Searching for Suitable Binary Fluid for an Ejector Heat Pump for Domestic Water Heating

In the United States, domestic water heaters are mainly electric-resistance or fossil-fired types, representing a significant portion of residential energy consumption and greenhouse gas emissions [1]. Heat pump water heaters (HPWHs) are an emerging technology that uses vapor compression refrigerant cycles for higher energy efficiency in terms of the coefficient of performance (COP). Properly selecting working fluids (i.e., refrigerants) for HPWHs is critical to achieving a high COP [2]. The suitable working fluids should have desired thermophysical and chemical properties, such as high latent heat of vaporization, high thermal conductivity, nontoxicity, nonflammability, and nonexplosiveness [3]. Moreover, the desired refrigerants should be eco-friendly, with a low global warming potential (GWP) and zero ozone depletion potential (ODP). The working fluids for vapor compression cycles can be classified as chlorofluorocarbons (CFCs), hydrocarbons (HCs), hydrochlorofluorocarbons (HCFCs), hydrofluorocarbons (HFCs), hydrofluoroolefins (HFOs), hydrofluoroethers (HFEs), fluorocarbons, and natural refrigerants. CFCs and HCFCs have a high ODP and were phased out under the Montreal Protocol. HFCs with a high GWP (>150) are being phased out. HCs are A3 flammable refrigerants, and only a few HCs, such as isobutane (R600a), are approved by the U.S. Environmental Protection Agency [3]. HFOs were developed as alternative low-GWP refrigerants to replace HFCs and HCFCs [4]. HFEs are nonflammable fluids with low toxicity, chemically inert, and have high-temperature stability [5]. Recently, HFOs and HFEs have been actively investigated as alternative refrigerants for HPWHs [6].

The U.S. market for HPWHs is dominated by electric-driven heat pumps, which operate with vapor compression cycles using mechanical compressors [7]. A thermally activated HPWH is an alternative heat pump technology that uses widely available thermal energy from waste heat or solar thermal energy [8]. Thermally activated HPWHs have good potential to boost primary energy utilization efficiency by avoiding the losses in electric generation and transmissions, and their overall energy efficiency could be further improved by integrating with space cooling [9]. There are currently no thermally activated HPWH commercially available in the United States. However, two gas-fired HPWHs based on the absorption cycle are under development. The Gas Technology Institute and Stone Mountain Technologies, Inc. demonstrated a fuel-fired, air-source HPWH, which operates with a single-effect vapor absorption cycle using NH₃/water as the working fluid [10,11]. At a return water temperature of 38 °C–49 °C and ambient temperature of 2 °C–44 °C, its gas-basis COP is in the range of 1.10–1.45 for water heating only and 1.30–1.90 for integrated water heating and air conditioning. This technology could reduce 40% of natural gas consumption and 20% of the building air conditioning load in full-service restaurants. The University of Florida and Oak Ridge National Laboratory are developing a semi-open, ionic liquid absorption HPWH for residential water heating, dehumidification, and space evaporative cooling [12–14]. This technology is enabled by a membrane-based absorption/desorption process in which hydrophobic, vapor-permeable membranes constrain the absorbent. It uses ambient water vapor as the working fluid, which is absorbed by the ionic liquid in the absorber. The process water is sequentially heated in the absorber by the latent heat of the ambient water vapor, in the solution heat exchanger by the sensible heat of the concentrated ionic liquid, and in the condenser by the latent heat of desorbed water vapor. This technology achieved a heating-cycle COP of 1.2 with a domestic hot water (DHW) delivery temperature of 56 °C when operated with ambient air at 19 °C and 49% relative humidity.

