Integration of solar latent heat storage towards optimal small-scale combined heat and power generation by Organic Rankine Cycle

Contents lists available at ScienceDirect Journal of Energy Storage journal homepage: www.elsevier.com/locate/est Integration of solar latent heat storage towards optimal small-scale combined heat and power generation by Organic Rankine Cycle Jesus Lizana a,⁎ , Chiara Bordin b , Talieh Rajabloo c,d,⁎ a Instituto Universitario de Arquitectura y Ciencias de la Construcción, Universidad de Sevilla, Avda. Reina Mercedes 2, 41012 Seville, Spain b Arctic University of Norway, Tromsø, Norway c Hasselt University, Martelarenlaan 42, 3500, Hasselt, Belgium d imec, Kapeldreef 75, 3000, Leuven, Belgium ARTICLE INFO Keywords: Thermal energy storage Solar thermal energy Phase change material ORC Evacuated tube collector ABSTRACT Thermal energy and distributed electricity demand are continuously increased in areas poorly served by a centralized power grid. In many cases, the deployment of the electricity grid is not economically feasible. Smallscale Organic Rankine Cycle (ORC) appears as a promising technology that can be operated by solar energy, providing combined heat and power (CHP) generation. Additionally, thermal energy storage can ensure stable and continuous operation in case of scarce thermal energy availability. This paper evaluates the potential application of latent heat storage to enhance solar ORC performance at operating temperatures between 80 °C and 140 °C, aiming at improving the efficiency and capacity of ORC for low-cost non-concentrating solar-thermal collectors. Three thermal energy storage scenarios are considered. Scenario 1 and 2 consist of reference cases based on a solar ORC system integrated with a conventional hot water tank and a pressurised water tank. Scenario 3 implements a storage unit based on a phase change material. The simulation was carried out through models developed in TRNSYS for solar energy balance and ASPEN for ORC system performance. The results show that solar latent heat storage tank can provide 54% of useful collector gains with a higher and narrower temperature range in the evaporator, increasing the annual thermal energy capacity by 19%, reducing annual heat losses by 66% and decreasing the investment cost by 50% in comparison with a pressurised water tank. It also allows increasing the efficiency of ORC cycle by approximately 18% (from 8.9% to 10.5%) with a higher net generated power than a conventional water tank integration, scaled up from 498 W to 1628 W. These results highlight the potential benefits that latent heat integration provides to improve the low-cost solar ORC performance for powering electricity and thermal energy supply. Abbreviations CHP combined heat and power CSHP combined solar heat and power ETC evacuated tube collectors MCH magnesium chloride hexahydrate mt melting temperature ORC Organic Rankine Cycle PCM phase change material PV photovoltaic PVT photovoltaic-thermal hybrid TES thermal energy storage 1. Introduction Around 2 billion people worldwide do not have access to electricity. Typically, these populations live in remote areas far from the centralized electricity grids and are characterized by a low income. Besides, for electricity transmission and distribution operators, the deployment of electricity grids is not economically advantageous, as they prefer to extend their activities in urban areas [1]. In recent years, Organic Rankine Cycle (ORC) has attracted considerable attention in the scientific researches and appears as a promising technology for conversion of heat into electricity [2,3], since it can be designed for operation at low temperatures with suitably-se- lected working fluids [4]. The heat for ORC can be provided from various sources: solar radiation, biomass combustion, geothermal heat or waste heat from industry [5,6]. Hence, ORC would be an appropriate https://doi.org/10.1016/j.est.2020.101367 Received 21 October 2019; Received in revised form 6 March 2020; Accepted 7 March 2020 ⁎ Corresponding author. E-mail addresses: [email protected] (J. Lizana), [email protected] (T. Rajabloo). Journal of Energy Storage 29 (2020) 101367 Available online 16 March 2020 2352-152X/ © 2020 Elsevier Ltd. All rights reserved. T option for remote areas. For distant regions with a lot of sunshine, a combination of ORC and photovoltaic panels (PV) can be a suitable alternative for electricity generation. Furthermore, a tailored model and fine management can provide both electricity and thermal energy for the local inhabitants. An overview of the ORC market evolution, considering the present installed capacity, historical data and macro-eco- nomic trends, was reported by Tartière et al. [7]. They highlighted the future perspectives and growth potential of the ORC market by putting a special focus on waste heat recovery applications. As a heat conversion technology, ORC is particularly suitable to increase the