Fresnel solar cooling plant for buildings: Optimal operation of an absorption chiller through inverse modelling

Energy Reports 8 (2022) 3189–3212 Contents lists available at ScienceDirect Energy Reports journal homepage: www.elsevier.com/locate/egyr Research paper Fresnel solar cooling plant for buildings: Optimal operation of an absorption chiller through inverse modelling MCarmen Guerrero Delgadoa, Jose Sánchez Ramosa, Daniel Castro Medinaa, Teresa Rocío Palomo Amoresa, Alberto Cerezo-Narváezb,∗, Servando Álvarez Domíngueza aGrupo Termotecnia, Escuela Técnica Superior de Ingeniería, Universidad de Sevilla, Camino de los Descubrimientos S/N, 41092, Seville, Spain bDepartamento de Ingeniería Mecánica y Diseño Industrial, Escuela Superior de Ingeniería, Universidad de Cádiz, Avenida Universidad de Cádiz N◦10, 11519, Puerto Real, Cádiz, Spain article info Article history: Received 16 October 2021 Received in revised form 31 December 2021 Accepted 10 February 2022 Available online xxxx Keywords: Solar cooling plant Absorption chiller Double effect Inverse modelling Renewable energy integration abstract Increasing comfort conditions in buildings imply higher energy demands. However, these needs can be mitigated by solar cooling solutions. These systems, such as absorption chillers, are complex and require stable operation, with strict control to maximise the solar fraction and minimise gas consumption. This is incompatible with the variability of renewable resources, so they are often coupled with auxiliary gas systems. Although gas-free operation is possible if these systems are optimally controlled, they would require special supervision. This paper aims to develop an experimental validation of an inverse model to manage an absorption chiller coupled with a solar cooling plant. To know its real behaviour, long-term experiments have been performed using this plant, which consists of a linear Fresnel solar collector and an auxiliary natural gas boiler. The inverse model is used as a predictive control tool to decide the auxiliary boiler commands of the absorption chiller to optimise its operation: maximum cooling production by minimising gas consumption and maximising solar contribution. It has been identified with data from two weeks and validated with data from one summer month. Results show that the model estimates, on a time base of fewer than 30 min, are acceptable with errors of less than 5%. In addition, the maximum error of the estimated seasonal COP and the renewable fraction are less than 6% per day. Therefore, the results prove the usefulness of the proposal as a predictive control for optimal operation. Furthermore, it could be used as a baseline for preventive maintenance. If the proposed model is used for optimal management of the absorption chiller, the thermal efficiency of the plant increases significantly, doubling the solar contribution. As a result, the gas consumption of the solar cooling plant is halved and the total cost of air conditioning the building decreases by 16%. ©2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction International Energy Agency estimates that a third of CO2 emissions come from buildings. If the population continues increasing, energy needs could rise to 50% in 2050 (International Energy Agency,2013). New energy-saving regulations and growing awareness in society lead to the need for decreasing energy consumption in buildings (European Commission,2018). In addition, the cooling consumption of buildings represents a considerable fraction of total energy consumption in the world (Santamouris et al.,2004). Therefore, emphasis must be placed on reducing cooling needs and meeting the remaining needs with highly efficient equipment powered by renewable energy sources. ∗Corresponding author. E-mail address: [email protected] (A. Cerezo-Narváez). Considering that peak cooling demand occurs when solar energy potential is at its highest level, solar cooling would be an attractive alternative to reduce the environmental impact of conventional cooling systems (Bataineh and Taamneh,2016). However, a review of the studies carried out shows that, although the technology exists in the market, its implementation in real buildings is still scarce. One of these limitations is the control of the absorption chiller operation (Ahn et al.,2019). In addition, the hybridisation of these chillers with solar cooling systems leads to changes in their performance due to the different characteristics of solar energy compared to auxiliary energy sources. However, many efforts have been made to discuss the performance of these systems from different points of view, including energy (Karki et al.,2019), economic (Leckner and Zmeureanu, 2011) and environmental impact (Chen et al.,2020). Therefore, different methodologies have been used to optimise operation https://doi.org/10.1016/j.egyr.2022.02.128 2352-4847/©2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Nomenclature A Area (m2) BMS Building Management System (-) Cp,w Specific heat of water (J/kg K) CAPnom Nominal power expressed in EUROVENT conditions (W) Coldreal Cooling capacity in real operating conditions (kW) Coldnom Cooling capacity in nominal operating conditions (kW) ConGasHreal Gas consumption in heating mode in real operating conditions (kW) ConGasHnom Gas consumption in heating mode in nominal operating conditions (kW) ConElecHreal Electrical consumption in heating mode in real operating conditions (kW) ConElecHnom Electrical consumption in heating mode in nominal operating conditions (kW) ConGasCreal Gas consumption in cooling mode in real operating conditions (kW) ConGasCnom Gas consumption in cooling mode in nominal operating conditions (kW) ConElecCreal Electrical consumption in cooling mode (kW) ConElecCnom Electrical consumption in cooling mode in nominal operating conditions (kW) COP Coefficient Of Performance (-) COP(t) Relationship between supplied and consumed power (-) DDt’ Adapted Characteristic Equation model (-) DEAM Double Effect Absorption Machine (-) DHW Domestic Hot Water (-) EER Energy Efficiency Ratio (-) EES Engineering Equation Solver (-) EHP Electric Heat Pump (-) EPBD Energy Performance of Buildings Directive (-) FCAP Correction factor that corrects the cooling capacity based on the operating temperature (-) FCON−1Correction factor of the efficiency curve as a function of the condenser inlet temperature (-) FCON−2Correction factor of the efficiency curve as a function of the partial load factor at which the machine works (-) FDD Fault Detection and Diagnosis (-) GEHP Gas Engine Heat Pump (-) GNA Adapted Gordon-Ng model (-) H2O Water (-) Heatreal Heat capacity in real operating conditions (kW) Heatnom Heat capacity in nominal operating conditions (kW) HTG(t) Generator temperature (◦C) HVAC Heating, Ventilation and Air Conditioning (-) LiBr Lithium bromide (-) ˙ mgen Flow rate (kg/s) ˙ mcond(t) Fluid flow rate through the condenser (kg/s) mrec Bypass flow (kg/s) mwWater flow (kg/s) MCp Thermal mass (J/K) MCpon Thermal mass found in the generator (J/K) MCpoff Thermal mass in the generator taking into account the calorific coefficient of that mixture at stop (J/K) MPR Multivariate Polynomial Regression model (-) NG Natural gas (-) NZEB Nearly Zero Energy Building (-) pCAPi Coefficient that defines the correction curve (-) pCONi−jCoefficient of the fit curve for the correction factors (-) PCM Phase-change material (-) PLR Part load ratio (%) Qcond(t) Power supplied by the condenser (W) Qcooling Cooling power delivered