The ejector heat pump water heater (EHPWH) is a novel, thermally driven cycle in which the ejector functions as a thermocompressor that replaces the electric-powered compressor in conventional vapor compression cycles [15,16]. Compared with electric-powered compressors, the ejector has many practical advantages, such as simple construction, no moving parts, high reliability, and low cost, making it attractive for residential heating/cooling systems [17]. One of the technical barriers in the ejector-based vapor compression cycle is its low energy efficiency, which can be addressed by using binary fluids as the working fluids. The binary fluid ejector (BFE) is a promising technology that was proposed by Zhadan et al. in the 1980s [18], and researchers have focused on its application in refrigeration systems. Recently, May-Ruben Technologies, Inc. (Canada), demonstrated the technical merits of the BFE using two chemically distinct fluids (i.e., Vertrel Sinera as the primary working fluid (PF) and Vertrel XF as the secondary working fluid (SF)). This work empirically showed an improvement of 300% in the cooling-cycle COP over a single-fluid steam ejector refrigeration system (ERS) [19]. Buyadgie et al. [20] theoretically showed that using multicomponent fluids in ERS (C₅F₁₂/R236ea, R11/R600a, and R11/R152a) could improve the cooling-cycle COP by 30–50%. Chen et al. [21] theoretically compared the binary fluid ERS using zeotropic mixtures (R32/R134a, R32/R152a, R134a/R142B, R152a/R142b, R290/R600a, and R600a/R600), which had higher cooling-cycle COPs than the single-fluid ERSs. Tan et al. [22] theoretically investigated the zeotropic refrigerant mixture (R32/R236fa) in auto-cascade ERS, which could deliver low evaporation temperatures (−30 °C), but the cooling-cycle COP was only 0.05. However, Lallemand et al. [23,24] showed that the use of binary fluids did not always increase the performance of ERSs; only a mildly zeotropic mixture (R134a/R142b) or almost-azeotropic mixture (R134a/R152a) improved the cooling-cycle COP of ERSs.

The authors’ previous research showed that the theoretical heating-cycle COP could be up to 2.0 with a low entrainment ratio of 0.1 using the binary fluid pair (BFP) of HFE7500 and water [25]. This finding confirmed that the proper selection of BFPs would improve the performance of BFEs. However, searching for the best-suitable binary fluids for EHPWHs is still challenging because of the high temperature lift (>40 K) required in producing DHW. Computational fluid dynamics (CFD) simulation on BFEs with artificial thermophysical properties demonstrated that the molecular weight (MW) and specific heat ratio dominate the performance of ejectors [26]. Buyadgie et al. [20] qualitatively carried out a theoretical analysis of the energy loss owing to the frictional force and the shock wave and suggested that the BFP should consist of PF with a high MW and SF with a low MW. Their work gave some guidelines for selecting working fluid pairs for BFEs. However, the general criteria for screening binary working fluids have not been well-established. In addition to the lack of research on binary fluid in ERS, screening binary fluids for ejector heat pumps is more challenging because of the requirement of a high normal boiling point (NBP) in working fluids.

This study screens different working fluids for EHPWHs that could be operated under positive pressure with a high heating-cycle COP. Criteria for working fluid pairs in BFEs are established, and working fluid candidates for EHPWHs are shortlisted. A thermodynamic model for an EHPWH is deployed to evaluate its heating-cycle COP. The performance of an EHPWH with the shortlisted working fluids is evaluated under various operating pressures of the PF in the high-temperature evaporator and the condensation temperature.

A typical ejector heat pump (EHP) consists of an ejector, a condenser, a low-temperature evaporator (LTE), and a high-temperature evaporator (HTE). Thermal energy is added to evaporate the PF in the HTE to drive the thermodynamic cycle. The SF evaporates in the LTE, and the mixture of PF and SF is exhausted into the condenser. The heat of the LTE or condenser can be harvested to produce cooling or heating energy simultaneously. These energies depend mainly on the working fluids. Criteria for shortlisting the binary working fluids are discussed in the following subsection.

The generalized criteria for screening working fluid pairs are established based on their ability to achieve high-performance operation above atmospheric pressure (i.e., no vacuum), easy separation of SF or PF from the mixtures, and other generally common considerations for refrigerants.

