Compressor

In order to accurately determine the performance of the thermodynamic cycle, isentropic efficiency of the compressor was carefully analyzed. As is well known, it represents the ratio between the work input to an isentropic process and the work input to the actual process, at the same inlet and exit pressures. Two approaches were considered to establish the influence of the isentropic efficiency on the cycle performance.

As a first step of the analysis, a constant value of the isentropic efficiency was assumed: this hypothesis ensures a fair comparison of the performances of different fluids, considering the same efficiency for all fluids and for all the working conditions.

The second step, more realistic, considered a variable isentropic e fficiency, depending on the fluid, the selected commercial compressor and the working conditions. A scroll compressor (Bitzer GSU60120VL, Sindelfingen, Germany) available in the market and optimized for refrigerants R410A, R32, R454B, (and also R452B) was chosen, depending on the typical size of the GSHPs studied in the project. Working with this compressor at given conditions, each fluid has a di fferent isentropic efficiency. The discharge temperatures were obtained through the Bitzer free online software 6.9 [33] and, indirectly, the isentropic e fficiency was derived. This second approach helps to understand the di fferent quality of the compression work and the heat losses occurring during the actual compression process, considering a real compressor present in the market.

Figure 2 shows the isentropic e fficiency values for each fluid as a function of the user outlet temperature, i.e. the temperature of the secondary fluid at condenser outlet, once all the other boundary conditions are fixed. Assuming a constant value of the isentropic e fficiency (Ref. line in Figure 2) for all fluids and all working conditions may lead to a not realistic evaluation of the compressor performance and therefore of the global performance of the thermodynamic cycle. Moreover, this assumption may cause a wrong selection of the most suitable refrigerant for a specific application with given working conditions.

**Figure 2.** Constant value and actual values of isentropic e fficiency at given boundary conditions calculated for the Bitzer GSU60120VL compressor.

#### *2.3. Regenerative Cycle*

With the aim to improve the performance of the base thermodynamic cycle, the addition of a liquid-line/suction-line heat exchanger (LLSL-HX) was evaluated. Thanks to the intra-cycle exchanger, the high-pressure refrigerant from the condenser is subcooled by the low pressure vapor entering the compressor [34]. This configuration is shown in Figure 3, with the description of the cycle in the P-h diagram.

As well known in the literature, high molecular mass fluids can take advantage from the regeneration because of their lower negative or their positive slope of vapor saturation curve in *T*-*s* diagram [35]. Subcooling of high-pressure liquid and superheating of low pressure vapor depend on the amount of heat transferred in the LLSL-HX. The maximum advantage is obtained considering a flooded evaporator, in which the refrigerant is not fully evaporated. Here, a two-phase mixture with 0.9 vapor quality has been considered as leaving low pressure fluid, where the vapor quality is defined from thermodynamics as the ratio between the vapor mass and the total mass of the mixture. The evaporation process is thus completed, together with the superheating of the vapor, in the LLSL-HX. The main benefit of this solution consists in a higher evaporation temperature and therefore in a reduction of the pressure ratio and of the compressor work.

**Figure 3.** Compression cycle with (**a**) liquid-line/suction-line heat exchanger (LLSL-HX) and (**b**) the corresponding pressure-enthalpy diagram [36].

#### *2.4. Energy Analysis*

To evaluate the performance of the thermodynamic cycle, an energy analysis was carried out. Isentropic efficiency was calculated as follows (see Figure 2):

$$\eta\_{\rm is} = (h\_{\rm 3is} - h\_2)/(h\_3 - h\_2)$$

where *h*3*is* and *h*3 are the enthalpies at compressor discharge, respectively for isentropic and real compression, *h*2 is the enthalpy at compressor suction.

Volumetric Heating Effect is another interesting parameter that represents the refrigerating effect per unit of swept volume. It provides information about the heat pump dimensions and about the required refrigerant charge.

$$\mathsf{VHE} = \Delta l \mathsf{i} / \upsilon$$

where Δ*h* is the enthalpy variation at the condenser and *v* is the refrigerant specific volume at compressor inlet.

The main energetic parameter used to compare the refrigerants efficiency is the coefficient of performance (COP) of the heat pump cycle, defined as the ratio between the heat supplied from the cycle to the hot reservoir ( . *Qcond*) and the required network input at the compressor ( . *Wc*):

$$\text{COP = \dot{Q}\_{\text{cond}}/\dot{W}\_{\text{c}}}$$

where *Qcond* represents the power absorbed by the working fluid at the condenser, exchanged with the user and set at 7000 W for every working condition.

