the 𝐺𝑂𝑅 but also increases the specific heat transfer area. It is needed a compromise solution between the size of the heat exchanger (and therefore its cost) and the thermal efficiency of the plant. For values lower than 4 ‒ 3 °C, the specific heat transfer area grows markedly. Figure 3.11. Gain output ratio and specific heat transfer area as function of the terminal temperature difference at the preheater of the first effect. Chapter 3 Opportunities of improving MED seawater desalination process by … Page 112 3.3.2.4 Vapour produced by boiling in each effect The distribution of the vapour produced by boiling in each effect, for different number of effects, is depicted in Figure 3.12. The vapour produced slightly decreases in each effect, starting from the first. As a first approximation and in the case of high number of effects, the vapour produced in each effect may be considered constant as its variation is small. However, the small decrease could be caused by the increase on the specific enthalpy of evaporation of the seawater and the thermodynamic losses along the effects. Figure 3.12. Mass flow rate of vapour produced by boiling in each effect, as function of the number of effects. 3.3.2.5 Internal diameter of the pipes connecting the effects In the base case considered the distillate production has been fixed as 1 kg/s, which is a typical value used in the literature for the analysis of MED models. Nevertheless, it is interesting to investigate the influence of this parameter on the pressure losses of the vapour in the connecting lines between effects, which will eventually affect to the efficiency and heat transfer area of evaporators. For this purpose, different internal diameter of the pipes between effects have been considered, from 1000 mm to 200 mm, and the distillate production has been varied from 1 kg/s to 35 kg/s (nearly 86 and 3024 m3/d, respectively), as shown in Figure 3.13. As the larger pressure loss inside the pipes of the connecting lines takes place on the last effect, where the vapour is driven to the end condenser, it has been selected for the simulation as the key design parameter. In addition, same features of the base case have been used along with the data presented in Table 3.3. Also, for this analysis, the presence of all the demisters has been considered and the thermodynamic losses have been accounted. Doctoral dissertation Bartolomé Ortega Delgado Page 113 Table 3.3. Features of demisters, pipes connecting lines and evaporators. Parameter Value Length of connecting lines, m 2 Length of evaporator tubes, m 5 External diameter of evaporator tubes, m 0.030 External diameter of evaporator tubes, m 0.029 Wire diameter of demisters, mm 0.28 Density of demisters, kg/m3 280 Mesh pad thickness of demisters, m 0.15 Diameter of the vessel, m 4.8 From Figure 3.13 it is concluded that the pressure losses greatly increase with the distillate production when the internal diameter of the connecting lines is lower than 400 mm. In fact, for an internal diameter of 300 mm and a daily production of 3000 m3/d, the pressure losses due to friction inside the pipe connecting the last effect and the end condenser are of 2145 Pa, which results in a saturation temperature drop of the vapour of almost 5 °C, as it can be seen in Figure 3.14. These are unfeasible conditions because the value of the thermodynamic losses in that case would rapidly increase the specific heat transfer area (see Figure 3.15), due to the decrease on the temperature difference between the condensing vapour inside the evaporator and the boiling brine in the corresponding effect. Typically, the total thermodynamic losses may vary in the range of 0.5 ‒ 3 °C (El-Dessouky and Ettouney, 2002), so the minimum diameter of the pipes connecting the effects should be of 400 mm in this particular case. Figure 3.13. Pressure drop in the pipe connecting effect 8 and end condenser as function of the distillate production and for different internal diameters. Chapter 3 Opportunities of improving MED seawater desalination process by … Page 114 Figure 3.14. Saturation temperature decrease of the vapour in the pipe connecting effect 8 and end condenser as function of the distillate production and for different internal diameters of the pipes. Figure 3.15. Specific heat transfer area as function of the distillate production and for different internal diameters of the pipes connecting the effects. Doctoral dissertation Bartolomé Ortega Delgado Page 115 3.3.2.6 Intake seawater salinity The salinity of the intake seawater is an important design parameter which depends on the location of the plant. Although the mean salinity of the seawater is typically considered as 35 g/L (South Atlantic ocean), some areas of the world present higher salinity values, like the Red Sea and Arabian Gulf region, with seawater salinity of around 40 and 50 g/L, respectively (Bower et al., 2000). Therefore, the influence of this parameter on the 𝐺𝑂𝑅, 𝑠𝐴 and specific flow rate of cooling seawater (𝑠𝑞𝑐𝑤), defined as the mass flow rate of cooling seawater per unit of distillate produced, has been analyzed and the results are presented on Figure 3.16. It can be seen how the 𝐺𝑂𝑅 and the 𝑠𝑞𝑐𝑤 decrease with the increase in the intake seawater salinity, while the 𝑠𝐴 varies slightly reaching a maximum for a specific value of the salinity. The raise of the intake seawater salinity reduces the recovery ratio, all other variables maintained constant, and therefore increases the feedwater flow rate. Moreover, it decreases the rejected cooling seawater as the distillate production does not vary and the amount of total vapour to be condensed at the end condenser is almost the same. Hence, the intake seawater entering the end condenser does not change significantly. The reduction of the 𝐺𝑂𝑅 may be explained by lower preheating of the feedwater flow rate, which is greatly reduced due to the increase of the feedwater flow rate. Because of that, the heat added in the first effect must be higher and more heating steam flow rate is consumed. Figure 3.16. Gain output ratio, specific heat transfer area and specific flow rate of cooling seawater as function of the feed salinity. Chapter 3 Opportunities of improving MED seawater desalination process by … Page 122 Chapter 4. Preliminary model of TVC-MED plants coupled to parabolic trough concentrating solar power plants Part of this chapter has been published as a scientific article in Desalination and Water Treatment, 2016. pp. 1-12. Article in Press. doi:10.1080/19443994.2016.1173377 Title: “Quasi-steady state simulations of thermal vapour compression multi-effect distillation plants coupled to parabolic trough solar thermal power plants” Authors: Bartolomé Ortega-Delgado. Affiliation: CIEMAT-Plataforma Solar de Almería, Ctra. de Senés s/n, 04200 Tabernas, Almería, Spain. E-mail: [email protected] Patricia Palenzuela. Affiliation: CIEMAT-Plataforma Solar de Almería, Ctra. de Senés s/n, 04200 Tabernas, Almería, Spain. E-mail: [email protected]. Corresponding author. Diego-César Alarcón-Padilla. Affiliation: CIEMAT-Plataforma Solar de Almería, Ctra. de Senés s/n, 04200 Tabernas, Almería, Spain. E-mail: [email protected] Lourdes García-Rodríguez. Department of Energetic Engineering, Seville University, ETSI, Camino de los Descubrimientos s/n, 41092 Sevilla, Spain. E-mail: [email protected]s Abstract The evaluation of the coupling of a 50 MWe Parabolic Trough Concentrating Solar Power plant (PT-CSP) and a 10,000 m3/d Multi-Effect Distillation plant with Thermal Vapour Compression (MED-TVC) was performed. To that end, a model for the entire system has been developed and implemented within Engineering Equation Solver and Matlab software