For the mixing process in a BFE with a PF at Ma = 4 and SF at Ma = 1, the velocity difference is only 40 m/s. Additionally, a PF with a high MW and SF with a low MW could reduce the kinetic energy loss, frictional loss, and shock wave loss, resulting in a high entrainment ratio.

The operation of an EHPWH independent of a vacuum requires the NBP of the SF to be lower than the operating temperature of the LTE (i.e., NBP_SF < T_LTE). Sealing is a practical challenge in ERSs using water as the working fluid, and auxiliary vacuum pumps are commonly used to prepare the vacuum before operation [15,25]. An SF with an NBP lower than the operating temperature of the LTE will allow the whole system to be operated at a pressure higher than the atmosphere pressure. The leakage protection in a high-pressure system is not as difficult as sealing a vacuum system.

The binary mixture must be separated into single-fluid components with high purity. For the non-miscible PF–SF pairs, a large density difference is desired for a gravity-driven separator—for example, the PF–SF pairs of HFE7500 (ρ_HFE7500 = 1.614 g/cm³) and water (ρ_HFE7500 = 0.997 g/cm³) [25]. For a miscible PF–SF pair, a larger difference in the NBP is required for the fractionating condensation process [19]. The NBP of the PF was suggested to be 10–20 °C higher than that of the SF [20].

Chemical compatibility requires no chemical reaction between the PF and SF and between the PF–SF mixture and wetted components. For example, HFE7500 dissolves or swells elastomers [29], causing the gaskets to fail. The ODP and GWP of working fluids are also considered during the screening.

Following the generalized criteria in the previous section, 24 working fluids for PFs and 34 working fluids for the SFs are narrowed down from 151 fluids available in the database of the Engineering Equation Solver (EES). Five PF candidates are shortlisted from working fluids, with the lowest latent heat of evaporation and the NBP in the range of −10.0 °C < NBP_PF < 140.0 °C. Four SF candidates are shortlisted from working fluids, with the highest latent heat of evaporation and the NBP in the range of −30.0 °C < NBP_SF < 20.0 °C. Five PF candidates and four SF candidates are selected following these established criteria, as listed in Tables 1 and 2. These shortlisted PFs and SFs provide 20 BFPs for EHPWHs. The thermophysical properties of working fluids are obtained from the database of the EES software (f-chart software [30]).

The presence of hydrogen on the HFE molecules makes HFEs miscible with HCs. The PF–SF mixture needs to be separated in fractional condensers. A fractional condenser is a column with a series of packing and heat transfer surfaces, which selectively condenses the PF to separate it from the vapor of the SF [31]. This study assumes that PF and SF are completely separated in the fractional condensers.

The shortlisted PF and SF candidates are evaluated in the EHPWH shown in Fig. 1 [32]. It consists of an HTE, a single-stage ejector, a fractional condenser for the PF (PFC), a fractional condenser for the SF (SFC), an LTE, an expansion valve, a circulation pump, and a flue gas heater (FGH).

The thermodynamic cycles and working fluids loop within the EHPWH are as follows:

• 1 → 2: isobaric heat addition process in the HTE;

• 2 → 3 ← 8: PF and SF vapor mixing within the ejector;

• 3 → 4: isobaric heat rejection process in the PFC;

• 4 → 1: isentropic compression process, where the PF liquid is pumped back to the HTE;

• 3 → 5: isobaric separation process, where the SF vapor is separated from the PF–SF mixture in the PFC;

• 5 → 6: isobaric heat rejection process, where the SF condenses in the SFC;

• 6 → 7: isenthalpic throttling process of the SF before it returns to the LTE; and

• 7 → 8: isobaric heat addition process, where the SF evaporates in the LTE.

