Solar Energy 79 (2005) 459–468

www.elsevier.com/locate/solener

Experimental investigation of a two-phase closed

thermosyphon solar water heater
Mehmet Esen *, Hikmet Esen
Faculty of Technical Education, Department of Mechanical Education, Fırat University, 23119 Elazığ, Turkey

Received 30 January 2004; received in revised form 20 December 2004; accepted 6 January 2005
Available online 23 February 2005

Communicated by: Associate Editor Brian Norton

Abstract

In this study, experiments were performed to find out how the thermal performance of a two-phase thermosyphon
solar collector was affected by using different refrigerants. Three identical small-scale solar water heating systems, using
refrigerants R-134a, R407C, and R410A, were constructed and tested side-by-side under various environmental and
load conditions. The performance of the system under clear-sky conditions has been investigated with and without
water load. Detailed temperature distributions and cumulative collection efficiencies were determined and presented.
The experimental results were compared to the results found in the literature and they showed good agreement.

2005 Elsevier Ltd. All rights reserved.

Keywords: Phase-change; Heat-pipe solar collector; Refrigerant-charged solar collector; Solar water heating

1. Introduction of the water on cold nights, pipe corrosion due to the use
of water and the limited quantity of heat transferred by
Usually, conventional solar collectors use pipes at- the heat transfer fluid. Heat pipes offer a promising solu-
tached to the collecting plate and a heat transfer fluid, tion to these problems. Heat pipes working under grav-
such as water, to transfer by natural or forced circula- ity with the condenser above the evaporator do not
tion the heat captured by the solar collector to a storage require external power or capillary action in order to re-
tank. Some of the shortcomings associated with conven- turn the heat transfer fluid to the evaporator. This type
tional solar collectors include extra expense of the forced of heat pipe is known as the gravity assisted heat pipe,
circulation system due to the pump and its extracted closed two-phase thermosyphon or wickless gravity as-
power, extra space required for the natural circulation sisted heat pipe. Closed two-phase thermosyphon solar
system due to the position limitations required, the night collectors consist of heat pipes (or thermosyphon tubes)
cooling due to the reverse flow of cooled water, freezing filled with a refrigerant and used in closed loop water
heating systems. Closed loop solar heating systems are
*
Corresponding author. Tel.: +90 424 2370000; fax: +90 424 suitable for domestic solar water heating, solar swim-
2367064. ming pool heating or solar space heating systems. They
E-mail addresses: mesen@firat.edu.tr (M. Esen), hikmetesen are also suitable for areas with questionable water qual-
@firat.edu.tr (H. Esen). ity and all climatic conditions. Furthermore, they are the

0038-092X/$ - see front matter
2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.solener.2005.01.001
460 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468

Nomenclature

A area of each collector absorber plate (m2) Ta,ave daily average ambient air temperature (C)
Cp,w water specific heat (J/kg C) Tw1 water temperature at the beginning of the
I global irradiance incident on solar collector day (C)
(W/m2) Tw2 water temperature at the actual time (C)
mw water mass inside each water storage tank Tp,ave daily average absorber plate temperature
(kg) (C)
Qcol thermal energy collected (kJ) Tw water temperature (C)
Qincident total solar irradiance incident on solar col- Tw,ave daily average water temperature (C)
lector (kJ) Tw,max daily maximum water temperature (C)
t time (s) g cumulative collection efficiency
Ta ambient air temperature (C) gmax daily maximum collection efficiency