The compressor power input *Wc* is calculated as follows:

.

$$\dot{\mathcal{W}}\_c = \dot{m}\_{ref}(h\_3 - h\_2)$$

where .*mref* is the refrigerant mass flow rate, which is known from the power exchanged at the condenser and the enthalpy difference at the condenser, *h*3 is the enthalpy at compressor discharge and *h*2 is the enthalpy at compressor suction.

#### *2.5. Exergy Analysis*

.

For a more comprehensive comparison of the refrigerant's performance, a detailed exergy analysis has been performed applying the general exergy theory described, e.g., in Reference [37]. This type of analysis allows to evaluate thermodynamic processes identifying the major sources of irreversibilities

.

and then inefficiencies in energy exploitation. The optimization of a thermodynamic process has the purpose to minimize exergy losses, whereas energy, according to the first law of thermodynamics, cannot be destroyed and then no information on the quality of each thermodynamic process can be derived by energy balances. The overall system exergetic efficiency can be defined as follows:

$$\eta\_{\rm ex} = \dot{E}\_{\rm useful} / \dot{W}\_{\rm c}$$

where *Euseful* is the output exergy flux, i.e. the exergy absorbed by the user secondary fluid, and *Wc* is the input exergy flux to the system, supplied through the compression work.

.

The exergy absorbed by the user secondary fluid *E ˙ useful* is obtained as:

$$
\dot{E}\_{useful} = \dot{m}\_{user} (k\_{u\\_out} - k\_{u\\_in})
$$

where .*muser* is the user flow rate of the secondary loop and *ku\_out* and *ku\_in* are the specific coenthalpies of the user fluid at the exit and at the entrance of the condenser. Coenthalpy is the potential of exergy flow and it is defined as:

$$k = h - T\_a \cdot s$$

where *Ta* is a reference temperature, set at 5 ◦C and *h* and *s* are respectively the enthalpy and the entropy of the working fluid at the heat exchanger.

For each thermodynamic process in the cycle, exergy losses were calculated to evaluate their relative contribute to system energy efficiency. Exergy losses were calculated using the following equations:


where .*mg* is the water flow rate of the ground loop, go<sup>t</sup> from the balances, and *kg\_out* and *kg\_in* are the specific coenthalpies of the geothermal fluid at the exit and at the entrance of the evaporator.

#### *2.6. Assumed System Parameters*

The following assumptions have been made during the analysis:


Simulations were run by varying both inlet and outlet temperature of the ground heat source (evaporator) and the user (condenser) secondary fluids (Table 2).


**Table 2.** Working conditions of the secondary fluids.

#### **3. Results and Discussion**

The main results of the analysis are summarized below and compare the performance of the thermodynamic cycle using R410A and those of the potential alternative fluids R32 and R454B. Diagrams are referred to the extreme working conditions for the secondary fluid circulating in the ground loop, i.e. inlet/outlet temperatures (*Tg*) at the evaporator:

$$\text{--}\qquad T\_{\mathcal{K}} = 0 \text{--} 3 \text{ } ^\circ \text{C}$$

$$- \qquad T\_{\mathbb{X}} = 10 \!\!/ \!\!\!/ \!\!\!/ \!\!\!/ \!\!\!\!/ \!\!\!\/,$$

For these two sets of secondary fluid temperatures, performance of the heat pump producing domestic water at 4 different user secondary fluid outlet temperatures (from 30 to 55 ◦C) are shown.

#### *3.1. Base Configuration*

#### 3.1.1. Isentropic Efficiency of the Compressor

Considering that isentropic efficiency is different for each fluid and working condition, as came out from previous analysis, it affects the performance of the thermodynamic cycle in different way. Figure 4 shows that compressor isentropic efficiency has a similar trend but, at the same time, different values for each fluid. R454B is the fluid that can guarantee the highest isentropic efficiency at all user temperatures.

**Figure 4.** Isentropic Efficiency. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C

It is also interesting to notice how the shape of trends is different for different temperature levels of the heat carrier fluid of the ground loop (*Tg*). This variation depends on the nominal design limits of the compressor and is due to the fact that when compressor works with too high or too small pressure ratio, the efficiency of the compression process decreases.

#### 3.1.2. Coefficient of Performance (COP)

Figure 5 shows that, for both extreme working temperature levels of the ground loop secondary fluid, COP decreases with the increasing of the user secondary fluid outlet temperature at the condenser, as expected due to the increase of pressure ratio. For the same reason, COP is lower when *Tg* = 0/−3 ◦C than when *Tg* = 10/7 ◦C, considering all other conditions fixed.