Tap water is sequentially heated in the SFC and PFC by harvesting the latent heat from the PF and SF, respectively. Then, the water is further heated to the supply temperature of the DHW (60 °C) using the flue gases exhausted from the HTE.

A thermodynamic model of an EHPWH is deployed to evaluate the component-level performance of BFEs and the system-level performance of EHPWHs. The model assumes that the EHPWH works under a steady-state condition, and the frictional pressure drop in the loop is negligible. The model was built by applying mass and energy conservation within each component and reported in the authors’ previous publications [25,32,33].

The mass flowrate of the SF, $m˙SF$⁠, is determined for a unit thermal energy input, and the mass flowrate of the PF, $m˙PF$⁠, is calculated from the entrainment ratio, ω_BFE, using Eq. (1). The subscripts “in” and “out” are for working fluids entering and leaving the heat exchangers.

The entrainment ratio of BFE, ω_BFE, under a specified operating condition, is predicted by a comprehensive, geometry-free, theoretical model of BFEs, which was developed from Huang’s [34] and Pounds’s models [35,36]. The theoretical model was developed with the following assumptions: (1) the SF is choked at a section “y-y” in the constant area of the mixing chamber; (2) the mixing of the PF and SF is an isobaric process where the mixing pressure, p_Mixing, equals the critical pressure of the choked SF, p_SF,cr, at the section “y-y”; and (3) a normal shock wave (NSW) occurs after the completed mixing of the PF and SF. Detailed information on the theoretical model and validation was reported in the authors’ previous publications [32,33].

In the following sections, the performance of the EHPWH is investigated at the component and system levels using the shortlisted working fluids under different operating conditions.

In the investigated EHPWH, the LTE is operated at an ambient temperature T_amb = 19.4 °C, as specified for rating air-source HPWHs by the U.S. Department of Energy [37]. The evaporation is assumed to approach the temperature in the LTE, which is 5.0 °C (i.e., ΔT_LTE = 5.0 °C). Therefore, the evaporation temperature of the SF in the LTE is T_LTE = 14.4 °C. The pressure limit of the HTE is 2.3 MPa, and the condensation temperature in the PFC, T_PFC, is in the range of 30.0 °C–50.0 °C. The component-level performance of the BFE with the shortlisted BFP is evaluated by the entrainment ratio, ω_BFE. The ratio is theoretically predicted using the developed theoretical model with ejector’s component isentropic efficiencies of η_N = 0.90, η_M = 0.95, η_D = 0.81, and η_S = 0.85 [33]. The effects of the BFPs with various PF and SF candidates; the operating pressure of the PF in the HTE, p_PF,HTE; and the condensation temperature of the PFC, T_Cond, are summarized in Figs. 3–6. Detailed information on BFEs and the EHPWH operating with shortlisted BFPs at various temperatures T_HTE and T_Cond available in the Supplemental Materials on the ASME Digital Collection.

Figure 2 shows the entrainment ratio of the BFE operating with HFE7100 as the PF. The evaporation pressure range of HFE7100 is 0.14–2.02 MPa, and the corresponding operating temperature range of the HTE is 70–190 °C. The BFEs can functionalize properly only when the pressure of PF, p_PF,HTE, is larger than the minimal functioning pressure, p_PF,min. If p_PF,HTE < p_PF,min, the SF cannot be entrained into the suction chamber, represented as “x—No functional point” in Fig. 2. For a specified PF, p_PF,min is closely related to the NBPs of the SF. Generally, an SF with a lower NBP requires a higher p_PF,min. An SF with a lower NBP has a higher stagnation pressure at the fixed operating temperature of the LTE (T_LTE = 14.4 °C), giving a higher critical pressure when the SF is chocked at the section “y-y.” The static pressure of the PF is the same as the choked pressure of the SF at the section “y-y.” For a BFE operating with HFE7100 as the PF, the p_PF,min is 0.57 MPa for RE170 (NBP_RE170 = −24.92 °C), 0.35 MPa for R600a (NBP_R600a = −11.68 °C), 0.20 MPa for R600 (NBP_R600 = −0.53 °C), and 0.15 MPa for R1234ze(Z) (NBP_R1234ze(Z) = 9.72 °C).