preferred option for extremely cold areas. The basic heat ing is eliminated. Akyurt (1984) designed and manuf-
pipe is a closed container consisting of a small amount actured numerous heat pipes. Each heat pipe was
of vaporizable fluid (e.g., a refrigerant). The heat pipe incorporated into a prototype solar water heater. An
employs an evaporating-condensing cycle, which accepts extensive testing program lasting for more than a year
heat from an external source, uses this heat to evaporate revealed that the heat pipes perform satisfactorily as
the liquid (latent heat) and then releases latent heat by heat transfer elements in solar water heaters. Fanney
reverse transformation (condensation) at a heat sink and Terlizzi (1985) conducted experiments to determine
region. This process is repeated continuously by a return the effect of irradiance level on the thermal performance
feed mechanism of the condensed fluid back to the heat- of the refrigerant-charged solar domestic hot water sys-
ing zone. In the solar collector, the condensation zone is tem, and concluded that it is negligible for the irradiance
at a higher level than the evaporation zone. The trans- levels considered. Radhwan et al. (1990) investigated
port medium condensed (in the condensation zone) experimentally the thermal performances of two R-11
returns to the evaporation zone under the influence of charged integrated solar water heaters for forced and
the gravity. The maximum operating temperature of a natural circulation water flows. The results showed that
heat pipe is the critical temperature of the used heat the inclination of the condenser integrated within the
transfer medium. The low freezing points of the refriger- collector frame had a remarkable effect for natural circu-
ants and the use of latent heat to transfer energy from lation water flow, while it had no significant effect for
one place to another are of special interest in solar forced circulation flow. Hammad (1995) designed and
applications. manufactured a flat plate solar collector cooled by a
A number of studies have been previously conducted set of heat pipes to work at low temperature conditions
in which various phase change fluids or refrigerants have equal to that of flat plate solar collectors. Efficiencies
been evaluated for use in a thermosyphon solar water and quantities of energy transferred to a water storage
heater. Soin et al. (1979, 1987) investigated the thermal tank were calculated and found to be comparable to
performance of a thermosyphon collector containing those obtained by using a water-cooled solar collector.
boiling acetone and petroleum ether, and presented the Pluta and Pomierny (1995) studied a two-phase solar
effect of insolation and the liquid level on the collector thermosyphon for domestic hot water, and indicated
performance. Schreyer (1981) experimentally investi- that proper construction and choosing suitable phase-
gated the use of refrigerant R-11 in a thermosyphon change medium play very important role in assuring
solar collector for residential applications. He found proper operating conditions of a phase-change ther-
that, for two identical collectors, the peak instantaneous mosyphon. Ghaddar and Nasr (1998) fabricated and
efficiency of a boiling refrigerant charged collector was tested an R-11 charged solar collector with an integrated
6% greater than that of a hydronic fluid circulating solar condenser for secondary-cooling water flow. Forced
collector. Bottum (1981) based upon engineering tests circulation flow demonstrated instantaneous system
and more than four years of observation in the field of efficiency values varying from 60% to 20%, which is in
both hydronic and phase change domestic water heating range of conventional water solar collectors. Chun et
systems, concluded that hydronic systems do work very al. (1999) developed experimentally a solar domestic
well when properly installed and maintained, but refrig- hot water system using heat pipes. A number of sys-
erant charged systems are more efficient, usually require tems with different configurations and working fluids
less maintenance, and the possibility of collectors freez- (such as acetone, methanol, ethanol, distilled water,
 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468 461

