Fluid Selection of Transcritical Rankine Cycle for Engine Waste Heat Recovery Based on Temperature Match Method

: Engines waste a major part of their fuel energy in the jacket water and exhaust gas. Transcritical Rankine cycles are a promising technology to recover the waste heat e ﬃ ciently. The working ﬂuid selection seems to be a key factor that determines the system performances. However, most of the studies are mainly devoted to compare their thermodynamic performances of various ﬂuids and to decide what kind of properties the best-working ﬂuid shows. In this work, an active working ﬂuid selection instruction is proposed to deal with the temperature match between the bottoming system and cold source. The characters of ideal working ﬂuids are summarized ﬁrstly when the temperature match method of a pinch analysis is combined. Various selected ﬂuids are compared in thermodynamic and economic performances to verify the ﬂuid selection instruction. It is found that when the ratio of the average speciﬁc heat in the heat transfer zone of exhaust gas to the average speciﬁc heat in the heat transfer zone of jacket water becomes higher, the irreversibility loss between the working ﬂuid and cold source is improved. The ethanol shows the highest net power output of 25.52 kW and lowest electricity production cost of $1.97 / (kWh) among candidate working ﬂuids.


Introduction
Internal combustion engines (ICEs) are important power sources and are responsible for around 60% of all oil produced [1]. However, the brake thermal efficiency and fuel economy of current engines are still quite low. One of the reasons that limit further improvement is that part of the heat energy from fuel combustion is wasted to the ambient through multiple media such as exhaust gas, jacket water, turbo-charge coolant, exhaust gas recirculation coolant, etc. [2]. Hence, great potential in engine waste heat recovery has been put forward and widely concerned if the waste heat can be transformed into mechanical or electricity power. The Rankin cycle (RC) is supposed to be a promising bottoming cycle technology when engine waste heat recovery is set as the topping cycle [3].
Many researches about RCs are devoted in biomass [4], solar power plants [5], geothermal [6], industrial waste heat [7], combined heat and power generation [8], etc. As for the RCs applied to the engine waste heat recovery, the first study proposed may be traced to the 1970s when Patel et al. [9] compounded the truck diesel engine with a bottoming cycle of the subcritical organic Rankine fluids under different heat source temperatures. Shu et al. [22] conducted a theoretic analysis on a dual-loop TRC system in an engine waste heat recovery to overcome the large temperature differences between the high-temperature exhaust gas and low-temperature engine jacket water. Six candidate working fluids were compared and analyzed. Li et al. [23] proposed a novel configuration coupling the supercritical and subcritical heat absorption processes to increase the heat-power conversion efficiency and to improve the adaptability to heat sources. Meng et al. [24] conducted a thermoeconomic performance comparison between the TRC and ORC in a low-temperature heat source utilization. The authors found that the net power output could be improved by 128.5% at maximum for the TRC and considered it a great potential in low-temperature heat source recovery due to its good environmental properties.
In addition to cycle comparison and configuration modification, a key design factor of the TRCs is the working fluid selection. Recently, a growing number of scholars have devoted their efforts to selecting the best working fluid from numerous candidates in the specific application area considering various optimization objectives. On one hand, when the operation stability and compatibility with materials of different working fluids are investigated as criteria, several indicators are included, mainly considering the working fluid properties. Liu et al. [25] firstly proposed concepts of dry, wet, and isentropic working fluids classified by the positive, negative, and infinite slope of the T-s curve, respectively. Further assertion was given by Hung et al. [26] that isentropic or dry fluids were suggested to avoid liquid droplet impingement in the turbine blades during the expansion. One solution to the use of wet working fluids was given by Desai et al. [27] to ensure the superheated state at the turbine inlet. On the other hand, when the thermodynamic performances of different working fluids are studied, a brief review is also given below. The critical temperature of different working fluids showed great impacts on system thermodynamic performances. Heberle et al. [28] compared the second law of efficiency of various fluids and recommended that the working fluids with high critical temperatures like isopentane were suitable for the ORCs. Similar results could be found when the boiling temperatures of fluids were compared. Actually, the relations between the critical temperature and boiling temperature were given by Joback [29] that the boiling temperature became higher if the critical temperature became higher for the fluids in the same fluid family. Besides the critical temperature, the ratio of latent heat to sensible heat is commonly used and modified to evaluate exergetic performances of working fluids. The specific heat of a working fluid is an important property, since the direct impact is caused by different values of specific heats on the pressurization and expansion processes. Maizza and Papadopoulos et al. [30,31] believed that the lower specific heat would decrease the work consumed by the pump and increase the power output indirectly. Furthermore, a higher specific heat was better-suited to obtain large expansion work. Stijepovic et al. [32] observed that working fluids with a low ratio of latent heat to sensible heat, which would be inclined to extract more heat from the heat source, showed better performances regarding exergy efficiency. A combination of the defined Jakob number and the average temperature between the evaporation temperature and condensation temperature was deduced by Su et al. [33]. They recommended it as more favorable for system thermal efficiency when a smaller Jakob number was achieved. The molecular complexity and molecular weight of different working fluids also has an impact on the system thermodynamic performance, especially on the expansion process [34]. As for the environmental aspects, the main concerns included the ozone depletion potential (ODP), global warming potential (GWP), and the atmospheric lifetime (ALT).
After a concise literature review about different performances of working fluids in TRCs and fluid selection instructions, it is concluded that the way to improve the temperature match between working fluids and thermal sources such as zeotropic mixtures and cycle modification shows a great impact on systems' performances. In addition, the reason why different thermodynamic performances of different working fluids are performed can be derived from the multiple properties of working fluids. Nevertheless, a current limitation of the reviewed working fluid selection is that a sufficient selection condition of what kind of properties the best working fluid possesses is proposed only after the comparison of their thermodynamic or economic performances. It seems to show a lack of active fluid selection instruction from the perspective of what kind of properties better-working fluids should possess. Additionally, the thermodynamic and economic performances of selected working fluids are served as a validation of the active fluid selection instruction. Hence, based on the temperature match method, which has been proved to be an effective and efficient way to achieve configuration modification, an optimal TRC layout is proposed to accomplish the active working fluid selection. The illustration and validation of the approach will allow the development of selecting working fluids actively from another dimension. Hence, the aim of the study is twofold: (i) set up a temperature match method-based working fluid selection instruction to obtain better working fluid properties with better temperature match performances between the bottoming system and multiple engine heat sources, as well as the cold source, and (ii) conduct a thermodynamic and economic analysis to validate the proposed method and optimize important operation parameters.

