Energy 64 (2014) 187e199

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Thermal performance of a solar water heater with internal exchanger

using thermosiphon system in Côte d’Ivoire
Paul Magloire E. Koffi a, *, Blaise K. Koua b, Prosper Gbaha a, Siaka Touré b
a
Laboratoire d’Energies nouvelles et Renouvelables, Institut National Polytechnique Félix Houphouet Boigny, B.P. 581, Yamoussoukro, Cote d’Ivoire
b
Laboratoire d’Energie Solaire, Université Félix Houphouet Boigny, 22 B.P. 582, Abidjan 22, Cote d’Ivoire

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

Article history: This study presents a theoretical and experimental analysis of the thermal performance of a solar water
Received 24 February 2013 heater prototype with an internal exchanger using thermosiphon system. The results focus mainly on the
Received in revised form levels of the heat fluxes temperatures recorded, mass flow rate and efficiency of collector. These tests are
10 August 2013
performed for a sunny day and a cloudy day. The daily solar intensities range from 300 to 1233 W/m2,
Accepted 26 September 2013
Available online 21 November 2013
with the daily ambient temperature ranging between 27
C and 33
C. Maximum temperatures at the flat
solar collector output are 88
C and 58
C for the sunny day and cloudy day, respectively. Maximum
instantaneous efficiencies are 68.33% and 50% for the sunny day and the cloudy day, respectively. The
Keywords:
Solar water heater
values of the thermal performances parameters FR (sa) and FRUL are 0.780 and 4.252 W/m2
C respec-
Thermal performance tively for the cloudy day and 0.777 and 4.689 W/m2
C respectively for the sunny day. The coefficient of
Efficiency exchange thermal of heat exchanger Ue found is 149.15 W/(m2 K) when, the average heat exchanger
Heat exchanger effectiveness obtained is 70%. The experimental results show that mean daily efficiency is near 50%. This
Temperature reveals a good compatibility of the system to convert solar energy to heat which can be used for heating
water.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction related to the performance of the collectors which convert it. For
the collectors whose caloporting fluid is water; thermal transfer is
Solar energy can be used in the industrial, commercial and do- made suitably because water is a good conductor of heat [2].
mestic sectors. In the domestic applications, households consume Many studies have been conducted on the SWHS with or
energy in air conditioning, heating, water heating, lighting and without a heat exchanger. Zerrouki et al. [3], Belessiotis and
other applications. An economic and efficient system is required to Mathioulakis [4] conducted their work on thermosiphon solar
encourage households to use solar water heating. Solar water water heaters without heat exchanger and took, as parameters of
heating technologies are well-known and contributing significantly their study, the storage tank average temperature and the water
to hot water production in several countries, lowering energy bills, mass flow rate in the system. Nahar [5] conducted their work on
and reducing the environmental pollution. SWHS (Solar water thermosiphon SWHS without heat exchanger. The particularity of
heating systems) are also attracting increasing interest as a solution this study is that the storage tank is filled beforehand with water.
to reduce fossil fuel consumption and greenhouse gas emissions of The collector is made of a selective surface absorber in MAXORB
residential buildings [1]. A SWHS saves energy and reduces designed with galvanized iron tubes interdependent with an
greenhouse emissions relative to conventional fossil fuel WHS aluminum foil. Roberts and Forbes [6] conducted their work on
(water heating system) in the use phase. The other important gain thermosiphon SWHS without heat exchanger and studied the in-
owing to the use of solar water heaters is the reduction of CO2 fluence of absorber plate absorptance and thermal emittance on
footprint. In SWHS, the solar collectors are devices which capture the efficiency. They also show that the impact on the efficiency of
the solar energy and transfer it into thermal energy that increases making changes on parameters, such as reducing heat loss co-
the internal energy in the fluids, and hence increases their tem- efficients, could be quickly evaluated. In particular, the role of the
perature. The success of the solar energy applications is closely absorptance and emittance of the absorber plate was studied to see
the influence of solar absorber selectivity on the overall efficiency.
Tang et al. [7] investigate the effects of water temperature in the
storage tank and height difference between collector loop con-
* Corresponding author. Tel.: þ225 (0)7 98 32 97. nections at the tank on freeze protection of flat-plate collectors at
E-mail address: paumagloirel@yahoo.fr (P.M.E. Koffi).