Figure 3 shows the entrainment ratios of the BFE operating with R600 as the SF. No relationship appears between the minimum functioning pressure of the PFs, p_PF,min, and the NBPs of the PF, NBP_PF. For the specified SF, its critical pressure at the section “y-y” in the ejector, p_SF,ii, is fixed and independent of selected PFs. The static pressure of the PF, p_SF,i, equals p_SF,ii, but its stagnation pressure is closely related to its gas-dynamic and thermophysical properties, such as the Mach number, Ma, and the specific heat ratio, k. For BFEs operating with R600 as the SF at T_Cond = 30.0 °C, the p_PF,min is 0.24 MPa for HFE7000, 0.14 MPa for HFE7100, 0.16 MPa for HFE7200, 0.15 MPa for Novec649, and 0.14 MPa for HFE7500.

Additionally, Figs. 2 and 3 show that the condensation temperature, T_Cond, has no apparent effects on p_PF,min. As discussed previously, p_PF,min is only related to the critical pressure of the SF and the static pressure of the PF at section “y-y,” which are independent of the downstream flow conditions within the ejector.

The entrainment ratios of BFE, ω_BFE, increase continuously with the increase of the HTE evaporation pressure, p_PF,HTE, at a specified condensation temperature in the PFC, T_Cond. However, this trend is affected by the proximity of p_PF,HTE to the critical point of PF, P_PF,cr, as shown in Figs. 2 and 3. When p_PF,HTE approaches p_PF,cr, further increasing p_PF,HTE leads to a decreased ω_BFE because of the dramatically deteriorated thermophysical properties of the PF [26], such as the extremely high value of the specific heat ratio, k. For the BFEs operating with HFE7100/R1234ze(Z) at T_Cond = 30.0 °C, ω_BFE increases from 0.374 at p_PF,HTE = 0.14 MPa to the maximum entrainment ratio of 1.703 at p_PF,HTE = 1.69 MPa, then reduces to 1.588 at p_PF,HTE = 2.20 MPa. Here, the critical pressure of HFE7100 is 2.23 MPa. These effects of p_PF,HTE on ω_BFE in BFEs are consistent with previous experimental and theoretical results for single fluid ejectors [15,32,35].

The optimum p_PF,HTE for the maximum entrainment ratio of a specified PF, ω_BFE,max, is not sensitive to the SF. For HFE7100 as the PF, the optimum p_PF,HTE is 1.69 MPa for RE170, R600a, and R600, but it is 2.02 MPa for R1234ze(Z). Additionally, the optimum p_PF,HTE is independent of the condensation temperature, T_Cond.

A lower condensation temperature, T_Cond, gives a higher value of the maximum entrainment ratio, ω_BFE,max, for a specified BFP, as shown in Figs. 2 and 3. A lower T_Cond corresponds to a lower saturated vapor pressure of the PF and SF in the condensers. If ω_BFE is fixed, a lower T_Cond results in lower back pressure for a BFE, p_MF,3. Therefore, a lower T_Cond could enable a higher ω_BFE of BFE. Table 3 lists the dynamic gas properties of the PF, SF, and mixed fluid (MF) within BFEs. It shows that a lower T_Cond requires a weaker compression effect, which can be generated by the MF flow with a smaller Mach number, Ma_MF,iii, before an NSW. With a given Mach number and static pressure of the PF and SF before mixing, a lower Ma_MF,iii could be achieved with a higher ω_BFE (i.e., more mass of SF) owing to the momentum of conservation within the mixing process. For BFEs operating with HFE7100/R1234ze(Z) at p_PF,HTE = 1.69 MPa, ω_BFE,max decreases from 1.703 to 0.609 as T_Cond increases from 30.0 °C to 50.0 °C. These effects of T_Cond on ω_BFE for the BFE are consistent with previous experimental and theoretical results for single fluid ejectors [15,31].