supply of renewable energy, mainly because of its ability to recover low-grade heat and the possibility to be implemented in decentralised low-capacity power plants [8]. A techno-economic survey of ORC systems was developed by Quoilin et al. [8], where was described as the state of the ORC technology with a particular emphasis on the temperature levels. In this study, a comparison with the traditional steam cycle revealed that ORC cycles are more appropriate for moderate power ranges and/or for low-temperature applications. Also, a compilation of the available market data showed that actual plant size is limited principally by a minimum power output of a few hundreds of kWe. For small-scale power range, there is no commercial solution developed of ORC systems. Low-capacity systems are currently under development or in the demonstration phase because they still require attractive markets to begin industrial production and reduce their cost [8]. Rahbar et al. [9] carried out a review of ORC for small scale applications, with a focus on ORC configurations, applications, working fluid selection, modelling and experimental study of the expansion devices. The results highlighted that most studies carried out about the ORC were mainly devoted to the selection of suitable working fluids for various applications, thermodynamic modelling and optimization of the ORC performance metrics. Manolakos et al. [10] studied a 2 kWe lowtemperature solar ORC with R134a as a working fluid and evacuated tube collectors. An overall efficiency below 4% was obtained. Wang et al. [11] studied a 1.6 kWe solar ORC using a rolling piston expander giving an overall efficiency of 4.2%. Kane [12] developed a model of a cascaded ORC using scroll expanders and coupled to a collector model. The ORC integration is also being explored as an alternative to photovoltaic (PV) systems. Freeman et al. [13] investigated the heating and power potential of a small-scale combined solar heat and power (CSHP) system based on an ORC for domestic use in the UK, in comparison with a PV-thermal hybrid (PVT) system (producing electricity and hot water). The cost per unit generating capacity of a CSHP-ORC system is found higher than a PVT system. However, the authors highlighted that the PVT alternative has a significantly reduced capacity for hot water provision. Small-scale solar organic cycles are well adapted for remote off-grid areas of developing countries. The performance and design optimization of a low-cost solar ORC for remote power generation was investigated by Quoilin et al. [14], showing that an overall electrical efficiency between 7% and 8% can be reached with conservative hypotheses. The main benefit of ORC power unit is that it can operate using “freely” available thermal energy obtained from renewables, that can be stored. Thus, storage is a key issue in solar ORC systems, since the implementation of thermal energy storage (TES) is arguably a key advantage over systems based on PV technologies [15]. TES can ensure stable and continuous operation in case of scarce solar radiation and enables dispatchability in the generation of electricity and home space heating requirements. Amongst the available TES alternatives, latent TES systems have recently received attention in the literature for ORC assemblies [4], as it is a more attractive solution than sensible storage because of the advantages of higher energy storage with a given volume, compactness and isothermal phase change [16]. Potential phase change material (PCM) candidates have demonstrated good performance for solar TES [17,18], showing benefits related to longer storage periods, reduced heat losses [19], and higher storage capacity for heating and hot water applications [20]. Bhagat et al. [21] developed a numerical analysis of latent heat TES using encapsulated PCM for a solar thermal power plant. It was found that the ability of the latent heat TES system to store and release energy is significantly improved by increasing mass flow rate and inlet charging temperature. Costa et al. [22] presented a solar salt latent heat thermal storage for a small ORC plant, operating at temperature levels between 220–230 °C. Numerical simulations indicated that the use of metal foams provide the required thermal performance. Cioccolanti et al. [23] proposed an innovative small-scale concentrated solar 2 kWe ORC plant coupled with a PCM tank equipped with reversible heat pipes. The work represents a general guide for the design and optimization of the mutual interactions of the different subsystems in small-scale concentrated solar ORC plants. Despite reported studies, efficiency numbers found in the literature of TES applications in ORC systems, at appropriate operating temperatures for the use of low-cost non-concentrating solar-thermal collectors, are reduced. Relevant studies must be accomplished to find an attractive solution for the PCM integration towards low-cost solar ORC configurations [24]. Aiming at overcoming the reported issues and defining criteria and strategies for the effective implementation of lowcost solar ORC units, this paper evaluates the performance of a smallscale solar ORC system integrated with evacuated tube collectors through three TES alternatives, two based on sensible storage and one in latent heat storage, with operating temperatures between 80 °C and 140 °C. The objective is to identify potential benefits and provide design criteria towards an optimal PCM implementation into low-cost solar ORC systems. Sensible heat storage alternatives consist of reference cases based on a conventional hot water tank and a pressurised water tank. They are compared with a latent heat storage unit, which implements a salt hydrate able to maximize the performance of ORC by storing and supplying high thermal energy amounts within a narrower and higher temperature range. Magnesium chloride hexahydrate was selected for the analysis due to a combination of favourable properties, such as high TES density, low-cost, availability and an appropriate melting temperature at 116 °C. The simulation was carried out through two models, developed in TRNSYS for solar energy balance and ASPEN for ORC system performance. The paper is structured as follows. First, the methodology for simulation is detailed. This was divided into two stages: solar energy balance in TRNSYS and ORC model in ASPEN. Second, the results of three schematic assemblies for solar ORC systems are compared. Finally, the viability of solar ORC alternatives is discussed. Nomenclature A area, m 2 c p specific heat, KJ/(kg.K) h heat of fusion, kJ/kg P pressure, bar Q heat transfer rate, W T Temperature, °C Subscript cond condensation evap evaporation in inlet l liquid out outdoor s solid J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 2 2. Methodology for numerical simulation The performance of small-scale solar ORC system is evaluated through three TES scenarios, characterised in Table 1. Scenario 1 and 2 consist of reference cases based on a solar ORC system integrated with a conventional hot water tank and a pressurised water tank, as sensible TES units. Scenario 3 implements a PCM tank based on magnesium chloride hexahydrate (MCH), a latent heat storage material with a melting temperature of 116 °C. Three storage tanks have the same TES capacity for a temperature difference of 20 °C. The numerical simulation was carried out through two models, developed in TRNSYS (Transient System Simulation) v18 [25] and ASPEN Plus simulation software [26], whose global schematic diagram is illustrated in Fig. 1. TRNSYS was used for the assessment of solar energy balance. It is a flowsheet simulator with a graphical interface which facilitates the decomposition of complex problems into various, interconnected model components. TRNSYS implements algebraic and first-order ordinary differential equations, describing physical components into software subroutines (called Types), with a standard interface. The TRNSYS model library includes components for the calculation of solar systems, thermal energy storage units or building thermal loads, most of them experimentally validated, as well as climatic data files, which make it a very suitable tool for a reliable energy modelling. ASPEN was used for the evaluation of ORC unit, using a validated ORC model previously reported in [27]. In this way, predefined components are settled in a close loop and simulation is done by defining required parameters, assumptions and strategies. Thermodynamic calculations through the program result in the amount of generated electricity. Then, cycle efficiency is calculated based on the results. Moreover, additional components are designed, integrated and analysed according to different requirements and alternative boundary conditions. Both numerical simulations were carried out separately, without a direct coupling. The results of the solar energy obtained in TRNSYS (Step 1) were implemented as input data in the ASPEN model (Step 2) to analyse the potential benefits and drawbacks of two solar ORC integrations. Pipe heat losses have been neglected in all scenarios. In the next subsections, the simulation procedures used in each energy model are further described. 