and ethanol–distilled water mixture) were constructed ing directly to the oven plate, to cook and/or to keep
and their performances were measured to elicit the most food warm in the late evening. It was demonstrated that
feasible model considering the various limiting condi- short cooking times can be obtained with a solar cook-
tions present. Joudi and Al-Tabbakh (1999) conducted ing system consisting of vacuum-tube collector with inte-
a theoretical analysis by computer simulation of a two- grated long heat pipes using refrigerant as working fluid.
phase thermosyphon solar domestic hot water system As seen from the literature review mentioned above,
using R-11. The computer program and calculation pro- very few environmentally friendly working fluids or
cedure were first validated by comparing the results with refrigerants (e.g., R-134a) have been tried and tested
established results of single phase systems. Then, calcu- for two-phase thermosyphon solar collectors. The result-
lations were performed for the two-phase thermosyphon ing environmental concerns about ozone depletion and
system. The results showed that, in the two-phase sys- global warming have stopped the production of CFC
tem, the collector efficiency did not reveal a serious refrigerants and the production of HCFCs will stop in
change with the loading condition. The collector effi- 2015. Therefore in the present study a two-phase closed
ciency of the two-phase system was approximately thermosyphon flat-plate solar collector with different
20% higher than a single phase collector. Also, the re- environmentally friendly refrigerants was designed, con-
sponse of the two-phase system was faster than a single structed and tested under real conditions with the system
phase system. Hussein et al. (1999a,b, 2001) and Hussein installed outdoors. The primary objective of the current
(2002) investigated theoretically and experimentally a work is to experimentally study the feasibility of using
thermosyphon flat-plate solar collector. The transient refrigerant as a working fluid for a two-phase thermosy-
thermal behaviour of wickless heat pipe flat plate solar phon solar collector and to compare the system perfor-
collectors was analyzed with regard to various parame- mance using three different environmentally friendly
ters such as global solar radiation intensity, inlet cooling refrigerants. The novelty of the system investigated in
water temperature, absorber plate material and thickness, the present experimental study consists in the test of dif-
ratio of pitch distance to wickless heat pipe diameter and ferent refrigerants simultaneously under same working
ratio of condenser section length to total wickless heat conditions such as solar radiation, ambient air temper-
pipe length. The pitch distance was defined as width of ature, wind velocity, collector materials and dimensions
absorber plate. The results revealed that the pitch dis- to determine the best refrigerant should be used as work-
tance limits the selection of an absorber plate having a ing fluid for a two-phase thermosyphon solar collector.
high value of thermal conductivity. Also, from the theo- The experimental results are analyzed and discussed in
retical analysis, it was concluded that the condenser sec- detail.
tion dimensionless ratio and heat pipe inclination angle
had a significant effect on the condensation heat-transfer
coefficient inside the inclined wickless heat pipes. Esen 2. Experimental set-up and test procedure
and Yüksel (2000) manufactured a two-phase thermosy-
phon solar collector with heat pipes, and studied expe- A scheme of the refrigerant-charged flat plate solar
rimentally various phase-change fluids (acetone, collector with wickless heat pipes under consideration
methanol, ethanol) to evaluate the effect of insolation is shown in Fig. 1. To evaluate the thermal performance
and mass of fluid on the collector performance. It was of two-phase thermosyphons with different refrigerants
found out that such a collector can be successfully used, under same working conditions such as solar radiation,
especially during cold, cloudy and windy days. Paya- ambient air temperature, wind velocity, collector materi-
karuk et al. (2000) studied the inclined thermosyphons. als and dimensions, three absorber plates with heat pipes
Thermosyphons were employed with R-22, R-123, R- were contained within the same collector case.
134a, ethanol, and water. The selected filling ratios were Each absorber plate contains two heat pipes filled
50%, 80%, and 100%. It was found from the experiments with the working fluid. Thus, three different refrigerants
that the filling ratio had no effect on the ratio of heat can be simultaneously employed under identical operat-
transfer characteristics at any angle to that of the verti- ing conditions. Two-phase thermosyphons (i.e., wickless
cal position, but the properties of the working fluid heat pipes) were made of copper tubes with an OD of
affected the mentioned ratio of heat transfer rate. 6.32 mm (1/4 in.) and a length of 1050 mm. Each absor-
Mathioulakis and Belessiotis (2002) presented the results ber plate is 0.8 mm thickness, 200 mm width and
of a theoretical and an experimental investigation of the 1000 mm length made of copper painted with dull black
energy behaviour of a new type of solar system employ- paint. The three absorber plates and associated heat
ing a heat-pipe filled with ethanol. The energy behaviour pipes were housed in a collector case of 1150 mm length,
of the system was characterised by high instantaneous 800 mm breadth, and 150 mm thickness. The collector
efficiencies up to 60%. Esen (2004) employed refrigerants consisted of a galvanised steel sheet case, a 100 mm thick
in a solar cooking system which consists of a vacuum- insulation of glass wool behind the absorber plates and a
tube solar collector with integrated long heat pipes lead- 4 mm glass cover with 30 mm air gap between the glass
462 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468

Fig. 1. Experimental set-up of thermosyphon two-phase solar collector together with the temperature measurement points.