SF with a higher NBP gives a higher value of ω_BFE,max for a specified PF. At a fixed evaporation temperature of SF in the LTE (i.e., T_LTE = 14.4 °C), SF with a lower NBP has a higher evaporation pressure, p_SF,LTE, as well as a higher mixing pressure, p_Mixing, leading to a smaller Mach number for the PF after the mixing process, Ma_MF,iii. It generates a weaker compression effect in NSW, giving a larger Mach number for the MF after the NSW, Ma_MF,iii, or Ma_MF,iv, but a smaller pressure ratio across the NSW, p_MF,iv or p_MF,iii. Although p_MF,iv or p_MF,iii is smaller, the backpressure of the BFE, p_SF,3, is higher for an SF with lower NBP owing to a higher mixing pressure. As a result, a higher entrainment ratio could be achieved with an SF with a higher NBP. Table 4 lists the gas-dynamic properties of the PF, SF, and MF within BFEs, which operate with HFE7100 as the PF at T_Cond = 30.0 °C. The ω_BFE,max significantly increases from 0.636 achieved with RE170 (NBP_RE170 = −24.92 °C) to 1.703 achieved with R1234ze(Z) (NBP_R1234ze(Z) = 9.72 °C).

The effects of SFs on ω_BFE,max can be further interpreted in terms of MWs because refrigerants with higher MWs generally have higher NBPs. Figure 4 shows that an SF with higher MW gives a higher ω_BFE,max in BFEs with HFE7000 and HFE7100 as the PF. In particular, R600 and R600a give almost the same ω_BFE,max because they have the same MWs, despite a large difference in NBPs (∼11.2 °C). The same results are achieved for the other PFs and are available in the Supplemental Materials on the ASME Digital Collection. These results agree with the numerical results of the CFD analysis on BFEs for ideal fluids, which showed that the entrainment ratio increased significantly with the increased MWs of the SF [26].

Figure 5 shows the maximum entrainment ratio, ω_BFE,max, achieved with various PFs. With RE170 as the SF, PFs with higher MWs give higher values for ω_BFE,max, but for the other three SFs, HFE7100 gives a higher ω_BFE,max. The shortlisted PFs can be briefly classified into the low-MW group, consisting of HFE7100, HFE7200, and HFE7300, and the high-MW group, consisting of Novec649 and HFE7500. The low-MW group gives a higher ω_BFE,max than the high-MW group, which agrees with the numerical results of the CFD analysis on BFEs using ideal fluids [26]. However, the effects of MWs on ω_BFE,max are not obvious because the properties of real fluids depend on various factors other than MWs.

In summary, HFE7100/R1234ze(Z) gives a higher entrainment ratio than the other investigated BFPs. The largest value of ω_BFE,max is 1.703 achieved with HFE7100/R1234ze(Z) operated at p_PF,HTE = 1.69 MPa and T_Cond = 30.0 °C.

In the investigated EHPWH, tap water is primarily heated in the SFC and PFC by the discharged PF/SF and is further heated in the FGH by the high-temperature flue gas. The performance of the EHPWH is evaluated by the heating-cycle COP, COP_EHPWH, as given in Eq. (4). Figures 6 and 7 show the performance of the EHPWH with various BFPs at T_Cond = 30.0 °C and 50.0 °C, respectively. For a specified BFP, COP_EHPWH has similar trends of change as ω_BFE in terms of T_HTE and T_Cond because a larger ω_BFE requires a lower Q_HTE for a specified Q_LTE.