2.1. Solar energy balance The evaluation of solar energy balance involves seven main components: weather data, solar collector, thermal energy storage, ORC heat exchanger, pump and controller. Figs. 2 and 3illustrate the two proposed scenarios of small-scale solar ORC system based on water tanks (scenario 1 and scenario 2) and PCM tank (scenario 3), respectively. Weather data were generated using Meteonorm and implemented within TRNSYS using the standard weather data reader component (Type 15–2). Seville was selected as a reference region for the simulation, with high solar availability (global average irradiance higher than 5.00 KWh/m 2 day) [28]. It is also characterised by a Mediterranean climate, with relatively mild winters and very warm summers. Solar collectors were modelled using Type 71. This component simulates the thermal performance of a variety of evacuated tube collectors (ETC). Solar collector data from standard tests of efficiency was used for simulation. Technical parameters of the solar system are summarised in Table 2. Moreover, the hot media from the solar collectors to the evaporator of ORC is a mixture of water and 30% ethylene glycol. As TES tanks, three alternatives are considered, whose technical parameters are defined in Table 1. Water storage tanks were modelled using Type 156. They have a volume of 750 L, with a loss coefficient of 0.7/m 2 K, according to reference commercial data. Heat loss rates of the modelled water tanks, considering 24 h temperature decay with an ambient temperature of 20 °C, are 6.38 and 8.77 kWh/day for scenario 1 and 2, respectively. Stratification was considered through three tank nodes. The thermal energy storage capacity of the water tanks for a temperature difference of 20 °C is 17.3 kWh. PCM tank was modelled through a new TRNSYS component. It is written in Fortran and compiled with the Fortran compiler. The developed model follows the structure of TRNSYS Type 156 based on a stratified fluid tank for sensible storage, in which sensible storage capacity was changed into latent heat storage. The latent heat performance is based on a mathematical procedure reported in the Report C5 of subtask C within the Task 32 of IEA Solar Heating and Cooling programme, about Simulation Models of PCM Storage Units [29], which has been tested and widely used in different PCM model proposals, such as the latent heat TES units based on PCM modules plunged into water tanks [30–32] or bulk PCM units [31,33]. This mathematical model is based on an enthalpy approach, in which the enthalpy is a continuous and invertible function of the temperature. It means that for a given volume and material, a continuous and reversible function can be calculated which will return the temperature (T) depending on the calculated enthalpy (h) [34]. Fig. 4 shows this function, which is modelled by a succession of three straight lines: two for the sensible heat in the solid and liquid phase, and one straight line in the phase change temperature region. This latent heat model requires six input parameters: specific heat capacity in solid and liquid stage (Cp s and Cp l , kJ/kg K), temperature range limit of phase change (T1 and T2, °C), heat of fusion (h, kJ/kg) and average thermal conductivity (λ, W/m K). Subcooling has been neglected in material performance. PCM selected for the latent heat storage unit is magnesium chloride hexahydrate (MCH, MgCl 2 •6H 2 O),tanking as reference the comparative review developed by Lizana et al. [35,36]. Thermo-physical properties of the selected PCM are reported in Table 3 [37,38]. It consists of a salt Table 1 Characteristics of three TES scenarios for small-scale solar ORC system. Characterisation Scenario 1: Water tank Scenario 2: Pressurised water tank Scenario 3: PCM tank Storage medium Water Water MCH Storage temperature ( °C) 80–100 °C 110–130 °C 110–130 °C (mt at 116 °C) Size (dm 3 ) 750 750 200 Tank height (m) 1.85 1.85 1.15 TES capacity (kWh)ª17.3 (80–100 °C) 17.3 (110–130 °C) 17.3 (110–130 °C) Fluid specific heat (kJ/kg.K) 4.19 4.19 2.0/2.4 (solid/liquid) Loss coefficient (W/m 2 .K) 0.7 0.7 0.7 Q loss (kWh/24 h) b ≈6.38 ≈8.77 ≈2.93 Tank pressure (bar) 1.0–3.0 >5.0 – a Storage capacity for a temperature difference of 20 °C. b Considering 24 h temperature decay from 100 °C for scenario 1, and from 130 °C for scenario 2 and 3, with an ambient temperature of 20 °C. J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 3 hydrate, which has high latent heat capacity per unit volume, moderate thermal conductivity, and little volume change during melting [39]. The selected inorganic component is characterised by a melting temperature of 116 °C, a heat storage capacity of 165 kJ/kg and a PCM cost of approximately 0.50 €/kg [40]. PCM tank is characterised by a volume of 200 L, with a loss coefficient of 0.7 W/m 2 K. Heat loss rate of modelled PCM tank is approximately 2.93 kWh/day, considering 24 h temperature decay from 130 °C with an ambient temperature of 20 °C. Stratification was considered through three tank nodes. The thermal energy storage capacity of PCM unit for a temperature difference of 20 °C during the phase change temperature is 17.3 kWh. Pump was modelled using Type 977, with a water flow of 320 kg/h. Other used components were diverting and mixing valves (Type 647 and 649), and different standard controllers and utilities. The operation modes of the solar systems are three: solar energy storage in TES tank, direct solar energy to ORC heat exchanger (excess of solar heat) and indirect solar energy to ORC heat exchanger (heat provided by TES tank). They are illustrated in Fig. 5. Programming of operational strategies and priority of operating modes are summarized in Table 4. The controllers update each parameter along each simulation time step (3 min), changing operation mode according to operational strategies implemented. The operating modes operate sequentially, is not allowed to operate simultaneously. 