and absorber plates. Absorber plates with heat pipes sible. All valves, shown in Fig. 1, were opened and when
placed in same plane do not contact each ones, and the maximum vacuum is reached the valves are closed.
act like an evaporator section of the two-phase ther- Once all the heat pipes of the collector were completely
mosyphon solar collector. Distance (spacing) between evacuated of air, a known amount of working fluid is
absorber plates is set to 50 mm. Distance between two admitted into the heat pipes. After the filling process,
thermosyphon tubes (heat pipes) is equal to 100 mm. the collector panel was covered by a radiation shield.
The tubes were connected to the absorber plates with Three thermosyphon working fluids were investigated
silver-solder. in this study, R-134a, R407C, and R410A. Characteris-
Three condensers associated with the absorber plates tic features of these are given in Table 1. Each heat pipe
are placed inside well insulated hot water storage tanks was filled with 70 g of the refrigerant. All three refriger-
separated from each other. The volume of each identical ants were tested simultaneously under same ambient
hot water storage tank is set to 10 l. In the case of hot conditions.
water withdrawals, half of volume of each hot water Further, it is important to note that side-by-side
storage tank, 5 l, was simultaneously drained, and then experiments of the three systems using the same refriger-
filled by the same volume of cold water. It was decided ant were conducted as a control test to ensure that the
that in order to increase the heat transfer area of the performance of the three systems is identical before
condenser tubes, each tube should be formed in coils. introducing the various refrigerants. It was therefore
Each condenser consists of two copper coils with an deduced that the three built systems have identical
OD of 6.32 mm and total length of 900 mm (four coils performance.
corresponding to each heat pipe). The water filling, Temperatures were measured using copper–constan-
and discharge valves were respectively connected to the tan (type-T) thermocouples mounted at different loca-
top and bottom of each hot water storage tank. tions shown in Fig. 1. Temperatures were recorded
The upper part of each condenser coil was equipped using a computer based datalogger. As shown in Fig.
with a vacuum seal valve for connection to a vacuum 1, the copper–constantan thermocouples were attached
pump and to the charging line for the working fluid. Be- to the backside of the absorber plates using cold metal
fore filling each thermosyphon heat pipe with working welds at three locations. Six type-T thermocouples were
fluid, vacuum was applied to remove as much air as pos- placed inside each hot water storage tank to measure the
 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468 463

Table 1
Characteristic properties of refrigerants used (Suva Refrigerants, 2004)
Properties R-134a R407C R410A
Molecular weight (kg/kmol) 102.03 86.2 72.58
Boiling point at 1 atm (C)
26.5
43.56
51.53
Latent heat of vaporization at boiling point (kJ/kg) 215.5 245.1 276.2
Liquid density at 25 C (kg/m3) 1210 1134 1062
Vapor pressure of satd. liquid at 25 C (kPa) 661.9 1174 1653
Heat capacity of liquid at 25 C (kJ/kg K) 1.42 1.54 1.84
Heat capacity of vapor at 1 atm, at 25 C (kJ/kg K) 0.854 0.829 0.833
Thermal conductivity of liquid at 25 C (W/mK) 0.0824 0.0819 0.0886
Thermal conductivity of vapor at 1 atm (W/mK) 0.0145 0.01314 0.01339
Dynamic viscosity of satd. liquid at 25 C (kg/ms) 2.03 · 10
4 1.588 · 10
4 1.19 · 10
4
Kinematic viscosity of satd. liquid at 25 C (m2/s) 0.155 · 10
6 0.140 · 10
6 0.112 · 10
6
Critical temperature (C) 101.1 86.74 72.13
Critical pressure (kPa) 4060 4619 4926.1
ODP, CFC-12 = 1 0 0 0
GWP, CO2 = 1 1.300 1.526 1.725

