Gas engine heat pump system: Experimental facility and thermal evaluation for 5 different units

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This is an Accepted Manuscript of an article published by Elsevier in Energy Conversion and Management, Vol. 199, on November 2019, , available at: https://doi.org/10.1016/j.enconman.2019.112060 © 2019 Elsevier. En idUS Licencia Creative Commons CC BY-NC-ND Gas engine heat pump system: experimental facility and 1 thermal evaluation for 5 different units 2 Authors: José Sánchez Ramosa, MCarmen Guerrero Delgadob, Servando Álvarez Domínguezb, 3 José Luis Molina Félixb, Luisa F. Cabezac 4 a Thermal machines and engines, School of engineer, Av. University of Cádiz, 10, 11519, Puerto Real, Spain 5 b Department of energy engineering. School of engineer. University of Seville. Camino de los Descubrimientos s/n, 6 41092 Seville, Spain. 7 c GREiA Research Centre, INSPIRES Research Centre, Universitat de Lleida, Pere de Cabrera s/n, 25001 Lleida 8 9 Abstract 10 Gas Engine Heat Pump requires simplified characterization models that allow evaluating its 11 economic and energy impacts. These simplified solutions can be implemented in complex 12 simulation algorithms for different commercial solutions with a low computational cost. 13 The study aims to develop a simplified characterization model highly accurate, easy to reproduce 14 and apply. The methodology carried out uses an experimental installation to analyze different 15 units tested under different operating conditions. With the results of the carried out 16 experimentation, it will be possible to know the real thermal response of these HVAC systems 17 and to develop the simplified characterization model based on operating curves. The thermal 18 behaviour of the systems could be evaluated using the defined curves under any operating 19 condition of the GEHP system. Moreover, they are based on parameters available in the 20 manufacturer datasheets. 21 The results of the validation show that this model is highly accurate. It has a maximum error of 22 25% and an average error of less than 7%. Also, the formulation shows that it is easy to 23 reproduce and to apply. Last, uncertainty analysis shows reasonable confidence in the identified 24 performance curves. 25 Keywords: Gas engine-driven heat pump, experimental model, waste heat recovery, thermal 26 characterization, inverse modelling 27 28 29 0 Nomenclature 30 Variable/Acronym/ Abbreviation Description Units GEHP Gas Engine Heat Pump - EPBD Energy Performance of Buildings Directive - NZEB Nearly Zero Energy Buildings - EHP Electric Heat Pump - EER Energy Efficient Ratio - COP Coefficient Of Performance - SEER Seasonal Energy Efficient Ratio - SCOP Seasonal Coefficient Of Performance - mW Water flow kg/s PLR Part load ratio % TLIM-REC Minimum outdoor temperature to put on recovery system of GEHP °C DHW Domestic Hot Water - TSETPOINT Setpoint temperature °C QRECOVERY Recovery heat from cooling system of engine kW Tow Outlet water temperature °C Tiw Inlet water temperature °C Tdr Outdoor dry-bulb temperature °C Twb Outdoor wet-bulb temperature °C 𝐻𝑒𝑎𝑡𝑟𝑒𝑎𝑙 Heat capacity in real operation conditions kW 𝐻𝑒𝑎𝑡𝑛𝑜𝑚 Heat capacity in nominal operation conditions kW 𝐶𝑜𝑛𝐺𝑎𝑠𝐻𝑟𝑒𝑎𝑙 Gas consumption in heating mode in real operation conditions kW 𝐶𝑜𝑛𝐺𝑎𝑠𝐻𝑛𝑜𝑚 Gas consumption in heating mode in nominal operation conditions kW 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐻𝑟𝑒𝑎𝑙 Electrical consumption in heating mode in real operation conditions kW 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐻𝑛𝑜𝑚 Electrical consumption in heating mode in nominal operation conditions kW 𝑅𝑒𝑐𝑄𝐻𝑟𝑒𝑎𝑙 Heat recoverable in heating mode in real operation conditions kW 𝑅𝑒𝑐𝑄𝐻𝑛𝑜𝑚 Heat recoverable in heating mode in nominal operation conditions kW 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙 Cooling capacity in real operation conditions kW 𝐶𝑜𝑙𝑑𝑛𝑜𝑚 Cooling capacity in nominal operation conditions kW 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑟𝑒𝑎𝑙 Gas consumption in cooling mode in real operation conditions kW 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑛𝑜𝑚 Gas consumption in cooling mode in nominal operation conditions kW 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐶𝑟𝑒𝑎𝑙 Electrical consumption in cooling mode in real operation conditions kW 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐶𝑛𝑜𝑚 Electrical consumption in cooling mode in nominal operation conditions kW 𝑅𝑒𝑐𝑄𝐶𝑟𝑒𝑎𝑙 Heat recoverable in cooling mode in real operation conditions kW 𝑅𝑒𝑐𝑄𝐶𝑛𝑜𝑚 Heat recoverable in cooling mode in nominal operation conditions kW Torec Outlet water temperature of recovery loop °C 𝑇𝑖𝑟𝑒𝑐 Inlet water temperature of recovery loop °C mrec Bypass flow kg/s HVAC Heating Ventilation and Air Conditioning - 31 1 Introduction 32 1.1 Background 33 Directive 2018/2001 / CE [1] has indicated to European countries a series of objectives related 34 to energy efficiency and use of renewable energy, such as reduction of greenhouse gases by 20% 35 and an increase in the contribution of renewable energy by 20% against total energy 36 consumption. In recent years the gas heat pump has attracted much attention due to its 37 environmentally friendly advantages and efficiency as well as energy savings [2,3]. Gas engine 38 heat pumps (GEHP) are essentially the same as electric heat pumps (EHP). The difference is that 39 in the GEHP the electric motor is replaced by an internal combustion engine based on an OTTO 40 cycle. The advantage concerning for to the electric heat pump is the use of residual energy 41 available for the production of domestic hot water or to prevent frost under extreme 42 temperature conditions in winter. Accordingly, the result is not only a reduction of CO2 43 emissions but also a reduction in energy consumption [4]. 44 The main disadvantage of GEHP systems is the limited optimal zone of operation. Then, they 45 require robust models for sizing with the variability of buildings energy needs. Besides, 46 manufacturers have few different units, because they propose to use thermal energy storage, 47 parallel installations and efficiency control systems to achieve high efficiency [5]. To ensure that 48 GEHPs operate within the best economic zone, researchers propose combining GEHPs with 49 hybrid power technology [6–9], energy storage technology [10,11], absorption heat pump [12] 50 and photovoltaic integration [13–15]. These studies demonstrate the potential of these systems 51 in combination with conventional installations. Therefore, it is necessary to develop new 52 simulation models for analysing this technology in new simulation schemes. Simplified models 53 will allow for optimal sizing of hybrid installations for heat and cold production. Therefore, it is 54 required to have robust models for this technology that can be integrated into more complex 55 simulation schemes. 56 Then, hybrid solutions could appear as solutions for distributed generation. Renewable gas 57 contributes to the discharge and improvement of the air quality of our cities, and facilitate the 58 energy transition towards a truly circular economy. Xiaofan et al. [9] got good economic and 59 energetic results of hybrid power gas with thermal and electric energy storage. These authors 60 [9] justified the need of suitable models to optimize control and operation of facilities. Recently, 61 Qiang et al. [15] designed a novel system where GEHP is integrated with solar energy, energy 62 storage and smart control of energy needs. Novel system [15] required an intelligent solution 63 using a neural network to guarantee optimal operation. So, the proposal of this paper could be 64 integrated into these complex algorithms [15–17] and will improve final results. 