Abstract
A unique hybrid cooling, heating, and power (HCHP) concept has been recently developed as an alternative to environmental control units. It combines a small-scale organic Rankine cycle (ORC) with a vapor compression cycle. The unique drive-train design flexibly and efficiently converts engine waste heat into useful energy in the form of cooling, heating, and power depending upon the energy needs. Compared to a standard military environmental control unit which puts an electric load on a diesel generator, the HCHP system uses engine exhaust heat as the primary energy input. Utilizing the exhaust heat can potentially provide 27% reduction on fuel consumption when operating in the cooling mode. When cooling is not needed, it is able to provide power and/or heating output using engine waste heat—a significant advantage over other heat activated cooling technologies. The prototype unit based on the HCHP design has been developed to demonstrate the concept. It leveraged the microchannel heat exchanger and scroll expander technologies to achieve high-performance, small-size, and low-cost design in order to meet the growing distributed energy applications.
1 Introduction
While the demand for energy continues to grow worldwide at a faster pace, a huge amount of energy is being wasted through various combustion and manufacturing processes. It has been estimated that industrial low-grade waste heat accounts for more than 50% of heat generated [1]. In order to address this increasing demand and consequence of global warming, improving energy efficiency and integrating renewable energy will be a key to energy sustainability into the future.
In recent years, renewable energy sources have played increasingly important roles in meeting growing energy demand while curbing carbon emissions. Among various technologies, the organic Rankine cycle (ORC) is a relatively mature technology and has been effective in converting thermal energy sources such as geothermal [2], solar thermal [3,4], and waste heat to power [5–7]. Dependent upon the source temperature, a wide range of working fluids have been studied by various research groups [8–11]. The desired working fluids not only present attractive thermodynamic advantages, but also provide thermal transport and economic benefits in designing compact systems. Given the large base of fluid choices including pure fluids and mixtures of various hydrocarbons and hydrofluorocarbons, ORCs have been demonstrated to achieve appreciable efficiency over a wide range of temperatures from 120 °C to 370 °C [12–14].
In this work, we demonstrate a unique design of using ORC to convert recovered thermal energy into mechanical energy. Instead of outputting electrical energy alone as a conventional ORC system in the aforementioned studies, the mechanical energy is then converted to other form of energy (i.e., cooling, heating and power) depending upon the energy needs. The flexible cooling, heating, and power functions are achieved through innovative coupling with a cooling cycle and an electric generator. Some of the system and component modeling, prototype development, and experiment validation are presented in the following sections.
2 A System Concept of Hybrid Cooling, Heating, and Power
So far, most of the studies related to waste heat recovery have been focusing on converting the thermal energy to either power or cooling. Essentially all the cited work related to ORC in the prior section are converting heat to power. However, being able to convert waste heat to cooling directly through advanced cycles (e.g., absorption and adsorption cycles) and components remains very attractive because of significant demand for space cooling. As a result, significant amount of work has gone into this field [15]. A comprehensive review of the state-of-art heat activated cooling technologies using solar thermal energy has been summarized by Refs. [16–19]. Although great progress has been made for absorption heat pumps, in a recent study led by Garimella et al. [20], the technology is still facing challenges in terms of COP, cost, and system complexity. On the other hand, with discovery of various metal organic frameworks (MOFs), the research effort of developing MOF-based adsorption heat pump systems for various cooling and heating applications has been on the rise in recent years [21,22]. While conventional refrigerants and water have been used as the working fluids, some studies also used alcohol as the adsorption cycle working fluids with MOFs [23,24]. Heat pump systems based on MOF have also been compared with the zeolite-based adsorption system [25].
In contrast, the hybrid system concept presented has flexible cooling, heating, and/or power outputs. The original intent of the hybrid system is to utilize waste heat from diesel generators as the energy input, but it certainly can be extended to renewable energy sources (e.g., solar thermal and geothermal). Given its flexible drivetrain design, it can deliver usable energy output year-round, which could significantly improve the levelized cost of energy (LCOE). Given the compact components and portable design, it manifests itself to be suitable for distributed energy system applications.