temperatures of the stored water (Tw) (see marked loca- Qcol ¼ mw C p;w ðT w2
T w1 Þ ð1Þ
tions in Fig. 1).
where mw is the mass of water within the storage tank,
A thermopile-based pyranometer integrated to a
kg, Cp,w is the specific heat associated with water, kJ/
computer based datalogger, mounted on a surface paral-
kg C, and (Tw2
Tw1) is the temperature rise of water
lel to the plane of the collector, was used to measure the
within the tank, C. Tw1 is the temperature of the water
intensity of the global solar radiation incident on the
within the tank at the beginning of the day while Tw2 is
collector surface. Ambient temperature was measured
the water temperature at the actual time. The total
by a mercury thermometer as (Ta). A manometer con-
amount of solar energy incident on the collector over
nected to the upper part of the heat pipe measured the
the test period,
operating pressure. The wind velocity was measured
Z 2
with a cup type anemometer in the plane of the solar col-
lectors. The wind velocities measured during experi- Qincident ¼ A I dt ð2Þ
1
ments were between 1 and 3 m/s.
All measured values during the experiments were where A is the area of each absorber plate, m2, I is the
taken at 5 min intervals and averaged every half hour. global irradiance on the solar collector, W/m2.
The two-phase thermosyphon solar collector was installed The cumulative collection efficiency (g) can be
and tested under outdoor field conditions of Elazığ, expressed as a ratio of amount of heat stored in the
Turkey (latitude 38.40N; longitude 39.13E; altitude tank to the total amount of solar energy incident on
1067 m). The collector used in experiments was set at a the collector for the same period of time, i.e.,
tilt from horizontal equal to Elazığ latitude, facing south g ¼ mw C p;w ðT w2
T w1 Þ=Qincident ð3Þ
to optimize the collection of solar radiation. The radia-
tion shield covering the solar collector panel was re- It is noted that the cumulative efficiencies are calculated
moved at 08:00 a.m. and data collection commenced at from the beginning of the day, and therefore these inte-
08:30 a.m. Day-long experiments were carried out from grated efficiencies based upon the initial tank temper-
8:30 a.m. to 7:00 p.m., and values were recorded at half ature are not instantaneous values.
hour intervals. The experiments were performed Three representative tests were chosen to analyze the
throughout April and May 2003. On each day, the hot system performance. The first was conducted on April 4,
water storage tanks were completely drained in the 2003, the second on May 10, 2003 and the third on May
morning and refilled with cold water. 11, 2003. Intermittent cloud conditions existed on April
4, while clear-sky conditions existed on May 10 and
May 11.
3. Results and discussion The results of this study are shown in Figs. 2–4. At
the beginning of test 1, the storage tanks were filled with
The quantity of heat transferred to the water within water at 10 C. During this test the systems are allowed
the tank can be calculated by the water temperature vari- to operate throughout the day without water draw-off.
ation, taking into account the water quantity and its The changes in solar radiation (I ), the ambient temper-
specific heat: ature (Ta) and water temperature (Tw) during test 1 for
464 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468

4 April 2003 (Cloudy conditions, No loading)

35 800

30

2
Solar radiation, W/m
600

Temperature,o C
25 Ta

20 Tw(R410A)
400 Tw(R407C)
15
Tw(R-134a)
10 I
200
5

0 0
08:30

09:30

10:30

11:30

12:30
13:30

14:30

15:30

16:30

17:30

18:30
Local time, hr

Fig. 2. Daily variations of ambient temperature (Ta), average water temperature (Tw) and solar radiation (I ) for different refrigerants
in test 1.

10 May 2003
(Clear-sky conditions, No loading)
80 1

70 Ta
0.8
60 2 Tw(R410A)
Solar radiation, kW/m
Temperature, oC

50 Tw(R407C)
Efficiency,

0.6
Tw(R-134a)
40
I
30 0.4
η (R410A)
20 η (R407C)
0.2
10 η (R-134a)

0 0
08:30

09:30

10:30

11:30

12:30

13:30

14:30

15:30

16:30

17:30

18:30

Local time, hr

Fig. 3. Daily variations of ambient temperature (Ta), average water temperature (Tw), solar radiation (I ) and collection efficiency (g)
for different refrigerants in test 2.

different refrigerants are shown in Fig. 2. The average erant charged systems operate well even when sunlight is
ambient temperature and cumulative (total) incident more or less continuous. The water temperatures in-
solar radiation on the solar collector throughout the test crease rapidly around solar noon, and then remain
period were 11.93 C and 13.24 MJ/m2, respectively. approximately constant. The mentioned phenomena
Moreover, the maximum ambient temperature during can be seen for all refrigerants studied. The maximum
the test is 15.5 C, while the instantaneous maximum collection efficiencies of 50.84%, 49.43% and 48.72%
solar radiation is 770 W/m2. The maximum water temper- for R410A, R407C and R-134a were reached through-
ature rises of 19.5, 17.5 and 16.1 C for R410A, R407C out the daytime, respectively. Due to the different ther-
and R-134a are reached at about 17:00 h, respectively. It mal characteristics between refrigerants used, different
is also clear from the graph that the water temperature performances were found. Best performance was ob-
in the early morning hours does not increase rapidly tained using R410A probably due to its higher latent
due to very low solar radiation level and low ambient heat, higher liquid thermal conductivity, and lower vis-
temperature. Although solar radiation input shows some cosity. The working fluids chosen for heat pipes need
fluctuations, the water temperature increases monotoni- high enthalpies of vaporization. Thus, a small vapor
cally with time. Thus, it can be concluded that the refrig- flow along the tube can transport a large amount of
 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468 465