65 1.2 Thermal evaluation 66 Thermal models are necessary to enable evaluation of the corresponding energy behaviour. In 67 this context, various models have been developed. These may be classified into two categories: 68 detailed models or models based on balance equations and simplified models or models based 69 on behaviour curves. Nonetheless, the main requirement for thermal characterization models is 70 that it is easy to reproduce, and can be adapted for any system of gas engine heat pump. 71 From a review of reference literature, the use of balance equations to evaluate the performance 72 of a GEHP is highlighted by analysis using energy and exergy balances on the complete unit 73 [3,10], or each element of the system [6]. These studies shed light on the internal behaviour of 74 these units and enable exergoeconomic analysis for decision making in environmental policy 75 definition algorithms, as shown in [18]. However, these require knowledge of internal operating 76 data of the units not usually provided by manufacturers. 77 To solve the issue of lack of information on about internal operation parameters, researchers 78 added mathematical components into physical models. This type of model is called hybrid 79 models. Here are exemplary studies such as [19,20]. The first [19] proposes a set of corrective 80 functions for the energy balances that characterise these machines. These functions are 81 modified from experimental data employing an algorithm based on high complexity neural 82 networks. However, the same author proposed a variant of this simplified model [21] with 83 parameters set at typical values. Nevertheless, there has been no experimental validation for 84 the validity of this hypothesis. The second study [20] coincided with the model typologies, but 85 it offered an original proposal of how to integrate control of these units using a neural network. 86 These solutions have a high mathematical complexity, which is a disadvantage when integrated 87 into simulation algorithms. Solving this limitation, simplifying the model and extrapolating it to 88 other types of machines is a challenge to be solved. This problem is more pronounced when a 89 multitude of heat, cold and electricity generation systems (with and without storage) are 90 integrated into the same decision-making algorithm. 91 The most standard simulation tools for building air conditioning systems are TRNSYS [22] and 92 EnergyPlus [23]. The proposal of these tools for simulation of GEHP systems is a routine based 93 on a set of operating curves. These curves are obtained for a given unit and, valid hypotheses 94 are generated for this technology. Additionally, TRNSYS [22], in corresponding TESS libraries, 95 offers a TYPE for the detailed simulation of GEHP systems. This TYPE goes into the 96 thermodynamic detail of the cooling cycle as well as the combustion stages of the engine. Thus, 97 it requires the definition of a set of detail parameters, such as volumetric performance of the 98 combustion chamber, cylinder size, etc. The main disadvantage of these models is the difficulty 99 in defining the parameters necessary for their implementation. Most parameters are not 100 provided in the manufacturer's data sheets and cannot be defined from the data available in 101 them. 102 The use of simplified models or models based on behaviour curves let designers and engineers 103 determine the energy performance of the system at different operating conditions. They have 104 the advantage of a significantly reducing computational cost in comparison with detailed 105 procedures. Instead, they require high experimentation of different units to demonstrate the 106 validity of them. 107 Along this line, researchers have developed the characterisation of energy behaviour curves of 108 the electric heat pump in relation to its thermal capacity, as well as the functional dependencies 109 of the curves, with external climatic conditions, using water distribution temperature and partial 110 load factor [24,25]. 111 However, in some cases, authors use very simplified models due to the uncertainty existing in 112 the real conditions of operation of the building. For example, Sheng et al. [26] studied the energy 113 performance of GEHP using simulation tools. The authors used a simplified model, but they 114 validated it using measurement from the real operation of buildings. Alternatively, Kamal et al. 115 [27] who had an experimental installation with several GEHP units. Subsequently, these same 116 authors [27]compared the results of the same with those obtained from a simulation in 117 EnergyPlus [28] using the default model for GEHP system. 118 For the GEHP system, there are experimental modelling methods where thermal efficiency of 119 the engine is established as a function of the corresponding torque and speed [16,29]. 120 On the one hand, Bin Hu et al. [29] proposed a model to simulate gas driven heat pumps. The 121 proposed methodology requires to know all technical details of these systems, for example, the 122 expansion valve. However, manufacturers do not provide this information, and it is very difficult 123 to approximate. 124 Also, QingkunMeng et al. [16] developed a mathematical model for optimizing the operation of 125 engine and compressor. The model requires to be detailed and thanks to it, it is possible to 126 optimize internal the operation parameters. However, it is not compatible to be integrated into 127 building energy simulation algorithms. 128 Experimental studies such as [30,31] demonstrated dependencies on thermal behaviour of gas 129 heat pump system with ambient air temperature, water temperature and partial load factor. 130 These studies facilitate understanding of system behaviour. But, mathematical models are 131 required to quantify them outside the range of experimental testing. This approach was not 132 validated in these studies. Feng et al. [32] established detailed experimental results of the 133 thermal behaviour of the motor coupled to these systems and the variations of the recoverable 134 heat in it. 135 Additionally, studies such as [33] demonstrated the potential of the use of curves to stablish 136 optimal strategies for heat recovery. The results were based on experimental data and they 137 were obtained using a detailed experimental facility. Furthermore, there are studies such as 138 [34,35], where, based on experimental data, operating laws for a unit were obtained enabling 139 evaluation in critical summer conditions [34] or combined with known demand [35]. Both of 140 them [34,35] are useful to understand this technology and define the best experiments to 141 analyse these units. Moreover, Fengguo et al. [12] experimentally demonstrated that it is 142 possible to define optimal operating instructions for GEHP systems and, the importance of 143 taking them into account in the design of this type of facilities. 144 Accordingly, experimental studies about GEHP have been well treated in the existing 145 publications. Shah et al. [36] developed detailed studies about diesel engine driven heat pump, 146 but they did not propose any model. Jiying et al. [37] who analysed the effect of renewable 147 biogas as a fuel for GEHP system. Huanwei et al. [30] integrated evaporative cooling with GEHP 148 system and studied energy performance in different operational conditions. Or, Liu et al. [35] 149 who investigated the possibilities of these systems for the production of domestic heat water. 150 Even, Mingtao et al. [38] defined an experimental facility where they tested different control 151 strategies. Before this work, they used simulation [39] for studying GEHP system. 