A schematic diagram of the hybrid cooling, heating, and power (HCHP) concept is shown in Fig. 1. In the near term, it can be used to improve energy efficiency of widely used military environmental control units (ECUs)—a device providing both cooling and heating functions in various military environment. The hybrid features enable the system to convert waste heat from the diesel generator to power when cooling (as the primary ECU function) is not needed. In the case of cooling being partially needed, the extra waste heat from the generator can be outputted as electric power. This allows more energy to be recovered and utilized through a single multi-functional system.
As a typical example for military application, in order to meet 5.3 kW (1.5 tons) of cooling for an existing military ECU while delivering 4.4 kW of electricity for various electric loads, the diesel generator has to consume 24.7 kW of diesel fuel (energy equivalent). The energy flow for this existing ECU case is shown in Fig. 2, which also included an assumption for the ECU coefficient of performance (COP) of 1.8.
For the same application, the hybrid cooling, heating and power (or a hybrid ECU) system uses diesel generator exhaust heat as the primary energy input, which is in contrast to electrical power as the sole energy input for generating either cooling or heating for current ECUs. The waste heat in engine exhaust is first recovered through a high performance, low impact microchannel heat exchanger. The heat is then converted to shaft power using a small-scale ORC. Detailed of the small-scale ORC can be found here [26]. The shaft power is then used to drive a compressor, providing required cooling, which otherwise would come from electricity produced by the diesel generator. This can greatly reduce the fuel consumption due to cooling loads, as shown in Fig. 3. Compared to current military ECU practice shown in Fig. 2, the hybrid ECU system can reduce fuel consumption by 27%. This kind of improvement on energy efficiency can significantly improve military operations and reduce logistic burdens.
This hybrid ECU concept also provides great flexibility when waste heat for a diesel generator is not enough in meeting the cooling demand. During those situations, the motor/generator within the hybrid ECU system can draw additional power from the generator to supplement the cooling need. In contrast, when there is significant amount of waste heat from the generator but there is no or reduced cooling need, this hybrid system can output power to offset the power needed from the generator, thus reducing overall fuel consumption. Given its flexibility, the following operational modes can be achieved with the hybrid ECU system.
Sufficient genset waste heat → Cooling
Insufficient waste heat + motor power → Cooling
Genset waste heat → Power (no cooling needs)
Genset waste heat → Cooling + Power (partial cooling needs)
3 Prototype Development and Key Component Performance
In order to demonstrate the hybrid cooling, heating, and power system concept as a military ECU, a flexible drivetrain using pulleys and timing belt to connect the rotating equipment was adopted initially and a schematic of it is shown in Fig. 4. Three electromagnetic clutches were used to engage or disengage the individual component to the drivetrain, achieving the flexible energy input and output functions as designed.
To complete the hybrid ECU concept, the other key feature built into the system was a high efficiency heating function. During the design phase, three different heating options were considered and evaluated, which included electrical resistance heating, traditional heat pump approach, and novel waste heater approach utilizing the condensing heat from the ORC for indoor space heating. Meanwhile, the indoor air provides the cooling capacity needed to condense the ORC working fluid. Thermodynamic analyses were performed to determine expected values for key engineering specifications for each of the options. For the resistance heating option, the resistant heater shown in Fig. 5(b) is turned on (both the ORC and ECU are turn off). For the heat pump option, the ECU is turn on while the ORC and resistant heater are turned off. For the waste heat option, the ORC is turn on while the ECU and the resistant heater are off. The results from the analysis are listed in Table 1. According to the analyses, the heating COP (i.e., heating capacity over electricity consumption) is much more superior for the waste heater design from the standpoint of energy efficiency. This is because the only electrical consumption for the waste heater design is from the fan attached to the condenser, while compressor draws much more power for the heat pump option. For resistant heaters, the highest COP would be unit. Thus, although the waste heater option adds significantly more weight to the system, utilization of exhaust heat through the ORC for heating provides substantial energy savings to achieve the same heating capacity. Furthermore, the ORC power cycle generates “free” power when the system is running in the heating mode.