11 May 2003
(Clear-sky conditions,
Intermittent loading of 5 litre at 11 a.m. and 2 p.m.)
110

Solar radiation, Wm-2/10

100
90
Temperature,oC 80 I
70 Ta
60
50
Tw(R410A)
40 Tw(R407C)
30
Tw(R-134a)
20
10
0
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
Local time, hr

Fig. 4. Daily variations of solar radiation (I ), ambient temperature (Ta) and average water temperature (Tw) for different refrigerants
in test 3.

thermal energy. In addition to a high enthalpy of vapor- 2 p.m. 5 l of cold water at 25 C were added to each
ization, other desirable characteristics of working fluids tank. The water filling rate was to the withdrawal rate.
include a high surface tension and a low viscosity, to System response was faster and more sensitive to the
improve capillary pumping, and a high liquid thermal incident solar radiation in early operation hours and
conductivity, to reduce the temperature drops in the after water withdrawals. R410A exhibits the highest effi-
evaporator and condenser (Mills, 1999). ciency due to its higher latent heat, higher liquid thermal
The experimental performance of the system for a conductivity, and lower viscosity compared to other
clear day (May 10, 2003) is shown in Fig. 3. The storage working fluids investigated in this study. The water tem-
tanks were filled up with water at 25 C the morning be- perature reached 46 C just before the first withdrawal
fore the beginning of the test. This figure gives the ambi- process while it reached 58 C just before the second
ent temperature, water temperature, solar radiation and withdrawal process for R410A. After the second with-
collection efficiency for different refrigerants on May 10, drawal process the water temperature reached 51 C
2003. No water draw-off took place over the test period. for R410A.
As seen from the figure, for R410A, the higher collector Fig. 5 shows the results obtained throughout April
efficiencies were obtained compared to other refrigerants 2003, characterised by low solar radiation level and by
due to the reasons previously discussed. Since all three low ambient temperature. At the beginning of all tests
systems are identical in thermal capacity, it can be noted performed in April, the storage tanks were filled with
that the efficiency curves are proportional to the temper- water at 10 C. The days with highly variable solar radi-
ature curves shown in Fig. 3. Both the effect of the solar ation were April 2, 4, 10, 15, 18, 19, 21, 22, 23, 24, 25, 26
intensity and ambient temperature combine to affect the and 27, 2003. For these days, the daily maximum water
water temperature daily distribution shown in Fig. 3, temperature varies from 20.4 to 49.5 C for R410A while
exhibiting the effect of the ambient temperature to shift the daily maximum collection efficiency ranges from
the maximum water temperature at least 4 h after the 46.72% to 55.31% for R410A. During all experiments
maximum solar intensity time. conducted in this month it was observed that the plate
In case of hot water withdrawal, for an understand- temperatures of all three systems are very uniform, thus
ing of the performance evaluation of the collector, Fig. producing minimum thermal losses. From this figure it
4 gives the instantaneous variation of the ambient tem- will be noted that water is heated faster and to a higher
perature, the water temperatures and the experimental temperature in the R410A charged system, the collector
measured global solar radiation intensity along the stan- absorber plate temperature in the R410A charged sys-
dard local time of the day for three different refrigerants. tem remains considerably lower than the plate temper-
The first hot water draw took place simultaneously in ature in the R407C, and R-134a charged system. This
each tank at 11 a.m. while second withdrawal was made is exactly what we would expect. For the same amount
simultaneously in each tank at 2 p.m. Processes of water of insolation from the sun, the system having the lower
filling and withdrawal for all tanks were performed plate temperature will be more efficient. This is readily
simultaneously at 2 min intervals. At 11 a.m. and understood because at a lower plate temperature, less
466 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468

April 2003
I Ta,ave Tw,ave(R410A) Tw,max(R410A)
Tp,ave(R410A) Tp,ave(R407C) Tp,ave(R-134a) ηmax(R410A)
ηmax(R407C) ηmax(R-134a)