Heat Sources
In this paper, a 6-cylinder diesel engine in-line with a 2-stage turbocharger and intercooler for heavy-duty trucks was selected for waste heat recovery [35]. The heat balance test of the diesel engine without a waste heat recovery system was conducted to investigate the amount of recoverable heat. The major parameters of the target diesel engine are listed in Table 1. The air fuel ratio for the diesel engine is usually a variable to optimize combustion, leading to the change of exhaust gas composition under different engine loads, as shown in Table 2. The constituent part and mass fraction of the exhaust compositions under the design condition of 1300rpm and 1279 N·m are set as: CO 2 = 15.2%, H 2 O = 6.0%, N 2 = 73.0%, and O 2 = 5.8%. Additionally, detailed parameters of waste heat sources and engine design conditions are shown in Table 3. According to the heat balance test, the exhaust gas temperature reached 469.4 • C, whereas the mass flow rate of the jacket water reached 2.42 kg/s under the operation conditions of 1300 rpm and 1279 N·m, which were dominated by various waste heat sources and should be recovered primarily.

TRC Configuration
In this research, a split dual regenerative transcritical Rankine cycle (SR-TRC) is investigated shown in Figure 1. Due to the design of split branches and two regenerators, the SR-TRC system can meet the requirement of recovering the jacket water and exhaust gas simultaneously and improve the temperature matching in the whole heat absorption process, and has been proved to be a novel cycle configuration to achieve better temperature match between the system and heat sources [36]. The concrete working process of SR-TRC is: the working fluid transferring heat from jacket water in the preheater (2-3) is split into two parts, namely, high-temperature branch (H-branch) and low-temperature branch (L-branch). The working fluid in H-branch completely absorbs waste heat from exhaust gas in the gas heater (3)(4)(5) and then generates power in the expander-generator (5-6), whereas the working fluid in L-branch streams through H-regenerator (3)(4) to absorb heat after expansion and then generates power (4)(5)(6)(7)(8)(9)(10)(11). After the working fluid of H-branch streams through the H-regenerator (6-7), there still exists recoverable heat to be utilized. Thus, the L-regenerator (7-8) is designed to heat the working fluid (1-2) after pressurization (10-1). The working fluid in both branches enters condenser (9-10) to regain the capability to work after their convergence (8-9 and 11-9).

TRC configuration
In this research, a split dual regenerative transcritical Rankine cycle (SR-TRC) is investigated shown in Figure 1. Due to the design of split branches and two regenerators, the SR-TRC system can meet the requirement of recovering the jacket water and exhaust gas simultaneously and improve the temperature matching in the whole heat absorption process, and has been proved to be a novel cycle configuration to achieve better temperature match between the system and heat sources [36]. The concrete working process of SR-TRC is: the working fluid transferring heat from jacket water in the preheater (2-3) is split into two parts, namely, high-temperature branch (H-branch) and lowtemperature branch (L-branch). The working fluid in H-branch completely absorbs waste heat from exhaust gas in the gas heater (3)(4)(5) and then generates power in the expander-generator (5-6), whereas the working fluid in L-branch streams through H-regenerator (3)(4) to absorb heat after expansion and then generates power (4)(5)(6)(7)(8)(9)(10)(11). After the working fluid of H-branch streams through the Hregenerator (6-7), there still exists recoverable heat to be utilized. Thus, the L-regenerator (7-8) is designed to heat the working fluid (1-2) after pressurization (10-1). The working fluid in both branches enters condenser (9-10) to regain the capability to work after their convergence (8-9 and 11-9).

Mathematical Modeling
The study is performed by means of a thermodynamic and economic analysis focusing on the energetic, exergetic, and economic performances. Several reasonable assumptions below are defined to simplify the computation process [37][38][39]: (1) each component and operating point is under equilibrium and in a steady-state condition; (2) heat losses and pressure losses are neglected in the pipes and in all components; (3) the isentropic efficiencies of the expander, pump, and generator are assumed to be 0.7, 0.8, and 0.9, respectively; (4) the changes in the kinetic and potential energy of the fluids can be neglected; (5) the pinch point temperature differences (Tpp) in the gas heater, preheater, and regenerator are set to be 30 °C, 5 °C, and 15 °C; and (6) temperatures of the condensation process and the ambient are set at 25 °C to guarantee that CO2 can be condensed stably during the phase change process. The properties of the working fluid are derived from the REFPROP-NIST [40].

Mathematical Modeling
The study is performed by means of a thermodynamic and economic analysis focusing on the energetic, exergetic, and economic performances. Several reasonable assumptions below are defined to simplify the computation process [37][38][39]: (1) each component and operating point is under equilibrium and in a steady-state condition; (2) heat losses and pressure losses are neglected in the pipes and in all components; (3) the isentropic efficiencies of the expander, pump, and generator are assumed to be 0.7, 0.8, and 0.9, respectively; (4) the changes in the kinetic and potential energy of the fluids can be neglected; (5) the pinch point temperature differences (T pp ) in the gas heater, preheater, and regenerator are set to be 30 • C, 5 • C, and 15 • C; and (6) temperatures of the condensation process and the ambient are set at 25 • C to guarantee that CO 2 can be condensed stably during the phase change process. The properties of the working fluid are derived from the REFPROP-NIST [40].