Analysis of heating options corresponding to key engineering specifications
Heater type | COPheating | Mass (kg) |
---|---|---|
Resistance heater | 0.95 | 0.9 |
Heat pump | 4.68 | 3.6 |
Waste heater | 22.77 | 5.2 |
Heater type | COPheating | Mass (kg) |
---|---|---|
Resistance heater | 0.95 | 0.9 |
Heat pump | 4.68 | 3.6 |
Waste heater | 22.77 | 5.2 |
Figure 5(a) highlights the waste heater approach within the ORC power cycle, and it uses the condensing heat of the power cycle working fluid to heat the indoor air. From the cycle thermodynamic perspective, it acts as a secondary condenser for the ORC. A set of two valves were used to switch between the main ORC condenser and the waste heater. In addition, a bank of three resistant heaters were added in order to provide more robust ECU design so that heating needs can still be met when waste heat is unavailable. Figure 5(b) shows the resistant heater bank providing roughly 2.7 kW heating capacity.
A rectangular design was selected for the waste heater/secondary condenser. The rectangular design allows the waste heater to fit around the current centrifugal fan that provides the indoor air circulation. When the centrifugal fan takes in room air it then pushes the air through the hot fins of the waste heater, heating up the air before being redistributed to the indoor environment. A CAD model depicting the condenser fan operation is shown in Fig. 6(a). The rectangle waste heater is composed of four pieces of aluminum flat tubes—A, B, C and D shown in Fig. 6(b). Special attention was given in designing the fin spacing on the air side, in order to balance high heat transfer and minimum pressure drop to maintain good air flow through both the evaporator and waste heater. Given the heat transfer area for the waste heater and current condenser heat duty, the air heating capacity of the new waste heater was estimated to be around 6.3 kW.

(a) Assembled waste heat condenser with arrows showing air flow paths and (b) computer-aided design (CAD) model of the waste heat condenser
At the component level, various high-performance microchannel heat exchangers were used during the prototype demonstration [27]. One of the key components in the ORC power cycle is the microchannel boiler used from a previous waste heat recovery project [28]. The primary focus during this project, however, was to design the vapor side manifold to ensure its mechanical integrity under power cycle pressure (300–400 psig) while achieving uniform flow distributions for the incoming working fluid. Both finite element analysis and computational fluid dynamics were used to aid the design phase by simulating the maximum stress under 600 psig design pressure and flow distributions of various internal features within the manifold. The plots showing the simulation results are included in Figs. 7 and 8. According to the stress analysis using the finite element approach, the designed wall thickness can provide a static safety factor of 2. After incorporating the features illustrated in Fig. 8(a), good flow distributions can be achieved based on the CFD analysis. The inlet manifold of the boiler is modeled in ansys fluent. The three-dimensional Navier–Stokes equations are solved using the SIMPLE algorithm with the velocity boundary condition at the inlet (the tube opening) and pressure boundary condition at the manifold outlet (the mating plane with the boiler channel inlet). No slip boundary condition is also applied at the manifold walls. The manifolds were made using SS316 with silicone O-ring grooves for sealing. Separate pressure and flow tests were conducted before assembled it into the system.

(a) Assembly of the microchannel ORC boiler, (b) Von Mises stress for the manifold under 600 psig, and (c) static displacement under 600 psig

(a) Internal flow features of the manifold, (b) vector plot of the side view, and (c) vector plot of the front view—the circle showing incoming fluid
A microchannel heat exchanger design was also used for the ORC recuperator as shown in Fig. 9(a). The recuperator is a counter-flow heat exchanger with the hydraulic diameter for an individual channel at 240 µm. Compared to conventional plate heat exchangers, microchannel recuperators have significantly higher heat transfer coefficients due to short diffusion length scale associated with thinner thermal boundary layers. Along with high surface area to volume ratio, the size of microchannel heat exchangers can be substantially smaller than conventional plate heat exchangers. As shown in Fig. 9(b), the clear advantage of using microchannel design over conventional plate heat exchanger design is observed through a comparison study [27].

(a) Picture of the counter-flow microchannel recuperator and (b) experimental study to compare the microchannel and plate recuperators
A high efficiency scroll expander was developed based on an earlier project [29]. A picture of it without thermal insulation is shown in Fig. 10(a). The measured isentropic efficiency as a function of rotating speed and pressure ratio is presented in Figs. 10(b) and 10(c). As shown, its efficiency is much less sensitive to rotating speed. Unlike turbines, this makes scroll expanders very suitable for distributed energy applications where heat sources and electric loads will likely be fluctuating. The data also shows its isentropic efficiency stays fairly consistent for a wide range of pressure ratios. Given this kind of performance and at least an order of magnitude less in manufacturing cost, the scroll expander will find wide applications for distributed energy systems.