2.2. ORC model One of the main effective parameters to run an ORC is the properties of available heat source such as temperature and mass flow rate. In this study, the solar heat from collectors (direct heat) and storage tanks (indirect heat), previously defined and characterised, provide the related inputs for the ORC simulation. Since properties of heat source change throughout the year, some adaptations of the ORC model are introduced to evaluate the seasonal performance in selected scenarios. The selection of working fluid also has a great effect on the operation, efficiency and environmental impact of ORC. In this way, many aspects need to be taken into account, namely, thermodynamic properties, global warming potential (GWP), thermal stability, safety and environmental aspects, toxicity, flammability, auto-ignition temperature, costs, and availability, as it was discussed in several studies [3,41–44]. As a suitable working fluid for the cycle, with the mentioned heat source,isopentane was selected for the analysis of water-cooled ORC. In addition, the mixture of isopentane and isobutane at 50% wt was selected for air-cooled cycles. Fig. 1. Schematic diagram of solar-ORC system with thermal energy storage. Fig. 2. Numerical simulation model for scenario 1 and 2 (small-scale solar ORC with water tanks) developed in TRNSYS v18. J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 4 The performance of ORC model is evaluated through three TES scenarios, considering all involved components from the heat exchanger (evaporator) for ASPEN simulation, as defined in Fig. 1. The seasonal boundary conditions for proposed scenarios, reported from TRNSYS simulation, are summarised in Table 5. Monthly mean hot source temperature provided from solar panels ranges from 124 to 155 °C. Mean hot temperature in evaporator considering the proposed TES tanks are 85 °C, 113 °C and 113 °C for water tank, pressurised tank and PCM tank, respectively. The provided mass flow rate of the hot sources was equal to 320 kg/hr. Fig. 6 depicts a scheme of the ORC model in ASPEN. Taking into account reported boundary conditions, several ORC models were created at on-design conditions for the mentioned working temperatures. The following calculation assumptions were implemented in the ORC model: the mass flow of the working fluid and the cooling water is varied in order to match the 5 °C minimum temperature at the evaporator and the condenser, respectively; and the evaporating pressure was optimized in order to maximize the net power output of the plant. The type of cooling system for ORC will depend on the available resources. Two concepts of cooling systems, namely water cooling and air cooling, were analysed. The water-cooled cycle was studied through two alternatives, one with a mean water temperature of 11 °C, which Fig. 3. Numerical simulation model for scenario 3 (small-scale solar ORC with PCM tank) developed in TRNSYS v18. Table 2 Characteristics of solar collector. Parameters Evacuated tube collectors (ETC) Model Vaillant VTK 1140 Number of collectors 4 Aperture area per collector (m 2 ) 2 Fluid specific heat (kJ/kg.K) 3.918 Tested flow rate (kg/h.m 2 ) 24 l/h/m 2 Optical efficiency (n 0 ) 0.642 Heat loss coefficient a 1 (W/m 2 .K) 0.885 Heat loss coefficient a 2 (W/m 2 .K 2 ) 0.001 Weight (when empty) (kg) 37 Stagnation temperature (°C) 272 Collector slope (°) 37.23 TRNSYS Type Type 71 Fig. 4. Characterization of enthalpy curve implemented in new TRNSYS PCM type. Table 3 Thermo-physical properties of magnesium chloride hexahydrate (MCH, MgCl 2 •6H 2 O). Properties (solid/liquid). Properties Characterisation Melting temperature (°C) 116 Latent heat (kJ/kg) 165 Thermal conductivity (W/m.K) 0.58–0.70 Specific heat (kJ/kg.K) 2.00/2.40 Density (kg/m 3 ) 1570/1450 Price (€/kg) ≈0.5 J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 5 can be an option with cold underground water, and another one with a mean water supply temperature of 21 °C, as a reference value in the region under study. Moreover, the air-cooling concept was evaluated considering a mean ambient temperature of 20 °C, for scenarios in which water is not accessible. The related results give a better understanding of the effect of cooling properties on the cycle efficiency. Overall, air coolers result in lower power generation and cycle efficiencies in comparison to water cooling at the same cooling temperatures and pressures. Based on the work of Rajabloo et al. [27], in the case of air cooling, which requires high amounts of fan power, a mixture of the working fluid can be beneficial. As their results showed, temperature glide at condenser leads to lower required mass flow rate of air