152 In the study presented here, the GEHP system behaviour curves are defined. Also, they allow to 153 characterising the thermal behaviour of these systems in terms of thermal capacity, 154 consumption and energy recovery, as well as the functional dependencies of the curves, on 155 external climatic conditions, partial load factor and water distribution temperature. It is 156 obtained using the experimental data. Besides, the model developed using this data, is valid for 157 heating and cooling, and it is possible to be implemented in energy buildings performance 158 simulation tools with control operation strategies. 159 Last, measurement and classification of energy efficiency of air conditioning systems, by 160 regulations [40] and a multitude of research projects, is based on basic EER and COP indicators, 161 or more sophisticated parameters such as SEER and SCOP [41]. Among recent publications, the 162 work of Alves et al. [42] where, under different climatic and operating conditions (partial load 163 factor), performance variability of several models of electric heat pumps is evaluated. This study 164 demonstrates the need to distinguish between nominal and average seasonal characteristics. 165 Since results of instant characteristic sensitivity of machines are presented and how average 166 seasonal performances of the same machine may be doubled according to operating conditions 167 even more so in real buildings with variable operating conditions. Zhang et al. [43] showed the 168 importance of the accumulation and the variations of operation of the buildings in the seasonal 169 benefits of GEHP. 170 Proper evaluation of the seasonal performance of these systems may be problematic for 171 designers due to the lack of precise data as regards the behaviour of heat pumps beyond 172 nominal conditions. In fact, the majority of heat pump manufacturers still avoid providing 173 detailed information concerning this point in their datasheets. Additionally, it is possible to 174 obtain seasonal performance ratios using simulation tools, but models need to reflect the real 175 performance of a commercial system with credible results. 176 1.3 Aims 177 Gas Engine Heat Pump requires simplified characterization models. These models allow 178 evaluating its economic and energy impact independently. Alternatevely, they could be 179 implemented in complex simulation algorithms for different commercial solutions with a low 180 computational cost. 181 In the existing literature, there are detailed solutions with high mathematical complexity, which 182 is a disadvantage when integrated into simulation algorithms. On the other hand, there are 183 simplified models but whose main disadvantage is the difficulty in defining the parameters 184 necessary for its implementation. Most of the parameters are not provided in the 185 manufacturer's data sheets and cannot be defined from the data available in them. 186 Solving these limitations is a challenge to solve. In order to solve this gap in the existing literature 187 of the technology, an inverse characterization methodology based on experimental 188 measurements is carried out. For this purpose, the main manufacturers in the sector defined a 189 sample of five units. This sample could be considered as a representation of the current state of 190 the market for this technology. 191 The methodology carried out uses an experimental installation to analyze the different units 192 tested under different operating conditions. With the results of the experimentation carried out, 193 it will be possible to know the real thermal behaviour of these HVAC systems and to develop the 194 simplified characterization model based on operating curves. These curves characterize the 195 thermal behaviour of these systems in terms of thermal capacity, energy consumption and 196 recovery, as well as the functional dependencies of the curves, in external climatic conditions, 197 partial load factor and water distribution temperature using experimental data provided by the 198 manufacturers. Therefore, experimental results and proposed curves are the most relevant 199 findings of this work. 200 The defined curves allow evaluating the thermal behaviour of these systems under any 201 operating condition of the GEHP system and based on parameters available in the manufacturer 202 datasheets. This developed model is valid for heating and cooling. Also it can be implemented 203 in performance simulation tools of energy buildings with control operation strategies. Then, 204 usability and replicability of this work is can be high. 205 The proposed model is highly accurate and easy to reproduce. The results of the validation 206 performed show a maximum error of 25% and an average error of less than 7%. 207 2 Characterization of the gas heat pump system 208 2.1 Basic operation of GEHP systems 209 Gas Heat Pumps (GEHP) are essentially the same as electric heat pumps (EHP), based on a 210 thermodynamic cycle in which energy is transferred from a cold source to a hot source by 211 application through the compressor shaft of the equipment. Its difference is that in the GEHP, 212 the electric motor coupled to the compressor in the electric heat pump is replaced by an internal 213 combustion engine based on an OTTO cycle. 214 This engine is fed with fossil gas fuel and must be cooled to remove the residual heat generated 215 during operation. The heat generated from cooling the engine block, the oil pan and exhaust 216 gases are transmitted to a hydraulic circuit to be used to produce hot water. This hot water is 217 produced from residual heat. Additionally, when external temperatures are excessively low, this 218 heat is used to prevent the formation of frost on the evaporator without the need for cycle 219 inversion. This functionality enables excellent performance from this type of machine at low 220 temperatures. Fig. 1 displays a schematic view of a gas engine heat pump (system) in cooling 221 mode. 222 223 Figure 1. Schematic diagram of a GEHP cycle in cooling mode 224 The right part of figure 1 demonstrates how the machine combustion engine consumes fuel and 225 air to drive the compressor of the heat pump. Cooling the engine (see IC BT in figure 1), totally 226 or partially, heat is recovered in the form of hot water; the remaining, if any, is evacuated to the 227 exterior. Besides, heat of exhaust gas is recovered (see IC AT in figure 1). If the left part of Figure 228 1 is analysed, the compression cycle of the system may be observed. In Figure 1, operation in 229 cooling mode is schematised. In this mode, the heat pump produces cold water in the 230 evaporator and heat dissipates in the condenser with the outside environment. The heating 231 mode is achieved by inverting the compression cycle on the left, as with conventional heat 232 pumps. Control of the machine is carried out on the rotational speed of the compressor, i.e., by 233 controlling the combustion engine. 234 GEHP systems are equipped with a 4-stroke engine and 3 or 4 cylinders in line, which moves 235 between 952-1998 cm3 rotating between 600 and 3000 rpm, depending on model, brand and 236 load. These motors rotate, employing a belt system, depending on the model, one or two open 237 scroll compressors. The management system of these motor-compressor assemblies decides 238 their optimum speed of rotation and eventually, the partial load ratio of the compressors. These 239 decisions depend on the needs in order to achieve maximum energy efficiency at all times, 240 especially at partial loads. 