Thermodynamic Analysis
Thermodynamic modeling of the combined system can be constituted by the energetic and exergetic terms. Based on the first law of thermodynamics, the energy balance equations of the system components are listed in the second column of Table 4. An exergy analysis based on the second law of thermodynamics is necessary to judge the irreversibility of each component during the investigation of the cycles. The exergy loss equations of the system components are listed in the third column of Table 4. To make full use of one heat source at least, the mass flow rate (

Components 1st Law of Thermodynamics 2nd Law of Thermodynamics
In order to ensure the stable operation of the heat exchanger, the outlet temperature of the exhaust gas from the gas heater should not be lower than the acid dew point temperature (120 • C) to prevent the corrosion of the heat exchanger [39]. In addition, the outlet temperature of the jacket water after the preheater should be higher than the returned temperature to avoid an unnecessary impact on the Energies 2020, 13, 1830 7 of 19 engine operation. Based on the aforementioned conditions, the heat recovery efficiency and system total efficiency are defined as:

Economic Analysis
Considering the characteristics of the heat exchanger-based split cycles, the heat transfer areas of the heat exchangers are calculated by: where K is the total heat transfer coefficient of the heat exchanger, and ∆t is the log mean temperature difference (LMTD) between the hot side and the cold side.
To determine the unit cost of the streams in the conversion system, the capital cost according to the economic situation in the year 2014 is expressed as: where CEPCI 2001 = 382 and CEPCI 2014 = 586.77 (CEPCI means chemical engineering plant cost index).
The purchased cost C p and the cost factor F bm are two main parameters that affect the base investment cost of each component. Here, the C p is calculated by the capacity of the components, and the relevant coefficient K, and the cost factors, including direct project expenses, contingency and contractor fees, indirect project expenses, and the auxiliary facilities, are taken into consideration by F bm . The relevant investment models and coefficients according to [41] are listed in Table 5. In addition, the capital recovery (CRF) is estimated as: where i is the interest rate, and the value is set to be 5%; time is the economic lifetime, and its value is set to 15 years. The annuity of the investment A nk can be expressed as: The electricity production cost (EPC) can be calculated by the equation: where f k is the operation, maintenance, and insurance cost factor, and its value is 1.65%; h is the operation hours of a year, and its value is set to be 2190 h considering the practical operation.
The aforementioned calculations were conducted through MATLAB. To help readers get a good grasp of the calculation and optimization procedure, the flow chart is given in Figure 2. Table 5. Calculation coefficients of the components.

Components
Heat exchanger

Fluid Selection Indicator
An active fluid selection instruction is illustrated in this section from the point of working fluid property to achieve better temperature match performance. The novel design of SR-TRC system is characterized by improving the temperature match between the working fluid and heat sources. To further improve the temperature match between the working fluid and cold source, the inlet temperature of the working fluid of the condenser should be kept as low as possible to make full use of the waste heat and reduce the heat transported to the cold source. The principle can be explained that, when the inlet temperature of the condenser is lower, the average temperature difference between the working fluid and cold source can be reduced to get closer to the ideal heat transfer process. Concretely, the temperature of the working fluid after the L-Regenerator in the H-branch, namely, state 8, should be kept low enough. In the meantime, the expander outlet temperature in the

Model Validation
The model of the transcritical Rankine cycle system with a preheater and a regenerator was verified with the same operating parameters as in reference [42], and the comparison results are listed in Table 6 based on the thermodynamic analysis as follows. A 25-MW gas turbine was used as the heat source under a condensing temperature T cond of 20 • C and a expander inlet temperature T expander of 391 • C. The RD (relative deviation) is described as: Energies 2020, 13, 1830 The results calculated in this paper are in a good agreement with those of the reference. Therefore, the models in this paper are sufficiently accurate for our investigation. Table 6. Comparison of the results calculated in this paper with the data from reference [42]. RD: relative deviation.