by the machine (W) Qext (t) Necessary heat to be dissipated from the generator for the correct operation of the absorption machine (W) Qgen,nom Nominal power of the generator (W) QRECOVERY Heat recovery from cooling system of engine (kW) RecQHreal Heat recovery in heating mode in real operating conditions (kW) RecQHnom Heat recovery in heating mode in nominal operating conditions (kW) RecQCreal Heat recovery in cooling mode in real operating conditions (kW) RecQCnom Heat recovery in cooling mode in nominal operating conditions (kW) RER Renewable Energy Ratio (-) SEER Seasonal Energy Efficiency Ratio (-) SCOP Seasonal Coefficient Of Performance (-) SCR Solar cooling ratio (-) SHF Solar heat fraction (-) τenf Time constant (s−1) τon Generator start time constant (s−1) τoff Generator shutdown time constant (s−1) Tdr Outdoor dry-bulb temperature (◦C) TwbOutdoor wet-bulb temperature (◦C) Tevap,iInlet water temperature to the evaporator (◦C) Tevap,oOutlet water temperature from the evaporator (◦C) Ton Starting temperature of the generator (◦C) TgGenerator temperature (◦C) Tg,min Minimum working temperature in the generator (◦C) Tgen,i,nom Nominal operating temperature of the generator (◦C) Tg(t-∆t) Generator temperature at a previous time (◦C) 3190 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Tgen,i(t) Fluid temperature at the generator inlet (◦C) Tgen,o(t) Fluid temperature at the generator outlet (◦C) Tm(t) Average fluid temperature between the inlet and outlet to the generator (◦C) Tcond,iInlet water temperature of the dissipation loop (◦C) Tcond,o(t) Condenser outlet temperature (◦C) Tirec Inlet water temperature of recovery loop (◦C) Torec Outlet water temperature of recovery loop (◦C) TiwInlet water temperature (◦C) TowOutlet water temperature (◦C) TLIM−REC Minimum outdoor temperature to put on recovery system of GEHP (◦C) TSETPOINT Set point temperature (◦C) TRNSYS Transient System Simulation Tool (-) UOverall heat transfer coefficient (W/m2 K) UA Overall heat transfer coefficient per generator area (W/K) UAon Overall transfer coefficient through generator area at start-up (W/k) UAoff Overall heat transfer coefficient through the generator area at stop (W/k) strategies (Araújo and Silva,2020) and system configurations,10 to improve the performance of hybrid solar cooling systems. Absorption chillers are being used in most of the solar cooling systems in operation today (Prieto et al.,2019). The typical range of cooling capacities for commercial absorption chillers is from 5 to 1500 kW (Xu and Wang,2017). Thermal compression of the refrigerant is achieved by using a liquid refrigerant (absorbent) solution and a heat source, which replaces the electrical consumption of a mechanical compressor. There are two configurations for absorption systems: single effect and double effect. The main difference between them is the operation temperature, which is lower in single effect ones. In addition, double-effect absorption chillers may have a higher COP due to possible energy recovery (Wang et al.,2017). As mentioned above, control of the absorption machine operation is required to optimise system performance. In first, the operation of the system should be controlled. For example, the weak point of this solution is the risk of crystallisation (Jradi and Riffat,2014). For that, the dissipation water temperature must be controlled. On the other hand, a solar cooling plant requires efficiency control to maximise the renewable contribution. Modelling can be considered an excellent solution for operating this chiller (Jradi and Riffat,2014) and predicting solar production (Rivarolo et al.,2013). When modelling the performance of an absorption chiller machine, the coupling between the device and the solar collector field must not be forgotten. This coupling condition consists of the thermal power produced by the solar field being greater than or equal to that required by the absorption chiller generator (Sarabia Escriva et al.,2011). This issue has also been studied by Evola et al. (2013), and they highlighted the relevance of this problem for absorption chillers. In addition, the high mass of the internal components and the accumulation of fluids inside the vessels usually make the transient period longer than in the case of mechanical compression. Therefore, authors as Sebastián et al. (2021) demonstrated the need for well-characterised inertia models in short time steps (less than 30 min) for decision making in solar cooling plants. Previous works showed the existing research gap in the literature on the lack of solution about the coupling between the absortium chiller model and solar sources. Shirazi et al. (2018) conducted a comprehensive literature review on the current status of this technology and its modelling. This work showed the lack of experimental validation of many publications and how the modelling streams are either TRNSYS types (Klein,2010) or models with detailed thermodynamic processes. According to Martínez et al. (2016), there are primarily two approaches to model an absorption chiller. The first option (detailed models) consists of adding all thermodynamic characteristics of the chiller into a set of simple algebraic equations. The second option (simplified models) uses heat transfer coefficients (U), heat exchange areas (A) and working fluid characteristics to simulate machine behaviour. In this second case, mass and energy balance equations are established for each cycle that considers the different streams that converge in each component, defining their thermodynamic behaviour. Table 1 summarises a review of absorption chiller characterisation models. The most recurrent option in the reviewed literature is to use simplified models for the characterisation of absorption chillers. However, the methodologies presented do not clearly explain the formulation that allows the integration of manufacturer data or the thermal parameters obtained through experiments into these models. For example, it does not appear how to consider a machine of better performance and therefore higher cost than another of low performance. The existing literature (Nikbakhti et al.,2020) also presents modelling proposals for absorption chillers but does not provide a solution for their integration with solar systems, avoiding connecting them to the management system for optimal operation. The present work provides a simplified model based on characteristic parameters that can be identified from manufacturer or experimental data. In addition, this proposal solves the integration with renewable sources in a simple way. Table 1 also shows how simplified models could present a lack of cohesion between the experimental data, the actual performance of the machines, and the input/output data of these models. To close this gap, seasonal values can be used as the key performance indicator to evaluate the acceptability of the estimation or the real benefits of facilities (Kylili and Fokaides, 2015). For example, the most frequent are the thermal efficiency of the chiller or the percentage of renewable contribution using the RER of the building (Nikbakhti et al.,2020). Recent publications, such as Bilardo et al. (2020), discuss the limitations of using simulations with the current state of models. Simplified models remain the most compatible with transient energy performance simulation tools (Castro et al.,2020). Many of these renewable technology models appear in the databases of TRNSYS (Klein, 2010) or EnergyPlus (Anon,2018). Although the