241 It should be noted that there are developments of two types of machines: hydronic systems that 242 produce cold or hot water to compensate for the demands of cooling and heating respectively, 243 and direct expansion systems. The latter pushes coolant in a gaseous state under different 244 pressure conditions for producing cold or heat in conditioned areas. This study focuses on 245 hydronic systems. 246 2.2 Assumptions 247 The proposed methodology aims to analyze and characterize GEHP systems through 248 experimental data. For this, detailed experiments have been carried out on different machines, 249 and a characterization model based on operating curves has been developed. The assumptions 250 considered in the development of this work after analysis of the technology and pooling with 251 manufacturers are the following: 252 - Operation is considered at a constant water flow rate (mW), i.e., the heat pump operates 253 by varying the TOW output temperature of the fluid streams with which it exchanges 254 heat. 255 - A gas internal combustion engine drives these machines. Their operation is carried out 256 with modulation between load levels (PLR): stopped (0%), half load (50%) and full load 257 (100%). 258 Therefore, a possible maximum and minimum load level has been established. 259 - The minimum load ratio is 30% according to manufacturers, with a maximum of 130%. 260 - The maximum heat production temperature in the condenser is 55°C. The minimum 261 evaporator output temperature is 4°C and the maximum recovery water temperature 262 may reach 65°C. 263 - 7°C is used as the minimum outside air temperature for recovery of residual heat from 264 the engine (TLIM-REC). Recovery is not possible at lower temperatures. 265 - The heat pump can operate producing heat for heating, DHW and cold water for cooling; 266 depending on the established set-point temperature (TSETPOINT). Additionally, hot water 267 will be generated from residual heat generated during engine cooling (QRECOVERY). 268 - The heat to be evacuated with the environment is carried out by air-coolers, from which 269 ventilator consumption is calculated. 270 The following sections describe the basis of the model and its formulation. 271 2.3 Description of the system characterisation 272 Characterisation of the GEHP system is based on behavioural equations obtained from 273 experimental data. Besides, algorithm contains s the sequence of required operations and 274 decisions regarding internal control. 275 This study adopts a computation approach for modelling the GEHP system and the installations. 276 Following the modelling philosophy, the system variables may be classified into three types: 277 It should be noted that the calculated energy flows relate to the desired operating 403 conditions. In point 3 (equations 11 and 12), the corresponding outlet temperatures are 404 calculated and analysed if valid within the operating limits of the machine. 405 Production of cold water in heat pump (Mode=2) 406 Cooling mode treatment is analogous to heating mode. Equation 7 enables estimation of 407 cooling power, equations 8 and 9 estimate energy consumption of the machine; and finally 408 equation 10 is equivalent to equation 4 but uses the behaviour curves of the machine in cold 409 mode to estimate recoverable heat: 410 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙[𝑘𝑊]= 𝑃𝐿𝑅 ·COLDCAPT·𝐶𝑜𝑙𝑑𝑛𝑜𝑚[𝑘𝑊] Eq. (7) 411 Where 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙[𝑘𝑊] is cooling capacity in real operation conditions, 𝐶𝑜𝑙𝑑𝑛𝑜𝑚[𝑘𝑊] is 412 cooling capacity in nominal operation conditions, COLDCAPT is variation of cooling capacity 413 (PLR=100%) as a function of outlet water temperature and outdoor wet-bulb temperature 414 and PLR is defined as the quotient between the energy delivered by the machine 415 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙[𝑘𝑊] and its nominal capacity 𝐶𝑜𝑙𝑑𝑛𝑜𝑚[𝑘𝑊]. 416 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑟𝑒𝑎𝑙[𝑘𝑊]= COLDCONPLR·COLDCONT·𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑛𝑜𝑚[𝑘𝑊] Eq. (8) 417 Where 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑟𝑒𝑎𝑙[𝑘𝑊] is gas consumption in cooling mode in real operation conditions, 418 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑛𝑜𝑚[𝑘𝑊] is gas consumption in cooling mode in nominal operation conditions, 419 COLDCONPLR is variation of gas consumption in cooling mode as a function of partial load 420 ratio and COLDCONT is variation of gas consumption in cooling mode (PLR=100%) as a 421 function of outlet water temperature and outdoor web-bulb temperature. 422 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐶𝑟𝑒𝑎𝑙[𝑘𝑊]= COLDECONPLR·COLDECONT·𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐶𝑛𝑜𝑚[𝑘𝑊] Eq. (9) 423 Where 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐻𝑟𝑒𝑎𝑙[𝑘𝑊] is the electricity energy consumption in heating mode in real 424 operation conditions, 𝐶𝑜𝑛𝐸𝑙𝑒𝑐𝐻𝑛𝑜𝑚[𝑘𝑊] is the electricity energy consumption in heating 425 mode in nominal operation conditions, HEATECONPLR is the variation of electricity 426 consumption in heating mode as a function of partial load ratio and HEATECONT is the 427 variation of electricity consumption in heating mode (PLR = 100%) as a function of outdoor 428 wet-bulb temperature. 429 𝑅𝑒𝑐𝑄𝐶𝑟𝑒𝑎𝑙[𝑘𝑊]= 𝑅𝐸𝐶COLDPLR·𝑅𝐸𝐶COLDT·𝑅𝑒𝑐𝑄𝐶𝑛𝑜𝑚[𝑘𝑊] Eq. (10) 430 Where 𝑅𝑒𝑐𝑄𝐶𝑟𝑒𝑎𝑙[𝑘𝑊] is heat recoverable in cooling mode in real operation conditions, 431 𝑅𝑒𝑐𝑄𝐶𝑛𝑜𝑚[𝑘𝑊] is heat recoverable in cooling mode in nominal operation conditions, 432 𝑅𝐸𝐶COLDPLR is the variation of heat recovery in cooling mode as a function of partial load 433 ratio and RECCOLDCAPT is the variation of recovery heat in cooling mode (PLR=100%) as a 434 function of outlet water temperature and outdoor web-bulb temperature. 435 3. Output flow temperatures are eliminated from heat fluxes: 436 Production of hot water in heat pump (Mode=1 or 3) 437 In heating mode, the water comes out hotter from both the condenser (equation 11) and 438 the heat recovery circuit of the engine (equation 12): 439 Tow[°C]=Heatreal[kW] CpW[kJ kg·K]·mW[kg s]+𝑇𝑖𝑤 Eq. (11) 440 Tow is condenser outlet water temperature, 𝑇𝑖𝑤 is condenser inlet water temperature, 441 Heatreal is heat capacity in real operation conditions, CpW is specific heat of water and mW 442 is condenser water flow. 443 Torec[°C]=𝑅𝑒𝑐𝑄𝐻𝑟𝑒𝑎𝑙[kW] CpW[kJ kg·K]·mrec[kg s]+𝑇𝑖𝑟𝑒𝑐 Eq. (12) 444 Torec is outlet water temperature of recovery loop, 𝑇𝑖𝑟𝑒𝑐 is inlet water temperature of 445 recovery loop, 𝑅𝑒𝑐𝑄𝐻𝑟𝑒𝑎𝑙is heat recoverable in heating mode in real operation conditions, 446 CpW is specific heat of water and mW is recovery circuit water flow. 447 Production of cold water in heat pump (Mode=2) 448 Conversely, in cooling mode the water is cooled in the evaporator according to equation 13 449 and the heated water temperature in the recovery circuit according to equation 14: 450 Tow[°C]= 𝑇𝑖𝑤 −𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙[𝑘𝑊] CpW[kJ kg·K]·mW[kg s] Eq. (13) 451 Tow is condenser outlet water temperature, 𝑇𝑖𝑤 is condenser inlet water temperature, 452 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙 is cooling capacity in real operation conditions, CpW is specific heat of water and 453 mW is condenser water flow. 454 Torec[°C]=𝑅𝑒𝑐𝑄𝐶𝑟𝑒𝑎𝑙[kW] CpW[kJ kg·K]·mrec[kg s]+𝑇𝑖𝑟𝑒𝑐 Eq. (14) 455 Torec is recovery circuit outlet water temperature, 𝑇𝑖𝑟𝑒𝑐 is recovery circuit inlet water 456 temperature, 𝑅𝑒𝑐𝑄𝐶𝑟𝑒𝑎𝑙is heat recoverable in cooling mode in real operation conditions, 457 CpW is specific heat of water and mW is recovery circuit water flow. 458 Accordingly, the calculated temperatures referenced the expected values according to 459 previously estimated flows and desired operating conditions. However, in Table 2, the 460 machine parameters are defined, where the operating limits are indicated. If the water flow 461 temperature calculated with equation 11 (heating mode) or 12 (cooling mode) is out of 462 limits, an iterative process is carried out to establish a suitable partial load factor close to 463 that desired. Moreover, the recovery temperature calculated with equations 13 and 14 464 cannot exceed a maximum value. Otherwise, the maximum value would be established, and 465 the remaining recoverable heat would dissipate. 