0360-5442/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.energy.2013.09.059
188 P.M.E. Koffi et al. / Energy 64 (2014) 187e199

Nomenclature UL overall loss coefficient (W/m2 K)

k thermal conductivity of plate material (W/m K)
Ac collector area (m2) Ut top loss coefficient (W/m2 K)
Tf2 outlet working fluid temperature (
C) kINS thermal conductivity of glass wool insulation (W/m K)
Ae outside heat exchanger heat transfer area (m2) v wind speed (m/s)
Tf3 working fluid temperature at inlet of heat LINS thickness of insulation (m)
exchanger (
C) W distance between riser tubes (m)
Cb bond conductance N number of glazing
Tf4 working fluid temperature at outlet of heat
exchanger (
C) Greek letters
Cf specific heat of water through heat exchanger (J/kgK) m_ mass flow rate (kg/s)
Ti average water temperature in storage tank at start of DTm log-mean temperature difference (K)
test (
C) Q_ total heat transfer rate (W)
D outer diameter of riser tube (m) εhx heat exchanger effectiveness
Tw temperature of hot water at the exit of the Qu useful energy (W/m2)
storage tank (
C) εp emittance of plate surface
F0 collector efficiency factor S absorbed flux (W/m2)
Tpm mean absorber plate temperature and (
C) εc emittance of glass cover
FR collector heat removal factor t fin thickness (m)
T_ w average temperature of the storage fluid at time t. s StefaneBoltzmann constant (W/m2 K4)
hFI heat transfer coefficient between fluid and tube wall Ta ambient temperature (
C)
(W/m2 K) (sa) fraction of the solar radiation absorbed
(UA)hx overall heat transfer coefficient of the heat exchanger Te temperature of cold water network entering in storage
(W/K) tank (
C)
hw heat transfer coefficient (W/m2 K) h collector instantaneous efficiency (%)
Ub bottom loss coefficient (W/m2 K) Tf average water temperature in storage tank at the end
H total solar radiation on collector slope (J/m2) of test (
C)
Ue thermal coefficient exchange of heat hd mean daily efficiency (%)
exchanger (W/m2 K) Tf1 inlet working fluid temperature (
C)
IT incident radiation (W/m2)

