Energy and Buildings 40 (2008) 2022–2027

Contents lists available at ScienceDirect

Energy and Buildings journal homepage: www.elsevier.com/locate/enbuild

Performance of a new refrigeration cycle using refrigerant mixture R32/R134a for residential air-conditioner applications Jianyong Chen, Jianlin Yu * Department of Refrigeration and Cryogenic Engineering, School of Energy and Power Engineering, Xi’an Jiaotong University, Xi’an 710049, China

A R T I C L E I N F O

A B S T R A C T

Article history: Received 10 March 2008 Received in revised form 5 May 2008 Accepted 12 May 2008

In this paper, a new refrigeration cycle (NRC) using the binary non-azeotropic refrigerant mixture R32/ R134a is presented, which can be an alternative refrigeration cycle applied in residential air-conditioner. In the NRC, refrigerant circuit of the evaporator is separated into two branches. Because the nonazeotropic mixture has the characteristic of temperature glide, an important benefit of such configuration is that the Lorentz cycle can be realized. Compared with that of conventional cycle configuration, the new cycle efficiency can be improved. The calculating results show that, in the conventional refrigeration cycle (CRC), the mixture R32/R134a has a close performance to that is obtainable with pure refrigerant R22. However, the mixture R32/R134a in the NRC will result in a better performance. The maximal COP can be improved in a range of 8–9% over that of the CRC, and the volumetric refrigerating capacity can be approximately increased by 9.5%. ß 2008 Elsevier B.V. All rights reserved.

Keywords: Refrigeration cycle Refrigerant mixture R32/R134a Performance

1. Introduction For the past decades, R22 has been predominantly used in the field of heat pump and air-conditioner because of its favorable characteristics. Even though the ozone depleting potential of R22 (HCFC) is not as high as CFCs, it still contains ozone depleting chlorine and brings on the environmental problems. According to the Montreal Protocol, R22 will be completely discarded in 2020. Thus, a replacement for this well-known refrigerant is apparently very urgent. Because of the obvious contribution of R22 to the ozone depletion, substitutes for R22 have been developed [1]. At present, alternative refrigerants available for R22 in residential airconditioners and heat pumps can be categorized into three types: (i) HFC (hydro-fluorocarbons) represented by R410A, R407C and R134a, (ii) HC (hydro-carbons) represented by R290, and (iii) natural substances CO2 and R717 [2]. However, many other researches have been done to find the alternative refrigerants [3,4]. Calm and Domanski [5] proposed 14 candidate refrigerants as the alternative refrigerants available for R22. Thermodynamic performance of two pure hydrocarbons and seven mixtures composed of propylene (R1270), propane (R290), HFC152a, and dimethylether (RE170, DME) was measured by Park and Jung [6] in an attempt to substitute for R22 in residential air-conditioners. The test results

* Corresponding author. Tel.: +86 29 82668738; fax: +86 29 82668725. E-mail address: [email protected] (J. Yu). 0378-7788/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2008.05.003

show that the coefficient of performance (COP) of these mixtures is up to 5.7% higher than that of R22, and these fluids provide a good thermodynamic performance with reasonable energy savings and no any environmental problem. Zhao et al. [7] investigated the performance of some new refrigerant mixtures R32/125/152a, R125/290, R32/290 and R32/125/290 as a replacement for R22 in theory and experiment. It is drawn that the compositions of the mixtures should be optimized according to the ranges of operating conditions of the unit applications. In general, all the papers mentioned above mainly focused on the studying of the alternative refrigerant properties and the present cycle system itself. However, various new cycle designs may promote the use of the alternative refrigerants and further improve the cycle performance. In this paper, a new refrigeration cycle (NRC) for binary nonazeotropic mixed refrigerants is presented in order to obtain better performance. The cycle takes advantage of the characteristic of temperature glide of the mixture to realize the Lorentz cycle, and then the cycle performance can be improved. The present study mainly focuses on a theoretical investigation on the performance of this NRC. The binary non-azeotropic mixture R32/R134a is employed as the working fluid and the cycle will be simulated at typical air-conditioning conditions to investigate the effects of main parameters, such as the composition of the used refrigerant mixtures, the subcooling and superheating degree. In addition, the performance comparison between the new refrigeration cycle (NRC) and the conventional refrigeration cycle (CRC) will also be discussed.