Working Fluid Screening
Working fluid selection is very important, since it directly influences performances such as net power output, thermal efficiency, and heat transfer area, etc. The choice of the working fluid is a complex problem, because it implies the selection of the appropriate working fluid among a large number of candidates. A concise literature review has been carried out in the Introduction section. In this study, 18 different and commonly used working fluids are listed in Table 7, and the primary selection was conducted by means of their pyrolytic decomposition temperatures. Considering the high temperature of the exhaust gas (469.4 • C), the candidates, which had higher pyrolytic decomposition temperatures than 300 • C, were selected and analyzed to avoid the decomposition problem. Hence, the final working fluids selected included CO 2 , R143a, R123, R11, and ethanol. It is worth noting that the mentioned working fluids, such as R11 (ODP = 1, GWP = 1), R143a (ODP = 0, GWP = 3800), and R22 (ODP = 0.055, GWP = 1700), may cause severe environmental impacts, including depletion of the ozone layer and global warming, and have been banned or limited in use in future equipment by the Montreal Protocol on Substances that Deplete the Ozone Layer and Kyoto Protocol. Nevertheless, in order to give a clear explanation of the proposed method of active fluid selection, the aforementioned candidates are only used for the validation of the fluid selection method from the temperature match standpoint, despite some of them being restricted from the perspective of environmental protection.