466 4. Finally, once the heat fluxes are known and the desired, valid operating conditions are 467 established according to the iterative process. Then the efficiency indicators can be 468 calculated. These indicators may be divided into two groups, those referencing the system 469 cooling cycle (COP equation 15 and EER equation 16); and those referencing the entire 470 system ( 𝐶𝑂𝑃𝑆𝑌𝑆𝑇𝐸𝑀 equation 17 and 𝐸𝐸𝑅𝑆𝑌𝑆𝑇𝐸𝑀 equation 18): 471 𝐶𝑂𝑃 = 𝐻𝑒𝑎𝑡𝑟𝑒𝑎𝑙[𝑘𝑊] 𝐶𝑜𝑛𝐺𝑎𝑠𝐻𝑟𝑒𝑎𝑙[𝑘𝑊] Eq. (15) 472 𝐻𝑒𝑎𝑡𝑟𝑒𝑎𝑙 is heat capacity in real operation conditions, and 𝐶𝑜𝑛𝐺𝑎𝑠𝐻𝑟𝑒𝑎𝑙 is gas consumption 473 in heating mode in real operation conditions. 474 𝐸𝐸𝑅 =𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙[𝑘𝑊] 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑟𝑒𝑎𝑙[𝑘𝑊] Eq. (16) 475 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙 is cooling capacity in real operation conditions, and 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑟𝑒𝑎𝑙 is gas 476 consumption in cooling mode in real operation conditions. 477 If the entire system is considered, the numerator must be added to the heat flux recovered 478 in the engine. In such a way that the returns calculated with equations 17 and 18 are higher 479 than those estimated with equations 15 and 16: 480 𝐶𝑂𝑃𝑆𝑌𝑆𝑇𝐸𝑀 =𝐻𝑒𝑎𝑡𝑟𝑒𝑎𝑙[𝑘𝑊]+𝑅𝑒𝑐𝑄𝐻𝑟𝑒𝑎𝑙[𝑘𝑊] 𝐶𝑜𝑛𝐺𝑎𝑠𝐻𝑟𝑒𝑎𝑙[𝑘𝑊] Eq. (17) 481 𝐻𝑒𝑎𝑡𝑟𝑒𝑎𝑙 is heat capacity in real operation conditions, 𝑅𝑒𝑐𝑄𝐻𝑟𝑒𝑎𝑙 is heat recoverable in 482 heating mode in real operation conditions and 𝐶𝑜𝑛𝐺𝑎𝑠𝐻𝑟𝑒𝑎𝑙 is gas consumption in heating 483 mode in real operation conditions. 484 𝐸𝐸𝑅𝑆𝑌𝑆𝑇𝐸𝑀 =𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙[𝑘𝑊]+𝑅𝑒𝑐𝑄𝐶𝑟𝑒𝑎𝑙[𝑘𝑊] 𝐶𝑜𝑛𝐺𝑎𝑠𝐶𝑟𝑒𝑎𝑙[𝑘𝑊] Eq. (18) 485 𝐶𝑜𝑙𝑑𝑟𝑒𝑎𝑙 is cooling capacity in real operation conditions, 𝑅𝑒𝑐𝑄𝐻𝑟𝑒𝑎𝑙 is heat recoverable in 486 cooling mode in real operation conditions and 𝐶𝑜𝑛𝐺𝑎𝑠𝑟𝑒𝑎𝑙 is gas consumption in cooling 487 mode in real operation conditions. 488 The GEHP operation depends on the partial load factor (PLR, namely partial load ratio) and in 489 this paper, it refers to PLR as the ratio between the actual power transferred by the GEHP and 490 the maximum power (see equation 3). The latter is the definition of the capacity ratio (CR), more 491 useful to use when the interaction between plant and building has to be determined for 492 evaluation of real operation conditions. Indeed, it represents one of the relevant parameters to 493 consider for the correct identification of the operational curves. 494 3 Results 495 3.1 Experimental facility 496 Two manufacturers of this technology have collaborated in the development and validation of 497 the results indicated: Panasonic and Aisin. These manufacturers have selected several of their 498 commercial models, and have collaborated to carry out the operational experiments in climatic 499 chambers. These tests are used in many applications related to the experimental analysis of 500 construction elements or equipment [45–47]. Testing consists of generating a chamber with 501 controlled temperature and humidity conditions. The experiment consisted of the assembly of 502 the installation presented in Figure 4. 503 504 Figure 4. Schematic diagram of the experimental setup 505 Figure 4 shows a schematic position of sensors. There is three type of monitors: temperature 506 [T], pressure [P] and water flow [W]. Sensor of temperature is a thin film platinum resistor - 507 Pt100 - trimmed at a laboratory to comply with the resistance values of the standard IEC 60751, 508 class A (error less than 1%). The pressure sensor is a 1/4'NPT Pressure Gauge Pressure 509 Manometer (Hydraulic Gauge 0-4 Bar 0-60 PSI). Water flow sensor is a high-accuracy [0.2 % ±1 510 mm/s] mass flow measurement [measuring range 0 to 10 m/s]. 511 Installation of Figure 4 enables: 512 - Real-time data results in every 60 s for water temperatures, flow rates and pressures for 513 the most representative points of the installation. Sampling has been carried out within 514 a short period time to ensure stabilised measurements. 515 - Control of the GEHP unit operating parameters: heat/cold flows, recovery mode and 516 load of the engine. 517 - Generation of heat and cold loads in water flow 1 (heating and cooling). Maintenance 518 of constant heat flux 2 temperature for heat recovery. 519 - Control of return temperature of water flows entering the machine. 520 On the other hand, it is interesting to note that the experimental installation has an estimated 521 cost of more than 40000€ (not including the price of GEHP systems and climatic chambers). 522 Accordingly, two types of experiments are carried out. Type 1 experiment consists of a full load 523 operation of the engine, maintaining the water return temperature constant while varying the 524 environmental conditions and flow temperature. Type 2 experiment consists of maintaining the 525 environmental conditions and flow temperature constant, in order to vary the workload of the 526 engine. These experiments are carried out under different conditions until obtaining the 527 functional maps of the machines. 528 It has been used five machines from different manufacturers. Table 6 and 7 show datasheet of 529 the tested units. This sample represents the most usually units in the market since these 530 manufactures. 531 Table 6. Studied GEHP units: nominal parameters of heating 532 ID List of GEHP units Capacity [kW] Nominal conditions: heating mode Gas consumption [kW] Electricity consumption [kW] COP Recovery heat [kW] 1 Panasonic ECO-g with Hidrokit (U-20GE2E5 + S-500WX2E5) 60 42.5 0.65 1.48 20 2 Panasonic ECO-g with Hidrokit (U-30GE2E5 + S-710WX2E5) 80 68.1 1.46 1.40 30 3 AISIN AXGP450E1-16HP + AWS 16HP-E1(J) P450 47.5 30.2 1.24 1.58 16.5 4 AISIN AXGP560E1-20HP + AWS 20HP-E1(J) P560 60 42 1.24 1.50 20 5 AISIN AXGP710E1-25HP + AWS 25HP-E1(J) P710 75 53.6 1.4 1.49 25 Table 7. Studied GEHP units: nominal parameters of cooling 533 ID List of GEHP units Capacity [kW] Nominal conditions: cooling mode Gas consumption [kW] Electricity consumption [kW] EER Recovery heat [kW] 1 Panasonic ECO-g with Hidrokit (U-20GE2E5 + S-500WX2E5) 50 39.1 1.03 1.43 20 2 Panasonic ECO-g with Hidrokit (U-30GE2E5 + S-710WX2E5) 71 67.9 1.71 1.25 30 3 AISIN AXGP450E1-16HP with AWS 16HP-E1(J) P450 41 32 1.28 1.45 16.5 4 AISIN AXGP560E1-20HP with AWS 20HP-E1(J) P560 52 41.5 1.32 1.38 20 5 AISIN AXGP710E1-25HP with AWS 25HP-E1(J) P710 63 55.1 1.59 1.29 25 It is important to note that the nominal recoverable heat for DHW production is identical in both 534 operation modes due to being linked to the internal combustion engine. 535 3.2 Experimental results 536 This section describes the experimental data obtained according to the details summarised in 537 section 3.1. The different experiments carried out are summarized in Table 8 and Table 9. 