and less fan power consumption consequently. However, all possible alternatives are considered for solar-ORC implementation in this work. Fig. 5. Operating modes of small-scale solar ORC system with TES. Table 4 The linear operating structure of controllers. Operating mode Thermostats and controllers Mode 1. Solar energy storage Solar outlet - Upper dead-band dT: +10 °C (Type 2) Storage tank – T set-point 100 °C or 130 °C (Type 166) Mode 2. Direct solar energy Solar outlet - Upper dead-band dT: +10 °C (Type 2) (Excess of solar heat to ORC) Storage tank – Completely charged (Type 166) Mode 3. Indirect solar energy Solar outlet - Temperature < 80 °C (Type 166) (Heat provide by the tank to ORC) Storage tank – Temperature >80 °C or >110 °C (Type 166) J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 6 3. Results and discussions 3.1. Solar energy availability through TES alternatives Monthly solar energy availability for solar ORC with three TES alternatives are shown in Tables 6–8, and annual performance with regard to final energy balance and operating temperature ranges are compared in Fig. 7 and Table 9. Table 6 summaries the results of the monthly performance of the direct and indirect solar heat in Scenario 1 (Solar ORC with water tank). This water storage scenario results in a seasonal collector efficiency of 59%. Water tank stores 71% of useful collector gains, of which 49% is provided with an average supply temperature in the evaporator of 85 °C. The water tank is also characterised by a high percentage of heat losses (21%), associated with the large tank size in comparison with PCM tank. Table 7 summaries the results of the monthly performance of the direct and indirect solar heat in Scenario 2 (Solar ORC with pressurised water tank). The pressurised water tank results in a seasonal collector efficiency of 58%, derived from the higher operating temperature. Pressurised water tank stores 77% of useful collector gains, of which 45% is provided with an average supply temperature in the evaporator of 113 °C. The pressurised tank is also characterised by a high percentage of heat losses (31%), associated with the large tank size in comparison with PCM tank, and the higher storage temperature in comparison with scenario 1. Table 8 summaries the results of the monthly performance of the direct and indirect solar heat of integration in Scenario 3 (Solar ORC with PCM tank). This integration shows the same seasonal collector efficiency than pressurised water tank (58%), derived from the same operating temperature, between 110 °C and 130 °C. PCM tank stores 70% of useful collector gains, of which 54% is provided with an average supply temperature in the evaporator of 113 °C. PCM tank is also characterised by a low percentage of heat losses (11%), associated with the smaller tank size in comparison with water tanks. Fig. 7 compares the annual solar energy balance in three solar ORC scenarios, and Table 9 summaries the results of final energy balance and operating temperatures. Operating temperatures were analysed according to the temperature evolution during discharging periods in evaporator, with the calculation of mean temperature value, standard deviation (SD), 25th percentile and 75th percentile. The results show that PCM tank has higher benefits in comparison with water tank assemblies to enhance solar ORC performance. PCM tank can provide a higher capacity of useful solar heat in comparison with a pressurised water tank (+25%) and conventional hot water tank (+7%), with small heat losses of 11%, Heat losses are reduced by more than 50% in comparison with sensible heat storage alternatives. In addition, the annual mean temperature in the evaporator for ORC cycle is higher than conventional water tank assembly, and within a higher and narrower output temperature range than with a pressurised water tank, which enhances final ORC cycle efficiency. The temperature in evaporator during 50% of discharge time ranges between 111 °C and 114 °C for PCM tank, between 109 °C and 117 °C for pressurised water Table 5 Summary of mean operating conditions for ORC models. Characterisation Scenario 1: Solar ORC – Water tank Scenario 2: Solar ORC – Pressurised water tank Scenario 3: Solar ORC – PCM tank Storage medium Water Pressurised water MCH Mean temperature in the evaporator with direct heat (Solar heat from collectors) 147 °C 136 °C 137 °C Mean temperature in the evaporator with indirect heat (heat from storage tanks) 85 °C 113 °C 113 °C Hot source flow rate (kg/hr) 320 kg/hr 320 kg/hr 320 kg/hr Fig. 6. A schematic model of low-temperature ORC in ASPEN. Table 6 Monthly performance of the direct and indirect solar energy availability for ORC unit with water tank. Month Total solar radiation(kWh) Useful collector gain(kWh) Heat provided by tank to ORC(kWh) Mean temperature a (°C) Excess of heat from collector to ORC(kWh) Mean temperature a (°C) Water tank losses(kWh) January 1034 610 299 84 104 135 157 February 1016 