The relative value of COP_EHPWH and ω_BFE for different BFPs changes because of the significant difference in the thermophysical properties of the PFs or SFs. Novec649/R600 gives a significantly higher COP_EHPWH than HFE7500/R600, although these BFPs have similar values of ω_BFE, as shown in Fig. 3. There is a large difference in increased enthalpy of the liquid PF within the HTE (i.e., Δh_l,PF) in Eq. (3). At p_PF,HTE = 1.08 MPa and T_Cond = 30.0 °C, Δh_l,Novec649 is 124.6 kJ/kg for a temperature lift of 105 °C, and Δh_l,HFE7500 is 264.7 kJ/kg for a temperature lift of 235 °C. Similarly, although the ω_BFE for R600 and R600a are almost identical, R600 gives a higher COP_EHPWH than R600a because R600 has a larger latent heat of evaporation. At T_LTE = 14.5 °C, h_lv is 371.0 kJ/kg and 340.5 kJ/kg for R600 and R600a, respectively. As a result, although HFE7100/R1234ze(Z) gives the highest entrainment ratio of BFEs, HFE7000/R600 gives the best performance for an EHPWH.

Figure 8 shows the COP_EHPWH achieved with HFE7000/R600 at various p_PF,HTE and T_Cond. At a given T_Cond, the trends of COP_EHPWH related to p_PF,HTE are similar to those of ω_BFE. However, for higher T_Cond, COP_EHPWH becomes more sensitive to p_PF,HTE. The COP_EHPWH changes in a range of 1.091–1.166 for T_Cond = 30.0 °C and 1.047–1.328 for T_Cond = 50.0 °C. These significant changes in COP_EHPWH are related to the thermal energy distributions within EHPWHs at different T_Cond. With a higher T_Cond, BFEs contribute more thermal energy in producing DHW, resulting in less thermal energy consumption in the FGH. Figure 9 shows that the contribution of the BFEs, operating with HFE7000/R600, increases from 23.1% to 67.0% as T_Cond increases from 30.0 °C to 50.0 °C. A higher T_Cond gives a higher heating-cycle COP of the EHPWH because the energy efficiency of directly heating using flue gas in the FGH, for further boosting DHW to 60 °C, is less than that of the BFE.

In summary, HFE7000 and R600 give the highest COP_EHPWH among the shortlisted PFs and SFs. An EHPWH with HFE7000/R600 operated at p_PF,HTE = 1.30 MPa (T_HTE = 130.0 °C) gives the highest heating-cycle COP of 1.328 at T_Cond = 50.0 °C.

This study searches for binary fluids used in EHPWHs to achieve high COP. The shortlisted PFs are HFE7000, HFE7100, HFE7200, Novec649, and HFE7500, and the candidate SFs are RE170, R600a, R600, and R1234ze(Z). A thermodynamic model of an EHPWH is utilized to evaluate the component-level and system-level performance of EHPWHs. The performance of an EHPWH for producing DHW at 60.0 °C is evaluated under p_PF,HTE ≤ 2.3 MPa, 30.0 °C ≤ T_Cond ≤ 50.0 °C, and T_amb = 19.4 °C. Among all the BFPs and operating conditions, an EHPWH with HFE7000/R600 operates at p_PF,HTE = 1.69 MPa gives the highest heating-cycle COP_EHPWH of 1.328 at T_Cond = 50.0 °C. However, an EHPWH with HFE7100/R1234ze(Z) operating at p_PF,HTE = 1.69 MPa gives the highest ω_BFE of 1.703 at T_Cond = 30.0 °C.

The conclusions for the investigated EHPWH are as follows:

1. There is a minimal evaporation pressure of the PF in the HTE, p_PF,min, to properly functionize the BFEs. A higher p_PF,min is required for PFs with higher NBPs and/or SFs with lower NBPs.

2. For a specified BFP, ω_BFE is related to p_PF,HTE and T_Cond. Higher p_PF,HTE and/or lower T_Cond gives higher ω_BFE. However, when p_PF,HTE approaches the critical point of the PF, further increasing p_PF,HTE reduces ω_BFE.