J. Chen, J. Yu / Energy and Buildings 40 (2008) 2022–2027

2023

Nomenclature COP h m p P Q rp t Dtc Dth z v

coefficient of performance specific enthalpy (kJ kg1) mass flow rate (kg s1) pressure (kPa) compressor power (kW) refrigerating capacity (kW) pressure ratio temperature (8C) subcooling degree (8C) superheating degree (8C) mass fraction specific volume (m3 kg1)

Greek letters h efficiency

Fig. 1. The NRC system. (A) Compressor; (B) condenser; (C) phase separator; (D) subcooler; (E) recuperator; (F) expansion valve; (G) evaporator.

Subscripts c e com 1–13

condenser evaporator compressor state points in Figs. 1 and 2

2. The new refrigeration cycle Most of the current refrigerants, such as R22, R410A, and R407C, are used in residential air-conditioner. In the NRC, the refrigerant mixture R32/R134a is proposed as the alternative working fluid for the residential air-conditioner. Some basic properties of these refrigerants are presented in Table 1, including molar mass, NBP (normal boiling point), critical temperature, critical pressure, ODP (ozone depletion potential), GWP100 (global warming potential) and the temperature glide at NBP [8,9]. It can be seen that all three alternative refrigerants has an ODP of zero, but the mixture R32/ R134a (0.3/0.7) has a low GWP (1200) while both R410A and R407C is above 1800 GWP. On the other hand, R410A is a near azeotropic mixture with a 0.08 8C temperature glide, but R407C and R32/R134a (0.3/0.7) are totally non-zeotropic with a 7.0 and 7.2 8C temperature glide. This special characteristic can be exploited for performance improvements in various refrigeration cycles by partially matching the temperature glide with the temperature profile of the source or sink fluid in counter-current heat exchangers. However, since R407C consists of three components, problems have arisen while system leakage and recharging happen. This would cause a change in the system characteristics. Thus, as a general replacement for R22, R407C will not be considered in the NRC of this paper.

The NRC system is shown in Fig. 1, which is composed of a compressor, a phase separator, a subcooler, two condensers, two evaporators, two recuperators and two expansion valves. The CRC system merely includes a compressor, a condenser, an expansion valve and an evaporator. Fig. 2 is the p–h (pressure–enthalpy) diagram with the NRC, which shows the working process of the NRC. All the points in Fig. 2 are corresponding to the points in Fig. 1. In this NRC, the binary mixture vapor with initial charge composition (at point 1) is sucked by compressor. After compressed to point 2, the refrigerant mixture is partially condensed to point 3 in the condenser 1. This two-phase stream flows into the phase separator, in which the vapor phase at point 3v will have a different composition from the liquid phase at point 3l. The vapor phase has higher mass fraction of the pure refrigerant with lower boiling point and the liquid phase has higher mass fraction of the pure refrigerant with higher boiling point, therefore, the vapor phase and the liquid phase can be called as low boiling point stream and high boiling point stream, respectively. The low boiling point stream will enter into the condenser 2, in which the vapor phase stream is condensed totally and cooled to point 4. In the recuperator 1, the refrigerant is further cooled to point 5 and then flows into the expansion valve 1, expending from point 5 to point 6. After that, the low boiling point stream will enter into the evaporator and partially evaporated to point 7. At last, it flows through the recuperator 1 again (at point 8). Similarly, the high boiling point stream enters into the subcooler and will be subcooled to point 9, then in turn flows through the recuperator 2 (at point 10), the expansion valve 2 (at point 11). It is partially evaporated to point 12 in the evaporator 2, and then flows through the recuperator 2 (at point 13). Eventually, the low boiling point stream from the recuperator 1 and the high boiling point stream

