Experimental performance analysis and optimization of a direct expansion solar-assisted heat pump water heater

ARTICLE IN PRESS

Energy 32 (2007) 1361–1374 www.elsevier.com/locate/energy

Experimental performance analysis and optimization of a direct expansion solar-assisted heat pump water heater Y.W. Lia,b, R.Z. Wanga,b,, J.Y. Wua,b, Y.X. Xua,b a

Institute of Refrigeration and Cryogenics, Shanghai Jiaotong University, 200240 Shanghai, China b Engineering Research Center of Solar Power & Refrigeration, MOE, 200240 Shanghai, China Received 25 February 2006

Abstract In this study, a direct expansion solar-assisted heat pump water heater (DX-SAHPWH) with rated input power 750 W was tested and analyzed. Through experimental research in spring and thermodynamics analysis about the system performance, some suggestions for the system optimization are proposed. Then, a small-type DX-SAHPWH with rated input power 400 W was built, tested and analyzed. Through exergy analysis for each component of DX-SAHPWH (A) and (B), it can be seen that the highest exergy loss occurs in the compressor and collector/evaporator, followed by the condenser and expansion valve, respectively. Furthermore, some methods are suggested to improve the performance of each component, especially the collector/evaporator. A methodology for the design optimization of the collector/evaporator was introduced and applied. In order to maintain a proper matching between the heat pumping capacity of the compressor and the evaporative capacity of the collector/evaporator under widely varying ambient conditions, the electronic expansion valve and variable frequency compressor are suggested to be utilized for the DX-SAHPWH. r 2006 Elsevier Ltd. All rights reserved. Keywords: Heat pump; Water heater; Coefficient of performance; Solar collector efficiency; Exergy

1. Introduction Solar energy is renewable and ‘‘free’’, which can be a heat source of heat pump like the air source. In order to improve the heat pump COP, the idea of combining the heat pump with solar energy application system has been proposed and developed by many researchers around the world. In a so-called direct expansion solar-assisted heat pump (DX-SAHP), the collector and evaporator are combined into one unit (collector/evaporator), where the refrigerant circulating in heat pump system gets evaporated by absorbing the incident solar energy (and/or ambient air energy). The DX-SAHP offers several advantages over the conventional SAHP, such as superior thermodynamic performance, lower system cost and longer life time of collector/evaporator. Corresponding author. Institute of Refrigeration and Cryogenics, Shanghai Jiaotong University, 200240 Shanghai, PR China. Tel.: +86 21 62933838; fax: +86 21 62932601. E-mail address: [email protected] (R.Z. Wang).

0360-5442/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.energy.2006.11.003

It is estimated that, in China, solar water heater have been marketed for about 10 billion RMB Yuan each year. It is also reported that China has 5 million solar water heaters installed in families in 2000, and it is still being developed more and more rapidly. In virtue of its advantages over conventional solar water heaters, the DX-SAHPWH is expected to have a giant potential market in China. The DX-SAHP concept was first considered by Sporn and Ambrose in West Virginia [1]. Following their work, many theoretical and experimental studies have been reported in the past 27 years [2–26]. A review paper in this field has indicated that the COP values of the DX-SAHP systems range from 2 to 9 and the collector/ evaporator efficiencies vary between 40% and 75% under different climatic conditions, experimentally [25]. With different matching between the system components, especially the collector/evaporator and the compressor, the COP and Zcoll values are very different each other. So the proper matching between each component has an important impact on the performance of the DX-SAHPWH.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1362

Nomenclature Ac COP CPw D E_ Qw E_ rad e F F0 H_ h hw I_rr IT _r m Mw N P Q_ cond Q_ eva S_ s T ULc ULcond Vd Vw W _i W _ comp W

solar collector/evaporator area, m2 coefficient of performance specific heat at constant pressure of water, kJ/ kg K outside diameter of the tube, m exergy rate of hot water, kW exergy rate of solar radiation, kW specific exergy, kJ/kg fin efficiency collector efficiency factor enthalpy rate, kJ specific enthalpy, kJ/kg wind heat transfer coefficient, W/m2 K exergy loss rate, kW total solar radiation intensity, W/m2 mass flow rate of refrigerant, kg/s mass of water in domestic water tank, kg compressor motor speed, rpm pressure, Pa condensing heat rate of refrigerant, kW evaporating heat rate of refrigerant, kW entropy rate, kJ/K specific entropy, kJ/kg K absolute temperature, K collector heat loss coefficient, U Lc ¼ hw þ 4sT 30 , W/m2 K condenser heat loss coefficient, W/m2 K displacement volume, cm3/rev wind velocity at ambient environment, m/s pitch of the tubes in collector/evaporator, m indicated power of compressor, kW total input electric power to compressor, kW

Except for the collector/evaporator, the rest of the DX-SAHPWH system employs ordinary materials and components currently available in the refrigerating and air conditioning industry. The study of the performance enhancement for collector/evaporator is significant for the development of the DX-SAHPWH. In this study, the focus aims at developing a DX-SAHPWH with higher performance suitable for Chinese potential market. At present, two prototypes (DX-SAHPWH (A) and (B)) were built in sequence in Engineering Research Center of Solar Power & Refrigeration, MOE, China (in Shanghai Jiao Tong University, latitude 31.22 ‘N, longitude 121.48 ‘E). Seasonal experiments, the first-law-oriented energy analysis and the second-law-oriented exergy analysis for the DX-SAHPWH (A) and (B) systems were conducted, respectively. A methodology for the design optimization of the collector/ evaporator was also introduced and applied. It should be noted that refrigerant R-22 was used in this study, even though it would be phase out by the year 2040

Greek symbols a w e Z ZV n1 s t

absorptivity of collector/evaporator exergy loss coefficient emissivity of collector/evaporator plate efficiency volumetric efficiency specific volume of suction refrigerant vapor, m3/kg Stefan–Bolzmann constant, W/m2 K4 time, or one data acquisition interval, s

Subscripts 0 comp coll cond eva ex f p rad r sa v w i

ambient air compressor collector/evaporator condensation or condenser evaporation or collector/evaporator exergy refrigerant fluid collector/evaporator plate solar radiation refrigerant sol–air expansion valve hot water in tank inlet, or ith sequence of each component of the DX-SAHPWH system j jth segment of data acquisition interval o outlet 1, 2, 3, 4 state

in developing countries, such as China, where R-22 is still widely used on heat pump systems. It is significant to deal with improving the heat pump systems’ thermal performance using R-22 as refrigerant in order to enhancing the energy saving in the developing countries. On the other hand, the analysis in this way can be applied in the same way to heat pump systems using other kinds of refrigerant.