environments. Two coupling arrangements between the PT-CSP plant and the MED-TVC unit were selected: one taking low pressure steam (at 1.224 bar) from the power block to feed the MED-TVC and the other one taking high pressure steam (at 20.6 bar), and the simulations of the electricity and fresh water production of the PT-CSP+MED-TVC plant to be located in Almería (Spain) were carried out during three days in summer (21st23rd June) and three days in winter (21st23rd December). Results obtained showed that the use of the low pressure steam to feed the MED- Chapter 4 Preliminary model of TVC-MED plants coupled to parabolic trough solar thermal power plants Page 124 TVC plant reduces the electricity penalization compared with the use of high pressure steam but also decreases the fresh water production. Since in Spain the electricity demand is lower in summer than in winter, and the contrary occurs with the fresh water demand, the optimum coupling arrangement in summer was using high pressure steam to feed the MED-TVC (enough steam available in the turbines) and that one in winter was to feed the MED-TVC with low pressure steam having the lower electricity penalization at the cost of the decrease of the fresh water production. Keywords: concentrating solar power, desalination, modelling, multi-effect distillation, parabolic trough Doctoral dissertation Bartolomé Ortega Delgado Page 125 Contents Chapter 4. Preliminary model of TVC-MED plants coupled to parabolic trough concentrating solar power plants ..................................................................................................................... 123 List of figures ....................................................................................................................... 126 List of tables ......................................................................................................................... 127 Nomenclature ....................................................................................................................... 128 Introduction ................................................................................................................ 131 4.1 Methodology .............................................................................................................. 132 4.2 4.2.1 Solar field ........................................................................................................... 132 4.2.2 Multi-effect distillation plant with thermal vapour compression ....................... 135 4.2.3 Power block ........................................................................................................ 137 Results ........................................................................................................................ 140 4.3 Conclusions ................................................................................................................ 148 4.4 Appendix 4-A ....................................................................................................................... 149 Appendix 4-B ....................................................................................................................... 151 References ............................................................................................................................ 154 Chapter 4 Preliminary model of TVC-MED plants coupled to parabolic trough solar thermal power plants Page 126 List of figures Figure 4.1. Scheme of the overall system where the different alternatives of feeding the thermocompressor (C1 to C6) are represented. ....................................................................... 134 Figure 4.2. Scheme of the power block. ................................................................................. 137 Figure 4.3. Comparison of the thermal efficiency of the power block as function of the load for electricity-only and electricity plus water operation modes (with the MED-TVC fed by the C2 and C5 extractions). ................................................................................................................. 141 Figure 4.4. Simulation of the solar field for three days in summer: 21st-23rd June. ................ 143 Figure 4.5. Electricity and water production for three days in summer (21st23rd June) (with the MED-TVC fed by the C2 extraction). ............................................................................... 144 Figure 4.6. Electricity and water production for three days in summer (21st23rd June) (with the MED-TVC fed by the C5 extraction). ............................................................................... 145 Figure 4.7. Simulation of the solar field for three days in winter: 21st23rd December. ........ 146 Figure 4.8. Electricity and water production for three days in winter (21st23rd December) (with the MED-TVC fed by the C2 extraction). ...................................................................... 147 Figure 4.9. Electricity and water production for three days in winter (21st23rd December) (with the MED-TVC fed by the C5 extraction). ...................................................................... 147 Appendix 4-B Figure 4-B.1. Daily comparison of the daily electric energy generation and fresh water production, using the C2 and C5 steam extractions to feed the MED-TVC unit, during 21st-23rd June (a) and (b) and during 21st23rd December (c) and (d). ................................................... 153 Doctoral dissertation Bartolomé Ortega Delgado Page 127 List of tables Table 4.1. Characteristics of the ET-150 solar collector (Llorente García et al., 2011). ......... 132 Table 4.2. Main inputs for the design of the MED-TVC plant. ............................................... 136 Table 4.3. Characteristics of the power block at nominal conditions (Montes et al., 2009). .. 138 Table 4.4. Parametric analysis of the 𝐺𝑂𝑅 as function of the motive steam pressure and thermo-compressor location. .................................................................................................... 140 Appendix 4-A Table 4-A.1. MED-TVC performance in nominal operation mode for the C2 and C5 steam extractions. ................................................................................................................................ 149 Table 4-A.2. Stream data of the power block in nominal conditions. ..................................... 150 Table 4-A.3. Power block performance in on-design and off-design for only-electricity mode. .................................................................................................................................................. 