Fluid Selection Indicator
An active fluid selection instruction is illustrated in this section from the point of working fluid property to achieve better temperature match performance. The novel design of SR-TRC system is characterized by improving the temperature match between the working fluid and heat sources. To further improve the temperature match between the working fluid and cold source, the inlet temperature of the working fluid of the condenser should be kept as low as possible to make full use of the waste heat and reduce the heat transported to the cold source. The principle can be explained that, when the inlet temperature of the condenser is lower, the average temperature difference between the working fluid and cold source can be reduced to get closer to the ideal heat transfer process. Concretely, the temperature of the working fluid after the L-Regenerator in the H-branch, namely, state 8, should be kept low enough. In the meantime, the expander outlet temperature in the L-branch, namely, state 11, should not be so high either.
An effective and efficient way to meet such challenges is to decrease the value of m f,e but increase the value of m f,j simultaneously. Based on the equations in Table 4, the aforementioned mass flow rate mainly depends on the specific heat c p . The average specific heat of the working fluid in the heat transfer zone of the jacket water is expressed as c p,j , whereas the corresponding working fluid average specific heat in the heat transfer zone of the exhaust gas is expressed as c p,e . Concretely, a higher value of average specific heat c p,e when absorbing heat from the exhaust gas and a lower value of c p,j when absorbing heat from the jacket water will increase the m f,j but decrease the m f,e simultaneously. As shown in Figure 3, the c p /c p,max of all five candidate fluids increase firstly and decrease with the increase of temperature. Nevertheless, the peak values are distributed in different heat transfer regions. It is found that the ethanol shows the highest c p,e and lowest c p,j , whereas the CO 2 performs a contrary tendency. That is to say, a greater potential in temperature match between the ethanol and cold source can be achieved compared to the other fluids. In brief, a higher value of c p,e / c p,j will provide a better temperature match performance between the system and cold source. The comparison of c p,max and c p,e / c p, j are listed in Table 8. An effective and efficient way to meet such challenges is to decrease the value of mf,e but increase the value of mf,j simultaneously. Based on the equations in Table 4, the aforementioned mass flow rate mainly depends on the specific heat c p . The average specific heat of the working fluid in the heat transfer zone of the jacket water is expressed as c p,j , whereas the corresponding working fluid average specific heat in the heat transfer zone of the exhaust gas is expressed as c p,e . Concretely, a higher value of average specific heat c p,e when absorbing heat from the exhaust gas and a lower value of c p,j when absorbing heat from the jacket water will increase the mf,j but decrease the mf,e simultaneously. As shown in Figure 3, the cp/cp,max of all five candidate fluids increase firstly and decrease with the increase of temperature. Nevertheless, the peak values are distributed in different heat transfer regions. It is found that the ethanol shows the highest c p,e and lowest c p,j , whereas the CO2 performs a contrary tendency. That is to say, a greater potential in temperature match between the ethanol and cold source can be achieved compared to the other fluids. In brief, a higher value of c p,e / c p,j will provide a better temperature match performance between the system and cold source. The comparison of cp,max and c p,e / c p,j are listed in Table 7.   Pinch analysis techniques are widely used to instruct temperature match analysis and optimize the energy conversion systems. The pinch analysis is deduced by straightforward thermodynamics, and the primary tasks are setting the energy target and approaching the target. Various system configurations and working fluids are designed and simulated to recover as much heat as possible and minimize external heating and cooling. Generally, the T-Q diagrams or T-h diagrams are utilized to visually show what kind of process is inherently capable of achieving the energy target as much as possible. In this research, the typical heat sources, including the jacket water and exhaust gas from the diesel engine, are combined into one composite curve by means of summing the heat load at each temperature. Comparison results of various working fluids are presented in Figure 4. The horizontal axis represents the heat flow rate (kW), whereas the vertical axis represents temperature ( • C). Different tendencies of the heat absorption procedure can be obtained. When the fluids such as CO 2 and R143a are investigated, it is found that there exists a smaller average temperature difference between the system and heat source, whereas a large temperature difference between the system and cold source is also found. The reason can be explained by the relatively lower value of c p,e / c p,j illustrated before in Table 8. On the contrary, the performance of the temperature match between the system and cold source is better when R123, R11, and ethanol are studied. Specifically, the irreversibility loss in the condensation process is decreased by 24.2 kW when the ethanol is chosen as the working fluid compared to CO 2 .