2. System description and experimental setup 2.1. Thermodynamic cycle of the DX-SAHPWH system Fig. 1 shows the schematic diagram of the DXSAHPWH (A) in the present study. An unglazed solar collector as evaporator, an R-22 rotary-type hermetic compressor, a hot water tank with an immersed copper tube coil heat exchanger as condenser, a thermostatic expansion valve (TEV), some accessories and connective copper pipes compose the DX-SAHPWH system.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1363

Solar Collector/Evaporator

Tp

Domestic Water Tank R-22 Hot Water Supply

City Water Make-up

Peva,o

T2

T1

T3 Compressor Pcond

Pressure Position TemperaturePosition

Peva,i

T4 Filter-Drier Thermostatic Expansion Valve

Fig. 1. Schematic diagram of the DX-SAHPWH system circuit.

2.2. Prototype design An experimental prototype of DX-SAHPWH (A) was designed and fabricated as shown in Fig. 3. The specification of each main component is listed in Table 1.

Heat release to the hot water tank

3

7

6

2

lg p

The condensed liquid refrigerant (R-22) from the condenser passes through the TEV directly into the solar collector/evaporator where it gets evaporated by incident solar energy (and/or ambient air energy) heating. The ambient air acts as an additional heat source or sink, depending on whether the refrigerant temperature Teva is lower or higher than the ambient temperature T0. The vaporized refrigerant then passes through the compressor, and finally the high temperature/pressure vapor is pumped into the condenser where it gets condensed. The energy rejected by the condenser is absorbed by water as cooling medium through a refrigerant-to-water heat exchanger (copper tube coil type) immersed in the water tank. The process undergone by the refrigerant during a certain period of cycles can be represented by an idealized heat pump cycle as shown in Fig. 2. Here, 1-2, 2-3, 3-4 and 4-1 represents compressing, condensing, throttling and evaporating process, respectively. The corresponding thermodynamic states of the refrigerant at 1, 2, 3 and 4 points are superheated vapor at evaporating pressure, superheated vapor at condensing pressure, subcooled liquid at condensing pressure and two-phrase fluid at evaporating pressure, respectively. Because of pressure drop in the solar collector/evaporator, the evaporating pressure at point 1 is smaller than that at point 4 as shown in Fig. 2. Similar sets of periods occur in turn and the water in the tank gets warmed up, as long as the refrigerant in the collector/ evaporator can be heated.

5

4

1

Absorbing solar (and ambient) energy h Fig. 2. Heat release to the hot water tank.

A series of solar collector (with total area 4.20 m2) without any glazing or back insulation was used as heat source device as well as evaporator for the refrigerant, R-22. It consists of 4 aluminum absorber plates in parallel 2 flow paths, which were made by a special process, in which the piping network design is laid between two sheets of aluminum plates and retained after the sheets are bonded by rolling them together, and then the tubes are formed by over-pressurizing the network so that the fluid circuit is made with the fin. As a result, the aluminum collector/evaporator is light in weight and very thin so that it can be mounted easily anywhere. In this experimental study, the collector/evaporator was fixed on a facing south roof with a tilted angle 31.221 (the latitude of Shanghai) to horizon, as shown in Fig. 3. The collector/evaporator

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1364

Fig. 3. Photo of the experimental prototype setup.

Table 1 Specification of the main components of the DX-SAHPWH (A) system Name

Type

Remarks

Compressor Domestic water tank

Rotary Pressure resistance and heat insulation Aluminum plate TEX-2 type, manufactured by Danfoss, Denmark

Rated input power: 0.75 kW, displacement volume: 13.40 cm3/rev 150 L water, immersed 60 m copper coil (f9.90  0.75 mm) as condenser

Solar collector/evaporator Thermostatic expansion valve

4 plates in parallel 2 flow paths, total collector/evaporator area: 4.20 m2 External balance type

surface was selective coated to improve its absorptivity. A R-22 rotary-type hermetic compressor with rated input power 750 W was used in the system. To avoid the overload, an overheat protector and low-high pressure cut-off switches were connected to the compressor. The condenser was made up of a copper tube (f9.90  0.75 mm) coil with total length about 60 m, which was immersed in the domestic water tank (with water volume 150 L and polyurethane insulation thickness 38 mm), and a filter-drier was installed downstream the condenser. The TEV (external balance type) regulates refrigerant flow through the solar collector/evaporator, and the power on–off for the DX-SAHPWH system was controlled by a thermometer located inside the water tank. 2.3. Instrumentation The ambient temperature was measured by a platinum resistance thermometer (with grade ‘‘A’’ accuracy). The collector surface temperature and the refrigerant temperature at various locations of the system were measured by copper-

constantan thermocouples (T-type, with uncertainty 70.5 1C). The water temperature was measured by a thermistor thermometer (with uncertainty 70.5 1C) located in the hot water tank. The low and the high pressure were measured by pressure transducers. There were two lowpressure transducers (with uncertainty 70.08 bar) at inlet and outlet of the collector/evaporator, respectively, in order to measure the pressure drop along the flow path in the solar collector/evaporator. A high-pressure transducer (with uncertainty 70.15 bar) was located at outlet of the condenser. A solar pyranometer (with sensitivity 7.464 mV/W m2, uncertainty about 710 W/m2) was mounted near the collector/evaporator to measure the instantaneous solar radiation intensity. The power consumption of the compressor was measured instantaneously by an electric power measuring meter (with grade 0.5 accuracy, uncertainty about 710–15 W). The above measuring processes were monitored and controlled by a personal computer-based D-A system. The data were recorded automatically at every 2 minutes interval through a data logger (Keithley Model-2700 Data Acquisition System) for later analysis.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1365

3. Thermodynamic analysis of the system

The throttling process is considered approximately being isenthalpic:

First-law analysis (i.e. energy analysis) and second-law analysis (i.e. exergy analysis) are simple thermodynamic methods. Exergy analysis is a part of the energy analysis actually. The theory of exergy analysis is essentially the available energy analysis. Exergy is a measure of the maximum useful work that can be done by a system interacting with the surrounding environment, which is at a constant pressure P0 and a constant temperature T0 [26]. The use of simple thermodynamic first- and second-law methodologies in DX-SAHPWH could contribute to determining a rational system design. The results can lead engineering designers toward their easier identification of the components where exergy loss occurs significantly, so as to provide a quantified guideline for further modifications.