(a) Picture of the scroll expander, (b) isentropic efficiency as function of rotational speed, and (c) isentropic efficiency as function of pressure ratio
Complete views of the first hybrid ECU prototype based on the flexible cooling, heating, and power concept are shown in Fig. 11. Its primary energy input is waste heat from internal combustion engines; however it can also operate like a standard military ECUs (i.e., use electric power to provide cooling and heating). The arrows show the indoor and outdoor air flows when the system is running at various modes. A process and instrumentation diagram for the current prototype is shown in Fig. 12 to further illustrate the overall process and measurements. In addition, the sensors used for the experiment, as well as their specifications and relative uncertainties for the measurements are shown in Table 2.

(a) complete CAD model of the hybrid ECU prototype showing the heating function and (b) the finished first prototype system using diesel generator waste heat for flexible cooling, heating and power
Instrumentation specifications and relative uncertainty
Description | Manufacturer | Model number | Accuracy | Relative uncertainty |
---|---|---|---|---|
Turbine flowmeter | Omega | FTB-938 | 1% reading 0.25% repeatability | 2% |
Turbine flowmeter | Omega | FTB-901T | 0.5% reading 0.05% repeatability | 2% |
Pressure transducers | Cole-Parmer | All | 0.25% FS | 1% |
Thermocouples | Omega | KMQSS-062U-6 | Greater of 2.2 °C or 0.75% of reading | 2.5% |
Description | Manufacturer | Model number | Accuracy | Relative uncertainty |
---|---|---|---|---|
Turbine flowmeter | Omega | FTB-938 | 1% reading 0.25% repeatability | 2% |
Turbine flowmeter | Omega | FTB-901T | 0.5% reading 0.05% repeatability | 2% |
Pressure transducers | Cole-Parmer | All | 0.25% FS | 1% |
Thermocouples | Omega | KMQSS-062U-6 | Greater of 2.2 °C or 0.75% of reading | 2.5% |
4 Experimental Setup and Testing Results
4.1 Functional Testing.
Given the main objective of the work is to demonstrate a hybrid/flexible cooling, heating, and power concept for distributed energy applications, after integrating the system the preliminary functional tests were conducted to verify its design features. A process oil heater was used to simulate waste heat input to the prototype system as shown in Fig. 13. In addition, a load bank was connected with the motor/generator controller, allowing the generated power measured by the controller to be absorbed by the power resistors. The flexible energy output features were demonstrated in various operation modes shown below in the screenshots. Figure 14 shows a screenshot of the labview control program while the system is outputting both cooling and power (All the electromagnetic clutches to the compressor and the generator were turned on.). As shown on the screen, only 27 W of power was generated while the rest of the power (on the order of 500–600 W) was delivered to the compressor in order to meet the cooling load.
In addition, the cooling-only and power-only output modes were demonstrated, and a screen shot of each is shown in Fig. 15. With the cooling-only case (a), the evaporating temperature (3.1 °C) was significantly lower than the previous case as the compressor and expander were running at much higher speed (3200 rpm vs. 2200 rpm). For the power-only case, all the expander power from the ORC delivered to the generator, with the total output power increased to 573 W. For cases of power-only (cooling cycle was turn off), a detailed pressure and temperature measurement at key cycle state points are provided in Table 3. Each row represents a unique run time (total six are shown). Although the experiment lacked accurate flow measurement during the initial functional testing (due to a broken turbine flowmeter), using the control speed of the feed pump and the steady-state system pressures (e.g., the expander inlet pressure, P_EI), the relative flowrate for each case in Table 3 can be determined. As shown, the Run 3 in the table had the highest system pressure, which corresponds to the highest R245fa flowrate. Consequently, it delivered the highest power output from the ORC power cycle as expected. Given the flowrate and overall heat input stayed relatively constant during the other runs, the power output was only fluctuating within a narrow band.