clear nights in terms of outlet water temperature of the thermo- heat exchanger, but higher than that of all-glass evacuated tubular
siphonic reverse flow from the collector. Sakhrieh and Al-Ghandoor solar water heater.
[8] studied the experimental investigation of the performance of The effect on the performance of the collector using a heat
five types of solar collectors on SWHS. The systems involved in this exchanger between the collector and the tank was analyzed. Wong
study are blue and black coating-selective copper, copper, and et al. [13] investigate the potential for shower water heat recovery
aluminum collectors in addition to evacuated tubes collectors. from bathrooms equipped with instantaneous water heaters in
Their results show that evacuated tube solar collector has the high-rise residential buildings of Hong Kong. In their works, there is
highest efficiency, followed by black and blue coated solar collec- no allusion is, however, made to the inlet and outlet water tem-
tors. Pierson and Javelas [9], Hussein [10] and Soo Too et al. [11] peratures from the collector and also the inlet and outlet temper-
conducted their studies on a SWHS with an external exchanger. atures of the storage tank. A simple single-pass counter-flow heat
They particularly studied the role of stratification of the tempera- exchanger installed horizontally beneath the shower drain is
tures in the storage tank. They developed a theoretical model that employed as a localized heat recovery measure for preheating cold
allowed study of the influence of the essential parameters of the water going to a water heater. Zbyslaw and Wlodzimierz [14]
operation on the installation. Hussein [10] conducted a theoretical conducted their studies on a SWHS with an internal exchanger
and experimental study of a conventional solar water heater plan and undertook an experimental and theoretical study of a solar
with biphasic natural circulation. The experimental results show a water heater using an internal vertical heat exchanger. The authors
satisfactory agreement in the theoretical and experimental results. use R22 in a liquidevapor mixture as primary fluid. They specify
Soo Too et al. [11] studied the characteristics of a SWHS incorpo- that the heat transfer coefficients are, for phase changes (boiling
rating a vertical mantle heat exchanger with a narrow annular and condensation), on average, two times higher than those in the
spacing of 3 mm and a two-pass arrangement. Their comparison case of free convection in the traditional thermosiphon system.
between predictions of the annual solar contribution for mantle- Ayompe and Duffy [15] conducted their studies on a SWHS with an
tank systems and direct-coupled systems shows that the direct- internal exchanger and studied the thermal performance of a solar
coupled systems are found to provide slightly higher annual en- water heating system with heat pipe evacuated tube collector using
ergy saving than mantle-tank systems for standard domestic hot data obtained from a field trial installation over a year in Dublin,
water demand in Australia. Huang et al. [12] studied the thermal Ireland. An automated sub-system was developed and incorporated
performance of thermosiphon flat-plate SWHS with a mantle heat to control the hot water draw-offs and electric immersion heater to
exchanger. They investigated to show its applicability in China. mimic the operation of solar water heating systems in domestic
Their experimental results show that mean daily efficiency of the dwellings. Haltiwanger and Davidson [2] conducted their studies
thermosiphon flat plate solar water heater with a mantle heat on a SWHS with an internal exchanger and studied a discharge of a
exchanger with 10 mm gap can reach up to 50%, which is lower thermal storage tank using an immersed heat exchanger with an
than that of a thermosiphon flat-plate solar water heater without annular baffle. They studied and measured the temperatures at the
 P.M.E. Koffi et al. / Energy 64 (2014) 187e199 189

sources of energy. It will help at the environmental protection and

will reducing the poverty.

2. Solar energy resources in Côte d’Ivoire

Côte d’Ivoire, the country where the works are performed is

situated in Sub-Saharian Africa between 5
and 11
north latitude.
The country emitted 5.8 million tones of CO2 in 2010. This corre-
sponds to 0.29 t CO2/capita [16]. In this country, carbon emissions
and air pollution must be reduced and this can be achieved by
increasing the percentage of energy generated from clean resources
like solar, wind, geothermal and hydraulic energy. Several potential
solutions to the current environmental problems must be devel-
oped, including harnessing renewable energy as well as the utili-
zation of energy conservation technologies.
The annual solar energy received in Côte d’Ivoire lies between
1650 and 1950 kWh/m2 [17]. Côte d’Ivoire lies within a tropical
region and hence experience tropical climate. The country has two
main distinct seasons: the rainy season (from March to August) and
the dry season (from November to March). The other months are
the boundaries of the two seasons. The temperatures throughout
the year respectively range from a minimum average of 22
C to a
maximum average of 32
C. The average sunshine, which is the
duration of the bright sunshine between the sunrise and the sunset,
range from 5 to 8 h per day. This gives an annual average solar
insulation of about 5.0 kWh m2 day1, with a peak of sunshine
being received March and April according to regions [18,19] From
Fig. 1. A picture of the installation system.
the information provided, it is obvious that Côte d’Ivoire as a whole
is well endowed in terms of solar energy (Natural resources) and
these are adequate for any type of application. The climate condi-
tions do not in anyway hinder the utilization of solar energy.
inlet and outlet of the heat exchanger, the temperature difference
across the heat exchanger, the mass flow rate through the heat 3. Materials and methods
exchanger storage water temperature distribution in order to
calculate the heat exchanger effectiveness and heat transfer to the The SWHS realized includes a solar collector of a 2 m2, heat
heat exchanger. exchanger arranged in diagonal in the storage tank, a storage whose
The thermal performance on some of SWHS with or without capacity is 95 L and the piping of connections as shown in Fig. 1.
heat exchanger has been improved by many researchers by design With an outside diameter of 12 mm and a length of 6 m, the surface
improvement of absorber plate, fin efficiency, storage tank, selec- of the heat exchanger is 0.226 m2. The system has a glass wool
tive coating of absorber, working fluid, thermal insulation etc. insulation of thermal conductivity 0.040 W m1 K1. The solar
However, no more information is given on the collector thermal collector includes an absorber composed of 12 tubes separated
performances and on the solar system thermal performance. But, from 12 mm and painted in matt black. Below these tubes, there is a
studies on improvement of the thermal performance of thermosi- leaf of aluminum which reflects the thermal radiation received
phon solar water heaters using an internal exchanger aspect are towards the absorber. The collector is covered with a glass which
limited. For these reasons, an experimental study of solar water surface is 2 m2 and 4 mm in thickness. This collector is thermally
heater system with natural circulation by thermosiphon, including insulated at the back and on the lateral sides by a 5 cm thickness
an innovative internal heat exchanger made of rolled copper tube glass wool.
placed diagonally in the storage tank. The use of the exchanger in The works were performed in Yamoussoukro, the political
the primary circuit has the advantage of lowering the temperature capital city of Côte d’Ivoire. Yamoussoukro is located at 6.54