_ r h4 . _ r h3 ¼ m m

3.1. First-law analysis In this study, the energy balance equation for the DXSAHPWH can be expressed as _ i, Q_ cond ¼ Q_ eva þ W (1) where Q_ is the heat transfer rate, and the subscript ‘‘cond’’ stands for condenser and ‘‘eva’’ for collector/evaporator. _ i is the indicated power of compressor. W The heat absorbed from the ambience for the collector/ evaporator (Q_ eva ) can be expressed in terms of the enthalpy change of the refrigerant from the inlet to the outlet of the collector/evaporator as _ r ðh1  h4 Þ. Q_ eva: ¼ m

(2)

For the compressor under a constant speed operation, the mass of refrigerant pumped and circulated by the compressor is given as _r ¼ m

NV d ZV . 60n1

(3)

The motor RPM N, displacement volume rate Vd and volumetric efficiency ZV of the rotary-type hermetic compressor for the DX-SAHPWH (A) system are provided by the manufacture as 2830, 13.40 cm3/rev and 91%, respectively. The specific volume at the inlet of compressor n1 can be determined by definitive pressure Peva, o and T1 through refrigerant R-22’s properties table. The indicated power for compression can be calculated in terms of the enthalpy change of the refrigerant from the inlet to the outlet of the compressor _i¼W _ comp  Zcomp ¼ m _ r ðh2  h1 Þ, W

(4)

_ comp is the total input electric power to the where W compressor, Zcomp is the general efficiency of the compressor, a Zcomp of 0.75 was used in the calculations for the DX-SAHPWH (A) system in the following text. The condensing heat rate Q_ cond can be expressed in terms of the enthalpy change as follows: _ r ðh2  h3 Þ. Q_ cond ¼ m

(5)

(6)

As the first-law evaluating indicator, the solar collector efficiency Zcoll and the coefficient of performance COP of the DX-SAHPWH system is defined as P _ Qeva tj j Zcoll ¼ P , (7) A c I T tj j

P _ Qw  tj COP ¼ P

j

_ comp  tj W

,

(8)

j

where t stands for the 2 minutes data acquisition interval mentioned above, The subscript ‘‘j’’ represents the jth segment of data acquisition interval, Ac is the solar collector area, IT is the total solar radiation intensity, Q_ w is the heat rate gained by water from condenser and it can be considered approximately as Q_ cond since the heat loss from the hot water to the ambience through water tank wall with polyurethane insulation thickness 38 mm can be neglected. 3.2. Second-law analysis A thermodynamic system should satisfy the exergy balance equation as _ E_ in ¼ E_ out þ I_rr þ DE, (9) where E_ in is the input exergy rate, E_ out is the output exergy rate, I_rr is the exergy loss rate and DE_ is the exergy change rate. The exergy balance equation for the DX-SAHPWH system can be expressed as X _ comp ¼ E_ _ þ E_ rad þ W I_rr;i , (10) Qw i

where E_ is the exergy rate, and its subscript ‘‘rad’’ stands for heat transfer for solar radiation and subscript ‘‘Q_ w ’’ P rate released from condenser to water. i I_rr;i stands for the total exergy loss rate of main components of the DX-SAHPWH system. The subscript ‘‘i’’ represents the ith component of the DX-SAHPWH system. The exergy loss rate calculation equations of each component are given as follows. The exergy loss rate in the collector/evaporator can be concluded as I_rr;eva ¼ E_ rad þ ½E_ 4;inf  E_ 1;outf      ¼ E_ rad þ H_ 4  H_ 1  T 0 S_ 4  S_ 1 ,

ð11Þ

where E_ rad is the exergy rate of the incoming solar radiation, E_ rad ¼ Ac I T Zcoll  ð1  T 0 =T p Þ [14], IT is the total solar radiation intensity, Tp is the temperature of the collector/evaporator panel surface.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1366

The exergy loss rate in the compressor can be concluded as _ comp þ E_ 1;inf  E_ 2;outf I_rr;comp ¼ W     _ comp þ H_ 1  H_ 2  T 0 S_ 1  S_ 2 . ¼W

ð12Þ

The exergy loss rate in the condenser can be concluded as I_rr;cond ¼ E_ 2;inf  E_ 3;outf  E_ Qw       T0 ¼ H_ 2  H_ 3  T 0 S_ 2  S_ 3  Q_ w 1  . ð13Þ Tw The exergy loss rate in the expansion valve can be concluded as     I_rr;v ¼ E_ 3;inf  E_ 4;outf ¼ H_ 3  H_ 4  T 0 S_ 3  S_ 4   ð14Þ ¼ T 0 S_ 4  S_ 3 . As for the second-law evaluating indicator, exergy efficiency does not appear to have been standardized. In this paper, the exergy efficiency of the DX-SAHPWH system is defined as   E_ Q_ w Q_ w 1  ðT 0 =T w Þ Zex ¼ ¼ . (15) _ comp þ E_ rad _ comp þ E_ rad W W The exergy loss rate in each component of the system can be examined more clearly by using the concept ‘‘exergy loss coefficient’’ w, which is defined as wi ¼

I_rr;i . _ comp þ E_ rad W

(16)

The correlation of the Zex and the wex can be concluded as Zex ¼ 1  wex ¼ 1 

X

wi :