151 Chapter 4 Preliminary model of TVC-MED plants coupled to parabolic trough solar thermal power plants Page 128 Nomenclature Variables 𝑝 Pressure, bar 𝐸 Thermal energy, kWh 𝐸𝑏 Direct normal irradiance, W/m2 ℎ Specific enthalpy, kJ/kg 𝑃 Thermal power, kW or MW 𝑞 Mass flow rate, kg/s 𝑇 Temperature, °C 𝑈𝐴 Heat exchanger constant, kW/m2°C 𝑊 Specific energy, kJ/kg Acronyms and abbreviations AF Anti-Freeze System BRICS Brazil, Russia, India, China and South Africa C Cold Tank CP Condensate Pump CSP Concentrating Solar Power CSP+D Concentrating Solar Power and Desalination CTP Cooling Tower Pump DSH Desuperheater EES Engineering Equation Solver EV Expansion Vessel FWH Feedwater Heater G Electric Generator GOR Gain Output Ratio H Hot Tank HCE Heat Collection Elements HP High Pressure Turbine HTF Heat Transfer Fluid HX Heat Exchanger LP Low Pressure Turbine MED Multi-Effect Distillation PB Power Block Doctoral dissertation Bartolomé Ortega Delgado Page 129 PC Parallel-Cross PH Preheater PT Parabolic Trough RH Reheater SCA Solar Collector Assembly SCE Solar Collector Element SF Solar Field SG Steam Generator SH Superheater CSP Concentrating Solar Power TES Thermal Energy Storage TMY Typical Meteorological Year TVC Thermal Vapour Compression UT Universal Time Subscripts b Direct normal D Distillate gen Generator m Motive opt Optical ref Reference s Isentropic T Total t Turbine th Thermal u Useful v Vapour Greek symbols 𝛼 Fraction of the total mass flow rate used in each extraction 𝛽 Reduction of evaporator areas after the thermo-compressor extraction 𝜂 Efficiency 𝜏 Receiver glass transmissivity Chapter 4 Preliminary model of TVC-MED plants coupled to parabolic trough solar thermal power plants Page 130 Doctoral dissertation Bartolomé Ortega Delgado Page 131 Introduction 4.1 The increase of the global population and the rise in the agrarian and industrial activities is leading to a continuous growth of the electricity and water demands. This has become a significant issue in developing countries with emerging economies, like those within the BRICS (Brazil, Russia, India, China and South Africa) group or those located in the Middle East, which will represent a major share of the worldwide population raise in the next decades. Conventional power production systems based on fossil fuels are known to cause the global warming, mainly due to the CO2 emissions to the atmosphere. Moreover, these systems rely on a limited source of energy (coal, oil, etc.) that will eventually run out. In this context, it is necessary the use of a mix of energy sources (conventional and renewable) to produce the power supply in the near future. Concentrating Solar Power plants (CSP) have been proved as reliable systems to produce electricity using solar irradiation as the energy source (NREL, 2015; Torresol Energy, 2015). Their use has sense in regions of the world with high direct normal solar irradiation levels. Sometimes these zones also suffer from severe water stress (caused by the physical scarcity of fresh water or by the absence of facilities to extract the water from the natural sources) and they are located close to the sea. In these scenarios, the integration of CSP and Desalination plants, concept known as CSP+D, represents an opportunity to partially solve the energy and water supply problems of these areas. This chapter analyses the integration of Multi-Effect Distillation plants with Thermal Vapour Compression (MED-TVC) into Parabolic Trough Concentrating Solar Power plants (PT-CSP), based on the electricity and fresh water demands in Spain, which are variable during the year. The integration of the MED-TVC plant was made by taking steam from one of the extractions of the Power Block (PB) to feed the thermo-compressor. The comparison of the electricity production penalties and the fresh water production in different periods of the year, along with the suitability of using one coupling arrangement or another, is presented and discussed in this paper. Chapter 6 Operational analysis of the coupling between a MED-TVC unit and a Rankine cycle … Page 234 has to be decreased. At this point, it is observed a trend change and the decrease in the power production becomes smaller. It can be due to the fact that the penalty rate in the power production decreases with the mass flow rate of the motive steam for the thermocompressor. For instance, a reduction of the thermal power load from 42.5 to 40% leads to a decrease in the power production from 15,624 to 14,238 kW (8.9%), while if the thermal power load is reduced from 40 to 37.5%, the decrease in the power production goes from 14,238 to 13,553 kW (4.8%). On the other hand, from a comparison between the coupling arrangements, it is seen that the use of high pressure steam from HP2 to feed the MED-TVC unit leads to the maximum amount of freshwater for all the loads of the power block but also to the lowest power production. On the contrary, as expected, the use of steam from the low pressure extraction LP3 generates the lowest freshwater production but the maximum power for all the load range. Figure 6.11 depicts the variation of the motive steam mass flow rate and the Gain Output Ratio (𝐺𝑂𝑅) with the load of the power block. As it can be seen, the 𝐺𝑂𝑅 follows an evolution roughly constant from 120% to the limit load of the power block, when the motive steam mass flow rate has to be decreased. More importantly, variable area thermocompressors help to even increase the 𝐺𝑂𝑅 when the motive steam flow is strongly decreased (power block load below 50%). As observed, the highest 𝐺𝑂𝑅 was achieved using high pressure steam from HP2 extraction, while the lowest 𝐺𝑂𝑅 was obtained using steam from the lowest pressure extraction LP3. Figure 6.11. 𝐺𝑂𝑅 and motive steam mass flow rate and as function of the power block load. 0 2 4 6 8 10 12 14 16 18 20 20 30 40 50 60 70 80 90 100 110 120 GOR or qm (kg/s) Load (%) GOR HP2 GOR HP1 GOR DEA GOR LP3 qm HP2 qm HP1 qm DEA qm LP3 Doctoral dissertation Bartolomé Ortega Delgado Page 235 The variation of the feedwater mass flow rate and the brine salinity in the first effect as function of the load of the power block, for each steam extraction, is represented in Figure 6.12, where the control algorithm developed has been used. Figure 6.12. Feedwater mass flow rate and brine salinity in the first effect as function of the power block load. On the other hand, as observed in Figure 6.13, the control loop also allows to maintain the last effect temperature (𝑇12) (which is equivalent to the end condenser temperature) around the nominal value (37 °C), as long as the mass flow rate of feedwater was enough to maintain the brine salinity in the first effect under the maximum limit. In the same figure it has been represented the variation of the heating steam temperature with the power block load. In this case, while the motive steam mass flow rate is equal to the nominal, the heating steam temperature is kept constant and equal to its nominal value (70 °C), but when the motive steam mass flow rate decreases, the heating steam temperature is reduced according to Eq. 6.2). 60000 65000 70000 75000 80000 85000 90000 95000 0 200 400 600 800 1000 1200 1400 20 30 40 50 60 70 80 90 100 110 120 X1 (ppm) qF (T/h) Load (%) qF HP2 qF HP1 qF DEA qF LP3 X1 HP2 X1 HP1 X1 DEA X1 LP3 Chapter 6 Operational analysis of the coupling between a MED-TVC unit and a Rankine cycle … Page 236 Figure 6.13. Last effect temperature and heating steam temperature as function of the power block load. 6.4 Conclusions In this paper a parametric analysis of the coupling between a MED-TVC unit and a Rankine cycle power block with variable nozzle thermocompressors has been carried out, in order to investigate the operational limits of the integration of both systems. Four different coupling arrangements have been considered, corresponding with four steam extractions from the power block feeding the thermocompressor of the MED-TVC unit: two from the high pressure turbine (HP2, HP1) and two from the low pressure turbine (DEA, HP3). On one hand, the load of the power block has been decreased up to the technical limit. On the other hand, the MED-TVC unit has been simulated at part load conditions when there was not enough steam available in the power block. The main conclusions reached in this study are the following:  The use of variable nozzle thermocompressors in a MED-TVC unit coupled to a Rankine cycle power block would allow us to maintain the motive steam mass flow rate constant when the power load decreases (with sliding pressure regulation), thus operating the MED-TVC unit near to nominal conditions as long there is enough steam available in the power block . 