Table 1 The basic properties of various refrigerants Refrigerant

1

Molar mass (kg kmol ) NBP (8C) Critical temperature (8C) Critical pressure (kPa) ODP GWP100 Temperature glide at NBP (8C)

R22

R410A

R407C

R32/R134a (0.3/0.7)

86.47 40.8 96.1 4990 0.050 1810 0

72.58 51.4 70.5 4810 0 2100 0.08

86.20 43.6 85.8 4600 0 1800 7.0

79.19 41.7 91.6 4860 0 1200 7.2

J. Chen, J. Yu / Energy and Buildings 40 (2008) 2022–2027

2024

In the calculating program written with Fortran Language, the NIST routines were used to calculate these properties, which is an industry standard currently [8]. When the property of each point is calculated, the equations for the cycle performance can be found based on the mass and energy conservation. In the NRC, refrigerating capacity is given as Q 01 ¼ ðh7  h6 Þm˙ 7

(1)

Q 02 ¼ ðh12  h11 Þm˙ 11

(2)

Q 0 ¼ Q 01 þ Q 02 where Q01 is the refrigerating capacity of evaporator 1, Q02 is the refrigerating capacity of evaporator 2 and Q0 is the total refrigerating capacity of the cycle. Volumetric refrigerating capacity is obtained by Qv ¼ Fig. 2. The p–h diagram with the NRC.

from the recuperator 2 are mixed (at point 1). The mixed vapor is sucked by the compressor, and forms a complete cycle. Because the non-azeotropic refrigerant mixture has a significant characteristic of temperature glide, the heat transfer temperature difference between the refrigerant and air medium can be reduced by alternative configuration for path of refrigerant and air flow directions, and thus the irreversible loss also can be reduced. In the parallel-cross flow type configuration, refrigerant circuit of the evaporator has two separate paths. It is known that at the same pressure, the temperature of the high boiling point stream is higher than the low boiling point stream. As a result, the temperature at point 11 is higher than at point 6, and the temperature at point 12 is higher than at point 7. Moreover, due to that the refrigerant is partially evaporated in both evaporators, the temperature at point 11 can be higher than at point 7 and the temperatures of point 6, 7, 11, 12 can be gradually increased. When the air flows through the evaporators in terms of cross flow type, a smaller average temperature difference across the evaporators can be obtained. Similar configurations for the condensers and the subcooler are also adopted as shown in Fig. 1. Therefore, the Lorentz cycle can be realized in this new refrigeration cycle and the cycle efficiency may be improved. It should be noted that although the above new cycle configurations show increasing complexity for refrigerant circuits, the two evaporators can be assembled in a whole. In the same way, condensers and the subcooler are also fabricated into a whole. 3. Theoretical computation model

Q0 v1

(3)

The compressor power is given as P com ¼

ðh2  h1 Þm˙ 1

hcom

The coefficient of performance (COP) of the cycle can be determined by COP ¼

Q0 P com

(5)

where h1–h13 is the enthalpy and m˙ 1 m˙ 13 is the mass flow rate of state point 1–13 in the cycle. In the CRC, refrigerating capacity is given as Q 0 ¼ ðheout  hein Þm˙

(6)

where heout is the enthalpy of the evaporator outlet and hein is the enthalpy of the evaporator inlet. Volumetric refrigerating capacity is given as Qv ¼

Q0 vcin

(7)

where ncin is the specific volume of the compressor inlet. The compressor power of the compressor is given as P com ¼

ðhcout  hcin Þm˙

hcom

(8)

where hcout is enthalpy of the compressor outlet and hcin is enthalpy of the compressor inlet. The coefficient of performance (COP) of the cycle can be determined by Q0 P com

In order to make the theoretical calculation simple, the following assumptions are made for cycle performance calculation.