models before have not been updated for more than 20 years, and they are not taking into account the progress of these technologies. These simplified models characterise HVAC systems by the use of performance curves (Ray et al.,2021). Sánchez Ramos et al. (2019) demonstrated the validity of simplified models based on operating curves for equipment characterisation without high thermal inertia, such as natural gas heat pumps. However, these proposals should not be used for real-time optimal control due to their low accuracy in simulation with short time steps (Lubis et al.,2018). Nevertheless, the scientific community is accepting the use of simplified models if their estimation of main key performance indicators is acceptable. According to the above, there is a lack of research on this topic. However, this research addresses the experimental quantification 3191 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Table 1 Review of absorption chiller characterisation models. Reference Year Type of model Summary Martínez et al. (2016) 2016 Simplified They proposed a simulation model in the EES software environment that considered the internal characteristics of the absorption chiller. With this information, a weak solution flow rate and UA values for each chiller heat exchanger were estimated to reproduce consistent results that agreed with the available experimental data. Florides et al. (2002) 2002 Simplified They used the TRNSYS software environment using a new Type to simulate this system. The authors stressed the importance of providing measured input data, which the model should consider in order to account for measurable data. Zhao et al. (2020) 2020 Detailed They proposed a detailed model from which the different elements of a refrigeration machine could be extracted. The elements were modelled by their thermodynamic formulation. Chen et al. (2017) 2017 Simplified They developed a TRNSYS type, which was validated with experimental data. This comparison was performed by exciting the model with the input data measured on the experimental setup. The authors commented on the difficulty of establishing a proper definition of the absorption machine based on models proposed in the literature. Yu et al. (2019) 2019 Simplified They worked with the dynamic characterisation of the COP of an absorption machine in an experimental facility. The measured and estimated instantaneous results were different. However, the estimation of seasonal performance was acceptable. This work consolidates the idea that simplified models are required but considering the dynamics of these machines. Ibrahim et al. (2017) 2017 Simplified They studied an experimental facility attached to a building with thermal storage. Their absorption system model was limited to assessing the machine COP as a function of operating conditions. The authors commented on the non-validity of the model on a shorter time basis, highlighting the need to consider the start-up and shutdown dynamics of this type of chillers. Castro et al. (2020) 2020 Detailed They used a detailed modelling pathway to perform sensitivity studies for enabling the optimal design of this type of chillers (manufacturer). They obtained the requirements for the future modelisation and the monitoring system to develop an internal control for these chillers. of this control and management. For this purpose, the first goal of this research is to develop an inverse model for the absorption chiller, which enables its calibration with real operating data. It should be noted that the coupling between the plant elements (without optimising) has been solved in previous research (Gallego et al.,2020). In this context, the accuracy of the model is a critical parameter, as it is linked to the ability to analyse the transient effect of these technologies. Therefore, this is considered a point of great importance in order to optimise their operation. Another key element in a solar cooling plant is the absorption chiller (Shirazi et al.,2018), as it is connected to the natural gas consumption of the auxiliary boiler and the renewable energy source. Consequently, if it is not optimally operated, the solar contribution decreases significantly (Jalalizadeh et al.,2021). Recent reviews, Settino et al. (2018) have concluded that the validation of an absorption technology in a controlled environment is different from that of the same technology but coupled to a solar cooling plant. It is because solar plants are exposed to many variations in renewable resources, which generates a greater need for control in a short time, as Song et al. (2020) proved. Traditionally, linear and nonlinear models have been considered the best option (Shirazi et al.,2018) to calibrate using experimental data. Both are based on statistical errors when comparing real production with theoretical statistical production (Gallego et al., 2020). At present, artificial intelligence techniques are undergoing massive development to analyse complex technologies such as solar systems (Mao et al.,2020). In this context, Amiri Rad and Davoodi (2021) and Ali and Ratismith (2021) worked with simplified models, in which these systems are coupled into higher-order entities (in their case, thorny plants). Sharafi et al. (2015) optimise this renewable ratio and conclude that transient models with low computational overhead are needed to achieve seasonal results and generate robust transient results such as required peak powers. This research proposes a new model that combines equipment performance curves from experimental data provided by manufacturers with ‘‘in situ’’ experiments that include fast start up and shut down effects for the absorption chiller. Nevertheless, these authors did not solve the model for the coupling between performance curves and transient effects. Furthermore, the second goal of this research is to contribute to the existing literature by proposing an optimised model for the characterisation of absorption chillers coupled to solar systems. In addition, the inverse model is integrated into the management system as a predictive control to support the operator decision. The validation of the model is based on an experimental installation of solar cooling in a Fresnel collector attached to a university building for air conditioning. An entire solar cooling plant is operated to achieve these research goals. These experimental results are interesting and will be fully analysed throughout the paper. 