(a) A screenshot showing the current system operation mode of outputting cooling only, and (b) a screenshot showing the current system operation mode of outputting power only
Key state points measurement and system performance at various run time
Runs | T_RLI [C] | T_BI [C] | T_EI [C] | T_EO [C] | T_PCI [C] | P_RLI [psia] | P_EI [psia] | P_EO [psia] | P_PCI [psia] | Speed [RPM] | Power [W] |
---|---|---|---|---|---|---|---|---|---|---|---|
1 | 35.9 | 85.5 | 133.9 | 101.1 | 58.3 | 277.6 | 274.7 | 66.3 | 62.2 | 2602 | 529 |
2 | 36.0 | 86.2 | 134.8 | 101.4 | 59.5 | 276.3 | 272.6 | 68.9 | 64.9 | 2804 | 606 |
3 | 43.1 | 76.6 | 129.1 | 86.6 | 58.2 | 295.0 | 293.9 | 66.2 | 60.2 | 2799 | 650 |
4 | 43.7 | 75.4 | 119.9 | 84.0 | 57.2 | 258.7 | 256.0 | 64.8 | 59.6 | 2797 | 554 |
5 | 39.6 | 84.2 | 133.3 | 101.4 | 52.4 | 272.2 | 269.7 | 56.3 | 51.5 | 2797 | 585 |
6 | 30.4 | 85.1 | 143.6 | 107.6 | 48.6 | 241.1 | 239.8 | 50.5 | 46.6 | 2803 | 564 |
Runs | T_RLI [C] | T_BI [C] | T_EI [C] | T_EO [C] | T_PCI [C] | P_RLI [psia] | P_EI [psia] | P_EO [psia] | P_PCI [psia] | Speed [RPM] | Power [W] |
---|---|---|---|---|---|---|---|---|---|---|---|
1 | 35.9 | 85.5 | 133.9 | 101.1 | 58.3 | 277.6 | 274.7 | 66.3 | 62.2 | 2602 | 529 |
2 | 36.0 | 86.2 | 134.8 | 101.4 | 59.5 | 276.3 | 272.6 | 68.9 | 64.9 | 2804 | 606 |
3 | 43.1 | 76.6 | 129.1 | 86.6 | 58.2 | 295.0 | 293.9 | 66.2 | 60.2 | 2799 | 650 |
4 | 43.7 | 75.4 | 119.9 | 84.0 | 57.2 | 258.7 | 256.0 | 64.8 | 59.6 | 2797 | 554 |
5 | 39.6 | 84.2 | 133.3 | 101.4 | 52.4 | 272.2 | 269.7 | 56.3 | 51.5 | 2797 | 585 |
6 | 30.4 | 85.1 | 143.6 | 107.6 | 48.6 | 241.1 | 239.8 | 50.5 | 46.6 | 2803 | 564 |
RLI—Recuperator Liquid In; BI—Boiler In; EI—Expander In, EO—Expander Out; and PCI—Power Condenser In.
Due to the insufficient heat input from the oil heater during most of the run times, the vapor coming out of the boiler was only slightly superheated. This led to low sensible heat recovery rate from the recuperator, as it is shown low effectiveness in Table 4. As a result, the vapor began to condense within the recuperator causing pinch point. Once more heat can be added for future experiment, significant higher recuperator effectiveness should be expected.
Recuperator effectiveness at various run time
T_RLI [C] | T_BI [C] | T_EO [C] | T_PCI [C] | ɛ_recup |
---|---|---|---|---|
35.9 | 85.5 | 101.1 | 58.3 | 76% |
36.0 | 86.2 | 101.4 | 59.5 | 77% |
43.1 | 76.6 | 86.6 | 58.2 | 77% |
43.7 | 75.4 | 84.0 | 57.2 | 79% |
39.6 | 84.2 | 101.4 | 52.4 | 72% |
30.4 | 85.1 | 107.6 | 48.6 | 71% |
T_RLI [C] | T_BI [C] | T_EO [C] | T_PCI [C] | ɛ_recup |
---|---|---|---|---|
35.9 | 85.5 | 101.1 | 58.3 | 76% |
36.0 | 86.2 | 101.4 | 59.5 | 77% |
43.1 | 76.6 | 86.6 | 58.2 | 77% |
43.7 | 75.4 | 84.0 | 57.2 | 79% |
39.6 | 84.2 | 101.4 | 52.4 | 72% |
30.4 | 85.1 | 107.6 | 48.6 | 71% |
In addition, the heating function through the waste heater was also demonstrated. Both inlet and outlet air temperatures were measured along with air flowrate. All measured values and thermo-physical properties of air for a representative case are listed in Table 5. The calculated heating capacity of the waste heater was 6.4 kW, significantly above standard ECU heating capacities.