60 65 70 75 80 25 30 35 40 45 20 30 40 50 60 70 80 90 100 Ts ( C) T12 ( C) Load (%) T12 HP2 T12 HP1 T12 DEA T12 LP3 Ts HP2 Ts HP1 Ts DEA Ts LP3 Doctoral dissertation Bartolomé Ortega Delgado Page 237  There are bottom limits of the load in each coupling arrangement considered for which the steam entering the last feedwater heater of the low pressure turbine (LP P1) is near to zero. In those cases, the motive steam mass flow rate has to be decreased and consequently the fresh water production is considerably reduced. However, variable area thermocompressors help to even increase the 𝐺𝑂𝑅 in these cases. As an example, in the case of using steam from HP2, when the load decreases from 100% to 40% the freshwater reduction is of 2.6% and the 𝐺𝑂𝑅 decreases 2.66%; when the load decreases from 100% to 25%, the freshwater reduction is of 47%, while the 𝐺𝑂𝑅 decreases increases 28.8%. The operation limits of the power block are 40, 42.5, 42.5 and 45% of the load for the HP2, HP1, DEA and LP3 steam extractions of the high pressure and low pressure turbines, respectively. The nominal 𝐺𝑂𝑅 are 14.67, 14.2, 13.78 and 12.86, and they are reduced only to 14.28, 13.89, 13.07 and 11.97 for those lowest operation points maintaining the nominal motive steam mass flow rate.  When the motive steam mass flow rate of the thermocompressor is reduced from its nominal value, the MED-TVC unit works in off-design conditions and the brine salinity increases. In order to control the maximum value of the brine salinity in the unit and the brine temperature in the last effect, the feedwater mass flow rate and heating steam temperature need to be properly adjusted.  Regarding the coupling arrangements considered in the study case, the use of steam from the highest pressure extraction considered (HP2) leads to the maximum freshwater production, ranging from 9823.2 to 5228.2 m3/d for a variation of the power block load between 100% and 25% (steam extraction pressures of 45.4 and 11.4 bar, respectively), and the smallest power generation, from 44.094 to 8.979 MWe in all the load range. The contrary occurs when using steam from the lowest pressure extraction selected (LP3), with water and power productions ranging from 9384.6 to 5050.8 m3/d, and from 47.409 to 11.662 MWe, respectively, for power block loads between 100% and 27.5% (steam extraction pressures of 3.63 and 0.97 bar, respectively). Acknowledgments The authors wish to thank the European Commission (DG for Research & Innovation) for its financial assistance within the Integrated Research Programme in the field of Concentrated Solar Power (CSP) (STAGE-STE Project; Grant Agreement No. 609837). Chapter 6 Operational analysis of the coupling between a MED-TVC unit and a Rankine cycle … Page 238 References Cipollina, A., Micale, G., Rizzuti, L., 2005. A critical assessment of desalination operations in Sicily. Desalination 182, 1–12. doi:10.1016/j.desal.2005.03.004 Darwish, M.A., El-Dessouky, H., 1996. The heat recovery thermal vapour-compression desalting system: A comparison with other thermal desalination processes. Appl. Therm. Eng. 16, 523–537. doi:10.1016/1359-4311(95)00034-8 Desportes, C., 2006. Sea water desalination for the Bo-Hai Gulf: Case studies of MED desalination plants installed in Tianjin and Huanghua, in: IDA Desalination and Water Reuse International Forum & Exhibition. Tianjin (China). Efrat, T., Haimiao, Y., 2013. Design Challenges and Operational Experience of a Mega MED Seawater Desalination Plant in Tianjin, in: IDA 2013 World Congress. Tianjin, China. El-Dessouky, H.T., Ettouney, H.M., 2002. Thermodynamic Losses, in: El-Dessouky, H.T., Ettouney, H.M. (Eds.), Fundamentals of Salt Water Desalination. Elsevier Science B.V., Amsterdam, pp. 565–583. doi:10.1016/B978-044450810-2/50014-2 Hassan, A.S., Darwish, M.A., 2014. Performance of thermal vapor compression. Desalination 335, 41–46. doi:10.1016/j.desal.2013.12.004 Klein, S.A., 2013. Engineering Equation Solver Software (EES). MHPS, 2016. Reduction of Minimum Load | MITSUBISHI HITACHI POWER SYSTEMS, LTD. [WWW Document]. URL https://www.mhps.com/en/technology/business/power/service/simulation/minimum_load. html (accessed 8.4.16). Montes, M.J., Abánades, A., Martínez-Val, J.M., Valdés, M., 2009. Solar multiple optimization for a solar-only thermal power plant, using oil as heat transfer fluid in the parabolic trough collectors. Sol. Energy 83, 2165–2176. doi:10.1016/j.solener.2009.08.010 Ortega-Delgado, B., Palenzuela, P., Alarcón-Padilla, D.-C., 2016. Parametric study of a multieffect distillation plant with thermal vapor compression for its integration into a Rankine cycle power block. Desalination 394, 18–29. doi:10.1016/j.desal.2016.04.020 Power, R.B., 1994. Steam jet ejectors for the process industries. McGraw-Hill, New York. Shemer, H., 2011. Sliding Pressure Turbine Integrated with Seawater Desalination Facility (Multi-Effect Distillation - MED), in: IDA 2011 World Congress. Perth (Australia). Tamburini, A., Cipollina, A., Micale, G., Piacentino, A., 2015. CHP (combined heat and power) retrofit for a large MED-TVC (multiple effect distillation along with thermal vapour compression) desalination plant: High efficiency assessment for different design options under the current legislative EU framework. doi:10.1016/j.energy.2016.03.066 Temstet, C., Canton, G., Laborie, J., Durante, A., 1996. A large high-performance MED plant in Sicily. Proc. 1st Symp. Eur. Desalin. Soc. ’Desalination Eur. 105, 109–114. Doctoral dissertation Bartolomé Ortega Delgado Page 239 doi:10.1016/0011-9164(96)00064-1 Thermoflow, 2016. THERMOFLEX [WWW Document]. URL http://www.thermoflow.com/convsteamcycle_TFX.html (accessed 3.15.16). Yang, Y., Shen, S., Zhou, S., Mu, X., Zhang, K., 2013. Research for the adjustable performance of the thermal vapor compressor in the MED–TVC system. Desalin. Water Treat. 1–10. doi:10.1080/19443994.2013.855946 Zhang, B., Yang, L., Shen, S., Liu, X., Zhang, K., 2013. Analysis of adjusting method for load performance of TVC-MED desalination plant. Desalin. Water Treat. 51, 857–862. doi:10.1080/19443994.2012.714579 Chapter 6 Operational analysis of the coupling between a MED-TVC unit and a Rankine cycle … Page 240 Chapter 7. Yearly simulations of the water and power productions in CSP+D plants Contents Chapter 7. Yearly simulations of the water and power productions in CSP+D plants ............ 241 List of figures........................................................................................................................ 242 List of tables ......................................................................................................................... 245 Nomenclature........................................................................................................................ 246 7.1 Introduction ................................................................................................................ 247 7.2 Solar Field .................................................................................................................. 247 7.2.1 Characteristics of the solar field ......................................................................... 247 7.2.2 Operation strategy .............................................................................................. 250 7.3 Power block ............................................................................................................... 252 7.4 Desalination unit ........................................................................................................ 252 7.5 Yearly simulations ..................................................................................................... 257 7.5.1 Methodology ....................................................................................................... 