538 Table 8. Experiment of heating mode: organization 539 WATER SYSTEM HEATING MODE Dry-bulb temperature (°C) -10 -5 0 5 10 15 20 Water Temperature (inlet-outlet) °C 30-35 File 1 File 2 File 3 File 4 File 5 File 6 File 7 40-45 File 8 File 9 File 10 File 11 File 12 File 13 File 14 Table 9. Experiment of cooling mode: organization 540 WATER SYSTEM COOLING MODE Dry-bulb temperature (°C) 0 10 20 25 30 35 40 45 Water Temperature (inlet-outlet) °C 12 - 7 File 15 File 16 File 17 File 18 File 19 File 20 File 21 File 22 23 - 18 File 23 File 24 File 25 File 26 File 27 File 28 File 29 File 30 The parameters chosen to perform the experiments in table 8 and ¡Error! No se encuentra el o 541 rigen de la referencia. have been established with the manufacturers. The environmental 542 conditions used analyse the minimum and maximum expected temperatures in the climatic 543 regions of the Mediterranean [48]. Furthermore, flow temperatures in heating mode (Table 8) 544 are 35°C and 45°C, which indicate a low flow temperature. However, these machines may reach 545 temperatures of 55°C - 60°C in the condenser flow with a poor performance [35]. On the other 546 hand, in cold mode (table 9) operation is outlined under usual flow conditions at 7°C and a return 547 at 12°C, with a high flow temperature of 18°C. This second alternative is designed so that the 548 GEHP system works as a base in a conventional air conditioning scheme, for example, for pre- 549 treatment of ventilation air. 550 Table 10 shows the variation of machine performance for a constant flow temperature, two 551 different environmental conditions, and a partial load factor varying between 30% to 100%. Data 552 of Table 10 is corresponding with experiment of file 10 (Table 10) 553 Table 10. Experiment of cooling mode: example file 18 554 Unit ID Mode Water Outlet T / Indoor T [°C] Outdoor T [°C] Load [%] Capacity [kW] Gas Consumption [kW-LHV] Electricity Consump tion [kW] Engine heat recovery [kW] COP/EER [Capacity / Gas consumption] Specific Engine heat recovery [Heat recovery / Gas consumpt ion] 3 Cool 7 25 30 14.7 7.2 0.1 3.6 2.04 0.51 3 Cool 7 25 40 17.3 8.8 0.2 4.4 1.97 0.51 3 Cool 7 25 50 21.9 11.9 0.2 6.0 1.84 0.50 3 Cool 7 25 60 26.5 15.1 0.3 7.6 1.75 0.50 3 Cool 7 25 70 31.0 18.2 0.4 9.1 1.70 0.50 3 Cool 7 25 80 35.2 21.1 0.4 10.6 1.67 0.49 3 Cool 7 25 90 39.5 24.1 0.5 12.1 1.64 0.49 3 Cool 7 25 100 44.0 27.2 0.5 13.6 1.62 0.49 However, the wealth of experimental data available enables detailed analysis. Given that the 555 measurements indicated in the table are obtained when the machine reaches a stationary 556 operating regime for given conditions. 557 As an example of the experimental results obtained, the behaviour of the Panasonic machine 558 U20 (ID 1, table 3 and table 4) is displayed in Figure 5 and Figure 6. Figure 5 displays the 559 operation in heating mode (production of hot water in the condenser) at 100% load and for two 560 flow temperatures (35°C and 45°C). 561 562 Figure 5. Analysis of Unit ID 1: influence of set-point in heating mode 563 Figure 5 shows that although heat generation in the cycle condenser is stable with the 564 temperature operating at 100%, gas consumption is affected. This is because the behaviour of 565 the gas engine worsens at low temperatures [35]; even manufacturers recommend that these 566 machines not be operated when the outdoor dry temperature is below -5°C. 567 In Figure 6, the experimental results are displayed for the same machine studied in Figure 5, but 568 in cooling mode (production of cold water in the evaporator). As in the previous case (Figure 5), 569 the experiments are carried out at 100% engine load and at two flow temperatures (7°C and 570 18°C). 571 572 Figure 6. Analysis of Unit ID 1: influence of set-point in cooling mode 573 The results displayed in Figure 6 are stable. The higher the flow temperature in the evaporator, 574 the better the performance, as is postulated in Carnot's theorem [49]. However, as the outdoor 575 temperature increases, it is more difficult for the condenser to dissipate the heat, lowering the 576 performance of the cooling cycle, thus lowering the cooling capacity to 100% of the engine. 577 Likewise, for the engine to maintain the operating range, it must increase its consumption to 578 ensure the flow temperature. 579 3.2.1 Operation maps: unit 1 580 In this research, the performance characteristics of the GEHP system for cooling and heating 581 application are analysed. 582 Experiments in cooling mode were carried out with an engine load from 100% to 25%, set-point 583 at 7°C, and ambient air temperature from 0°C to 45°C. Furthermore, experiments in heating 584 mode were carried out with an engine load from 100% to 25%, set-point at 45°C, and ambient 585 air temperature from -10°C to 20°C. 586 These experimental results coincide with the trends of other published studies such as 587 [24,27,30]. Below, the main results obtained from the experiments carried out are analysed and 588 described: Figure 7 displays how the cooling capacity (left) and the EER (right) vary at different 589 load levels and environmental conditions for a fixed flow temperature of 7°C. 590 591 Figure 7. Analysis of Unit ID 1: influence of partial load ratio in cooling mode (set-point 7°C) 592 Likewise, the heating mode is represented in Figure 8. Heat capacity appears in the left of Figure 593 8 and the COP performance on the right, at different load regimes and environmental conditions 594 for a flow temperature of 45°C. 595 596 Figure 8. Analysis of Unit ID 1: influence of partial load ratio in heating mode (set-point 45°C) 597 The other useful heat flux of the system is heat recovered in the engine (jackets and exhaust 598 gases). Variation of this heat flux with the machine load regime and environmental conditions 599 are displayed in Figure 9. 600 601 602 Figure 9. Analysis of Unit ID 1: recovery heat in cooling mode (left, set-point 7°C) and heating mode (right, set-point 603 45°C) 604 On the left of Figure 9 is the cold mode, with a flow temperature of 7°C. On the right is the 605 heating mode with a flow temperature of 45°C. It should be noted that when environmental 606 conditions are cold, engine heat is no longer recovered. That is why in heating mode in Figure 9 607 (right), the power recovered is much lower than that recoverable in cooling mode (left). 608 Additionally, the drop in heat recovered for temperatures above 40°C (left, cooling mode, figure 609 9) is due to the engine safety systems activating the dissipation mode. This mode operates 610 discreetly at various speeds and at these temperatures, the sensor activates the maximum 611 speed in order to reduce the amount recovered by the water flow in the engine jackets. 612 Finally, figure 10 displays the electrical consumption of the system in cold (left) and heat (right) 613 modes. 614 615 Figure 10. Analysis of Unit ID 1: electricity consumption in cooling mode (left, set-point 7°C) and heating mode 616 (right, set-point 45°C) 617 The electrical consumption (Figure 10) of these units is low, as it only references the motor 618 dissipation fans and internal electronics of the system. 619 Effect of ambient air temperature in cooling mode 620 When the ambient air temperature increases, the variation of energy is illustrated in Figure 7. 621 Also, cooling capacity decrease (Figure 7 - left) while energy consumption increases (Figure 10 - 622 left) when the ambient air temperature increases. Cooling capacity decrease but the specific 623 compressor power increases. Therefore, EER decreases by as shown in Figure 7 (right). The 624 recovery heat to full charge increased from 32 kW to 44 kW as the ambient air temperature 625 increased from 0 °C to 45 °C (Figure 9 - left). 626 Effect of gas engine speed in cooling mode 627 When the gas engine speeds increases, the working fluid mass flow, rate load of compressor and 628 compressor speeds are also increased, therefore the cooling capacity, energy consumption and 629 compressor power increase with the increase of gas engine speeds. 630 The gas engine recovery heat increased from 9.3 kW to 20.6 kW as the gas engine speed 631 increased 25 % to 100 % of charge (figure 9 - left). The reason is that the cooling capacity and 632 shaft power of the compressor increase with the gas engine speeds increasing. 