Heating functional test measurement
Volume flowrate | 0.261 (554) | m3/s (cfm) |
---|---|---|
Density, ρ_air | 1.156 | kg/m3 |
Specific heat capacity, Cp_air | 1.005 | kJ/(kg-K) |
Volume flowrate, V_air | 0.261 (554) | m3/s (cfm) |
Inlet temperature, T_AI | 20 | °C |
Outlet temperature, T_AO | 42.99 | °C |
Heating capacity, Q_h | 6.4 | kW |
Volume flowrate | 0.261 (554) | m3/s (cfm) |
---|---|---|
Density, ρ_air | 1.156 | kg/m3 |
Specific heat capacity, Cp_air | 1.005 | kJ/(kg-K) |
Volume flowrate, V_air | 0.261 (554) | m3/s (cfm) |
Inlet temperature, T_AI | 20 | °C |
Outlet temperature, T_AO | 42.99 | °C |
Heating capacity, Q_h | 6.4 | kW |
4.2 Alternative Organic Rankine Cycle Power Measurement and Working Fluid.
After replacing the broken turbine flowmeter with a Coriolis mass flowmeter in the ORC, additional experiment was conducted to further evaluate the scroll expander power output using enthalpy change across it. This is due to some concerns were raised during the initial functional testing that the power readings from the controller of the motor/generator may have grossly underestimated the actual power output from the expander due to excessive mechanical losses in the drivetrain, especially at higher expander speeds. The enthalpies before and after the expander were calculated from the pressure and temperature measurements. The expander work was also calculated by measuring the voltage and current across the load bank connected to the expander, which helped assess the mechanical losses within the powertrain. Again, data were acquired and recorded using ni labview and daq.
For quantification of uncertainty, each sensor’s uncertainty included both the bias/instrumentation errors and precision errors. The bias/instrumentation errors were determined from our calibration process which included the manufacturer’s provided errors along with any errors in our data acquisition system. The precision errors were determined from the sampled experiment data, which estimated the population standard deviation. The bias and precision errors were then aggregated together using the square root of the sum-of-squares. To derive the uncertainty of derived variables such as expander power, uncertainty propagation of the Kline-McClintock method was used.
In addition, because of the high GWP of 245fa, alternative working fluids have been proposed with similar fluid properties but much lower GWP. One specific fluid is HFO trans-1-chloro-3,3,3-trifluoropropene (1233zd(E)). 1233zd(E) has similar fluid properties to 245fa, but with a much lower GWP. The data collection consisted of five different heat input temperatures (150 °C–190 °C, 10 °C intervals) for R245fa and one temperature for R1233zd(E). Each test allowed the system to reach steady-state before data was recorded for 5 min at a frequency of 0.2 Hz. During the recording of data, the pump and fan were kept at the same constant control input.
Figure 16 shows the expander power calculated using fluid properties and load bank voltage and current (from the controller). It shows a large difference between the values for each input temperature—much higher expander power is associated with the approach assessing enthalpy change. The reason for this is twofold: first, the expansion process in the scroll expander is a polytropic process, and the pressure and temperature measurements for the expander are not inside of the expander. This means there could be significant heat loss through the expander to the surroundings which manifests as a larger enthalpy change (higher power output) than in an adiabatic system. The second cause for the large discrepancy between values confirmed the suspected significant mechanical losses on the within the powertrain between the expander and load bank. Nevertheless, the expander power calculated from enthalpy change for both fluids at the comparable input temperature is almost identical, and it falls within the uncertainty of each point. However, the expander work calculated from the controller is significantly different and does not fall within the calculated uncertainty. Reasons for this is not completed clear, which suggests that more data is required before a reasonable determination can be made on the performance differences of these two fluids.
5 Discussions and Potential Improvement
A flexible, multi-function distributed energy system is demonstrated. Compared to other heat activated cooling systems, the current HCHP system has demonstrated an attractive function of converting waste heat to power when cooling is not needed or supplementing cooling when waste heat is insufficient. Although the work here serves the purpose of proving the concept, a number of issues and improvements were identified. For example, the current drivetrain design was not particularly efficient, which led to significant power losses at high speeds. The current motor/generator itself also had significant power lost at high speeds, which was not ideally suitable for efficiency-sensitive applications. In addition, measurement used for calculating enthalpy change across the scroll expander should be improved, which will also help mitigate the discrepancy of the expander output power.