2. Method This research intends to couple an absorption chiller with the energy management model of a solar cooling plant through an inverse model. This model is validated using a solar cooling plant. The experimental facility is coupled with a real building and operated in actual conditions. Furthermore, this research provides the identification, design and development of an inverse model for the chiller. In addition, a decision support system performs this coupling as a predictive control for the plant. Therefore, the main contributions of this paper are: the proposed model to characterise the absorption chiller coupled to the solar cooling plant, the experimental results obtained during the validation of the model, and the results linked with the application of the model as a predictive control for the plant. Fig. 1 shows the context of this research: the renewable thermal production system with storage and the building that requires this energy for cooling. For this purpose, a double-effect absorption machine receives heat both from renewable resources and a backup boiler coupled to it. This support boiler operates in accordance with the availability of renewable resources and the working instructions of the absorption machine. Therefore, if the control management is not performed intelligently to maintain the operating instructions of the absorption machine, uncontrolled entry from the natural gas support system may occur. For this reason, a characterisation model is needed to infer the response of the absorption machine to variations in building cooling 3192 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 1. Scheme of the methodology. needs and renewable resources. Consequently, gas consumption can be minimised and the use of renewable resources maximised because the inverse model allows deciding the best command to operate the absorption chiller. The proposed model is validated using available experimental data of real system operation. As mentioned above, a model of the behaviour of the absorption machine from the behaviour curves is defined. Both working temperatures and parameters of the absorption machine are used as variables. The developed model requires data from the manufacturer (curves and nominal parameters) and flow rates and temperatures of the input flows (solar generator, dissipation system and building cooling circuit). Then, the model estimates the outlet temperature of each flow and the possible gas consumption required, according to the control signal received (predictive control). The inverse model is executed every 30 min to evaluate the operation strategy for the absorption chiller. The inputs and outputs of the model will be detailed in the next section. 3. Experimental solar cooling plant 3.1. Overview of the solar cooling plant A pilot solar cooling plant was developed to test a costefficient alternative to grid power consumption, taking advantage of the high insolation levels in the Spanish region of Andalusia. The plant was coupled to an existing compression chiller cooling system. Both systems are located on the roof of a university building. Its main components are summarised in Fig. 2. The main aim of this plant is to prove the feasibility of significantly reducing the high electrical energy consumption of buildings for air conditioning by using solar energy. The plant has been operating since 2008. It has been tested and monitored during this period to obtain data to evidence the feasibility of the concept, refine the design and operating practices, and assess its potential economic and environmental value. The double effect absorption chiller (see Fig. 2 left) uses LiBr–H2O as a refrigerant. The generator is powered by the heat coming from a linear Fresnel solar collector (see Fig. 2 right), which is able to reach a water temperature of up to 180 ◦C, thanks to a solar tracking system implemented in its mirrors. If the solar field cannot reach the temperature required for the operation of the absorption machine, a PCM storage tank (see Fig. 2 centre) is available for use. The cooling capacity of the system is 174 (kW). The proximity of the Guadalquivir river provides the source of cooling water used in an open-loop since the river water temperature during the summer is about 27 ◦C. Fig. 3 shows the different subsystems within the plant that can be distinguished, which require a detailed description. Their synergy provides the energy direction of the plant design. Table 2 shows the parameters that characterise the Fresnel solar collector. In addition, the Fresnel solar collector starts tracking the sun when direct solar radiance exceeds 200 W/m2. This system is shown in detail in Fig. 4. The highest allowable temperature in the tube is 180 ◦C. If the temperature exceeds 190 ◦C, 3193 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 2. Components of the solar pilot cooling plant installed at the Engineering School of Seville (Spain) (left: Absorption Chiller, centre: PCM Storage Tank, right: Linear Fresnel Solar Collector). Fig. 3. Diagram of the pilot solar cooling plant installed at the Engineering School of Seville (Spain). Table 2 Fresnel solar collector parameters. Ground surface 512 m2 Solar field 352 m2 Orientation East–West (azimut of solar collector is 18◦ from North) Absorber tube length 64 m Receiver height 4 m Secondary width 0.3 m Absorber tube model SCHOTT PTR©70 Thermal fluid Liquid water Operating temperature 180 ◦C Operating pressure 13 bar Mirror dimension 2 ×0.5 m2 Mirror reflectivity 0.92 Collector concentration 25 all mirrors are automatically turned upside down to protect the system. The temperature is controlled to be maintained at a set point which, in the case shown (see Fig. 4 right), was 170 ◦C, although it is normally 180 ◦C. For a water flow of 12 m3/h, the average temperature gap between inlet and outlet was 7.2 ◦C, which corresponds to a solar thermal power of 100 kW. This power is used to provide heat to the chiller generator. The Fresnel solar collector and the absorption chiller are connected through a 150 m long pipeline and a 6 m3PCM storage tank. This tank is 18 m long and 1.31 m in diameter. It is a shell and tube heat exchanger with a theoretical thermal storage capacity of 275.5 kWh, corresponding to a temperature range of 145–180 ◦C. The substance used in the tank is hydroquinone (3500 kg), with a melting point of 170 ◦C. The absorption chiller of the plant only covers 30% of the total cooling needs of the building. However, it should be noted that this is a pilot plant. The absorption machine installed corresponds to the BROAD-BZH15 model, which has a nominal power of 174 kW and a theoretical COP of 1.34. The high-temperature generator receives hot water from the solar field, with a temperature range that is usually between 140–170 ◦C, with a nominal flow of 13 m3/h. If the temperature is lower than 145 ◦C, the absorption machine starts the auxiliary system, which uses natural gas as fuel. The evaporator is an exchanger that cools the water from 14 ◦C to 7 ◦C with a nominal flow rate of 30 m3/h. In the nominal operating regime, the temperature of the high-pressure generator is close to 145 ◦C, and most of the thermal power supplied to the machine comes from the solar collector water. The cooling capacity is maintained at around 135 kW (75% of its nominal capacity). 3194 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 4. Linear Fresnel solar collector installed at the Engineering School of Seville (Spain). The plant runs in different operating modes. In Fig. 5, those used in the experiments performed during a full cooling season of the building (from May to October) are shown: – Mode 1. Plant start-up. When direct solar irradiation is higher than 200 W/m2, the solar collector starts in recirculation mode. This recirculation is activated until the water temperature in the primary circuit is higher than 100 ◦C. At this point, Mode 2 is activated. – Mode 2. PCM tank preheating. The solar loop opens, and water flows into the PCM tank. The thermal storage system is discharged by heating the water until it reaches a temperature higher than 155 ◦C. – Mode 3. Chiller start-up. The absorption machine is turned on when the water temperature is higher than or equal to 155 ◦C. Then, the water access to the PCM tank is closed. On a typical day, the absorption chiller starts up at midday. When the water temperature of the solar collectors is enough (at least 155 ◦C), the control system opens the valve, allowing the generator to preheat. However, if the generator temperature drops below 145 ◦C, the natural gas burner starts up. It should be noted that the internal control of the chiller system prioritises the temperature level of the generator. This is why it is so important to predict its temperature in order to minimise the use of the gas burner. – Mode 4. Energy storage. If the outlet temperature of the solar collector is higher than 165 ◦C or the cooling needs of the building decrease, the PCM tank opens to be charged with the excess heat from the solar collector. There is also a Mode 5 that operates on weekend. The system is switched on to store heat in the PCM tank. For that, the water from the solar field is recirculated until it reaches 180 ◦C in a closed loop. At that point, access to the PCM tank is opened. This mode remains active as long as the inlet temperature to the PCM tank is 2 ◦C higher than the outlet temperature. On the other hand, it can be noted that the chiller start-up is possible before Mode 3 is activated. For this purpose, the support of the gas boiler is required. It is usually used in the early and late afternoon. This additional option is a further advantage to dispose of a model for predicting the behaviour of the chiller that improves its performance. Amiri Rad and Davoodi (2021) conclude how critical the operation of absorption chillers is when different energy sources are involved. In the case of this research, the internal control of the chiller attempts to ensure stable performance by operating the auxiliary boiler when the conditions of the renewable energy source vary. 3.2. Description of test plan and fault detection and diagnosis To develop the test plan and fault detection and diagnosis (FDD), the plant was operated and data was recorded for the results evaluation. From this point of view, two options were considered: – Type 1: Normal operation. The system operates in nominal values and automatic mode, satisfying the cooling/heating demand according to operating conditions. – Type 2: Parameter deviation. One or more parameters that are available to researchers are modified, such as flow and temperature setpoints, energy management modes of the absorption machine, valve position to force an operating mode, weekend operation to reduce load, etc. In the test plan and FDD, Type 1 experiments are performed from Monday to Friday. These are used to evaluate the efficiency of the installation and its equipment in normal operation, check the correct operation of all the elements (equipment, control system and strategies) and finally provide information to achieve possible improvements in both the elements and the whole. Type 2 tests are performed from Saturday to Sunday. The purpose of these is to operate the equipment at different points in order to obtain its operating curves. In this regard, it should be emphasised that the test plan and FDD should include several parameter deviation tests where several setpoint values and modes could be scanned on the same day. In particular, the following parameters are analysed to characterise the operation of the double effect absorption machine: absorption machine inlet and outlet temperatures in all circuits; flow rates through the absorption machine in all circuits; natural gas consumption; exhaust gas outlet temperature; and percentage of CO2 in the exhaust gases. These last two parameters are used to verify the good performance of the natural gas burner. In this context, the behaviour of the machine was studied for these nominal flow rates: cold water flow rate (30 m3/h), hot water flow rate (14.5 m3/h), condensing water flow rate (36.6 m3/h) and solar circuit water flow rate (7.6 m3/h). A partial load analysis was also performed by taking flow values of ±(25, 50, 75) %. In this way, how the COP varies at different loads was studied, comparing the results obtained with the values provided by the manufacturer. From the analysis of the days of operation in cooling mode, the following findings can be taken: – The collector operated with an average daily efficiency of more than 30%, reaching values around 37%. – The average pipe loss rate was 16.47%. This value is high due to the existence of thermal bridges. – The average solar factor obtained in the cooling period was over 60%, due to the cold starts and the direct radiation decrease in the late afternoon. The solar factor even dropped to 50% of previous values on cloudy days. Further details of these results are discussed in Section 5.1. In addition, the most probable measured results when the machine operated at 100% load being powered by the renewable resource were as follows: a cooling capacity of 174.4 kW was 3195 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 5. Operation modes of the pilot solar cooling plant. delivered, with the condenser dissipating a capacity of 296.4 kW. Under these conditions, the efficiency of the machine was 1.34. After the most relevant Type 1 findings have been shown, Type 2 ones are now considered. For this purpose, the failures that occurred in the solar cooling plant during a period of operation of 2 years are analysed (a first year of commissioning of the experimental installation and a second one in which the results presented in this work were carried out). To this end, a database was set up in which 433 orders were collected during operation, making it possible to know the place where they occurred, the start and end date and the type of order. Among all of them, 133 errors were distinguished, which can be divided according to the type of error, the zone in which they occurred or the equipment from which they were produced. Fig. 6 shows the frequency of faults divided by type. In the absorption solar cooling plant, the most frequent ones were the control system with 30.75%, mostly due to errors in the data acquisition system or in the reading/writing of signals. The next most frequent faults were electromechanical, mostly from pumps and solenoid valves, with 28.50%. These were followed by exclusively electrical faults in protections and other switchboard elements with 15.75%. The errors classified as ‘‘Others’’, with 12.00%, corresponded to errors related to issues such as corrosion, insulation, low battery in equipment, broken mirrors, etc. Fig. 7 shows the frequency of shutdowns and faults recorded in the solar cooling plant divided by zone. As can be checked, the zone where most errors occurred was the solar circuit, with 44.85%, as this is the zone of the installation with the highest technical requirements. In second place, the cooling circuit with a frequency of 18.38%, mainly due to errors in the start-up or shutdown due to low pressure in the cooling circuit. In third place, the control system, with a frequency of 13.97%, followed by those related to the machine room, the raw water circuit, the distribution panel, the heat circuit and, finally, the exhaust gases and the gas train. Fig. 8 shows the frequency of shutdowns and faults recorded in the solar cooling plant divided by the equipment on which this occurs and, therefore, on which it will be necessary to act, in accordance with the contingency plan. To this end, the errors found are divided according to the equipment from which they occurred. The equipment with the highest number of failures was the Fresnel collector with a frequency of 24.81%, mainly due to connection issues with the control system, failures in the row motors and control failures in the mirror rows. This was followed by failures in the pipe installation, due to instabilities in pressure and flow levels, as well as the occurrence of leaks. In third place, the control system was with a frequency of 12.78%. A corrective measure is formulated for each of the errors. Therefore, by way of example, each of the faults found in the double effect absorption machine (DEAM) are compiled in Table 3, supported by a brief description and the consequence of the fault, as well as the corrective measure applied. 