633 Effect of ambient air temperature in heating mode 634 When the ambient air temperature increases from -10◦C to 20◦C, the total heating capacity 635 (Figure 8 - left) and electricity consumption (Figure 10 - right) increase by 7% and 6% 636 respectively. 637 As evaporating temperature rises, the compression ratios decrease. Also, the refrigerant mass 638 flow increases with the evaporate temperature increases. Therefore, analysis shows that both 639 condenser heat and energy consumption increase, as the ambient air temperature increases. 640 Since the variation trend of energy consumption is less than the trends of total heating capacity 641 and condenser heat, the performance of GEHP system would improve. Just as it is shown in 642 Figure 8 (right), the COP and of the system go up by 0.8 and 1.3 with the ambient air temperature 643 changing from -10◦C to 20◦C. 644 Effect of gas engine speed in heating mode 645 The total heating capacity (Figure 8 - left), the waste heat (Figure 9 - right) and the energy 646 consumption of the GEHP system (Figure 10 - right) increased as the gas engine speeds 647 increased. Figure 9 (right) indicates that the waste heat increases slowly with the rise of speed 648 when part load ratio is higher than 25%. 649 The COP indicates two different trends (Figure 8 - right): for PLR greater than or equal to 80%, 650 the warmer the environment, the lower the COP. On the other hand, this drop is compensated 651 as the load factor of the machine is reduced, therefore, at 70% the COP remains constant at high 652 temperatures. 653 3.2.2 Units comparison 654 For these comparisons, all the machines from the same manufacturer have been selected, i.e., 655 units ID 3, 4 and 5 according to table 1 and ¡Error! No se encuentra el origen de la r 656 eferencia.table 2. For these comparisons, EER and COP efficiency parameters referencing the 657 cooling cycle, normalised with respect to the nominal value provided in the manufacturer´s 658 catalogue, have been selected. 659 Figure 11 displays the comparison in cooling mode for a flow temperature of 7°C, an external 660 environment dry temperature of 30°C and different load levels from 30% to 100%. 661 662 Figure 11. Variation of EER in different partial load ratios (set-point=7°C, Dry-bulb temperature=30°C) 663 Figure 11 demonstrates that at a low load level, the performance of the machines is better than 664 the nominal value. Along this line, the machines behave the same. However, the larger unit (unit 665 id 5) offers significantly better performance. 666 Likewise, the heating mode analysis is displayed in Figure 12. In this case, a flow temperature of 667 35°C with ambient humid temperature conditions of 5°C is used. 668 [21] Sanaye S, Meybodi MA, Chahartaghi M. Modeling and economic analysis of gas engine 845 heat pumps for residential and commercial buildings in various climate regions of Iran. 846 Energy Build 2010;42:1129–38. doi:10.1016/j.enbuild.2010.02.004. 847 [22] Trnsys n.d. http://www.trnsys.com/. 848 [23] DOE. EnergyPlus - Engineering Reference 2010. 849 [24] Naldi C, Dongellini M, Morini GL. Climate influence on seasonal performances of air-to- 850 water heat pumps for heating. Energy Procedia 2015;81:100–7. 851 doi:10.1016/j.egypro.2015.12.064. 852 [25] Dongellini M, Naldi C, Morini GL. Seasonal performance evaluation of electric air-to- 853 water heat pump systems. Appl Therm Eng 2015;90:1072–81. 854 doi:10.1016/j.applthermaleng.2015.03.026. 855 [26] Shang S, Li X, Wu W, Wang B, Shi W. Energy-saving analysis of a hybrid power-driven heat 856 pump system. Appl Therm Eng 2017;123:1050–9. 857 doi:10.1016/j.applthermaleng.2017.04.151. 858 [27] Kamal R, Narasimhan AK, Wickramaratne C, Bhardwaj A, Goswami DY, Stefanakos EK, et 859 al. Field performance of gas-engine driven heat pumps in a commercial building. Int J 860 Refrig 2016;68:15–27. doi:10.1016/j.ijrefrig.2016.04.019. 861 [28] Department of Energy - United States of America n.d. 862 http://apps1.eere.energy.gov/buildings/energyplus/. 863 [29] Hu B, Li C, Yin X, Cao F, Shu P. Thermal modeling and experimental research of a gas 864 engine-driven heat pump in variable condition. Appl Therm Eng 2017;123:1504–13. 865 doi:10.1016/j.applthermaleng.2017.05.189. 866 [30] Liu H, Zhou Q, Zhao H. Experimental study on cooling performance and energy saving of 867 gas engine-driven heat pump system with evaporative condenser. Energy Convers 868 Manag 2016;123:200–8. doi:10.1016/j.enconman.2016.06.044. 869 [31] Liu FG, Tian ZY, Dong FJ, Cao GZ, Zhang R, Yan A Bin. Experimental investigation of a gas 870 engine-driven heat pump system for cooling and heating operation. Int J Refrig 871 2018;86:196–202. doi:10.1016/j.ijrefrig.2017.10.034. 872 [32] Liu FG, Tian ZY, Dong FJ, Cao GZ, Zhang R, Yan A Bin. Experimental investigation of a gas 873 engine-driven heat pump system for cooling and heating operation. Int J Refrig 874 2018;86:196–202. doi:10.1016/j.ijrefrig.2017.10.034. 875 [33] Elgendy E, Schmidt J. Optimum utilization of recovered heat of a gas engine heat pump 876 used for water heating at low air temperature. Energy Build 2014;80:375–83. 877 doi:10.1016/j.enbuild.2014.05.054. 878 [34] Zhang X, Yang Z, Wu X, Su X-C. Evaluation method of gas engine-driven heat pump water 879 heater under the working condition of summer. Energy Build 2014;77:440–4. 880 doi:10.1016/j.enbuild.2014.03.067. 881 [35] Liu FG, Tian ZY, Dong FJ, Yan C, Zhang R, Yan A Bin. Experimental study on the 882 performance of a gas engine heat pump for heating and domestic hot water. Energy Build 883 2017;152:273–8. doi:10.1016/j.enbuild.2017.07.051. 884 [36] Shah NN, Huang MJ, Hewitt NJ. Performance analysis of diesel engine heat pump 885 incorporated with heat recovery. Appl Therm Eng 2016;108:181–91. 886 doi:10.1016/j.applthermaleng.2016.07.123. 887 [37] Wu J, Ma Y. Experimental study on performance of a biogas engine driven air source heat 888 pump system powered by renewable landfill gas. Int J Refrig 2016;62:19–29. 889 doi:10.1016/j.ijrefrig.2015.08.023. 890 [38] Wang M, Chen Y, Liu Q. Experimental study on the gas engine speed control and heating 891 performance of a gas Engine-driven heat pump. Energy Build 2018;178:84–93. 892 doi:10.1016/j.enbuild.2018.08.041. 893 [39] Wang M, Yang Z, Su X, Zhang B, Wu X, Shi Y. Simulation and experimental research of 894 engine rotary speed for gas engine heat pump based on expert control. Energy Build 895 2013;64:95–102. doi:10.1016/j.enbuild.2013.04.003. 896 [40] Commission THEE. L 239/136 2013:136–61. 897 [41] Pérez-Lombard L, Ortiz J, Maestre IR, Coronel JF. Constructing HVAC energy efficiency 898 indicators. Energy Build 2012;47:619–29. doi:10.1016/j.enbuild.2011.12.039. 899 [42] Alves O, Monteiro E, Brito P, Romano P. Measurement and classification of energy 900 efficiency in HVAC systems. Energy Build 2016;130:408–19. 901 doi:10.1016/j.enbuild.2016.08.070. 902 [43] Zhang Q, Yang Z, Li N, Feng R, Gao Y. The influence of building using function on the 903 operating characteristics of the gas engine driven heat pump with energy storage system 904 (ESGEHPs). Energy Build 2018;167:136–51. doi:10.1016/j.enbuild.2018.02.039. 905 [44] DOE2.com Home Page n.d. http://www.doe2.com/ (accessed December 28, 2018). 906 [45] Riederer P. Thermal Room Modeling Adapted To The Test Of HVAC Control Systems 907 2001;37:777–90. 908 [46] Peris B, Navarro-Esbrí J, Molés F, Martí JP, Mota-Babiloni A. Experimental 909 characterization of an Organic Rankine Cycle (ORC) for micro-scale CHP applications. Appl 910 Therm Eng 2015;79:1–8. doi:10.1016/j.applthermaleng.2015.01.020. 