Nevertheless, the technology was shown to be viable and has the potential to be relevant as a versatile distributed energy system in achieving higher energy efficiency. Because its energy output can be tailored for specific needs (cooling, heating, and power), a distributed energy system built with this functionality can be used more often so that more otherwise wasted energy can be recovered to significantly improve the “seasonal energy efficiency or recovery rate” and its associated levelized cost of energy.
In addition, a preliminary study using zeotropic fluid mixtures was conducted in order to gain further thermodynamic advantages for waste heat recovery applications. With better matching of the temperature profiles between the diesel exhaust stream and the working fluid in the ORC boiler, both heat recovery rate and cycle conversion efficiency can be further improved. This improvement comes from the resultant higher averaged input temperatures. A number of studies have conducted recently to demonstrate the thermodynamic advantages of using zeotropic blends. Wang et al. [30] used a solar collector to provide heat input to an ORC. They demonstrated higher thermal efficiency with zeotropic mixtures. Heberle et al. [31] indicated that, with better matching of geothermal heat sources temperature profile, the exergy destroyed in the ORC was reduced leading to higher thermal efficiencies. Compared to pure working fluid, Chys et al. [32] also achieved significant improvement on cycle efficiency with zeotropic mixtures.
The baseline ORC cycle efficiency using R245fa was established to be 10.9% in this study. Using REFPROP, the compositions of selected fluid mixtures can be changed in order to find desirable zeotropic mixture candidates. R152a was chosen as the second component to R245fa for the initial investigation. Li et al. [33] also presented many graphs representing net-work output for various compositions of R245fa and R152a as function of operating temperature. Similarly, Wang and Zhao [34] investigated three different mixtures of R245fa and R152a for low temperature solar thermal application. In the study, various fluid compositions were studied for potential performance gains. Each composition contained a different mass fraction of R152a, and it was varied from 0.1 to 1.0 in an interval of 0.1. The preliminary model result of the cycle efficiency for each composition is shown in Fig. 17. According to the plot, 10% improvement on ORC efficiency could be achieved with 70% mass fraction of R152a from the baseline using pure R245fa.

Results of the efficiency analyses performed for various compositions of R152a. The dashed line represents the baseline efficiency for pure R245fa.
6 Conclusions
The unique hybrid cooling, heating, and power system was developed in place of current military ECUs. Under the design condition, the energy model predicts 27% of diesel fuel saving. To demonstrate the concept, a prototype unit was developed and functional testing was performed at both system and component levels. All designed functions (operating modes) have been demonstrated successfully using an oil heater as the heat source. Its flexible energy outputs show clear advantages in converting engine waste heat to cooling, heating and/or power. Meanwhile, the functional testing has also exposed some issues and potential improvement areas for the current system. While being flexible, the current drivetrain using pulleys and electromagnetic clutches introduced significant mechanical losses which require serious design improvement. The oil heater was not able to deliver the targeted heat input to the ORC, which led to relatively poor effectiveness for the microchannel recuperator. The initial testing of R1233zd(E)—a low GWP fluid, did not show significant difference from R245fa. However, more data points will be needed to provide more comprehensive assessment. Given the potential of achieving higher thermal efficiency for ORCs and matching desired temperature profiles of a given waste heat stream, the initial assessment of using R245fa and R152a mixture indicates its potential improvement.
Besides serving the military applications as an environmental control unit, the HCHP system can also be valuable in recovering waste heat from industrial processes, as well as in converting low-grade solar thermal and geothermal energies to cooling and/or power, thus meeting the needs of the growing distributed energy applications.
Acknowledgment
The authors would like to extend the sincere gratitude to US Army CERDEC and Wyle for providing the funding and management on this project.
Conflict of Interest
There are no conflicts of interest.
Nomenclature
Subscripts
- AI =
inlet air of the power (ORC) condenser
- AO =
outlet air of the power (ORC) condenser
- BI =
boiler inlet
- EI =
expander inlet
- EO =
expander outlet
- PCI =
power condenser inlet
- RLI =
recuperator liquid inlet
- Q_h =
waste heater heating capacity =ρair × Vair × Cpair × (TAO − TAI)
- ɛ_recup =
effectiveness of the recuperator