**Figure 5.** Coefficient of performance (COP). (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C

R454B has slightly higher (around 5%) COP than R32 and R410A (similar each other) in all working conditions.

#### 3.1.3. Volumetric Heating Effect (VHE)

The results of the VHE are summarized in Figure 6. As it can be seen, R32 has the highest VHE. Thus, it needs lower volumetric flow rate than the other fluids to exchange the same thermal power. R454B, vice versa, has the smallest value of the volumetric heating effect because it is a mixture of HFC R32 and HFO R1234yf having relatively low density. R410A has an intermediate behavior. It is interesting to note that the trend of VHE for each fluid is almost independent from the user secondary fluid outlet temperature at the condenser, that is from the pressure ratio. The trend of the volumetric heating effect is very important because it gives information about the required volumetric mass flow rate and therefore about the size of the heat pump and pipes.

**Figure 6.** Volumetric heating effect. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C

#### 3.1.4. Exergetic Efficiency

Exergetic efficiency trend is shown in Figure 7. As for the isentropic efficiency, the value is higher for higher temperatures of the ground loop fluid. R454B always stands as the refrigerant with the best performance, as from first law analysis. The trend of the performance is not linear because exergetic efficiency of components influences the heat transfer process differently according to the working temperatures.

**Figure 7.** Exergetic Efficiency. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C.

3.1.5. Exergetic Losses

Exergetic analysis is useful because it allows to evaluate the distribution of exergetic losses and thus the inefficiencies of the heat pump components. Figure 8 shows what are the exergy losses generated in each component of the heat pump at a specific condenser condition for the lowest and the highest working temperature levels at the evaporator. In the case of *Tg* = 0/−3 ◦C, it is evident that compression is the thermodynamic process with the highest exergy losses, followed by throttling. Even for the case with the highest evaporation temperatures (*Tg* = 10/7 ◦C), and thus the lowest temperature difference between evaporation and condensation temperatures, compression and throttling are the worst processes, even if with lower differences from condensation and evaporation. Amongst the three refrigerants here considered, R454B showed total exergy losses lower than the other fluids: this result is mainly due to the much lower exergy losses in the compression process. In general, to improve the overall efficiency of the heat pump, efforts should be addressed to improve the efficiency of compression and throttling processes.

**Figure 8.** Exergy losses generated in the heat pump components considering inlet condenser temperature 40 ◦C and outlet condenser temperature 45 ◦C. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C.

(**a**) (**b**)

#### *3.2. Regenerative Configuration*

Regenerative configuration can increase the performance of the base cycle, depending on fluid and working conditions. In this configuration, a vapour quality of 0.9 was assumed for the working fluid at evaporator outlet. Then, the refrigerant moves to the LLSL heat exchanger, where it completes the evaporation and exits 5◦C superheated.

The consequences at the evaporator are:



When the conditions are favorable, regeneration process allows to decrease exergetic losses and compression work. At the same time, it can improve the performance of thermodynamic cycle. In relation to isentropic e fficiency, regeneration is not necessarily beneficial, because the reduction of the pressure ratio induced by the evaporating temperature increase is not always positive for the compression e fficiency. However, compression e fficiency worsening is less significant than compression work reduction and thus the overall e ffect can be positive.

Regenerative cycle performance is represented in terms of percentage deviation from the base cycle performance. Figures 10 and 11 highlight that regenerative configuration is beneficial for all fluids in every working condition in terms of both COP and exergetic e fficiency. This means that using LLSL-HX improves energy performance in all cases and then is strongly suggested for these fluids and operative conditions. While R32 and R410A have similar improvement with respect to the base case, the increase of COP and exergetic e fficiency is more evident for R454B with respect to the other fluids.

**Figure 9.** Temperature profiles in the evaporator for the two thermodynamic cycles. (**a**) Base Configuration; (**b**) Regenerative Configuration.

**Figure 10.** Percentage deviation of coe fficient of performance (COP). (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C.

**Figure 11.** Percentage deviation of the Volumetric Heating Effect. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C.

The regenerative configuration leads also to increased VHE, as represented in Figure 11. In relation to VHE, R32 has the highest benefit with respect to the other refrigerants.

#### 3.2.1. Exergetic Efficiency

As for the previous factors, exergetic efficiency benefits from the regenerative configuration as well. The increase of exergetic efficiency compared to the base cycle occurs for every fluid and every working temperature (Figure 12). Once again, R454B is the refrigerant that gets the most profit.