(17)

i

4. Experimental results and discussions At present, a series of seasonal sunny day experiments were done in April 2005 under the meteorological conditions at Shanghai, to study the performance of DX-SAHPWH (A) system during the spring period. The experimental data are summarized in Table 2. It is indicated that it takes 90–98 min for heating 150 L water

from about 14–20 to 501C and the total electric consumption for compressor was 0.98–1.06 kWh. The average COP and Zcoll was 5.21–6.61 and 88–105%, respectively. So, it was proven that the Zcoll can be more than 1.0 when the evaporating temperature is lower than the ambient temperature. This characteristic is helpful to improve the performance of the solar collector/evaporator. A set of experimental data under typical working condition on April 22, 2005 are selected to analyze the time-dependent performance of the DX-SAHPWH (A) system. From data shown in Fig. 4, it can be concluded that the average value of solar radiation I¯ T and ambient temperature T¯ 0 are 812 W/m2 and 24.4 1C during the experimental period, respectively. And there was no large variation about the instantaneous ambient temperature, but the instantaneous solar radiation varied unpredictably because of the cloud. 4.1. COP and Zcoll of the DX-SAHPWH system It can be concluded from the experimental results on April 22, 2005, that the COP and the Zcoll of the DX-SAHPWH (A) system is 5.21 and 88%, respectively. It was found that after about 30 minutes, the running DX-SAHPWH (A) system comes into a so-called ‘‘transsteady-state working condition period’’ stage for about an hour until the end of measurements. The system running parameters, such as Tcond,i, Tcond,o, Teva,o, Teva,i, Tw, Tp and so on, are all increasing almost linearly during this period. To study time-varying performance of the DX-SAHPWH system, the entire trans-steady-state working condition period is divided into seven intervals in this paper. They are (a) former 1/8 period, (b) former 1/4 period, (c) former 1/2 period, (d) whole period, (e) later 1/2 period, (f) later 1/4 period, (g) later 1/8 period. The timeintegrated average COP and Zcoll of the DX-SAHPWH (A) system at (a)–(g) seven intervals were calculated using Eqs. (7) and (8) and plotted at seven mid-points of the corresponding intervals, respectively, as shown in Fig. 5. From Fig. 5, it is noted that Zcoll values of the DX-SAHPWH (A) system almost reach about 90%, they are much more than that of the conventional solar collector

Table 2 Experimental data of DX-SAHPWH (A) taken in spring No. Date, Local time, mm/dd/yy hh:mm

1 2 3 4 5

04/04/05 04/05/05 04/15/05 04/20/05 04/22/05

10:30–12:04 9:58–11:32 11:39–13:17 9:58–11:28 10:21–11:51

P

Dt, Tw,1, Tw,2, T¯ p , T¯ 0 , 1C I¯ T , min. 1C W/m2 1C 1C

¯ eva;out , P¯ cond , Thermodynamic states P¯ eva;in , P MPa MPa MPa (abs) (abs) (abs) T¯ 2 , T¯ 3 , T¯ 4 , T¯ 1 , 1C 1C 1C 1C

Kwh

94 96 98 90 90

1.039 0.928 0.888 0.980 0.963

0.9870.02 1.0070.02 1.0670.02 1.0070.02 1.0170.02

20.6 22.1 25.1 25.7 24.4

955 858 663 812 812

13.4 14.3 17.4 20.3 20.5

50.5 50.6 49.3 50.4 50.5

31.9 30.0 25.1 29.6 31.4

0.884 0.802 0.767 0.844 0.830

1.680 1.589 1.683 1.746 1.738

33.0 29.2 23.4 28.6 30.4

76.8 72.9 75.0 79.6 80.0

40.8 40.6 45.0 46.8 48.9

26.1 25.7 21.7 26.0 27.3

_ comp  tj , COP W

j

Zcoll, %

6.6170.34 9174.0 6.3670.33 9774.5 5.2670.29 10575.8 5.2670.29 8874.8 5.2170.29 8874.8

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1367

900

28

800

26

700

24

IT (W/m2)

30

IT T0

600 500

T0 (°C)

April 22, 2005 1000

22

10:16 10:26 10:36 10:46 10:56 11:06 11:16 11:26 11:36 11:46 11:56 Local time (hh:mm)

20

Fig. 4. Variation of IT and T0 with time for the DX-SAHPWH (A) system.

April 22, 2005

12

55 50 45

8

40

6

35

4

ηcoll

30

2

COP Tw

25

0

30

40

50

60 Time (minutes)

70

80

90

Tw (°C)

10·coll & COP

10

20

Fig. 5. Variation of Zcoll, COP and Tw with time for the DX-SAHPWH (A) system.

April 22, 2005

Pi (MPa)

3 Pcond,out Peva,in Peva,out

2

1

0

0

10

20

30

40 50 Time (minutes)

60

70

80

90

Fig. 6. Variation of Pcond, Peva,in and Peva,out with time for the DX-SAHPWH (A) system.

and its variation range is not great. But the COP values generally decrease mainly because of the gradual increase of Tw and Tcond. It is worth noticing that Tw in Fig. 5 is transient measured value. As to experimental operation, once the measured value of Tw reached 55 1C, the compressor is switched off and the circulating water pump is switched on to mix the stratified hot water in tank. According to experimental results, the actual equilibrium water temperature Tw,2 was about 4.8 1C lower than the measured one. As above mentioned, in order to measure the pressure drop in the collector/evaporator, there were two low-

pressure transducers at inlet and outlet of the collector/ evaporator, respectively. A high-pressure transducer was located at outlet of the condenser to measure the condensing pressure. These three pressure values were plotted in Fig. 6. It is noticed from Fig. 6 that both Pcond and Peva increase with the increase of Tw, and the increasing rate of Pcond is higher. The compressor exhaust temperature is higher and higher (e.g. the discharge temperature is 90.4 1C when the actual Tw,2 reaches 50.5 1C). So, the excessive terminal setting value of Tw will do harm to the stability and reliability of the DXSAHPWH system.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1368

χeva χ comp χ

χv  ex

cond

30

40

50

60 Time (minutes)

70

80

90

50 45 40 35 30 25 20 15 10 5 0

100·ex

100·χi

April 22, 2005 100 90 80 70 60 50 40 30 20 10 0

Fig. 7. Variation of wi and Zex with time for the DX-SAHPWH (A) system.