COP ¼

(1) Neglecting the pressure losses and the heat losses to the environment in the condensers, phase separator, subcooler, recuperators and evaporators. (2) The low boiling point stream (vapor phase) from the separator is saturated vapor and the high boiling point stream (liquid phase) from the separator is saturated liquid, too. (3) The flow across the expansion valves is isenthalpic. (4) Taking the isentropic efficiency into account for the compressor.

4. Results and discussion

The corresponding thermodynamics properties of the cycle states can be determined from the equation of states of refrigerant.

(4)

(9)

In the present study, the refrigerant mixture R32/R134a is selected as typical working fluid for simulating NRC performance, which is also used in CRC. Furthermore, the refrigerant R22 and R410A, R407C are used in CRC for general comparison. To investigate the characteristics of the cycles, the following basic operating conditions are firstly assumed: (1) For the condenser, in CRC, tc is the condenser outlet temperature. In the NRC, tc is the outlet temperature of condenser 2. The value of tc in the simulation is fixed at 55 8C.

J. Chen, J. Yu / Energy and Buildings 40 (2008) 2022–2027

(2) In the cycles with the pure refrigerants, te is the evaporator outlet temperature and is fixed at 5 8C. In the cycles with the mixed refrigerant, because the mixed refrigerant has a characteristic of temperature glide, the lowest temperature (the evaporator inlet temperature) tlow is fixed at 5 8C. Furthermore, temperature of point 6, which is the lowest, is fixed at 5 8C in the NRC. (3) According to Ref. [5], the refrigerant mixture R32/R134a with four different mass fractions of 0.2/0.8, 0.25/0.75, 0.3/0.7 and 0.4/0.6 may be used. In the present study, the mass fraction 0.3/ 0.7 of R32/R134a is selected to analyze the cycle. For both the NRC and the CRC, the mass fraction of R32 as the initial charged refrigerant mixture at point 1 is 0.3. (4) In the CRC, the subcooling degree Dtc at condenser outlet is held 5 8C. The superheating degree Dth is zero. In the NRC, the subcooling degree is held 5 8C at the outlet of condenser 2, the decreasing temperature in the subcooler is also held 5 8C. The superheating degree is kept as zero in the two recuperators. (5) In the NRC, the Q01 and Q02 is adjusted to approximate equal for making the heat load allocation of two evaporators more reasonable. (6) The compressor is assumed to have an isentropic efficiency hcom = 0.8, which is constant and does not vary with the pressure ratio in all cases. And the mass flow ratio of refrigerant into the compressor is set at 1 kg/s in all cases. It is noted that a performance comparison for the NRC and CRC is made under the same operating conditions, which is shown in the following corresponding figures. 4.1. The comparison for the NRC and the CRC performance The working fluid of the NRC is the refrigerant mixture R32/ R134a with the mass fraction of 0.3/0.7, the working fluids of the CRC are R22, R410A, R407C and the mixture R32/R134a with the same mass fraction of 0.3/0.7, respectively. The following result is obtained when the condenser (or the condenser 2 in the NRC) outlet temperature is tc = 55 8C, the evaporator inlet temperature tlow = 5 8C, the mass fraction of R32 is z = 0.3. And the subcooling degree is 5 8C. Table 2 shows the performances comparison between the NRC and the CRC, which is based on the ratio relative to the values of the CRC with R22. From Table 2, the NRC gives the highest COP. R410A has the highest volumetric refrigerating capacity Qn, but the lowest COP. In addition, R410A also has the highest pressure of the compressor outlet, p2. R407C gives a higher COP than R410A, but it is a ternary non-azeotropic refrigerant mixture. Compared to the CRC with R32/R134a, the NRC has higher COP and Qn, but the values of Q0, p2, and rp in both CRC and NRC are very close. Thus, the refrigerant mixture R32/R134a with the mass fraction of 0.3/0.7 is a better alternative refrigerant for replacing R22 in the CRC. However, the R32/R134a with the mass fraction of 0.3/0.7 has a better performance in the NRC than that of CRC.