599 348 84 100 140 152 March 1430 853 384 85 300 153 170 April 1326 774 390 85 223 155 165 May 1529 892 418 85 300 153 173 June 1513 879 416 85 299 149 168 July 1689 996 447 85 377 152 174 August 1606 958 447 85 342 151 174 September 1490 895 418 85 311 154 167 October 1272 762 384 85 212 147 169 November 898 521 294 84 70 141 159 December 916 530 296 85 78 140 163 Total/Mean 15,720 9270 4539 mean: 85 2716 mean: 147 1990 a Mean temperature in the evaporator under the same boundary conditions in all scenarios. J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 7 tank and between 80 °C and 89 °C for conventional water tank. Final useful solar heat availability of three solar storage scenarios, considering direct and indirect energy supply in comparison with useful collector gains, is 78%, 68% and 85%, for water, pressurised water and PCM, respectively. These results allow highlighting the potential benefits that latent heat integration provides to improve the ORC performance using solar energy. 3.2. Electricity availability of ORC through TES alternatives The results of ORC models carried out in ASPEN according to the boundary conditions reported from TRNSYS simulations for Scenario 1 (Solar ORC – Water tank) and Scenario 2 (Solar ORC - PCM tank) are summarised in Tables 10-12. The ORC models consider an average direct and indirect solar source temperatures according to the results previously provided in Table 6,7and 8. Table 10 shows the performance of ORC water-cooled model for Scenario 1 (Solar ORC – Water tank) with a parametric analysis concerning the evaporation pressure. It only considers the use of indirect heat from the storage tank with a mean hot source temperature in the evaporator of 85 °C and cooling water at 11 °C. The results show that although optimum net power occurs at 7 bar, the return temperature was too low. The recoverable heat duty and evaporation pressure have a constraint on the return temperature, which should be higher than 70 °C to have stable functionality of the storage tank. The optimum achievable evaporation pressure was 10.85 bar for the hot source temperature of 85 °C, which provides a final ORC efficiency higher than 11%. Therefore, the optimum evaporation pressure and heat duty are implemented in this margin. The same boundaries were applied to the ORC model with a hot temperature of 113 °C, associated with Scenario 2 and 3 (indirect heat from pressurised water tank and PCM tank), and considering an average temperature in the evaporator of 130 °C, as a reference value for direct solar heating in three scenarios. In addition, as a conservative value of water-cooled cycle, the model was run with a mean water supply temperature of 21 °C, as a reference value in the region under study. The results at optimum evaporation pressure regarding solar tank constraints are available in Table 11. This parametric analysis, with the optimum evaporation pressure for each scenario, highlights that with higher hot source temperature and higher evaporation pressures, higher net generated power and cycle efficiencies are obtained, ranging for Scenario 1 (Water tank) from 8.9 to 11.7%, and for Scenario 2 and 3 (Pressurised water tank and PCM tank) from 10.5 to 11.7%. Comparing the ORC operation by indirect heat from TES tanks, the results show that pressurised water and latent heat storage can increase the ORC efficiency by up to 18% (from 8.9% to 10.5%) and the net generated power from 498.4 W to 1682.7 W, scaled up by 238%. The air-cooling possibility for ORC working with the same working fluid and heat resource properties was investigated, with a mean ambient temperature at the studied zone of 20 °C. Moreover, the possibility of air cooling of ORC was also studied by implementing a mixture of isobutane and isopentane at 50% wt as working fluid. The results are available in Table 12. Since the mixture of working fluids causes temperature glide during the evaporation and condensation, evaporation pressures were less. Hence, less power was generated in the turbine leading to fewer cycle efficiencies. It should be also considered that despite reduced Table 7 Monthly performance of the direct and indirect solar energy availability for ORC unit with pressurised water tank. Month Total solar radiation(kWh) Useful collector gain(kWh) Heat provided by tank to ORC(kWh) Mean temperature a (°C) Excess of heat from collector to ORC(kWh) Mean temperature a (°C) Water tank losses(kWh) January 1034 598 239 113 61 129 220 February 1016 585 302 112 64 128 217 March 1430 842 358 113 240 143 243 April 1326 762 349 113 175 140 236 May 1529 881 398 113 235 139 246 June 1513 870 397 113 233 135 239 July 1689 987 430 113 308 139 248 August 1606 949 424 113 277 139 247 September 1490 886 407 114 240 140 238 October 1272 750 339 113 167 138 243 November 898 507 239 112 39 131 228 December 916 518 259 113 31 126 234 Total/Mean 15,720 9133 4142 mean: 113 2070 mean: 136 2840 a In evaporator under the same