4.2. Irreversibilities in the system components Exergy loss coefficient of each main component and exergy efficiency of the DX-SAHPWH (A) system at above-mentioned seven mid points can be calculated using Eqs. (9)–(17) with state parameters determined by experimental data. The results are plotted in Fig. 7. In Fig. 7, it is shown that the Zex of the DX-SAHPWH (A) system increases with the increase of Tw. This is due to that the lower-grade thermal energy of the surrounding environment is upgraded by the DX-SAHPWH (A) system to the higher-grade thermal energy in the form of hot water. However, wi of each main component (expect for that of the TEV) decreases gradually. It is shown from Fig. 7 that the highest exergy loss occurs in the compressor, followed by the collector/evaporator, condenser and expansion valve, respectively. These calculation results justify the views of Chaturvedi et al. [5,6], Hawlader et al. [22], and Kuang et al. [25], that the proper matching between components of the DX-SAHP system, especially between the heat pumping capacity of the compressor and the evaporative capacity of the collector, is vital for improving its performance. The highest exergy loss occurs in the compressor and this fact firstly indicates that the compressor selection is not optimal restricted by the specifications of available products. Secondly, since compressor power consumption depends strongly on the inlet and outlet pressures, any methods will reduce compressor exergy loss by bringing the condensing and evaporating temperatures closer each other [26]. So, the higher solar radiation, the higher ambient temperature and the lower terminal setting value of Tw are all advantageous for improving the performance of the compressor. It is suggested, therefore, that the running period should be set in a period near the noon and the terminal setting value of Tw should not exceed more than 60 1C. Considering the DX-SAHPWH system is for supplying domestic hot water, excessive high water temperature is not necessary, 45–50 1C terminal setting value Tw,2 is recommended. The collector/evaporator operating at relatively lower temperatures results in more exergy loss. The higher evaporating temperature should be pursued if possible.

An obvious advantage of the DX-SAHP system is its higher evaporating temperature caused by incident solar energy. The refrigerant flow path in the collector/evaporator of the DX-SAHPWH (A) system is complicated as shown in Fig. 1. It can be noted from Fig. 6 that the pressure loss between the inlet and the outlet is about 0.133 MPa (corresponding temperature drop is about 5.5 1C). This pressure loss also results in exergy loss. Methods to decrease the pressure loss can improve the exergy performance of the collector/evaporator. In the following text, a new type collector/evaporator with lower pressure loss used in the DX-SAHPWH (B) system is introduced. The heat loss between the hot water in tank and the ambient air must result in exergy loss. In this study, the heat loss is ignored. So the calculated exergy loss of condenser in Fig. 7 is a bit lower than the actual value. We may decrease the exergy loss by increasing thermal insulation performance of water tank wall, enhancing heat exchange and optimizing the configuration and position of the condenser in the water tank. With increasing Tw, the compression ratio increases gradually as shown in Fig. 6, and the throttling loss increases too. The lower terminal setting value Tw,2, therefore, is advantageous. The calculated exergy loss coefficient of TEV is about 2–14%. 4.3. The optimum evaporating temperature The heat absorbed from the ambience for the collector/ evaporator (Q_ eva ) also can be calculated as [20,21] Q_ eva ¼ Ac ½S  U Lc ðT p  T 0 Þ ¼ Ac F 0 ½S  U Lc ðT eva  T 0 Þ, (18) where Ac is the collector area, S is the difference between the solar radiation absorbed by the collector/evaporator and the net radiation heat loss from the collector/ evaporator surface at the ambient temperature, S ¼ aI T  q0 , a is the absorptivity of collector/evaporator, e is the emissivity of the collector/evaporator plate, q0 is the difference between the emissive power per unit area

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

from a black body at the ambient air temperature and the emissive power from sky, q0 ¼ sT 40  q1 , qN is the emissive power from sky q1 ¼ sT 4Sky ; T Sky is the sky temperature, T Sky ¼ 0:0552T 1:5 [23], ULc is collector/ 0 evaporator heat loss coefficient, U Lc ¼ hw þ 4sT 30 , hw is the wind heat transfer coefficient, hw ¼ 5:7 þ 3:8V w [20,21], Vw is the wind velocity at ambient environment, s is the Stefan–Bolzmann constant, T is the temperature, the subscript ‘‘p’’ stands for collector plate, ‘‘eva’’ for evaporating refrigerant and ‘‘0’’ for the ambience, F0 is the collector efficiency factor. The exergy gain by the collector/evaporator is given by the heat absorbed from the incoming solar radiation times ZCarnot [14]. The exergetic efficiency of the collector/ evaporator can be defined as Zexcoll ¼ F 0 ½S  U Lc ðT eva  T 0 Þ ½1  T 0 =T eva Þ=I T ½1  T 0 =T p .

ð19Þ

Taking the derivative of Zex-coll with respect to Teva, we find the maximum exergetic efficiency can be obtained at 1=2

T eva;optimum ¼ T 0 ½ðS=U Lc Þ þ T 0 1=2 .

(20)

The second term in the right-hand side of Eq. (20) will be referred to as sol–air temperature T sa ¼ ðS=U Lc Þ þ T 0 .

(21)

Combining Eqs. (20) and (21), we obtain T eva;optimum ¼ ðT 0 T sa Þ1=2 .

(22)

As for the experimental data on April 22, 2005, the average ambient temperature and the average evaporating temperature is 24.4 and 19.8 1C, respectively. The average optimum evaporating temperature is calculated to be 39.1 1C. As shown in Fig. 8, the difference between the Teva,optimum and the Teva decreases with increasing Tw. This is accordant to the fact that the exergy loss coefficient of the collector/evaporator is dropping gradually in Fig. 7.

1369

4.4. The design optimization of collector/evaporator As calculated above, the optimum evaporating temperature is much higher than the actual evaporating temperature. So the higher evaporating temperature should be required. But the actual evaporating temperature depends upon the design configuration and the operating conditions. However, the twin goals of efficiency and reliability of this system have a conflicting object Teva. The compressor exhaust temperature may be unacceptable with too high value of Teva. As Chaturvedi et al.’s point of view, Teva may be higher or lower than T0, depending upon the design (system matching) and the weather condition [4]. Chaturvedi et al. further shows that an SAHP using a bare collector and a variable-frequency compressor has an optimum performance (based on maximization of exergy efficiency of the collector/evaporator) provided the evaporator temperature is maintained in a temperature range of 5–10 1C above ambient [6]. A methodology for determination of an ‘‘optimal evaporation temperature’’ proposed by Torres-Reyes et al. was also based on maximization of exergy efficiency of the collector/evaporator [14]. Ito [20,21] and Hawlader [22] also designed an SAHP with a bare collector operating at Teva4T0, respectively. But Huang [18] consider that an SAHP operating at TevaoT0 has an advantage of having lower compressor exhaust temperature and dual heat source from both solar radiation and ambient air (with higher Zcoll). In this experimental study, Teva is designed as about 5 1C below ambient (TevaoT0) according to the limit of the compressor suction pressure. As shown in Fig. 6, the pressure loss between the inlet and the outlet is large because of the complicated refrigerant flow path in the collector/evaporator. So, the flow path in the collector/evaporator would be designed properly so as to reduce the large pressure loss of the refrigerant flow. The procedure of design calculation for the collector/ evaporator is described as follows. Provided the average ambient temperature is 25 1C, the average solar radiation intensity is 800 W/m2 and the