70 0.7
Total radiation, MJ/m day

60 0.6
2
Temperature, oC

50 0.5

Efficiency
40 0.4

30 0.3

20 0.2

10 0.1

0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Days of April 2003

Fig. 5. Variations of daily total solar radiation, daily average ambient temperature, daily average water temperature, daily maximum
water temperature, daily average plate temperature, and daily maximum collection efficiency for different refrigerants throughout April
2003.

heat is emitted or radiated from the collector plate and neously at 2 min interval. At 11 a.m. and 2 p.m. 5 l of
there is also less heat loss through the collector box insu- cold water at 10 C were loaded to each tank. For these
lation. The hot water withdrawal experiments were con- days, the daily maximum water temperature ranges from
ducted on April 5, 10, 14, 17 and 30. First water 42.5 to 57.5 C for R410A while the daily maximum
withdrawal was performed simultaneously in all tanks collection efficiency varies from 53.75% to 58.96% for
at 11 a.m. while the second withdrawal was done simul- R410A.
taneously in all tanks at 2 p.m. Processes of water filling Finally, Fig. 6 summarises the results obtained
and withdrawal for all tanks were performed simulta- throughout May 2003, characterised by moderate solar

May 2003
I Ta,ave Tw,ave(R410A) Tw,max(R410A)
Tp,ave(R410A) Tp,ave(R407C) Tp,ave(R-134a) ηmax(R410A)
ηmax(R407C) ηmax(R-134a)
80 0.7

70
Total radiation, MJ/m2 day

0.6

60
Temperature, oC

0.5
Efficiency

50
0.4
40
0.3
30
0.2
20

10 0.1

0 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31

Days of May 2003

Fig. 6. Variations of daily total solar radiation, daily average ambient temperature, daily average water temperature, daily maximum
water temperature, daily average plate temperature, and daily maximum collection efficiency for different refrigerants throughout May
2003.
 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468 467