4. Model 4.1. Description Absorption machines are based on the principle of condensation and evaporation of a refrigerant to obtain heat or cold, as required. They are systems capable of generating cold from the use of a medium/high-temperature heat source. Fig. 9 shows the main elements that compose this type of machine. The heat 3196 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 6. Frequency of faults by type (in %). Fig. 7. Frequency of faults by zone (in %). Fig. 8. Frequency of faults by equipment (in %). flows (colour orange in Fig. 9) are explicitly shown in the machine model. If the information provided by the manufacturer is analysed together with the experimental data obtained, then the behaviour of the machine under study can be simplified into three operating modes, as shown in Fig. 10. The orange line represents the temperature of the generator, whereas the blue one represents the temperature of the evaporator. According to Fig. 10, Mode 1 corresponds to start up. The hot heat transfer fluid enters the generator to increase its temperature in this mode. However, this temperature (colour orange in Fig. 10) is not enough for the absorption machine operates. Once the temperature of the generator is adequate, the unit control starts the absorber–evaporator and condenser of the chiller sequentially. This Mode 2 refers to the nominal operation of the device, which ends with the shutdown mode. This model concerns the operation after the absorption machine stops due to the dilution effect. After turning off the device, the generator gradually cools down. Simultaneously, the evaporator produces cold water (blue colour in Fig. 10), although not at the nominal 3197 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 14. First results of the model validation: Detailed results (Day 4). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) the absorption chiller generator is provided by solar hot water (approximately 80%). Fig. 17 shows the evolution of the COP of the machine as the generator reaches this temperature. It is an interesting result because it allows knowing the real performance of the absorption chiller, being the maximum COP during the summer 1.45 and its average value 1.24. Cooling needs of the building change along with cooling requirements throughout the day, with the highest levels corresponding to the 130–140 kW evaporator capacity. Similarly, the 3204 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 15. First results of the model validation: Detailed results (Day 17). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) COP also varies throughout the day. The initial two hours correspond to the transition period after start-up until the temperatures of the different chiller sections reach a stable status and the maximum COP value is achieved. The Sankey diagram shown in Fig. 18 provides all the energy flows on a typical sunny day. Pipeline losses accounted for 20% of the total energy absorbed in the collector tubes. The absorption chiller COP was 1.16, and the SHF was 0.75, with the remaining 0.25 corresponding to the energy supplied by the natural gas for the HTG. In addition, the SCR reached 0.44. The validation has been performed using 33 consecutive days of operation. The absorption machine model is running during these days, and its daily estimate is compared with the values measured. Representative daily measured data are summarised in Table A.1 in Appendix. It should be noted that the only reliable data are the operating flow rates and inlet temperatures of each 3205 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 16. Operation data of solar cooling plant during cooling season. Fig. 17. Operation data of solar cooling plant during a typical day. 3206 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 18. Sankey diagram of a typical day validation of the plant operator. Table 4 Results of validation of the solar cooling plant. Solar cooling plant Solar cooling plant Solar energy efficiency ratio Error Solar energy efficiency ratio Error – % – % Day 1 0.27 −0.1 Day 18 0.3 6 Day 2 0.26 1.6 Day 19 0.35 1.6 Day 3 0.29 −2.2 Day 20 0.33 5 Day 4 0.24 −3.3 Day 21 0.41 −2.5 Day 5 0.28 −3.2 Day 22 0.33 1.4 Day 6 0.23 4 Day 23 0.27 −3.4 Day 7 0.25 −2.2 Day 24 0.24 4 Day 8 0.24 −1.7 Day 25 0.16 −5.1 Day 9 0.25 0.3 Day 26 0.43 5.2 Day 10 0.39 −1.4 Day 27 0.37 −2.5 Day 11 0.38 −0.3 Day 28 0.36 2.7 Day 12 0.38 −6.1 Day 29 0.01 25.7 Day 13 0.18 −4.3 Day 30 0.28 −6.2 Day 14 0.27 0.4 Day 31 0.41 −0.6 Day 15 0.11 −0.4 Day 32 0.36 1.5 Day 16 0.36 5.3 Day 33 0.41 −2.8 Day 17 0.29 −2.5 loop. These loops are the condensation circuit, evaporator cooling water, and solar plant generator. Variables calculated are the thermal energy consumption of the solar installation, the natural gas consumption and the cooling energy production. Therefore, the main key performance indicators are the COP and the solar energy efficiency ratio. The COP is calculated as the quotient between the cooling energy produced and the thermal energy consumption of the absorption machine. In addition, the solar energy efficiency ratio is calculated as the quotient between the cooling production and the maximum solar energy captured in the primary receiver. Table A.2 in Appendix shows all the validation results. However, Table 4 and Fig. 19 summarise the most relevant results. Relative errors obtained are low (which are shown in Table 4), proving the usefulness of the model for real-time predictive control or planning operating strategies for the system. In addition, the results are interesting because a plant efficiency higher than 40% is observed. The solar fraction of the solar cooling production system is more significant than 60% on average. It is a very useful result. The simplified methodology also fits nicely, being an easy and fast calculation solution. 5.2. Application of the inverse model for the plan operator The inverse model is integrated into the management system of the plant as a predictive control tool. The model is executed using the actual data of the plant as its inputs. So, it estimates the behaviour of the chiller and decides the optimum partial load ratio of the auxiliary boiler to maximise the cooling production 3207 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Fig. 19. Relative error of the model estimation (monthly and annual values). Table 5 Operational costs with & without pilot plant and with & without predictive control. Year Scenario Electricity consumption (MWh) Natural gas consumption (MWh) Cooling production (MWh) 1 Conventional system (measured) 171.41 0 720 2 Conventional system & pilot plant (measured) 108.72 264.59 822.79 2 Conventional system & pilot plant & predictive control (estimated) 106.91 134.52 822.79 and the solar fraction. So, it operates as a primary decision support system for the management system. The application allows validating how the proposed model improves the control of the plant by increasing the physical component of the system. Table 5 justifies the interest of this proposal, showing the measured consumption (MWh and = C) of the building operation in three scenarios: without the pilot solar cooling plant (year 1), and with the solar cooling plant but without predictive control of the absorption system (year 2), as well as the estimation when the proposed model is implemented in the system management of the pilot solar cooling plant (year 2 too). Table 5 shows the results of two different operation years. The first of them (year 1) does not include the solar cooling plant. The second one (year 2) integrates the solar cooling plant into the conventional system of the building. Regarding the results presented, although the conventional chiller achieves a very high COP (higher than 4, as shown in Table 5, year 1), the improvement of the whole system thanks to the pilot plant can be checked. However, despite this improvement, it is still possible to save an additional 49% of the natural gas consumption and 16% of the total cost if the absorption chiller operates optimally (third scenario). This is thanks to the predictive control of the auxiliary boiler to heat the generator when hot water from the Fresnel solar collector varies. 