257 7.5.2 Solar energy resource quantification .................................................................. 258 7.5.3 Yearly estimation of the power generation and fresh water production of the CSP+D plant ..................................................................................................................... 262 7.5.4 Daily simulations for representative months on summer and winter ................. 264 7.6 Conclusions ................................................................................................................ 290 References ............................................................................................................................ 291 Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 242 List of figures Figure 7.1. Scheme of the complete CSP+D plant with the three subsystems considered: solar field, power block and desalination plant. ............................................................................... 248 Figure 7.2. Scheme of a generic loop of the PT solar field. .................................................... 249 Figure 7.3. Diagram flow of the algorithm used for simulating the PT solar field. ................ 251 Figure 7.4. Scheme of the power block in nominal conditions. .............................................. 254 Figure 7.5. Scheme of the power block and MED-TVC unit coupled using the E1 (HP2) steam extraction, in nominal conditions. ............................................................................................ 255 Figure 7.6. Scheme of the power block and MED-TVC unit coupled using the E4 (LP3) steam extraction, in nominal conditions. ............................................................................................ 256 Figure 7.7. Monthly power demand profile for Andalusia region during 2015 (REE, 2016). 257 Figure 7.8. Yearly DNI as function of the time of the day. .................................................... 259 Figure 7.9. Yearly incidence angle as function of the time of the day. .................................. 260 Figure 7.10. Yearly Sun’s altitude as function of the time of the day. ................................... 261 Figure 7.11. Yearly thermal power absorbed by a loop as function of the time of the day. ... 262 Figure 7.12. Daily electric energy generation. ........................................................................ 263 Figure 7.13. Daily fresh water production. ............................................................................. 263 Figure 7.14. Solar field output, power and fresh water productions during July 1st-7th. ........ 266 Figure 7.15. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 1st-7th. .................................................................................................................... 267 Figure 7.16. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 1st-7th. 268 Figure 7.17. Solar field output, power and fresh water productions during July 8th-14th. ...... 269 Figure 7.18. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 8th-14th. .................................................................................................................. 270 Doctoral dissertation Bartolomé Ortega Delgado Page 243 Figure 7.19. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 8th-14th. .................................................................................................................................................. 271 Figure 7.20. Solar field output, power and fresh water productions during July 15th-21st. ..... 272 Figure 7.21. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 15th-21st. ................................................................................................................. 273 Figure 7.22. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 15th-21st. .................................................................................................................................................. 274 Figure 7.23. Solar field output, power and fresh water productions during July 22nd-31th. .... 275 Figure 7.24. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 22nd-31th. ................................................................................................................ 276 Figure 7.25. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 22nd-31th. .................................................................................................................................................. 277 Figure 7.26. Solar field output, power and fresh water productions during December 1st-7th.278 Figure 7.27. Electric energy and fresh water productions, for 10-min periods and daily periods, during December 1st-7th. ........................................................................................................... 279 Figure 7.28. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during December 1st7th. ............................................................................................................................................. 280 Figure 7.29. Solar field output, power and fresh water productions during December 8th-14th. .................................................................................................................................................. 281 Figure 7.30. Electric energy and fresh water productions, for 10-min periods and daily periods, during December 8th-14th. ......................................................................................................... 282 Figure 7.31. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during December 8th14th. ........................................................................................................................................... 283 Figure 7.32. Solar field output, power and fresh water productions during December 15th-21st. .................................................................................................................................................. 284 Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 250 7.2.2 Operation strategy The algorithm developed by Llorente García et al. (2011) and adapted in this research work calculates for each day the HTF temperatures in a generic collector loop and uses them to obtain the rest of operation variables (thermal power absorbed, thermal losses, etc.). The whole solar field is considered as a number of consecutive and equivalent loops, each one composed of five sections: four pipes corresponding to the four SCAs and another for the insulated outlet pipe of the solar field. The HTF temperatures are calculated applying an energy balance to a pipe portion and assuming linear approximation on the resulting differential equations. The validity of the approximation is maintained for time steps below 10 sec. The operation strategy followed, similar to that one of Llorente García et al. (2011), consists in considering four periods every day: a night time before sunrise where the solar field operates in recirculation mode (1 kg/s per loop) checking the status of the TES system, then, after the sunrise a HTF warm-up period begins with two stages: during the first one, the HTF is warmed up with a constant mass flow rate of 2.5 kg/s and by-pass the heat exchangers train of the power block, until the temperature of the HTF in the first SCA of the loop reaches a certain value, in this case 296 °C, as reported in the mentioned reference. After that, in the second stage, the HFT circulates through the heat exchangers until the temperature of the HTF in the insulated pipes reaches 310 °C, with the resulting mass flow rate calculated by the