radiation level and by moderate ambient temperature. tains side-by-side performance comparisons of three
The storage tanks were filled up with water at 25 C identical systems using three different refrigerants, this
the morning before the beginning of all tests conducted study can be useful to system designers in the selection
in May. Appreciable fluctuations of solar radiation were of an appropriate refrigerant.
observed on May 12, 14, 15, 18, 19, 21, 23, 28 and 31,
2003. For these days, the daily maximum water temper-
ature varies from 35.5 to 64.4 C for R410A while the Acknowledgment
daily maximum collection efficiency ranges from
50.86% to 57.13% for R410A. The hot water withdrawal The authors gratefully acknowledge the financial
experiments were performed on May 6, 11, 13, 20, 26 support from the Scientific Research Projects Adminis-
and 29. During each of these days, the two processes tration Unit of Fırat University for this study performed
of water filling and withdrawal were made as previously under project with grant no. 308.
described for April, except the temperature of the filling
water was at 25 C. For these days, the daily maximum
water temperature ranges from 54 to 60 C for R410A References
while the daily maximum collection efficiency varies
from 54.06% to 55.98% for R410A. Akyurt, M., 1984. Development of heat pipes for solar water
In summary, based on all results in Figs. 2–6, a rea- heaters. Solar Energy 32 (5), 625–631.
sonable similarity is obtained between the present experi- Bottum, E.W., 1981. Refrigerant charged phase change solar
water and space heating systems. ASHREA Transactions 87
mental results and the results of two-phase collector
(Pt. 2), 397–404.
studied by Hussein et al. (1999b), Chun et al. (1999),
Chun, W., Kang, Y.H., Kwak, H.Y., Lee, Y.S., 1999. An
Joudi and Al-Tabbakh (1999), Akyurt (1984) and Soin experimental study of the utilization of heat pipes for solar
et al. (1979). water heaters. Applied Thermal Engineering 19, 807–817.
Esen, M., 2004. Thermal performance of a solar cooker
integrated vacuum-tube collector with heat pipes containing
4. Conclusion different refrigerants. Solar Energy 76 (6), 751–757.
Esen, M., Yüksel, T., 2000. Performance of thermosyphon flat
The basic aim of the experiments conducted was to plate solar collector with heat pipes. Fırat University
determine the thermal performance of thermosyphon Journal of Science and Engineering 12 (2), 201–207 (in
Turkish).
heat-pipe solar collector under real operating conditions
Fanney, A.H., Terlizzi, C.P., 1985. Testing of refrigerant-
using three different refrigerants. A compact solar collec-
charged solar domestic hot water systems. Solar Energy 35
tor using different refrigerants to transfer heat collected (4), 353–366.
by a flat plate with heat pipes to heat water was designed Ghaddar, N., Nasr, Y., 1998. Experimental study of a
and produced, using medium-level technical abilities and refrigerant charged solar collector. International Journal
working at low temperature conditions. An extensive of Energy Research 22, 625–638.
two-phase closed thermosyphon solar collector testing Hammad, M., 1995. Experimental study of the performance of
program lasting for two months revealed that the proto- a solar collector cooled by heat pipes. Energy Conversion
type heat pipe solar water heaters performed satisfacto- and Management 36 (3), 197–203.
rily. The heat pipes did not present start up difficulties. It Hussein, H.M.S., 2002. Transient investigation of a two phase
closed thermosyphon flat plate solar water heater. Energy
was found that the refrigerant of the highest solar ther-
Conversion and Management 43, 2479–2492.
mal energy collection performance among the working
Hussein, H.M.S., Mohamad, M.A., El-Asfouri, A.S., 1999a.
fluids used in this experimental investigation was Transient investigation of a thermosyphon flat-plate solar
R410A. In future studies, the effect of the number of collector. Applied Thermal Engineering 19, 789–800.
the thermosyphon tubes of the collector on the water Hussein, H.M.S., Mohamad, M.A., El-Asfouri, A.S., 1999b.
temperature rise and the collection efficiency should be Optimization of a wickless heat pipe flat plate solar
investigated. Theoretical study and more experimental collector. Energy Conversion and Management 40, 1949–
data are needed to optimize the collector parameters. 1961.
The results shown in the present work may be used in Hussein, H.M.S., Mohamad, M.A., El-Asfouri, A.S., 2001.
the design and construction of a compact SDHWS. Theoretical analysis of laminar-film condensation heat
transfer inside inclined wickless heat pipes flat-plate solar
Generally speaking, refrigerant charged domestic hot
collector. Renewable Energy 23, 525–535.
water systems combine good energy behaviour with sim-
Joudi, K.A., Al-Tabbakh, A.A., 1999. Computer simulation of
plicity in manufacture, the use of common flat-plate a two phase thermosyphon solar domestic hot water heating
solar collectors without significant alterations of the system. Energy Conversion and Management 40, 775–793.
existing production line, good behaviour in freezing Mathioulakis, E., Belessiotis, V., 2002. A new heat-pipe type
and overheating, and an absence of moving parts or solar domestic hot water system. Solar Energy 72 (1), 13–20.
external pumping power. Also, as the present paper con- Mills, A.F., 1999. Heat Transfer. Prentice Hall, New Jersey.
468 M. Esen, H. Esen / Solar Energy 79 (2005) 459–468

Payakaruk, T., Terdtoon, P., Ritthidech, S., 2000. Correlations Schreyer, J.M., 1981. Residential application of refrigerant
to predict heat transfer characteristics of an inclined closed charged solar collectors. Solar Energy 26, 307–312.
two-phase thermosyphon at normal operating conditions. Soin, R.S., Sangameswar Rao, K., Rao, D.P., Rao, K.S., 1979.
Applied Thermal Engineering 20, 781–790. Performance of flat plate solar collector with fluid under-
Pluta, Z., Pomierny, W., 1995. The theoretical and experimen- going phase change. Solar Energy 23, 69–73.
tal investigation of the phase-change solar thermosyphon. Soin, R.S., Raghuraman, S., Murali, V., 1987. Two-phase water
Renewable Energy 6 (3), 317–321. heater model and long-term performance. Solar Energy 38
Radhwan, A.M., Zaki, G.M., Jamil, A., 1990. Refrigerant- (2), 105–112.
charged integrated solar water heater. International Journal Suva Refrigerants, 2004. Available from: <www.dupont.com/
of Energy Research 14, 421–432. suva/na/usa/literature/>.