Ti (°C)

April 22, 2005 45 40 35 30 25 20 15 10 5 0

Teva,optimum T0 T eva 0

10

20

30

40 50 Time (minutes)

60

70

80

90

Fig. 8. Variation of Teva,optimum, Teva and T0 with time for the DX-SAHPWH (A) system.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1370

average wind velocity is 3.1 m/s in spring at Shanghai, and a and e were both about 0.9 for the collector/evaporator used in this experiment. Then the S and ULc can be calculated using Eq. (18) as 650 W/m2 and 22.9 W/m2 K, respectively. Combing Eqs. (7) and (18) gives Zcoll ¼

F0 ½S  U Lc ðT eva  T 0 Þ. IT

(23)

Assumed the (TevaT0) and Zcoll is 5 1C and 90%, respectively, then F0 can be calculated as about 0.94. The collector efficiency factor can be expressed as [20] F 0 ¼ F þ ð1  F ÞðD=W Þ,

(24)

where F is the fin efficiency, F ¼ tanh U b =U b , pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi U b ¼ ½ðW  DÞ=2 U Lc =ðlP  dP Þ, W is the pitch of the tubes, D is the outer diameter of the tube, lP is the conductivity of the collector/evaporator plate, the material of the collector/evaporator is recommended as aluminum which can be easily shaped, so lP ¼ 236 W/m K, dP is the thickness of the collector/evaporator plate, 1.60 mm is suggested as same as that in the DX-SAHPWH (A) system. Then, D and W can be confirmed as 12.0 and 130.0 mm, respectively. Lastly, Ac can be determined by Eq. (18) once Qeva is requested. In theory, the schematic of the optimum designed collector/evaporator is shown in Fig. 9. However, under the limitation of the available products in the market, the actual collector/evaporator utilized in present study is made of 8 solar collector slab cores as shown in detail in Fig. 10. The total collector area is 2.08 m2. The collector/evaporator plate is a 0.18 mm thick copper sheet, with its surface black painted by solar selective coating. A copper tube, having inner and outer diameters of 11.0 and 12.0 mm, respectively, was soldered to the backside of the

copper sheet, with a pitch between tubes of 140.0 mm. The collector/evaporator was installed on the roof, facing south with an inclination angle of 31.221. Based on the experiment and analysis on the DX-SAHP (A), a small type DX-SAHPWH (B) system with 400 W input power is developed to optimize the system performance and economic cost. In this study, Zcomp of 0.60 was used in the calculations for the DX-SAHPWH (B) system. The specification of each main component of the DX-SAHPWH system (B) is listed in Table 3. 4.5. Experimental comparison between the DX-SAHPWH (A) and (B) system To examine the performance of the DX-SAHPWH (B) system during autumn period and compare it with that of the DX-SAHPWH (A) system, series of sunny day experiments were done in October 2005 under the meteorological conditions at Shanghai. The experimental data are summarized in Table 4. With the contrast of the data in Tables 2 and 4, it can be noted that the small type DX-SAHPWH (B) system need much longer time than the DX-SAHPWH (A) system at similar ambient conditions and hot water request. However, both DX-SAHPWH (A) and (B) need electric energy consumption no more than 1 kWh. At this similar running cost, the lower capital cost and smaller collector/evaporator area, which is advantageous for integrating it with buildings roof or walls, makes the benefits of the small type DXSAHPWH (B) system over the DX-SAHPWH (A) system. A set of experimental data under typical working condition on October 15, 2005 are selected to analyze the time-dependent performance of the DX-SAHPWH (B) system so as to compare it with the data for the DX-SAHPWH (A) system on April 22, 2005, for further analysis.

δp

D

W

Fig. 9. Schematic diagram of the design optimization for collector/evaporator.

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1371

R22

Copper sheet R22 Copper tube Fig. 10. Schematic diagram of collector/evaporator in the DX-SAHPWH (B) system.

Table 3 Specification of the main components of the DX-SAHPWH (B) system Name

Type

Remarks

Compressor Domestic water tank

Rotary Pressure resistance and heat insulation Slab cores made of copper sheets and tubes TEX-2 type, manufactured by Danfoss, Denmark

Rated input power: 0.40 kW, displacement volume:7.40 cm3/rev 150 L water, immersed 50 m copper coil (f9.90  0.75 mm) as condenser

Solar collector/evaporator Thermostatic expansion valve

8 slab cores (0.18 mm thick copper sheet soldered on f12.0  0.50 mm copper tube) connected in series, total collector/evaporator area: 2.08 m2 External balance type

Table 4 Experimental data for the DX-SAHPWH (B) system taken in autumn No. Date, Time, mm/dd/yy hh:mm

1 2 3 4 5

10/15/05 10/16/05 10/17/05 10/18/05 10/19//05

10:30–13:14 10:33–13:31 10:37–13:58 10:30–13:19 10:00–12:52

Dt, T¯ 0 , min. 1C

164 178 201 169 172

24.7 26.0 27.3 30.6 29.9

Tw,1, Tw,2, T¯ p , I¯ T , 1C W/m2 1C 1C

P¯ eva;in , P¯ eva;out , P¯ cond , Thermodynamic states MPa MPa MPa (abs) (abs) (abs) T¯ 2 , T¯ 3 , T¯ 4 , T¯ 1 , 1C 1C 1C 1C

795 689 557 829 859

0.751 0.724 0.756 0.848 0.855

24.4 24.0 23.6 23.4 23.4

49.9 50.0 50.0 50.0 50.1

35.7 32.7 30.4 38.8 38.7

0.668 0.642 0.672 0.760 0.767

From data shown in Fig. 11, it can be concluded that the average value of the solar radiation I¯ T and ambient temperature T¯ 0 is 795 W/m2 and 24.7 1C during the experimental period, respectively. As similarly to Fig. 4, the ambient temperature was almost unchanged. However, the instantaneous solar radiation varied with a sinusoid rule without the effect of cloud.