boundary conditions in all scenarios. Table 8 Monthly performance of the direct and indirect solar energy availability for ORC unit with PCM tank. Month Total solar radiation(kWh) Useful collector gain(kWh) Heat provided by tank to ORC(kWh) Mean temperature a (°C) Excess of heat by collector to ORC(kWh) Mean temperature a (°C) PCM Tank losses(kWh) January 1034 585 343 113 119 144 79 February 1016 583 376 113 107 129 74 March 1430 840 420 113 307 142 83 April 1326 760 413 113 237 138 80 May 1529 876 450 113 308 140 84 June 1513 863 443 113 309 137 81 July 1689 985 481 113 384 140 84 August 1606 947 480 113 351 139 84 September 1490 884 454 113 318 140 81 October 1272 747 415 113 220 138 82 November 898 503 320 113 78 129 77 December 916 518 327 113 85 124 79 Total/Mean 15,720 9090 4921 mean: 113 2823 mean: 137 968 a In evaporator under the same boundary conditions in all scenarios. J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 8 efficiencies by using air-cooled ORC, the higher working temperature of PCM tank provides higher efficiencies throughout the operation. 3.3. Economic analysis The total investment cost of three proposed small-scale solar ORC alternatives are evaluated as shown in Table 13, using cost data gathered from different cost database [45–47], market survey and scientific publications. Solar collectors based on ETC costs 6360€, with a specific cost ratio of approximately 795 €/m 2 , higher than flat plate collectors situated in 260 €/m 2 [47]. In this case, as temperatures higher than 130 °C are needed, ETC was selected instead of flat plate collectors to increase the seasonal efficiency of the system, increasing the final cost. The investment cost of the water tank and pressurised water tank, for a tank size of 750dm 3 , is approximately 2300€ and 3000€, respectively [47]. The higher cost of scenario 2 is mainly associated with the need for pressure requirements higher than 5 bar in order to remain water in liquid phase up to 150 °C [48,49]. In the case of PCM tank, the investment cost is reduced due to a smaller size of 200dm 3 . However, it has additional requirements in comparison with a water tank, associated with the need for a higher heat transfer ratio between the PCM and the heat transfer fluid [50], corrosion protection for the specific use of salt hydrates [51] and the PCM cost, which in this case is quite competitive, approximately 0.50 €/kg [40]. This configuration results in a PCM tank very competitive in comparison with water scenarios, with a reduction of investment cost by 35% and 50% compared to the water tank and pressurised water tank, respectively. In the case of ORC unit, the specific cost (€/kW) for small-scale applications is still too high [52]. The specific cost of ORC units is hardly below 5000 €/kW for power sizes below 10 kW [8]. Previous studies showed an investment cost ratio of 5000 €/kW [53], 5775 €/kW [54] and 5833 €/kW [15]. According to Tocci et al. [52], the specific cost of ORC should not exceed the value of 3500€/kW in the power range below 10 kW to consider it competitive. The remaining cost is Fig. 7. Comparison of annual thermal energy availability of solar ORC schemes with water tanks and PCM tank. Table 9 Summary of results in three solar TES scenarios. Performance indicators Scenario 1:Solar ORC – Water tank Scenario 2:Solar ORC – Pressurised water tank Scenario 3:Solar ORC - PCM tank Storage medium Water Pressurised water MCH Performance of solar collectors Useful collector gain 9270 kWh (100%) 9133 kWh (100%) 9090 kWh (100%) Seasonal collector efficiency 59% 58% 58% Useful solar heat (direct and indirect) 7255 kWh (78%) a 6211 kWh (68%) a 7745 kWh (85%) a Energy balance of TES tanks Indirect useful heat (heat provided by tank) 4539 kWh (49%) a 4142 kWh (45%) a 4921 kWh (54%) a Tank losses 1990 kWh (21%) a 2840 kWh (31%) a 968 kWh (11%) a Temperature evaluation in the evaporator (during discharge) b Average evaporator temperature °C 85 °C 113 °C 113 °C Standard deviation (SD) 5 5 3 25th percentile 80 °C 109 °C 111 °C 75th percentile 89 °C 117 °C 114 °C a Percentages calculated in relation with useful collector gains. b Comparison of temperatures in the evaporator under the same boundary conditions in all scenarios. Table 10 Results of water-cooled ORC with T hot source = 85 °C, T cold water = 11 °C, P cond = 2.7 bar, and, M hot source = 320 kg/hr. P evap. (bar) W net (kW) Q in,evap (kW) Efficiency (%) T out,evap ( °C) 11.0 586 5285 11.09 70.8 10.9 613 5568 11.01 70.0 10.0 774 7390 10.48 65.1 9.0 938 9596 9.77 59.2 8.0 1043 11,645 8.95 53.6 7.0 1104 13,814 7.99 47.8 6.0 1097 16,074 6.82 41.7 Table 11 Results of water-cooled ORC with T cold water = 21 °C, P cond = 3.7 bar, and, M hot source = 320 kg/hr. Hot source temperature ( °C) P evap. (bar) W net (W) Q in,evap (W) Thermal efficiency (%) 85 10.9 498.4 5597.9 8.9 113 13.6 1682.7 16,049.7 10.5 130 16.8 2637.5 22,462.1 11.7 J. Lizana, et al. Journal of Energy Storage 29 (2020) 101367 9