algorithm. After the warm-up period, a start-up of the turbine should be considered adding 20 minutes more of delaying, but in this work it has been neglected for simplicity. A full operation period is then applied until the sunset, where the power block generates electricity if there is enough thermal power provided by the solar field. The maximum thermal power absorbed by a loop is considered of 1.8 MW, with leads to HTF temperatures in the last SCA of 390 °C. Then, during this period, the thermal power delivered to the PB is calculated taking into account the resulting HTF solar field outlet temperature and 296 °C at the exit of the heat exchangers train. If there is an excess of thermal power, it is sent to the TES system. The maximum thermal energy absorbed by the TES is assumed to be 1010 MWh (which corresponds to 7.5 h of additional operation). The temperature of the hot tank has been established at 386 °C, and the cold tank temperature at 292 °C. In addition, a maximum limit of 1100 kg/s for the HTF mass flow rate has been imposed, which leads to 7 kg/s per loop as maximum. The last period considered, namely night period 2, takes place after the sunset and lasts until the beginning of the night period 1. Basically, the performance model adapted for the parabolic trough solar field simulation has the algorithm structure shown in Figure 7.3: firstly, the geographical, meteorological data of the location considered and the features of the solar field are used to obtain the solar time and Doctoral dissertation Bartolomé Ortega Delgado Page 251 incidence angle of the solar irradiation on a collector. Then the useful thermal power provided by the solar field is determined with the thermal power absorbed by a collector loop, the thermal losses and solar piping losses. This variable, together with the solar field temperature and state of the plant (storage tanks, irradiance levels, etc.), are used to decide the plant operation mode: night time period 1, HTF warm up, full operation and night time period 2. Also, it is calculated the thermal power sent to the power block and the thermal power delivered or added by TES system. Finally, the gross power output of the solar thermal power plant is obtained after solving the power block with the corresponding input of useful thermal power provided by the solar field. The meteorological data have been obtained in form of a Typical Meteorological Year (TMY), with the software Meteonorm for Almería (southern of Spain). The time step between data point was 10 minutes, but the solar field model is valid for time steps below 10 s, therefore interpolation of the DNI has been performed between those intervals. The main variables extracted were the date, UTC time, Direct Normal Irradiance (DNI), ambient temperature and wind velocity. In general the methodology used to calculate the solar time and incidence angle has been taken from Stine and Geyer (2001). Figure 7.3. Diagram flow of the algorithm used for simulating the PT solar field. INPUT DATA GEOGRAPHICAL DATA -Latitude, Longitude, time zone SOLAR FIELD DATA -Collector orientation -Tilt angle -Space between rows -Number of loops, etc COLLECTOR CHARACTERISTICS -Dimensions -Mirror properties -HCE receiver properties HEAT TRANSFER FLUID -Therminol VP-1 STORAGE FLUID -Molten salts TECHNICAL CHARACTERISTICS OF EQUIPMENT -Turbine, heat exchangers, pumps, flow limits, etc. METEOROLOGICAL DATA -Local time, DNI, ambient temperature, etc. SOLAR TIME AND ANGLE OF INCIDENCE CALCULATION -Collector s peak optical efficiency -Indicende angle modifier -Row shadowing factor -End loss factor COLLECTOR OPTICAL EFFICIENCY THERMAL POWER ABSORBED BY COLLECTOR LOOP USEFUL THERMAL POWER AT COLLECTOR LOOP HCE LOSSES IN A LOOP USEFUL THERMAL POWER SOLAR FIELD SOLAR FIELD HTF TEMPERATURES SOLAR FIELD PIPING LOSSES PLANT OPERATION MODE -Thermal power to PB - Thermal power to TES - Thermal power from TES GROSS ELECTRIC POWER GENERATED STATE OF THE PLANT - Storage tanks - Irradiance level - HTF temperatures - Night time period 1 - HTF warm-up & turbine startup - Full operation - Night time period 2 Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 252 7.3 Power block The power block subsystem has been described in Chapter 4, which has been developed in order to simulate the part load conditions with the methodology proposed by Montes et al. (2009). The design characteristics of the power block at nominal conditions are presented in Table 4.3 of Chapter 4. Notice that in this analysis the heat exchangers of the steam generation train have been supposed with constant 𝑈𝐴 and 𝐶𝑚𝑖𝑛, in order to simplify the solving of the problem. The power block is thermodynamically solved for nominal conditions in order to obtain the required thermal power by the steam generation train. In Figure 7.4, the diagram output from the model of the power block implemented in EES software is shown, in nominal conditions. Also, the two different MED-TVC coupling arrangements considered, using a HP and LP steam extractions of the power block as motive steam, have been solved for nominal conditions (see Figs. 7.5 and 7.6), which is needed to simulate the off design operation of the integrated system. It is illustrative to compare the penalty in the thermal efficiency of the power block as result of the fresh water production, in nominal conditions. In the electricity only mode operation, the power block has a thermal efficiency of 37.7%, which is decreased to 33.6% if the HP steam extraction is used to feed the MED-TVC unit, and to 34.8% in the case of using the LP steam extraction. 7.4 Desalination unit From Chapter 5 it was obtained the optimum coupling of a MED-TVC unit, based on the commercial Trapani plant, with a Rankine cycle power block with similar characteristics to that one of Andasol-1 (which are similar to Andasol-2), in terms of minimum specific heat transfer area and maximum 𝐺𝑂𝑅. The analysis was done for four different steam extractions of the power block: 45.4, 20.6 8.75 and 3.63 bar. Table 6.3 of Chapter 6 shows the main variables of the four different MED-TVC designs considered, corresponding to each steam extraction. The HP steam extraction of 45.4 bar (HP2) and the LP steam extraction of 3.63 bar (LP3) were identified as the most suitable steam extraction to feed the MED-TVC unit depending on the monthly power and water demands during the year on the location selected. In periods of the year with high electricity demand, the optimal coupling arrangement was found to be using the low pressure steam extraction of 3.63 bar to feed the MED-TVC unit, as the penalty on the electricity production was lower, although less water was produced. On the contrary, in periods of the year with low electricity demand, the high pressure steam extraction of 45.4 bar resulted the most suitable for increasing the water production and increase the efficiency of the desalination process, although the electricity generation was further penalised. In the simulations performed in this chapter, the MED-TVC operational model developed in Chapter 6 has been used, in order to maintain the mass flow rate of motive steam in nominal Doctoral dissertation Bartolomé Ortega Delgado Page 253 values at part load operation of the power block, as long as there is enough steam available in the cycle (which, similar to what was done in Chapter 6, has been considered when the steam mass flow rate entering the feedwater heater LP1 was below 0.1 kg/s). This could be done by using variable nozzle thermocompressors, as explained in Chapter 6. Also, the control algorithm implemented in the operational model of the MED-TVC unit, and described in detail in Chapter 6, has been used. This control algorithm adjusts the feedwater mass flow rate and the heating steam temperature so that the maximum brine salinity is always below 70,000 ppm and the end condenser temperature around its design value, 37 °C. Besides, the operation of the MED-TVC unit has been considered between 100 and 50% of the nominal load. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 254 Figure 7.4. Scheme of the power block in nominal conditions. Doctoral dissertation Bartolomé Ortega Delgado Page 255 Figure 7.5. Scheme of the power block and MED-TVC unit coupled using the E1 (HP2) steam extraction, in nominal conditions. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 256 Figure 7.6. Scheme of the power block and MED-TVC unit coupled using the E4 (LP3) steam extraction, in nominal conditions. Doctoral dissertation Bartolomé Ortega Delgado Page 257 7.5 Yearly simulations 7.5.1 Methodology The location considered is Almería, in the southeast of Spain (longitude 2.22W and latitude 37.06N), which has high levels of solar irradiation over the year and access to the sea. The base year selected has been 2015, a non-leap year. Two integrated PB+MED-TVC models have been considered, for the HP2 and LP3 steam extractions. The selection of the former or the latter coupling arrangement has been done taking into account the monthly power demand in Almería. For this purpose, the monthly data for 2015 provided by Red Eléctrica Española (REE, 2016) for Andalusia have been taken, which are shown in Figure 7.7. Therefore, according to this power demand, the following coupling arrangement has been established for every month (see Table 7.2): Table 7.2. Selection of the monthly coupling arrangement between the MED-TVC unit and CSP plant as function of the monthly power demand in Andalusia, during 2015. J F M M J J A S O N D LP3 HP2 HP2 LP3 LP3 LP3 LP3 HP2 HP2 HP2 LP3 Figure 7.7. Monthly power demand profile for Andalusia region during 2015 (REE, 2016). Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 258 The simulations have been carried out as follows:  Firstly, the solar field model uses the TMY to obtain the useful thermal power delivered to the power block among other key variables (HTF temperatures in each SCA of the loop and insulated pipes, HTF mass flow rate, useful thermal power absorbed by the solar field, thermal power losses, energy stored in the TES, etc.) for each time step selected (below 10 s) and during the period of time considered. In this case, weekly simulations have been done for each month, and the last week has been considered of 9 or 10 days, except in the case of February, that has 7 days. To that end, as the TMY provides 10-min data, the DNI has been interpolated. Therefore, 8640 points have been simulated for each day, which means 60,480 data points for each 7-day week and 3,153,600 points in the whole year. Of these, data points every 10 min have been extracted to create a vector file containing the useful thermal power transferred to the steam generator train of the power block, in order to reduce the calculation time.  Secondly, the useful thermal power in 10-min intervals is introduced in the corresponding integrated PB+MED-TVC model, for every steam extraction, which calls to an external procedure where the MED-TVC model has been implemented. Therefore, for each step time the model solves the power block calculating in each point the main thermodynamic variables (pressure, temperature, enthalpy…) and also the integrated MED-TVC unit fed by the corresponding steam extraction, determining in this way the power and fresh water production. During the calculations a control algorithm implemented in the MED-TVC model determines the feedwater mass flow rate and heating steam temperature, as explained in Chapter 6, so that the maximum brine salinity is always below 70,000 ppm and the temperature of the end condenser closer to 37 °C. All the outputs are stored in an Excel file.  Finally, the previous Excel file generated after solving the PB+MED-TVC integrated system, which contains the selected variables of interest, are passed to a MATLAB code for the elaboration of the different graphs. 7.5.2 Solar energy resource quantification It is interesting to show the solar energy availability in the particular case studied. To this respect, firstly, the yearly DNI in the location considered is depicted in a colour map graph as function of the time of the day, using the data provided by the TMY file (see Figure 7.8). It can be seen how, although high values of the DNI are reached in different periods of the year, during the summer season there are more days with high values of the DNI in comparison with Doctoral dissertation Bartolomé Ortega Delgado Page 259 the rest of the seasons. Also, the daylight hours are longer, which is the typical case of the latitude selected in the northern hemisphere. Figure 7.8. Yearly DNI as function of the time of the day. Two important parameters are the sun incidence angle on the collectors and the sun’s altitude. The first one is defined as the angle between the central ray of the sun and the normal vector to the aperture of the collector. The lower this angle the higher the thermal power absorbed by the collector, as more rays reach the receiver and transfer their energy to the heat transfer fluid. In Figure 7.9 the incidence angle versus the time of the day has been represented, for each day of the year. It is clear that the lower angles are obtained during the summer season in daylight hours, while during the winter season this angle is higher, which lead to a decrease in the thermal power absorbed by the HTF. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 266 Figure 7.14. Solar field output, power and fresh water productions during July 1st-7th. Doctoral dissertation Bartolomé Ortega Delgado Page 267 Figure 7.15. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 1st-7th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 268 Figure 7.16. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 1st-7th. Doctoral dissertation Bartolomé Ortega Delgado Page 269 Figure 7.17. Solar field output, power and fresh water productions during July 8th-14th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 270 Figure 7.18. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 8th-14th. Doctoral dissertation Bartolomé Ortega Delgado Page 271 Figure 7.19. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 8th-14th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 272 Figure 7.20. Solar field output, power and fresh water productions during July 15th-21st. Doctoral dissertation Bartolomé Ortega Delgado Page 273 Figure 7.21. Electric energy and fresh water productions, for 10-min periods and daily periods, during July 15th-21st. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 274 Figure 7.22. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during July 15th-21st. Doctoral dissertation Bartolomé Ortega Delgado Page 275 Figure 7.23. Solar field output, power and fresh water productions during July 22nd-31th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 282 Figure 7.30. Electric energy and fresh water productions, for 10-min periods and daily periods, during December 8th-14th. Doctoral dissertation Bartolomé Ortega Delgado Page 283 Figure 7.31. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during December 8th-14th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 284 Figure 7.32. Solar field output, power and fresh water productions during December 15th-21st. Doctoral dissertation Bartolomé Ortega Delgado Page 285 Figure 7.33. Electric energy and fresh water productions, for 10-min periods and daily periods, during December 15th-21st. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 286 Figure 7.34. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during December 15th-21st. Doctoral dissertation Bartolomé Ortega Delgado Page 287 Figure 7.35. Solar field output, power and fresh water productions during December 22nd-31th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 288 Figure 7.36. Electric energy and fresh water productions, for 10-min periods and daily periods, during December 22nd-31th. Doctoral dissertation Bartolomé Ortega Delgado Page 289 Figure 7.37. Motive steam pressure and mass flow rate, specific energy consumption, heating steam temperature, last effect temperature and brine salinity in1st effect, during December 22nd-31th. Chapter 7 Yearly simulations of power and water production in CSP+D plants Page 290 7.6 Conclusions In this chapter, the CSP+D simulation tool developed in previous chapters, comprising the models of the solar field, power block and desalination plant, has been used to estimate the yearly power and freshwater production for a hypothetic CSP+D plant located in Almería (Spain). Two coupling arrangements previously analysed (using a high pressure or low pressure steam extractions to feed the MED-TVC unit) have been alternatively used, depending on the monthly electric energy demand in the location considered. The features of the power block in nominal conditions, for three different scenarios: only electricity generation, electricity and water production using a high pressure steam extraction, and electricity and water productions using a low pressure steam extraction, have been also shown. Results obtained for the particular case studied showed that the maximum electricity and fresh water productions estimated were simultaneously obtained during the summer period, being July the month with the higher production. In particular, 24,122.7 MWh and 221,765.3 m3, for the electric energy and water productions, that may meet the needs of 72,369 households and 47,691 inhabitants, respectively. Finally, two representative months of the summer and winter periods (July and December) have been selected in order to show the daily fresh water and electricity productions. In addition, other significant variables regarding the solar field, power block and desalination unit in operation have been represented, such as the HTF temperatures of the collectors, the useful thermal power absorbed in the solar field and transferred to the power block, the motive steam pressure and mass flow rate entering the thermocompressor, the specific energy consumption of the MED-TVC unit, or the brine salinity in the first effect and temperatures of the heating steam and in the last effect. Doctoral dissertation Bartolomé Ortega Delgado Page 291 References Eck, M., Barroso, H., Blanco, M., Burgaleta, Juan-Ignacio Dersch, J., Feldhoff, J.-F., GarciaBarberena, J., Gonzalez, L., Hirsch, T., Ho, C., Kolb, G., Neises, T., Serrano, J.A., Tenz, D., Wagner, M., Zhu, G., 2011. guiSmo: Guidelines for CSP performance modeling – present status of the SolarPACES Task-1 project, in: Proceedings. 17th SolarPACES Conference. Granada, Spain. Klein, S.A., 2013. Engineering Equation Solver Software (EES). Llorente García, I., Álvarez, J.L., Blanco, D., 2011. Performance model for parabolic trough solar thermal power plants with thermal storage: Comparison to operating plant data. Sol. Energy 85, 2443–2460. doi:10.1016/j.solener.2011.07.002 Meteonorm, 2015. Meteonorm: Irradiation data for every place on Earth [WWW Document]. URL http://meteonorm.com/ (accessed 7.4.15). Montes, M.J., Abánades, A., Martínez-Val, J.M., Valdés, M., 2009. Solar multiple optimization for a solar-only thermal power plant, using oil as heat transfer fluid in the parabolic trough collectors. Sol. Energy 83, 2165–2176. doi:10.1016/j.solener.2009.08.010 REE, 2016. Red Eléctrica de España | Series estadísticas por comunidades autónomas [WWW Document]. URL http://www.ree.es/es/estadisticas-del-sistema-electrico-espanol/seriesestadisticas/series-estadisticas-por-comunidades-autonomas (accessed 7.15.16). Stine, W.B., Geyer, M., 2001. Power from the sun [WWW Document]. URL http://www.powerfromthesun.net/book.html Conclusions and future works Page 298 kind of plants. In this regard, the use of different coupling schemes depending on the power and water profile demands during the year increases the efficiency of both the power and water production, and therefore reduces the production costs. As a particular case study, the yearly power and fresh water production for a hypothetic MEDTVC+PT-CSP plant located in Almería (Spain), with same features of the subsystems described previously, has been estimated using the developed models of the solar field, power block and desalination unit. Two integration schemes previously analysed have been alternatively used, depending on the monthly electric energy demand: using steam from the high pressure turbine to feed the MED-TVC unit, and using steam from the low pressure turbine. Results obtained showed that the maximum electricity and fresh water productions were achieved during the summer period. In particular, July was the best month, with 24,122.7 MWh and 221,765.3 m3, which could supply electric energy and fresh water to 72,369 households and 47,691 inhabitants, respectively. Finally, the total yearly amount of electric energy and fresh water production were 150,663.2 MWh and 1,505,725 m3. If it is compared with the published annual electricity generation of 175,000 MWh in Andasol-1, it would mean a 14% of decrease due to the fresh water production. 8.2 Future works As a result of the different analyses performed in this thesis, further investigation on the improvement of the MED process and its integration with CSP plants are proposed. Particularly, reliable variable nozzle thermocompressor models are needed in order to make realistic estimations of their performance. To that end, a test bed facility recently installed at Plataforma Solar de Almería will be useful for obtaining empirical performance curves and for validating theoretical models. This facility consists in a train of four different thermocompressors operating in a wide range of motive and entrained vapour pressures and flow rates, which are fed by a steam generator powered by the thermal energy provided by a parabolic trough solar field. Another investigation line suggested is to increase the operation temperature at the outlet of the ejectors in MED-TVC units, which could improve the thermal efficiency of the desalination process although a higher exergy motive steam is expected to be required. Regarding the modelling of the MED process, the presence of non-condensable gases should be accounted in future works, along with a more exact calculation of the overall heat transfer coefficients. Doctoral dissertation Bartolomé Ortega Delgado Chapter 8 Page 299 Also, assessments of the integration of MED-TVC processes into higher temperature power cycles, such as Brayton, which are suitable for using in central receiver towers, may be carried out in order to investigate the potential of this kind of systems comparing with the ones already studied. Finally, with the simulation tool developed for the detailed calculation of the annual power and water production in CSP+MED-TVC systems, the completion of thermo-economic analyses are proposed in order to provide realistic estimations of the levelised energy and water costs, if reliable cost data of the different components are obtained. The accurate assessment of these costs is fundamental for project feasibility analyses. Conclusions and future works Page 300 Doctoral dissertation Bartolomé Ortega Delgado Chapter 8 Page 301 Conclusions and future works Page 302