To estimate the temperature match performance clearly, the irreversibility loss in the heat exchangers is commonly used. Hence, the comparison for various working fluids of exergy destruction in the preheater, gas heater, and condenser under the 1.1-times critical pressures are presented in Figure 5. It is observed that the maximum total exergy destruction can be obtained when CO 2 is investigated as the working fluid. Moreover, the exergy destruction varies among different exchangers. The maximum destruction can be obtained in the condenser when CO 2 and R143a are studied. The gas heater shows the largest irreversibility loss with R11 and R123, whereas the exergy destruction in the preheater reaches the highest with ethanol. In addition, the irreversibility loss of the system and cold source is significantly reduced in R11, R123, and ethanol, which proves that the method to select the working fluid actively with a higher value of c p,e / c p,j to improve the temperature match between the system and cold source is instructive in TRCs couple with the engine waste heat recovery. In summary, the ethanol provides the best temperature match performance between the system and cold source compared with the other candidates. Although the proposed selection method of the higher value of c p,e / c p,j provides effective results to improve the temperature match performance, further thermodynamic and economic analyses are also necessary to comprehensively verify the instructions.
Energies 2020, 13, x FOR PEER REVIEW 11 of 19 illustrated before in Table 7. On the contrary, the performance of the temperature match between the system and cold source is better when R123, R11, and ethanol are studied. Specifically, the irreversibility loss in the condensation process is decreased by 24.2 kW when the ethanol is chosen as the working fluid compared to CO2.
(e) ethanol To estimate the temperature match performance clearly, the irreversibility loss in the heat exchangers is commonly used. Hence, the comparison for various working fluids of exergy destruction in the preheater, gas heater, and condenser under the 1.1-times critical pressures are presented in Figure 5. It is observed that the maximum total exergy destruction can be obtained when CO2 is investigated as the working fluid. Moreover, the exergy destruction varies among different exchangers. The maximum destruction can be obtained in the condenser when CO2 and R143a are studied. The gas heater shows the largest irreversibility loss with R11 and R123, whereas the exergy destruction in the preheater reaches the highest with ethanol. In addition, the irreversibility loss of  To estimate the temperature match performance clearly, the irreversibility loss in the heat exchangers is commonly used. Hence, the comparison for various working fluids of exergy destruction in the preheater, gas heater, and condenser under the 1.1-times critical pressures are presented in Figure 5. It is observed that the maximum total exergy destruction can be obtained when CO2 is investigated as the working fluid. Moreover, the exergy destruction varies among different exchangers. The maximum destruction can be obtained in the condenser when CO2 and R143a are studied. The gas heater shows the largest irreversibility loss with R11 and R123, whereas the exergy destruction in the preheater reaches the highest with ethanol. In addition, the irreversibility loss of the system and cold source is significantly reduced in R11, R123, and ethanol, which proves that the method to select the working fluid actively with a higher value of c p,e / c p,j to improve the temperature match between the system and cold source is instructive in TRCs couple with the engine waste heat recovery. In summary, the ethanol provides the best temperature match performance between the system and cold source compared with the other candidates. Although the proposed selection method of the higher value of c p,e / c p,j provides effective results to improve the temperature match performance, further thermodynamic and economic analyses are also necessary to comprehensively verify the instructions.

