Energy Conversion and Management 51 (2010) 265–270

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Conventional versus storage domestic solar hot water systems: A comparative

performance study
Abdul Jabbar N. Khalifa *, Raid A. Abdul Jabbar
Nahrain University, College of Engineering, Jadiriya, P.O. Box 64040, Baghdad, Iraq

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents an experimental comparative performance study between a storage solar domestic
Received 12 February 2009 hot water system and a conventional one. The storage system differs from the conventional one by inte-
Accepted 22 September 2009 grating the collector and storage tank into one piece of equipment. The investigated storage collector con-
Available online 21 October 2009
sists of six 80-mm-outside diameter copper tubes connected in series to act as an absorber and a storage
tank. Tests under different flow patterns are conducted to simulate different patterns of hot water con-
Keywords: sumption. The overall heat loss coefficient, useful heat gain, instantaneous efficiency, efficiency factor,
Domestic hot water system
heat removal factor and the temperature distribution along the pipe are determined using modified equa-
Solar collector
Storage
tions derived for the storage collector from the basic derivation of Hottel–Bliss–Whillier. A comparison
Performance between the experimental and theoretical results for the storage collector showed good agreements
and that between the two systems showed the superiority of the storage system.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction tional SDHW system. The storage system differs from the conven-
tional one by its simplicity of structure because of no divided
Solar domestic hot water (SDHW) systems can be designed in collector, no storage tank, no connection pipes and small area
many ways and manufactured with a variety of techniques. Con- needed for installation. The thermal performance is evaluated
ventional (SDHW) systems consist generally of one or more solar using modified equations derived for the storage system from the
collector connected to a thermal storage tank via suitable piping. basic derivation of Hottel–Bliss–Whillier [12].
A storage system differs from the conventional one by integrating
the collector and storage tank into one piece of equipment. Some of 2. Experimental work and procedure
the manufacturing techniques and simulations reported in the lit-
erature include adding phase change material modules at the top The storage collector is constructed from six 80-mm-outside
of the water tank to give the system a higher storage density and diameter copper tubes, 1-mm-thick, and 150 mm long each with
compensate heat loss in the top layer [1], storage solar water heat- a center to center distance of 120 mm. The copper tubes are con-
ers with phase change material storage units [2], rectangular cross nected by 15-mm-outer diameter, 1-mm-thick pipes to form a ser-
section solar collector/storage systems filled with water and vari- ies flow pattern. The connecting pipes are welded from the upper
ous concentrations of phase change material slurries [3], a rectan- part of one tube to the lower part of the following tube. All sides,
gular box incorporating the solar collector and storage tank into a bottom and space between tubes have 40-mm-thick rigid Styro-
single unit [4,5], indirect heating of the service hot water through a foam insulation board that has a density of 40 kg/m3 to maximize
heat exchanger incorporated into front and back major surfaces of heat retention. The single glazed frame of the collector is made of
the heater [6], a prism shaped storage solar collector [7,8], a flat aluminum with final dimensions of 0.90 by 1.80 m and with 3 cm
plate integrated collector storage device with recirculation of the spacing between the cover and the absorber. The collector is tilted
storage water [9], vacuum solar collectors connected to warm at an angle of 45°. Type T thermocouples are used to measure the
water storage [10], and an integrated rock bed and solar collector temperatures at the locations shown in Fig. 1 by a digital electronic
system [11]. thermometer. Point (19) in the figure represents the location of the
In this work, a storage solar collector that acts as a solar collec- thermocouple used for measuring the temperature of the outside
tor and a storage tank integrated into one piece is manufactured; surface of glass cover. Fig. 2 shows a photograph of the storage sys-
its thermal performance is evaluated and compared to the conven- tem. The collector is supplied with water by gravity from a con-
stant level tank located at the top of a nearby building. A ball
* Corresponding author. Tel.: +964 7901 688086. valve is used between the water source and the inlet to the collec-
E-mail address: ajkhalifa2000@yahoo.com (A.J.N. Khalifa). tor to control the flow rate of water. The estimated errors in the

0196-8904/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2009.09.021
266 A.J.N. Khalifa, R.A. Abdul Jabbar / Energy Conversion and Management 51 (2010) 265–270