6. Conclusions In this research, an absorption chiller has been coupled with the energy management model of a solar cooling plant by means of an inverse model. The model is connected to the manufacturer data and the experimental data by identifying its parameters. This allows the replicability of the model for other chillers or conditions, as demonstrated by the validation performed. Furthermore, the usability of this model is validated by its integration into the control system of the plant. Therefore, the paper eliminates the performance gaps and solves the integration with the renewable source. The research also shows an experimental analysis to optimise an absorption chiller. With these experimental data, it has been possible to know the real operation of this system. In addition, a characterisation model is proposed, whose parameters use readily available experimental data. This model predicts the behaviour of the system in the case of changes in the input energy flows or in the control system. The model is robust, as demonstrated by the validation performed, whose experiments have been shown to be 3208 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Table A.1 Representative daily measured data. very accurate, easy to reproduce, and adaptable to any absorption chiller system. Moreover, the results prove that: – To model the behaviour of this technology in three sections is interesting in order to be able to adjust the dynamics of the system. The three stages are start-up, regular operation and shutdown. The detailed validation results indicate a good approximation of the estimates made with acceptable errors. – The application performed shows that the coupling of the model in higher-order energy simulation tools may be interesting. Because of the daily and seasonal estimates of renewable consumption, yields or fractions match those obtained in the experiments. This evidences that it is still 3209 M.C.GuerreroDelgado,J.SánchezRamos,D.CastroMedinaetal. EnergyReports8(2022)3189–3212 Table A.2 Results of validation: Absorption chiller. possible to save almost 50% of natural gas consumption with the same cooling production. Finally, the proposed procedure for characterising an absorption chiller and providing a solution for modelling this type of technology may be used as a methodology for evaluating and diagnosing these units. If these systems are in real operation, the values obtained by monitoring them can be compared with the estimates made with the curves proposed under the same conditions. This model can also be integrated into real-time predictive control for solar cooling facilities. This inverse model would allow solving the definition of optimal operating strategies. The results prove the interest of the model. Furthermore, the proposed methodology can be applied to any hybrid solar system for whatever control requirements. 3210 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 CRediTauthorshipcontributionstatement MCarmen Guerrero Delgado: Methodology, Investigation, Formal analysis, Writing – original draft. Jose Sánchez Ramos: Software, Validation, Data curation, Investigation, Conceptualization. Daniel Castro Medina: Experimental facility, Investigation. Teresa RocíoPalomoAmores:Simulation,Investigation,Visualization.AlbertoCerezo-Narváez:Conceptualization,Writing,Resources.Servando Álvarez Domínguez: Conceptualisation, Methodology, Supervision,Formalanalysis. Declarationofcompetinginterest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influencetheworkreportedinthispaper. Acknowledgements The European Commission funded this study under the project MedEcoSuRe - ‘‘Mediterranean University as Catalyst for Eco- Sustainable Renovation’’ (A_B.4.3_0218) to eliminate Fresnel Solar’s barriers Cooling Plants as a potential energy-efficient alternative in university buildings. Also, this study has been cofinanced by the European Regional Development Funds (ERDF). Appendix See Tables A.1 and A.2. References Ahn, H., Rim, D., Pavlak, G.S., Freihaut, J.D., 2019. Uncertainty analysis of energy and economic performances of hybrid solar photovoltaic and combined cooling, heating, and power (CCHP + PV) systems using a Monte-Carlo method. Appl. Energy 255, 113753. http://dx.doi.org/10.1016/j.apenergy.2019.113753. Ali, D., Ratismith, W., 2021. A semicircular trough solar collector for airconditioning system using a single effect NH3–H2O absorption chiller. Energy 215, 119073. http://dx.doi.org/10.1016/j.energy.2020.119073. Amiri Rad, E., Davoodi, V., 2021. Thermo-economic evaluation of a hybrid solar-gas driven and air-cooled absorption chiller integrated with hot water production by a transient modeling. Renew. Energy 163, 1253–1264. http: //dx.doi.org/10.1016/j.renene.2020.08.157. Anon, 2018. DOE2.com home page. http://www.doe2.com/. (Accessed 28 December 2018). Anon, 2021. Third party certification |Certita Eurovent. https://www.euroventcertification.com/es. (Accessed 19 February 2021). Araújo, A., Silva, R., 2020. Energy modeling of solar water heating systems with on-off control and thermally stratified storage using a fast computation algorithm. Renew. Energy 150, 891–906. http://dx.doi.org/10.1016/j.renene. 2020.01.026. Bataineh, K., Taamneh, Y., 2016. Review and recent improvements of solar sorption cooling systems. Energy Build. 128, 22–37. http://dx.doi.org/10. 1016/j.enbuild.2016.06.075. Bermejo, P., Pino, F.J., Rosa, F., 2010. Solar absorption cooling plant in Seville. Sol. Energy 84, 1503–1512. http://dx.doi.org/10.1016/j.solener.2010.05.012. Bilardo, M., Ferrara, M., Fabrizio, E., 2020. Performance assessment and optimization of a solar cooling system to satisfy Renewable Energy Ratio (RER) requirements in multi-family buildings. Renew. Energy http://dx.doi.org/10. 1016/j.renene.2020.03.044. Castro, J., Farnós, J., Papakokkinos, G., Zheng, J., Oliet, C., 2020. Transient model for the development of an air-cooled LiBr-H2O absorption chiller based on heat and mass transfer empirical correlations. Int. J. Refrig. 120, 406–419. http://dx.doi.org/10.1016/j.ijrefrig.2020.08.030. Chen, J.F., Dai, Y.J., Wang, R.Z., 2017. Experimental and analytical study on an air-cooled single effect LiBr-H2O absorption chiller driven by evacuated glass tube solar collector for cooling application in residential buildings. Sol. Energy 151, 110–118. http://dx.doi.org/10.1016/j.solener.2017.05.029. Chen, Y., Wang, J., Lund, P.D., 2020. Sustainability evaluation and sensitivity analysis of district heating systems coupled to geothermal and solar resources. Energy Convers. Manag. 