Table 2 The comparison between the NRC and the CRC

R410A in CRC R407C in CRC R32/R134a in CRC R32/R134a in NRC

COP

Qn

Q0

p2

rp

0 894 0.941 0 974 1.011

1.39 1.01 1.00 1.10

0 95 0.95 1.11 1.10

1 58 1.14 1.07 1.12

0 98 1.07 1.08 1.01

2025

Fig. 3. Variation of COP and Qn with the z.

4.2. Effect of the refrigerant mixture composition on cycle performance In the cycle using binary non-azeotropic mixed refrigerants, the pressure–temperature relationship of the refrigerant mixture in the condensers or evaporators is dependent on the composition. At the given temperature, the pressure changes as the composition of the mixture varies, which affects the cycle performance. So the composition of refrigerant mixture is important variable in the cycle. The mass fraction of R32 z as the initial charged refrigerant mixture at point 1 is treated as a variable operating condition. In this paper, both NRC and CRC with the mixture R32/R134a as working fluid are evaluated, and the variation of the cycle performance with the change of z is represented. As for composition of the refrigerant mixture, the z was selected in a range of 0.2–0.6, so that the maximum operating pressure was below 3MPa. The operating conditions are kept as tc = 55 8C, tlow = 5 8C and the subcooling degree is kept as 5 8C. The variations of the COP and Qn in the NRC and the CRC for the different are z displayed in Fig. 3. For the given conditions, the COPs in the two cycles decrease with an increase in the composition. By contraries, in two cycles the Qn increases with an increase in the composition. This is because the thermodynamic property of the refrigerant mixture changes as the composition of the mixture varies. Furthermore, it can be seen that the NRC has higher COP and Qn than the CRC over the entire range. Because of the phase separator in the NRC, there are three different mass fractions of R32: the initial charged refrigerant mixture at point 1, the low boiling point stream at point 3v and the high boiling point stream at point 3l, whereas there is only one mass fraction of R32 in the CRC. For example, in the NRC, when the mass fraction of R32 at point 1 is 0.3, the mass fraction of R32 at point 3v and 3l is 0.354 and 0.25, respectively. Compared with the CRC, when the tlow = 5 8C is fixed, the NRC has higher pressure because of the higher mass fraction of R32. So the power consumption of compressor is lower. At the same time, though the outlet of the evaporators is not saturated, the refrigerating capacity is not decreased because of applying the recuperators in the NRC. As a result, the NRC has higher COP. Especially, the NRC has a smaller specific volume vcin at the compressor inlet. So the Qn is increased by 9.5% as compared to that of the CRC. 4.3. Effect of the subcooling and superheating degree on the NRC performance It is known that the subcooling and superheating have great effects on the cycle performance. Thus, the subcooling and

2026

J. Chen, J. Yu / Energy and Buildings 40 (2008) 2022–2027

Fig. 4. Variation of COP and Qn with theDtc. Fig. 6. Variation of Qo with the Dth.