911 [47] Velasco Gómez E, Tejero González A, Rey Martínez FJ. Experimental characterisation of 912 an indirect evaporative cooling prototype in two operating modes. Appl Energy 913 2012;97:340–6. doi:10.1016/j.apenergy.2011.12.065. 914 [48] Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. World map of the Köppen-Geiger climate 915 classification updated. Meteorol Zeitschrift 2006;15:259–63. doi:10.1127/0941- 916 2948/2006/0130. 917 [49] Yunus A. Çengel & Michael A. Boles. Termodinámica McGraw-Hill. n.d. 918 [50] Söderström T. System identification. UK: Prentice Hall International; 1989. 919 [51] Coleman HW, Steele WG. Appendix B: Taylor Series Method (Tsm) for Uncertainty 920 Propagation. Exp Validation, Uncertain Anal Eng 2018:311–23. 921 doi:10.1002/9781119417989.app2. 922 [52] Refsgaard JC, van der Sluijs JP, Højberg AL, Vanrolleghem PA. Uncertainty in the 923 environmental modelling process - A framework and guidance. Environ Model Softw 924 2007;22:1543–56. doi:10.1016/j.envsoft.2007.02.004. 925 [53] Uusitalo L, Lehikoinen A, Helle I, Myrberg K. An overview of methods to evaluate 926 uncertainty of deterministic models in decision support. Environ Model Softw 927 2015;63:24–31. doi:10.1016/J.ENVSOFT.2014.09.017. 928 929 APPENDIX: Operation Curves A. Mathematical formulation Bi-quadratic 𝐶𝑢𝑟𝑣𝑒𝑖= 𝑃0+𝑃1·(𝑋11·𝑋20)+𝑃2·(𝑋10·𝑋21)+𝑃3·(𝑋12·𝑋20)+𝑃4·(𝑋11·𝑋21)+𝑃5·(𝑋10·𝑋22) Cubic 𝐶𝑢𝑟𝑣𝑒𝑖= 𝑃0+𝑃1·(𝑋1)+𝑃2·(𝑋2)+𝑃3·(𝑋3) Quadratic 𝐶𝑢𝑟𝑣𝑒𝑖= 𝑃0+𝑃1·(𝑋1)+𝑃2·(𝑋2) B. Coefficients of experimental performance curves Name Form Description Independient Variables COOLCAPT biquadratic Variation of cooling capacity (PLR=100%) as a function of outlet water temperature and outdoor dry-bulb temperature f(X1=Tow , X2=Tdr) Unit P0 P1 P2 P3 P4 P5 1 0.60980 0.01367 0.01189 -0.00045 0.00035 -0.00023 2 0.64560 0.01674 0.01680 -0.00070 0.00047 -0.00035 3 0.83434 0.00000 0.01841 0.00095 -0.00014 -0.00039 4 0.77431 0.00000 0.02122 0.00095 -0.00015 -0.00043 5 0.79190 0.00578 0.00893 -0.00019 0.00015 -0.00018 Generic 1.05716 0.00000 -0.00178 0.00061 -0.00004 -0.00001 Name Form Description Independient Variables HEATCAPT biquadratic Variation of heating capacity (PLR=100%) as a function of outlet water temperature and outdoor wet-bulb temperature f(Tow , Twb) Unit P0 P1 P2 P3 P4 P5 1 0.50540 0.02584 0.00328 -0.00036 0.00000 0.00000 2 0.34350 0.03674 0.00696 -0.00050 0.00000 0.00000 3 1.03489 0.00000 -0.00288 0.00000 0.00000 -0.00002 4 1.02542 0.00000 -0.00143 0.00000 0.00000 -0.00001 5 0.44300 0.03067 0.01718 -0.00040 -0.00026 0.00000 Generic 1.07058 0.00000 0.00357 -0.00004 -0.00001 -0.00006 Name For m Description Independient Variables COLDCON T biquadrati c Variation of gas consumption in cooling mode (PLR=100%) as a function of outlet water temperature and outdoor dry-bulb temperature f(Tow , Tdr) Unit P0 P1 P2 P3 P4 P5 1 0.81222 0.00000 0.00069 0.00042 0.00000 0.00012 2 0.75870 0.01873 -0.00076 -0.00012 0.00000 0.00022 3 0.25240 0.03885 0.00274 -0.00068 0.00020 0.00000 4 1.21713 0.00000 -0.02508 0.00000 0.00000 0.00054 5 0.71900 0.00469 -0.00997 -0.00033 0.00015 0.00023 Generic 0.81222 0.00000 0.00069 0.00042 0.00000 0.00012 Name Form Description Independient Variables COLDCONPLR cubic Variation of gas consumption in cooling mode as a function of partial load ratio f(PLR) Unit P3 P2 P1 P0 1 -1.59370 3.51510 -1.32740 0.41130 2 -5.22440 10.82100 -5.72920 1.13180 3 -0.79557 1.73773 -0.12780 0.18333 4 -0.54861 1.47396 -0.11711 0.19021 5 -2.56890 5.71440 2.18880 0.46660 Generic 0.70654 -3.30249 7.03230 -3.44537 Name Form Description Independient Variables HEATCON T biquadrati c Variation of gas consumption in heating mode (PLR=100%) as a function of outlet water temperature and outdoor webbulb temperature f(Tow , Twb) Unit P0 P1 P2 P3 P4 P5 1 1.85600 -0.02670 -0.08942 0.00034 0.00124 0.00029 2 1.65600 -0.00736 -0.08347 0.00013 0.00118 0.00000 3 1.34289 0.00000 -0.02702 -0.00005 -0.00025 0.00124 4 1.64300 -0.01897 -0.05980 0.00038 0.00017 0.00058 5 1.43900 -0.01052 -0.04892 0.00013 0.00049 0.00032 Generic 1.34289 0.00000 -0.02702 -0.00005 -0.00025 0.00124 Name Form Description Independient Variables HEATCONPLR cubic Variation of gas consumption in heating mode as a function of partial load ratio f(PLR) Unit P3 P2 P1 P0 1 -1.05590 1.60310 0.15110 0.28970 2 -1.06650 1.69790 0.22120 0.16560 3 -0.85200 1.87325 -0.21929 0.22296 4 -0.36694 1.08476 0.13034 0.14888 5 -0.87400 1.37640 0.24390 0.22730 Generic -0.02904 2.16531 -2.92832 1.79735 Name Form Description Independient Variables REC COLDCAP T biquadrati c Variation of heat recovery in cooling mode (PLR=100%) as a function of outlet water temperature and outdoor dry-bulb temperature f(Tow , Tdr) Unit P0 P1 P2 P3 P4 P5 1 0.99243 0.00000 -0.00867 0.00020 0.00001 0.00026 2 0.99243 0.00000 -0.00867 0.00020 0.00001 0.00026 3 1.26277 0.00000 -0.02602 0.00000 0.00000 0.00056 4 1.26277 0.00000 -0.02602 0.00000 0.00000 0.00056 5 1.26277 0.00000 -0.02602 0.00000 0.00000 0.00056 Generic 0.72208 0.00000 0.00868 0.00039 0.00002 -0.00003 Name Form Description Independient Variables RECCOLDCAPPLR cubic Variation of heat recovery in cooling mode as a function of partial load ratio f(PLR) Unit P3 P2 P1 P0 1 -0.41546 2.78635 -2.70191 1.32965 2 -0.41546 2.78635 -2.70191 1.32965 3 -0.54688 1.47017 -0.11443 0.18961 4 -0.54688 1.47017 -0.11443 0.18961 5 -0.54688 1.47017 -0.11443 0.18961 Generic -0.28404 4.10252 -5.28938 2.46970 Name Form Description Independient Variables RECHEATCA PT biquadrati c Variation of recovery heat in heating mode (PLR=100%) as a function of outlet water temperature and outdoor webbulb temperature f(Tow , Tdr) Unit P0 P1 P2 P3 P4 P5 1 0.04191 0.00000 0.03641 0.00002 0.00008 -0.00102 2 0.04191 0.00000 0.03641 0.00002 0.00008 -0.00102 3 -0.10189 0.00000 0.07282 0.00005 0.00016 -0.00233 4 -0.18330 0.00000 0.11172 0.00013 -0.00076 -0.00256 5 -0.18330 0.00000 0.11172 0.00013 -0.00076 -0.00256 Generic 0.18571 0.00000 0.00000 0.00000 0.00000 0.00029 Name Form Description Independient Variables RECHEATCAPPLR cubic Variation of heat recovery in heating mode as a function of partial load ratio f(PLR) Unit P3 P2 P1 P0 1 0.07458 0.09393 2.61711 -16.93921 2 0.11173 0.16955 1.73482 -8.59481 3 0.21281 -0.06572 1.52274 -0.67266 4 0.14887 0.24517 0.85252 -0.25041 5 0.14887 0.24517 0.85252 -0.25041 Generic 0.02334 1.05579 0.02194 -0.11154 Name Form Description Independient Variables COLDECONT quadratic Variation of electricity consumption in cooling mode (PLR = 100%) as a function of outdoor dry-bulb temperature f(Tdr) Unit P2 P1 P0 1 -0.46112 0.06886 -0.00109 2 -1.69531 0.16130 -0.00269 3 0.64459 -0.01644 0.00028 4 0.85369 -0.02124 0.00040 5 -1.63750 0.13656 -0.00219 Generic 0.45367 0.02297 -0.00015 Name Form Description Independient Variables COLDCONPLR cubic Variation of electricity consumption in cooling mode as a function of partial load ratio f(PLR) Unit P3 P2 P1 P0 1 0.07332 0.53436 0.43348 -0.04095 2 -0.40693 2.96228 -3.39177 1.84245 3 -0.19078 1.95915 -1.59652 0.83342 4 0.07332 0.53436 0.43348 -0.04095 5 -0.19078 1.95915 -1.59652 0.83342 Generic 0.52562 0.67958 0.05207 -0.24550 Name Form Description Independient Variables HEATECONT quadratic Variation of electricity consumption in heating mode (PLR = 100%) as a function of outdoor wet-bulb temperature f(Twb) Unit P2 P1 P0 1 0.85369 -0.02124 0.00040 2 0.89771 -0.01500 0.00019 3 0.64459 -0.01644 0.00028 4 0.85369 -0.02124 0.00040 5 0.64459 -0.01644 0.00028 Generic 0.73846 0.03282 -0.00100 Name Form Description Independient Variables HEATECONPLR cubic Variation of electricity consumption in heating mode as a function of partial load ratio f(PLR) Unit P3 P2 P1 P0 1 0.15004 0.12615 1.09094 -0.36977 2 0.05629 0.49903 0.62168 -0.17719 3 0.21993 -0.23053 1.85004 -0.84163 4 0.15004 0.12615 1.09094 -0.36977 5 0.21993 -0.23053 1.85004 -0.84163 Generic 0.27827 2.61989 -3.35744 1.47284