**Figure 12.** Percentage deviation of the exergetic efficiency. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C.

#### 3.2.2. Exergetic Losses

The regenerative configuration allows to reduce some exergetic losses of the system, represented in Figure 13. Comparing Figures 8 and 13, a considerable reduction during the throttling and evaporation phase is noticeable.

During the evaporation process, losses decrease depending on the change of position of the pinch point. Since the pinch point moves at refrigerant inlet in the evaporator, the area between the two temperature profiles is greatly reduced, and the losses decrease (see Figure 9).

Throttling losses are reduced both because of the increase of evaporation temperature and because of the higher subcooling of liquid at the entrance of the valve.

**Figure 13.** Exergy losses generated in the heat pump components, included liquid-line/suction-line heat exchanger (LLSL-HX), considering inlet condenser temperature 40 ◦C and outlet condenser temperature 45 ◦C. (**a**) *Tg* = 0/−3 ◦C; (**b**) *Tg* = 10/7 ◦C.

It also important to notice that the decrease varies according to the refrigerant and the complexity of the molecules.

Despite the fact that all the losses tend to decrease, it should be considered that in the regenerative configuration there is an additional exchanger, also characterized by exergetic losses. However, these losses do not give an important contribution to the total exergetic losses of the system.

## **4. Conclusions**

As there are many factors influencing the performance of a thermodynamic cycle, it is necessary to carry out an overall analysis to identify the best refrigerant, amongs<sup>t</sup> those with intermediate GWP (<1000), for the replacement of R410A in ground source heat pumps. Thus, a comprehensive analysis has been performed considering the first and the second laws of thermodynamics, as well as exergetic losses in the various components, compressor discharge temperature, and volumetric heating effect.

From the compressor point of view, the analysis has been performed considering both the case of fixed isentropic efficiency for all the considered refrigerants (R410A, R32 and R454B) and that of variable isentropic efficiency as calculated from a software reproducing the behavior of a commercial compressor working with all the considered refrigerants. It is interesting to note that the two approaches gave totally different results. In particular, in the more realistic case of variable isentropic efficiency, R454B showed to be the most promising refrigerant. Thus, assuming constant isentropic efficiency, even if in principle correct from the thermodynamic point of view, can be misleading with respect to the actual performance in a commercial compressor.

A comparison between base configuration and regenerative configuration of the cycle has also been performed, with the aim to evaluate the opportunity to install a LLSL-HX in the cycle. Significantly higher values of COP and exergetic efficiency than R410 and R32 are obtained with R454B in both cases, but with clear enhancements induced by the presence of the LLSL-HX with respect to the base configuration of the cycle.

Vice versa, R454B is characterized by a lower volumetric heating effect than R410A and R32. This can implicate a bigger size of components of the heat pump (e.g., compressor) and then higher costs. For sure, more studies and research have to be made in relation to R454B, which is basically unknown from the technical point of view, since the first compressors using this fluid have been produced in 2018. Moreover, R454B is a zeotropic mixture: therefore, it requests attention in heat exchangers optimization, since variations of the refrigerant temperature and composition can occur during the operation. Finally, in terms of flammability, it is classified as an A2L refrigerant: thus, the mass charged in the system should be limited, preventing use in large installations.

It is also important to underline that the analysis here performed is based on ideal assumptions (for example, effect of the pressure losses was not considered here) that could affect the results, if applied in real installations. The results found, however, provide a starting point for the selection of intermediate GWP replacement fluids.

**Author Contributions:** Conceptualization: S.B. and L.F.; Methodology: S.B., M.D.C. and M.C.; Software: M.C., A.T. and A.B. (Anna Bet); Validation: M.D.C., G.E. and D.M.; Formal Analysis: M.C. and A.B. (Anna Bet).; Investigation: M.C., A.B. (Anna Bet) and S.B.; Resources: F.P. and L.F.; Data Curation: G.M.; Writing-Original Draft Preparation: S.B., A.B. (Anna Bet) and M.C.; Writing-Review & Editing: S.B. and A.B. (Anna Bet); Visualization: S.B. and A.B. (Anna Bet); Supervision: M.D.C.; Project Administration: S.B. and L.F.; Funding Acquisition: A.B. (Adriana Bernardi).

**Funding:** This research was funded by Horizon 2020 (792355) and the APC was funded by ITC CNR.

**Acknowledgments:** The present study is realized within the project "Most Easy, Efficient and Low Cost Geothermal Systems for Retrofitting Civil and Historical Buildings" (Grant agreemen<sup>t</sup> ID: 792355) of the European Union's Horizon 2020 research and innovation program.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.