superheating degrees are considered in the simulation for the NRC. The operating condition is kept as tc = 55 8C, tlow = 5 8C and the mass fraction of R32 at point 1 is kept as 0.3. To ensure that the temperatures of point 6, 7, 11, 12 are gradually increased, the subcooling degrees at point 4 and at point 9 should be properly arranged. In the same way, the superheating degrees at point 8 and point 13 also need to be chosen at appropriate values. In this paper, the effect of the subcooling and superheating degree are calculated respectively, both subcooling degree and superheating degree are kept in a range of 0–15 8C. The variations of the COP, Q0 and Qn in the NRC for the different superheating and subcooling degrees are displayed in Figs. 4–6. Dtc1 is the subcooling degree of subcooler, Dtc2 is the subcooling degree at point 4 in the condenser 2, Dth1 is the superheating degree at point 13, and Dth2 is at point 8. As shown in the Fig. 4, both the COP and Qn increase greatly with an increase of the subcooling degree. Because the tlow = 5 8C is constantly fixed and the refrigerant mixture R32/R134a has a characteristic of temperature glide, the evaporator pressure increases with an increase of the subcooling degree. As a result, the refrigerating capacity increases, and the Pcom, decreases. Thus the COP increases greatly, and the maximal COP can be up to 3.837, which is improved in the range of 8–9% over that of the CRC. In addition, the specific volume decreases and the Qn increases. The

Figs. 5 and 6 show the variations COP and Q0 for the different superheating degrees, respectively. It is seen that the Q0 increases but the COP varies slightly. This is because the compressor inlet temperature increases, and the compressor power Pcom increases. Thus, the COP increases slightly. It is obvious that the subcooling has a greater effect on the cycle performance than the superheating does. 5. Conclusion The NRC for non-azeotropic refrigerant mixture was presented in this paper. Theoretical model was constructed to evaluate the performances of the cycle with the refrigerant mixture R32/R134a. The effects of main parameters on the performance of the cycle were investigated by using the theoretical model. The performance of the NRC is then compared with that of the CRC at the same operating conditions. From the above theoretical computation results, it can be found that the non-azeotropic mixed refrigerant R32/R134a with the mass fraction of 0.3/0.7 can be a candidate alternative refrigerant available for R22. In the conventional refrigeration cycle (CRC), the mixture R32/R134a has a close performance to that is obtainable with pure refrigerant R22. However, the mixture R32/R134a in the NRC will result in a better performance. The maximum COP can be improved in the range of 8–9% over that of the CRC, and the volumetric refrigerating capacity increase by approximately 9.5%. The subcooling and superheating have great effect on the cycle performance, and the effect of the subcooling is much more than the superheating. Acknowledgements The present study was supported by the Program for Changjiang Scholars and Innovative Research Team in University, through Grant No. IRT0746. References

Fig. 5. Variation of COP with the Dth.

[1] H.M. Hughes, Test results of HCFC-22 in a unitary air conditioner, in: Proceedings of the International CFC and Halon Alternatives Conference, Washington, DC, (October 1994), pp. 38–42. [2] W. Chen, A comparative study on the performance and environmental characteristics of R410A and R22 residential air conditioners, Applied Thermal Engineering 28 (2008) 1–7. [3] R. Radermacher, D. Jung, Theoretical analysis of replacement refrigerants for R22 for residential uses, ASHRAE Transactions 99 (1) (1993) 333–343. [4] A. Cavallini, Working fluids for mechanical refrigeration, International Journal of Refrigeration 19 (1996) 485–496. [5] J.M. Calm, P.A. Domanski, R-22 replacement status, ASHRAE Journal (2004) 29–39.

J. Chen, J. Yu / Energy and Buildings 40 (2008) 2022–2027 [6] K.-J. Park, D. Jung, Thermodynamic performance of HCFC22 alternative refrigerants for residential air-conditioning applications, Energy and Buildings 39 (2007) 675–680. [7] Z. Yang, Y. Ma, Y. Li, Z. Chen, L. Ma, The performance of some substitutes for HCFC22 under varying operating conditions, Applied Thermal Engineering 19 (1999) 801– 806.

2027

[8] NIST Standard Reference Database 23, NIST Thermodynamics and Transport Properties of Refrigerants and Refrigerant Mixtures, REFPROP, Version 6.01, 1998. [9] J.M. Calm, G.C. Hourahan, Refrigerant date update, Heating/Piping/Air Conditioning Engineering 79 (1) (2007) 50–52, 63–64.