Nomenclature

Ac surface area of the cover (1.36 m2) S net solar energy absorbed by the collector (W/m2)
AP projected area of the pipes = D  L (0.72 m2) Ta ambient temperature (K)
Cb bond conductance (W/m2 K) Tb temperature of the bottom of the collector (K)
Cp specific heat of water (J/kg °C) Tc temperature of the cover of the collector (K)
D outside diameter of the pipe (0.080 m) Tf water temperature (K)
Di inside diameter of the pipe (0.078 m) Tfm mean water temperature (K)
F collector efficiency factor Ti inlet water temperature (K)
F fin efficiency To outlet water temperature (K)
FR collector heat removal factor Tp pipe surface temperature (K)
HT total solar radiation on collector plane (W/m2) Tpm mean pipe surface temperature (K)
hi heat transfer coefficient between water and the tube Ts sky temperature (K)
wall (300 W/m2 °C) T average of the cover and mean pipe surface tempera-
hp-c convection coefficient between the pipes and the cover tures (°C)
(W/m2 °C) Ub back heat loss coefficient (W/m2 °C)
hr,c-s radiation coefficient from the cover to the sky (W/ UL overall heat loss coefficient (W/m2 °C)
m2 °C) Ut top heat loss coefficient (W/m2 °C)
hr,p-s radiation coefficient from the pipes to the cover (W/ V wind speed (m/s)
m2 °C) W distance between collector’s tubes (m)
hw wind convection coefficient (W/m2 °C) Wi width of the side insulation (m)
k insulation thermal conductivity (0.028 W/m °C) X insulation thickness (m)
L total length of the pipes (9 m) y distance along the pipes from the water entrance (m)
m_ collector mass flow rate (kg/s)
Qu total useful energy gain of the collector (W) Greek symbols
qu useful gain of the collector per unit length (W/m) a absorptivity of surface of the pipes (0.9)
R1 resistance to heat flow from back insulation (m2 °C/W) g hourly efficiency of the collector during working hours
R2 convection and radiation resistance from back of the (%)
collector to the ambient (m2 °C/W) c emittance of the cover
R3 convection and radiation resistance from the pipes to p emittance of the pipe surface
the cover (m2 °C/W) s transmissivity of cover (0.82)
R4 resistance to heat flow through the side insulation (s a) transmittance–absorptance product
(m2 °C/W) r Stefan–Boltzmann constant (5.668  108 W/m2 K4)
Rp-f heat resistance between the pipes and water (°C/W)

measurements of temperature and water flow rate and the esti- The top loss coefficient, Ut is:
mated uncertainties in the calculated parameters are given in  1
Table 1. 1 1
Ut ¼ þ ð3Þ
hpc þ hr;pc hw þ hr;cs

3. Mathematical analysis The values of the heat transfer coefficients of Eq. (3) are given as:
" #
Measurements of the inlet and outlet water temperatures, glass, DT 0:31
hpc ¼ ½1  0:0018ðT  10Þ 1:14 0:07
ð4Þ
pipe surface and ambient temperatures are conducted for different l
flow patterns. By employing the above measurements, the mean
pipe surface temperature, useful heat gain and instantaneous effi-
rðT 2p þ T 2c ÞðT p þ T c Þ
hr;pc ¼ ð5Þ
ciency are estimated. The overall heat loss coefficient, the useful ð1  ec Þ=ec þ 1=F pc þ ð1  ep ÞAc =ep Ap
heat gain, the efficiency factor, the heat removal factor and the hw ¼ 5:7 þ 3:8V ð6Þ
temperature distribution along the pipe are then evaluated using
 
Tc  Ts
the basic derivation by Hottel–Bliss–Whillier [12] with some mod- hr;cs ¼ ec r ðT 2c þ T 2s ÞðT c þ TsÞ ð7Þ
Tc  Ta
ifications where necessary.
The thermal network for a single cover system is shown in The useful gain of the collector per unit length is:
Fig. 3. Solar energy of amount S is absorbed by the pipes at some
typical location where the plate temperature is TP. This absorbed
qu ¼ D½HT ðsaÞ  U L ðT p  T a Þ ð8Þ
energy is distributed to losses through the top, bottom, and edges The useful heat gain transferred to the water can be expressed as:
and to useful energy gain. The thermal network of Fig. 3a may be
converted to that of Fig. 3b by calculating the different thermal qu ¼ FD½S  U L ðT f  T a Þ ð9Þ
resistances shown in the figure.
The overall loss coefficient from the storage collector may be gi- where F is the collector efficiency factor for the storage collector gi-
ven by: ven by:

UL ¼ Ut þ Ub ð1Þ 1=U L
F¼ h i ð10Þ
The back loss coefficient, Ub is: D DU L þ h p1D
1
i i

1 k The equivalent equation for the conventional flat plate collector is

Ub ¼ ¼ ð2Þ
R1 X [12]:
 A.J.N. Khalifa, R.A. Abdul Jabbar / Energy Conversion and Management 51 (2010) 265–270 267

Fig. 1. A Schematic diagram of the storage solar hot water system.