Thermodynamic Analysis
In this section, the net power output and exergy destruction are selected as the thermodynamic performance indicators, which are shown in Figure 6 and Figure 7, respectively. Parametric analysis

Thermodynamic Analysis
In this section, the net power output and exergy destruction are selected as the thermodynamic performance indicators, which are shown in Figures 6 and 7, respectively. Parametric analysis of the expander inlet temperature and pressure on the net power output is presented in Figure 6 with different map diagrams. When the working fluids of CO 2   jacket water, such as CO2 and R143a, the net power output of will increase firstly and then decrease with the increase of the expander inlet pressure. Contrarily, when the critical temperature is near the exhaust gas heat absorption zone, such as R123, R11, and ethanol, the net power output will decrease with the increase of the expander inlet pressure. Based on the aforementioned analyses, fluid selection with temperature match analysis is in favor of improving a system's thermodynamic performances.

Economic Analysis
The distribution of the electricity production cost of different working fluids is presented in Figure 8. Similar, but not the same, the trends of EPC look like the variation of net power output, whereas the specific values are distinct slightly. When the working fluid CO2 is studied, the optimal EPC reaches $2.72/(kWh) under the expander inlet temperature and pressure of 360 °C and 12.4 MPa. The expander inlet temperature brings about a slight effect on the EPC when the expander inlet pressure is kept unchanged. Moreover, the operation parameters when the maximum net power output is achieved are different from the values when the lowest EPC can be achieved, which means that the multiobjective optimization is necessary to provide competitive instruction for the decisionmaker. An optimal EPC of $2.08/(kWh) can be achieved by R143a when the expander inlet temperature and pressure is optimized under 370 °C and 4.4 MPa. For the working fluids R123, R11, and ethanol, they also present monotonicity within the preset operation ranges similar to the trend when the net power output is used as the function. The 1.1-times critical pressure is considered to be the best operation pressure with the minimum EPC of $2.06/(kWh), $2.14/(kWh), and $1.97/(kWh). Therefore, the working fluid ethanol, combining both the calculation results of the net power output and EPC, shows the best thermoeconomic performance with the highest net power output (25.52 kW) and lowest electricity production cost ($1.97/(kWh)) among the candidate working fluids. As presented in Figure 7, an apparent reduction of exergy destruction in the condenser with the increase of the inlet pressure are found especially in CO 2 . The reason can be explained that the value of c p,e / c p,j for CO 2 increases with the increase of the expander inlet pressure, leading to the decrease of m f,j /m f,e . Hence, the temperature match between the system and cold source is improved, and the irreversibility loss is reduced by 64%. From the point of dynamic components, both working fluids perform higher exergy destructions in pressurization and expansion processes under higher inlet pressures. In addition, the temperature rise in the L-Regenerator increases, since the specific heat near the critical temperature region of both fluids increases significantly. Thus, the average temperature difference between the working fluid and jacket water in the preheater decreases, and the irreversibility loss decreases.
In brief, when the critical temperature of the working fluid is near the heat transfer zone of the jacket water, such as CO 2 and R143a, the net power output of will increase firstly and then decrease with the increase of the expander inlet pressure. Contrarily, when the critical temperature is near the exhaust gas heat absorption zone, such as R123, R11, and ethanol, the net power output will decrease with the increase of the expander inlet pressure. Based on the aforementioned analyses, fluid selection with temperature match analysis is in favor of improving a system's thermodynamic performances.