3. The effects of the PF or SF on the ω_BFE,max are related to their MWs. A PF with a low MW or an SF with a high MW gives a high ω_BFE,max.

4. The COP_EHPWH is related to ω_BFE, the sensible heat of the PFs, and/or the latent heat of evaporation of the SFs. A PF with a smaller increased enthalpy within the HTE and/or an SF with a lower latent heat of evaporation gives a higher COP_EHPWH.

5. The COP_EHPWH depends largely on T_Cond. A higher COP_EHPWH is achieved with a higher T_Cond despite the lower ω_BFE.

Pengtao Wang: methodology, investigation, writing—original draft; Kashif Nawaz: supervision, investigation, resources, writing—review & editing; Ahmad Abu-Heiba: conceptualization, investigation, supervision, writing—review & editing; Ramy H. Mohammed: investigation, supervision, writing—review & editing; Jeremy Spitzenberger: investigation, writing—review & editing; Laith Ismael: investigation, writing—review & editing; Stephen Kowalski: investigation, resources, writing—review & editing; Hongbin Ma: conceptualization, methodology, supervision, investigation.

This material is based upon work supported by the U.S. Department of Energy Office of Science, Building Technologies Office. This research used resources from the Building Technologies Research and Integration Center, which is a U.S. Department of Energy Office of Science User Facility at Oak Ridge National Laboratory. Thanks also go to Mrs. Hames Wendy for her great technical editing.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The authors attest that all data for this study are included in the paper.

$h$ =

specific enthalpy, kJ/kg

k =

the ratio of specific heat

E =

kinetic energy of the working fluid

P =

pressure, kPa

Q =

thermal power, W

R =

gas constant, 8.314 J/(mol · K)

T =

temperature, °C

V =

flow velocity, m/s

W =

electric power, W

$m˙$ =

mass flowrate, kg/s

c_p =

specific heat capacity under constant pressure, kJ/(kg · K)

$hlv$ =

specific latent heat of vaporization, kJ/kg

Ma =

Mach number

α =

sound speed, m/s

η =

isentropic efficiency of ejector component

ρ =

density, kg/m³

ω =

entrainment ratio of an ejector

b =

backpressure

D =

diffuser section of an ejector

M =

mixing section of an ejector

N =

primary nozzle of an ejector

S =

suction chamber of an ejector

amb =

ambient

cr =

critical

max =

maximum

min =

minimum

out =

outlet

1,2,3, … =

state points in the thermal dynamics cycles of the EHPWH

i, ii, iii, … =

state points within the BFE

BFE =

binary fluid ejector

BFP =

binary fluid pair

CFC =

chlorofluorocarbon

CFD =

computational fluid dynamics

COP =

coefficient of performance

DHW =

domestic hot water

EES =

Engineering Equation Solver

EHP =

ejector heat pump

EHPWH =

ejector heat pump water heater

ERS =

ejector refrigeration system

FGH =

flue gas heater

GWP =

global warming potential

HC =

hydrocarbon

HCFC =

hydrochlorofluorocarbon

HFC =

hydrofluorocarbon

HFE =

hydrofluoroether

HFO =

hydrofluoroolefin

HPWH =

heat pump water heater

HTE =

high-temperature evaporator

LTE =

low-temperature evaporator

MF =

mixed fluid

MW =

molecular weight

NBP =

normal boiling point

NSW =

normal shock wave

ODP =

ozone depletion potential

PF =

primary working fluid

PFC =

fractional condenser for the PF

SF =

secondary working fluid

SFC =

fractional condenser for the SF

The following are the supplementary data to this article:

• S1. BFPs with HFE7000 as the PF

• S2. BFPs with Novec649 as the PF

• S3. BFPs with HFE7100 as the PF

• S4. BFPs with HFE7200 as the PF

• S5. BFPs with HFE7500 as the PF