1.661 1.611 1.605 1.591 1.592

17.2 14.7 14.3 19.1 19.3

63.4 58.5 48.4 51.4 50.8

39.3 39.5 41.5 41.5 41.6

15.5 14.0 15.2 19.5 19.7

P

_ comp  tj , COP W

j

Zcoll, %

kWh

0.9070.04 0.9870.04 1.1270.05 0.9170.04 0.9370.04

4.9770.42 8675.5 4.6570.39 9376.1 4.1370.34 10177.1 5.1270.43 8475.1 5.0170.42 8074.9

It can be concluded from the experimental results on October 15, 2005 that the COP and the Zcoll of the DX-SAHPWH (B) system is 4.97 and 86, respectively. As shown in Fig. 12, the variable characteristics of Zcoll, COP and Tw for the DX-SAHPWH (B) system was similar to that for the DX-SAHPWH (A) system. The main difference is that, after about only 10 minutes,

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1372

30

900

28

800

26 24

700

IT

600

T0

500

T0 (°C)

IT (W/m2)

October 15, 2005 1000

22 20

10:38 10:48 10:58 11:08 11:18 11:28 11:38 11:48 11:58 12:08 12:18 12:28 12:38 12:58 13:08

Local time (hh:mm)

Fig. 11. Variation of IT and T0 with time for the DX-SAHPWH (B) system.

October 15, 2005 55

10

50

8

45 40

6 4 2 0

35

ηcoll COP Tw

Tw (°C)

10·coll & COP

12

30 25

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time (minutes)

20

Fig. 12. Variation of Zcoll, COP and Tw with time for the DX-SAHPWH (B) system.

DX-SAHPWH (B) system comes into trans-steady-state working condition stage. The COP and Zcoll of the DXSAHPWH (B) is also smaller than that of the DXSAHPWH (A) system with contrast between Figs. 5 and 12. The Tw in Fig. 12 is also transient measured value. As to experimental operation, once the measured value of Tw reached 54 1C, the compressor is switched off and the circulating water pump is switched on to mix the stratified hot water in tank. According to experimental results, the actual equilibrium water temperature Tw,2 was about 4.0 1C lower than the measured one. The pressure (temperature) drop in the collector/ evaporator of the DX-SAHPWH (B) is much smaller than that of the DX-SAHPWH (A) by comparing the data in Tables 2 and 4. It can be noted from data shown in Fig. 13 that the pressure loss between the inlet and the outlet is about 0.083 MPa (corresponding temperature drop is about 3.9 1C). Contrast to the DX-SAHPWH (A) system, the pressure drop decreased by about 38% and corresponding temperature drop decreased by about 29% by comparing data shown in Fig. 6 and Fig. 13. This is helpful to improve the efficiency of the collector/evaporator. Using the same methodology as above, the exergy loss coefficient for each component and exergy efficiency for the DX-SAHPWH (B) system at seven mid points previously mentioned was calculated and potted in Fig. 14. Unlike to the DX-SAHPWH (A) system, the highest exergy loss occurs in the collector/evaporator, but no longer in the compressor.

As for the experimental data on October 15, 2005, the average ambient temperature and the average evaporating temperature is 24.7 and 8.2 1C, respectively (this Teva is 59% lower than 19.8 1C of the DX-SAHPWH (A)). The average optimum evaporating temperature is calculated to be 38.7 1C. As shown in Fig. 15, the difference between Teva,optimum and Teva also decreases gradually. Although the problems of the pressure drop along the refrigerant flow path in the collector/evaporator were improved, the structure of the collector/evaporator in the DX-SAHPWH (B) system is ‘‘sheet-on-tube’’ instead of ‘‘sheet-cross-tube’’ as design optimization because of the limitation of the available products in the market. This ‘‘sheet-on-tube’’ structure type makes the heat exchange between the collector/evaporator and the ambient is not good enough. It can be calculated that the F0 is only about 0.67 from experimental data on October 15, 2005 in Table 4. 4.6. Methods for further improvement With the variation in heating load and climate, seasonal and all-year working condition of the DX-SAHPWH system is continuously changing. The mismatching between each component (especially the compressor and the collector/evaporator) of the DX-SAHPWH system caused by the variable working condition will also results in the exergy losses. The exergy losses reduction can be achieved if the difference between theTeva,optimum and the Teva is reduced by

ARTICLE IN PRESS Y.W. Li et al. / Energy 32 (2007) 1361–1374

1373

October 15, 2005 3

Pcond,out

Pi (MPa)

Peva,in 2

Peva,out

1

0

0

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time (minutes)

Fig. 13. Variation of Pcond, Peva,in and Peva,out with time for the DX-SAHPWH (B) system.

χ eva χ

comp

χ cond

0

50 45 40 35 30 25 20 15 10 5 0

χv 

ex

100·ex

100·χi

October 15, 2005 100 90 80 70 60 50 40 30 20 10 0

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time (minutes)

Fig. 14. Variation of wi and Zex with time for the DX-SAHPWH (B) system.

Ti (°C)

October 15, 2005 45 40 35 30 25 20 15 10 5 0

Teva,optimum T0 Teva

0

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Time (minutes)

Fig. 15. Variation of Teva,optimum, Teva and T0 with time for the DX-SAHPWH (B) system.

the automatic control of the refrigerant flow into the collector/evaporator, using the ‘‘sol–air temperature’’ as the monitoring parameter during the operation [15]. A higher evaporating temperature is pursued if possible, without exceeding the desired design suction pressure limit of the compressor. However, the twin goals of moderately high COP and reliability of the DX-SAHPWH system have a conflicting object Teva. How to determine the moderate Teva is worth emphasizing according to the actual conditions. To maintain a high COP and system reliability, and to control the object Teva to a moderate value, a variable

speed compressor and an electronic expansion valve with a controller would be included in this DX-SAHPWH system in the future. Some proper control methods will be applied to effectively combine these two controllable components to maintain a proper matching between the heat pumping capacity of the compressor and the evaporative capacity of the collector/evaporator under widely varying ambient conditions all year. Then, the higher exergy loss occurring in the compressor and collector/evaporator can be also positively reduced. Further, the thermal performance and the reliability of the DX-SAHPWH system can be improved.