Economic Analysis
The distribution of the electricity production cost of different working fluids is presented in Figure 8. Similar, but not the same, the trends of EPC look like the variation of net power output, whereas the specific values are distinct slightly. When the working fluid CO 2 is studied, the optimal EPC reaches $2.72/(kWh) under the expander inlet temperature and pressure of 360 • C and 12.4 MPa. The expander inlet temperature brings about a slight effect on the EPC when the expander inlet pressure is kept unchanged. Moreover, the operation parameters when the maximum net power output is achieved are different from the values when the lowest EPC can be achieved, which means that the multiobjective optimization is necessary to provide competitive instruction for the decision-maker. An optimal EPC of $2.08/(kWh) can be achieved by R143a when the expander inlet temperature and pressure is optimized under 370 • C and 4.4 MPa. For the working fluids R123, R11, and ethanol, they also present monotonicity within the preset operation ranges similar to the trend when the net power output is used as the function. The 1.1-times critical pressure is considered to be the best operation pressure with the minimum EPC of $2.06/(kWh), $2.14/(kWh), and $1.97/(kWh). Therefore, the working fluid ethanol, combining both the calculation results of the net power output and EPC, shows the best thermoeconomic performance with the highest net power output (25.52 kW) and lowest electricity production cost ($1.97/(kWh)) among the candidate working fluids.

Conclusions
In this work, a novel temperature match method was defined for the active working fluid selection based on the relationship between the physical property of working fluid and heat sources' syncretizing pinch analysis techniques. The main objective of the method was to improve the temperature match between the system and cold source to achieve a higher net power output and lower exergy destruction and electricity production cost. The further fluid selection instruction and optimization procedure could be finished by means of the proposed method. In addition, the parametric analysis was presented when the expander inlet temperature and pressure were chosen to be optimized to verify whether the proposed method was effective or not. The main conclusions are listed below:

Conclusions
In this work, a novel temperature match method was defined for the active working fluid selection based on the relationship between the physical property of working fluid and heat sources' syncretizing pinch analysis techniques. The main objective of the method was to improve the temperature match between the system and cold source to achieve a higher net power output and lower exergy destruction and electricity production cost. The further fluid selection instruction and optimization procedure could be finished by means of the proposed method. In addition, the parametric analysis was presented when the expander inlet temperature and pressure were chosen to be optimized to verify whether the proposed method was effective or not. The main conclusions are listed below: (1) Considering the engine waste heat recovery, a better temperature match between the SR-TRC system and cold source could be achieved when the critical temperature of the working fluid was within or near the heat transfer zone of the exhaust gas. In addition, a new fluid selection indicator for multiple heat sources in the engine was defined. Concretely, the average specific heat of the working fluid in the heat transfer zone of the exhaust gas and jacket water were defined as c p,e and c p,j , respectively. The larger the ratio of c p,e / c p,j , the smaller the irreversibility loss of the heat exchange process between the working fluid and cold source mainly attributed to a better temperature match performance, which could be an effective instruction for active working fluid selection. (2) From the point of parametric analysis, which was utilized in a validation of the proposed fluid selection method, a higher expander inlet temperature led to a higher net power output. The different trends of thermodynamic performances were caused by different properties of working fluids. Concretely, the net power output and EPC of the SR-TRC system increased firstly and then decreased with the increase of the expander inlet pressure when the critical temperature of the working fluid was within or near the jacket water heat transfer zone, such as CO 2 and R143a. The net power output and EPC decreased monotonously with the increase of the expander inlet pressure if the critical temperature of the working fluid was within or near the heat transfer zone of the exhaust gas, such as R123, R11, and ethanol. (3) The results showed that the best thermodynamic and economic performances of the system could be achieved when the ethanol was utilized as the working fluid. The optimum net power output and EPC of the system reached 25.52 kW and $1.97/(kWh), which were improved by 47.86% and 27.24% compared with CO 2 . However, when both the objectives of the net power output and EPC are investigated, further multiobjective optimization is necessary, since the optimal operation parameters for both objectives are different from each other.
The presented fluid selection instruction and parametric analysis can be improved by further studies. On one hand, the current fluid selection standard mainly considers the temperature match performance affected by different properties of working fluids to achieve a higher net power output and lower EPC. Some important properties of working fluids from the point of environmental protection should be comprehensively evaluated. On the other hand, the multiobjective optimization to combine the thermodynamic and economic analyses is necessary to provide competitive instruction for the designers' reference.

Conflicts of Interest:
The authors declare no conflicts of interest.