1=U L m Cp 
F¼ h i ð11Þ _ p
1 1 1 FR ¼ 1  e½UL DFL=mC ð14Þ
W U L ½DþðWDÞF 0 
þ Cb
þ pDi hi Ap U L

If it is assumed that F and UL are independent of position, then Eq. (14) is a modified version of the equation reported for conven-
the solution for the temperature at any position y for the storage tional flat plate collector [12]:
collector (subject to the condition that the inlet water temperature 
m Cp  _ p
is Ti) is given by: FR ¼ 1  e½UL WFL=mC ð15Þ
Ac U L
T f  T a  S=U L _ p The total useful energy gain of the collector may be given also by:
¼ e½UL DFy=mC ð12Þ
T i  T a  S=U L
Q u ¼ Ap F R ½S  U L ðT i  T a Þ ð16Þ
T f  T a  S=U L _ p
¼ e½UL WFy=mC ð13Þ The mean pipes surface temperature is given as:
T i  T a  S=U L
 
Qu FR
From Eq. (13) with y = L, FR can be expressed for the storage collec- T fm ¼ T i þ 1 ð17Þ
Ap U L F R F
tor as:
268 A.J.N. Khalifa, R.A. Abdul Jabbar / Energy Conversion and Management 51 (2010) 265–270

The mean pipe surface temperature will always be greater than

the mean water temperature due to the heat transfer resistance be-
tween the absorbing surface and the water. This temperature dif-
ference is usually small for liquid systems.
T pm ¼ T fm þ Q u Rp-f ð18Þ

where Rp-f is the heat transfer resistance between the pipes and the
water which is given by:
1
Rp-f ¼ ð19Þ
hi pDi L
The instantaneous efficiency of the collector is given by:

Qu _ p ðT o  T 1 Þ
mC
g¼ ¼ ð20Þ
H T AP HT AP

4. Performance results

Fig. 2. Photograph of the storage solar hot water system. Seven tests, each with a different flow pattern, are conducted
during selected clear days of February–May; the results from only
Table 1
one test are presented here. The quantity of hot water withdrawn
Estimated error in measurement and uncertainties in the calculated and derived ranged from 22 to 59 l per test.
parameters. Good agreements can be noticed in Fig. 4 between the theoret-
Parameter Typical error/ Remarks
ical and experimental mean plate temperature and the tempera-
uncertainty ture difference across the storage solar collector. A temperature
Ti, To and Ta ±0.2 °C Measured by mercury in glass thermometer
difference of up to 30 °C (25 °C in the test presented) is noticed
Tp and Tc ±0.2 °C Measured by type K thermocouple compared to the typical maximum 10 °C for the conventional
m_ ±5% Measured by graduated flask and stop watch system.
hp-c 0.52% Function of Tp and Tc Fig. 5 shows, again, good agreements between the experimental
hr,p-s 0.15% Function of Tp and Tc
and theoretical values of the hourly efficiency and the useful gain.
hr,c-s 0.11% Function of Ta and Tc
UL 0.14% Function of hp-c, hr,p-s and hr,c-s The top and overall loss coefficients are shown in Fig. 6, while the
F 0.01% Function of UL collector efficiency factor and heat removal factor are shown in
FR 1.12% _ UL and F
Function of m, Fig. 7. The distribution of water temperature in the flow direction
g 10% _ Ti and To
Function of m, for different hours of the day is shown in Fig. 8. Some deviation be-

Fig. 3. (a) Thermal network for the storage system and (b) the equivalent network.
 A.J.N. Khalifa, R.A. Abdul Jabbar / Energy Conversion and Management 51 (2010) 265–270 269

Fig. 7. Time variation of collector efficiency factor (F) and heat removal factor (FR).

Fig. 4. Time variation of inlet, outlet, ambient, average pipe surface temperatures
and temperature difference across the collector (Ti, To, Tpm, and To–Ti, respectively).

Fig. 8. Distribution of water temperature along the pipe for different hours of the
day.

Fig. 5. Time variation of radiation (HT), useful energy gain (Qu) and the hourly
efficiency (g).

Fig. 9. Time variation of pipe surface temperature at different locations.