220, 113084. http://dx.doi.org/10.1016/j.enconman. 2020.113084. DOE, 2010. EnergyPlus - Engineering reference. European Commission, 2018. Directive (EU) U) 2018/844. Off. J. Eur. Union 2018, 75–91. Evola, G., Le Pierrès, N., Boudehenn, F., Papillon, P., 2013. Proposal and validation of a model for the dynamic simulation of a solar-assisted single-stage LiBr/water absorption chiller. Int. J. Refrig. 36, 1015–1028. http://dx.doi.org/ 10.1016/j.ijrefrig.2012.10.013. Florides, G.A., Kalogirou, S.A., Tassou, S.A., Wrobel, L.C., 2002. Modelling, simulation and warming impact assessment of a domestic-size absorption solar cooling system. Appl. Therm. Eng. 22, 1313–1325. http://dx.doi.org/10.1016/ S1359-4311(02)00054-6. Gallego, A.J., Sánchez, A.J., Berenguel, M., Camacho, E.F., 2020. Adaptive UKF- based model predictive control of a Fresnel collector field. J. Process Control 85, 76–90. http://dx.doi.org/10.1016/j.jprocont.2019.09.003. Ibrahim, N.I., Khan, M.M.A., Mahbubul, I.M., Saidur, R., Al-Sulaiman, F.A., 2017. Experimental testing of the performance of a solar absorption cooling system assisted with ice-storage for an office space. Energy Convers. Manag. 148, 1399–1408. http://dx.doi.org/10.1016/j.enconman.2017.07.001. International Energy Agency, 2013. Transition to sustainable buildings: Strategies and opportunities to 2050. http://dx.doi.org/10.1787/9789264202955-en. Jalalizadeh, M., Fayaz, R., Delfani, S., Mosleh, H.J., Karami, M., 2021. Dynamic simulation of a trigeneration system using an absorption cooling system and building integrated photovoltaic thermal solar collectors. J. Build. Eng. 43, 102482. http://dx.doi.org/10.1016/j.jobe.2021.102482. Jradi, M., Riffat, S., 2014. Tri-generation systems: Energy policies, prime movers, cooling technologies, configurations and operation strategies. Renew. Sustain. Energy Rev. 32, 396–415. http://dx.doi.org/10.1016/j.rser.2014.01.039. Karki, S., Haapala, K.R., Fronk, B.M., 2019. Investigation of the combined efficiency of a solar/gas hybrid water heating system. Appl. Therm. Eng. 149, 1035–1043. http://dx.doi.org/10.1016/j.applthermaleng.2018.12.086. Klein, S.A., 2010. TRNSYS 17: A Transient System Simulation Program. vol. 1, Sol. Energy Lab. Univ. Wisconsin, Madison, USA, pp. 1–5. Kylili, A., Fokaides, P.A., 2015. European smart cities: The role of zero energy buildings. Sustain. Cities Soc. 15, 86–95. http://dx.doi.org/10.1016/j.scs.2014. 12.003. Leckner, M., Zmeureanu, R., 2011. Life cycle cost and energy analysis of a Net Zero Energy House with solar combisystem. Appl. Energy 88, 232–241. http://dx.doi.org/10.1016/j.apenergy.2010.07.031. Lubis, A., Jeong, J., Giannetti, N., Yamaguchi, S., Saito, K., Yabase, H., Alhamid, M.I., Nasruddin, 2018. Operation performance enhancement of single-double- effect absorption chiller. Appl. Energy 219, 299–311. http://dx.doi.org/10. 1016/j.apenergy.2018.03.046. Mao, M., Cui, L., Zhang, Q., Guo, K., Zhou, L., Huang, H., 2020. Classification and summarization of solar photovoltaic MPPT techniques: A review based on traditional and intelligent control strategies. Energy Rep. 6, 1312–1327. http://dx.doi.org/10.1016/j.egyr.2020.05.013. Martínez, J.C., Martinez, P.J., Bujedo, L.A., 2016. Development and experimental validation of a simulation model to reproduce the performance of a 17.6 kW LiBr–water absorption chiller. Renew. Energy 86, 473–482. http://dx.doi.org/ 10.1016/j.renene.2015.08.049. Nikbakhti, R., Wang, X., Hussein, A.K., Iranmanesh, A., 2020. Absorption cooling systems – Review of various techniques for energy performance enhancement. Alex. Eng. J. 59, 707–738. http://dx.doi.org/10.1016/j.aej.2020.01. 036. Prieto, A., Knaack, U., Auer, T., Klein, T., 2019. COOLFACADE: State-of-the-art review and evaluation of solar cooling technologies on their potential for façade integration. Renew. Sustain. Energy Rev. 101, 395–414. http://dx.doi. org/10.1016/j.rser.2018.11.015. Ray, M., Samal, P., Panigrahi, C.K., 2021. The influencing factors on efficacy enhancement of HVAC systems – A review. Mater. Today Proc. http://dx. doi.org/10.1016/j.matpr.2021.07.264. Rivarolo, M., Greco, A., Massardo, A.F., 2013. Thermo-economic optimization of the impact of renewable generators on poly-generation smart-grids including hot thermal storage. Energy Convers. Manag. 65, 75–83. http://dx.doi.org/10. 1016/j.enconman.2012.09.005. Sánchez Ramos, J., Delgado, Mc.C. Guerrero, Álvarez Domínguez, S., Molina Félix, J.L., Cabeza, L.F., 2019. Gas engine heat pump system: Experimental facility and thermal evaluation for 5 different units. Energy Convers. Manag. 199, 112060. http://dx.doi.org/10.1016/j.enconman.2019. 112060. Santamouris, M., Adnot, J., Álvarez Domínguez, S., Klitsikas, N., Orphelin, M., Lopes, C., Sánchez de la Flor, F.J., 2004. Cooling the cities - Raifraichir les villes. Ecole des Mines de Paris. Les Presses, Paris. Sarabia Escriva, E.J., Lamas Sivila, E.V., Soto Frances, V.M., 2011. Air conditioning production by a single effect absorption cooling machine directly coupled to a solar collector field. Application to Spanish climates. Sol. Energy 85, 2108–2121. http://dx.doi.org/10.1016/j.solener.2011.05.019. Sebastián, A., Abbas, R., Valdés, M., Rovira, A., 2021. Modular micro-trigeneration system for a novel rotatory solar Fresnel collector: A design space analysis. Energy Convers. Manag. 227, 113599. http://dx.doi.org/10.1016/j.enconman. 2020.113599. 3211 M.C. Guerrero Delgado, J. Sánchez Ramos, D. Castro Medina et al. Energy Reports 8 (2022) 3189–3212 Settino, J., Sant, T., Micallef, C., Farrugia, M., Spiteri Staines, C., Licari, J., Micallef, A., 2018. Overview of solar technologies for electricity, heating and cooling production. Renew. Sustain. Energy Rev. 90, 892–909. http: //dx.doi.org/10.1016/j.rser.2018.03.112. Sharafi, M., ElMekkawy, T.Y., Bibeau, E.L., 2015. Optimal design of hybrid renewable energy systems in buildings with low to high renewable energy ratio. Renew. Energy 83, 1026–1042. http://dx.doi.org/10.1016/j.renene.2015. 05.022. Shirazi, A., Taylor, R.A., Morrison, G.L., White, S.D., 2018. Solar-powered absorption chillers: A comprehensive and critical review. Energy Convers. Manag. 171, 59–81. http://dx.doi.org/10.1016/j.enconman.2018.05.091. Song, Z., Ji, J., Cai, J., Li, Z., Gao, Y., 2020. Performance prediction on a novel solar assisted heat pump with hybrid Fresnel PV plus TEG evaporator. Energy Convers. Manag. 210, 112651. http://dx.doi.org/10.1016/j.enconman. 2020.112651. Wang, J., Li, X., Wang, B., Wu, W., Song, P., Shi, W., 2017. Performance comparison between an absorption-compression hybrid refrigeration system and a double-effect absorption refrigeration sys-tem. Procedia Eng. 205, 241–247. http://dx.doi.org/10.1016/j.proeng.2017.09.959. Xu, Z.Y., Wang, R.Z., 2017. Comparison of CPC driven solar absorption cooling systems with single, double and variable effect absorption chillers. Sol. Energy 158, 511–519. http://dx.doi.org/10.1016/j.solener.2017.10.014. Yu, J., Li, Z., Chen, E., Xu, Y., Chen, H., Wang, L., 2019. Experimental assessment of solar absorption-subcooled compression hybrid cooling system. Sol. Energy 185, 245–254. http://dx.doi.org/10.1016/j.solener.2019.04.055. Zhao, T., Chen, X., Chen, Q., 2020. Heat current method-based modeling and optimization of the single effect lithium bromide absorption chiller. Appl. Therm. Eng. 175, 115345. http://dx.doi.org/10.1016/j.applthermaleng.2020. 115345. 3212