ARTICLE IN PRESS 1374

Y.W. Li et al. / Energy 32 (2007) 1361–1374

5. Conclusions The performance of the DX-SAHPWH (A) system (with rated input power 750 W) was shown through experimental research in spring and thermodynamic analysis and some suggestions for the system design optimization are proposed. Then, a small-type DX-SAHPWH (B) system (with rated input power 400 W) with new type collector/ evaporator was built, tested, analyzed and then compared with the DX-SAHPWH (A) system. Both the DX-SAHPWH (A) and (B) system need electric energy consumption almost no more than 1 kWh. Under similar running cost expected, the smaller system with lower capital cost and smaller collector/evaporator area has the advantage so as to integrate it with buildings roof. Through exergy analysis for each component of the DX-SAHPWH (A) and (B) systems, it is concluded that the highest exergy loss occurs in the compressor and collector/ evaporator, followed by the condenser and expansion valve. Furthermore, some methods are suggested to improve the performance of each component, especially for the collector/evaporator. A methodology for the design optimization of the collector/evaporator was adopted and applied. Under variable working conditions and actual limitation of the available products in the market, the selection of the compressor and collector/evaporator usually could not be perfectly matching as design. In order to maintain a proper matching between the heat pumping capacity of the compressor and the evaporative capacity of the collector/ evaporator under widely varying ambient conditions, the variable frequency compressor and electronic expansion valve will be adopted for the DX-SAHPWH system to be built in the next work. Acknowledgments This work was partly supported by Ministry of Science and Technology of China under the contract no. 2005B A908B 07, Shanghai Commission of Science and Technology under the contract no. 05dz 05807. Authors are particularly grateful to Mr. Qiang Ma for his technical support and to Mr. Yunkang Sun, Mr. Taihua Wang for their helps in the experimental work. References [1] Sporn P, Ambrose ER. The heat pump and solar energy. In: Proceedings of the world symposium on applied solar energy 1955. November 1–5 [in Phoenix, Ariz]. [2] Freeman TL, Mitchell JW, Audit TE. Performance of combined solar-heat pump systems. Solar Energy 1979;22:25–125.

[3] Chaturvedi SK, Roberts AS, Mei V. Solar collector as heat pump evaporator. In: Proceedings of 13th intersociety energy conversion conference, Boston, MA, 1979. August 4–9. [4] Chaturvedi SK, Shen JY. Thermal performance of a direct expansion solar-assisted heat pump. Solar Energy 1984;33:155–62. [5] Chaturvedi SK, Abazeri M. Transient simulation of a capacitymodulated, direct-expansion, solar-assisted heat pump. Solar Energy 1987;39:421–8. [6] Chaturvedi SK, Chen DT, Kheireddine A. Thermal performance of a variable capacity direct expansion solar-assisted heat pump. Energy Convers Manage 1998;39(3/4):181–91. [7] Aziz W, Chaturvedi SK, Kheireddine A. Thermodynamic analysis of two-component, two-phase flow in solar collectors with application to a direct-expansion solar-assisted heat pump. Energy 1999;24: 247–59. [8] Morgan RG. Solar assisted heat pump. Solar Energy 1982;28: 129–35. [9] Krakow KL, Lin SA. solar source heat pump with refrigerant-cooled solar collectors for cold climates. Trans ASHRAE 1982;88:417–39. [10] O’Dell MP, Mitchell JW, Beckman WA. Solar heat pump systems with refrigerant-filled collectors. Trans ASHRAE 1983;89:519–25. [11] Macarthur JW. Theoretical analysis of the dynamic interactions of vapor compression heat pump. Energy Convers Manage 1984;24: 49–66. [12] Morrison GL. Simulation of packaged solar heat-pump water heaters. Solar Energy 1994;53:249–57. [13] Axaopoulos P, Panagakis P, Kyritsis S. Experimental comparison of a solar-assisted heat pump vs. a conventional thermosyphon solar system. Int J Energy Res 1998;22:1107–20. [14] Torres-Reyes E, Picon-Nunez M, Cervantes-De GJ. Exergy analysis and optimization of a solar-assisted heat pump. Energy 1998;23: 337–44. [15] Cervantes JG, Torres-Reyes E. Experiments on a solar-assisted heat pump and an exergy analysis of the system. Appl Thermal Eng 2002;22:1289–97. [16] Axaopoulos P, Panagakis P, Kyritsis S. Experimental comparison of a solar-assisted heat pump vs. a conventional thermosyphon solar system. Int J Energy Res 1998;22:1107–20. [17] Huang BJ, Chyng JP. Integral-type solar-assisted heat pump water heater. Renew Energy 1999;16:731–4. [18] Huang BJ, Chyng JP. Performance characteristic of integral type solar-assisted heat pump. Solar Energy 2001;71:403–14. [19] Chyng JP, Lee CP, Huang BJ. Performance analysis of a solarassisted heat pump water heater. Solar Energy 2003;74:33–44. [20] Ito S, Miura N, Wang K. Performance of a heat pump using direct expansion solar collector/evaporators. Solar Energy 1999;65(3): 189–96. [21] Ito S, Miura N, Takano Y. Studies of heat pumps using direct expansion type solar collectors. J Solar Energy Eng 2005;127(2):60–4. [22] Hawlader MNA, Chou SK, Ullah MZ. The performance of a solar assisted heat pump water heating system. Appl Thermal Eng 2001;21: 1049–65. [23] Swinbank WC. Long-wave radiation from clear skies. QJR Meteorol Soc 1963;89:339–40. [24] Kuang YH, Wang RZ, Yu LQ. Experimental study of solar assisted heat pump system for heat supply. Energy Convers Manage 2003;44(7):1089–98. [25] Kuang YH, Sumathy K, Wang RZ. Study on a direct-expansion solar-assisted heat pump water heating system. Int J Energy Res 2003;27:531–48. [26] Ozgener O, Hepbasli A. Experimental performance analysis of a solar assisted ground-source heat pump greenhouse heating system. Energy Buildings 2005;37:101–10.