Fig. 6. Time variation of top and overall loss coefficients (Ut and UL).

tween the experimental and theoretical values can be noticed

especially in the entrance region of the collector. This may come
from the assumptions that accompanied the derivation of the
equation used.
Fig. 9 shows the time variation of the measured plate tempera-
ture. The points shown are those at the middle of the six pipes of
the storage collector (see Fig. 1). The increase in temperature with
both distance and time is obvious. Fig. 10 shows the flow pattern
under which the above results are obtained which intended to sim-
ulate a certain pattern of domestic hot water usage. A comparison
between the values of the performance parameters of the storage
and the conventional systems is shown in Table 2. Fig. 10. The flow pattern used for water load.
270 A.J.N. Khalifa, R.A. Abdul Jabbar / Energy Conversion and Management 51 (2010) 265–270

Table 2 5. Like the conventional solar collector, the investigation showed

A comparison between the performance parameters of the storage and the conven- that the collector efficiency factor for the storage collector can
tional solar domestic hot water systems.
be taken as constant for a given collector design and flow rate.
Parameter Storage Conventional system 6. The collector overall heat loss coefficient has a comparable val-
system [13] ues for the two collectors; a range of 7.4–8.6 (an average of
Temperature difference across the 2–30 2–10 8 W/m2 °C) for the storage collector compared to 6.9–7.5 (an
collector, °C average of 7.2 W/m2 °C) for the conventional one.
Instantaneous efficiency 0.40–0.78 0.21–0.35
Heat removal factor 0.72–0.98 0.52–0.58
Collector efficiency factor 0.99 0.66
Overall heat loss coefficient, W/m2 °C 7.40–8.60 6.90–7.50 References

[1] Mazman M, Cabeza L, Mehling H, Nogues M, Evliya H, Paksoy H. Utilization of

phase change materials in solar domestic hot water systems. Renew Energy
The results from all seven tests showed that the theoretical and 2009;34:1639–43.
experimental average hourly efficiency during the test for the sys- [2] Tarhan S, Sari A, Yardim M. Temperature distributions in trapezoidal built in
storage solar water heaters with/without phase change materials. Energy
tem was in the range 0.53–0.70 and 0.51–0.69, respectively with a Convers Manage 2006;47:2143–54.
percentage deviation range between them of 0.58–5.04%. The rea- [3] Eames P, Griffiths P. Thermal behaviour of integrated solar collector/storage
sonable deviations between the theoretical and the experimental unit with 65 °C phase change material. Energy Convers Manage
2006;47:3611–8.
results validate the equations derived for the storage collector. [4] Garnier CJ, Currie J, Muneer T. Integrated collector storage solar water heater:
temperature stratification. Appl Energy 2009;86:1465–9.
[5] Currie J, Garnier C, Muneer T, Grassie T, Henderson D. Modelling bulk water
5. Conclusions
temperature in integrated collector storage systems. Build Ser Eng Res Technol
2008;29:203–18.
The experimental investigation on the solar storage collector [6] Gertzos K, Pnevmatikakis S, Caouris Y. Experimental and numerical study of
and the comparative performance study between the storage sys- heat transfer phenomena, inside a flat-plate integrated collector storage solar
water heater (ICSSWH), with indirect heat withdrawal. Energy Convers
tem and the conventional one yielded the following conclusions: Manage 2008;49:3104–15.
[7] Joudi K, Hussein I, Farhan A. Computational model for a prism shaped storage
1. Good agreements between the theoretical and experimental solar collector with a right triangular cross section. Desalination
2007;209:171–81.
values for the mean plate temperature, the temperature differ- [8] Joudi K, Hussein I, Farhan A. Computational model for a prism shaped storage
ence across the collector, the hourly efficiency and the useful solar collector with a right triangular cross section. Energy Convers Manage
gain are noticed for the storage collector. 2004;45:391–409.
[9] Gertzos K, Caouris Y. Experimental and computational study of the developed
2. A temperature difference of up to 30 °C for the storage collector flow field in a flat plate integrated collector storage (ICS) solar device with
is noticed compared to the typical maximum 10 °C for the con- recirculation. Exp Therm Fluid Sci 2007; 31:1133–45.
ventional system. [10] Georgiev A. Simulation and experimental results of a vacuum solar collector
system with storage. Energy Convers Manage 2005;46:1423–42.
3. The instantaneous efficiency of the storage collector was found
[11] Hamdan M. Investigation of an inexpensive solar collector storage system.
to be in the range 0.4–0.78 compared to 0.21–0.35 for the con- Energy Convers Manage 1998;39:415–20.
ventional one. [12] Duffie JA, Beckman WA. Solar energy of thermal processes. New York: John
Wiley and sons Inc.; 2006.
4. The heat removal factor for the storage collector was in the
[13] Khalifa AJN. Thermal performance of locally made flat plate collectors used as
range 0.72–0.98 compared to 0.52–0.58 for the conventional apart of a domestic hot water system. Energy Convers Manage
collector while the collector efficiency factor approached unity 1999;40:1825–33.
compared to 0.66 for the conventional one.