of the fluid at the outlet side of the exchanger, what makes possible Northern latitude.
the increasing in the effectiveness of the collector and consequently The solar collector is directed in the South direction and tilted by
that of the system. The objective required by laying out the 10
relative to the horizontal. The experiments consisted in the
exchanger in diagonal, which is also the subject of the work, is to measure of the period of sunshine received by the solar collector
increase the heat-transferring surface and especially to permit the and of the temperature in diverse places of the system. The period
hot fluid to transfer a significant part of its heat to the fluid in the of sunshine is measured by means of a KIPP and ZONEN pyran-
storage tank. This disposition of the exchanger permits to improve ometer with relative uncertainty of 2%. It is connected to a digital
the effectiveness of the system. It is wise to indicate that the integrator allowing the reading of the immediate received solar
absorber of the collector of the studied system is made of copper energy and the irradiation. The pyranometer is horizontally placed
tubes painted in black mat no selective. This solar water heater to get all the solar radiation. A data acquisition card, made, by our
realized will permit the greenhouse gas emission reduction if the care, allows recording the temperature in diverse places of the
electric water heaters for example are replaced by solar water system. To avoid perturbing the fluid flow, one uses probes of small
heaters in residential, hospitals and hotels sector in Cote d’Ivoire. As dimensions made of 1.6 mm diameter of diode 1N4148 in silicon,
a result, water heating is consuming a large portion of natural gas; 0.5
C precision. The entire collector temperatures (16 all in all)
therefore, resulting in polluting the environment. So, our system used is calibrated before using. The center of measure is presented
will also allow reducing the consumption of the wood and other at the Fig. 2. Besides the measure of the period of sunshine, of
190 P.M.E. Koffi et al. / Energy 64 (2014) 187e199

Fig. 2. Various points of measure of the temperatures and protocol of acquisition of data.

ambient temperature, of fluid temperature in the solar collector

(inlet and outlet collector), of fluid temperature in the heat
exchanger (heat exchanger inlet), of temperature of the absorber 2 31
and the temperature of the warm water of storage are measured in
6 N 1 7
the system. Ut ¼ 6
4   þ 7
344 Tpm Ta 0:31 hW 5
Tpm Nþf
4. Theoretical analysis 2 3
  
6
6
s Tpm
2 þ T2
a Tpm þ Ta 7
7
Thermal analysis of solar collectors is covered in many solar þ6 7 (1)
thermal engineering texts [6,20,21]. Therefore, only equations 4 1 þ 2Nþf 1
 N5
εp þ0:0425Nð1εp Þ εC
which describe the main parameters and their importance will be
described in this paper. The top heat loss coefficient Ut is evaluated
empirically by Ref. [22]: with
 P.M.E. Koffi et al. / Energy 64 (2014) 187e199 191

 
f ¼ 1  0:04hw þ 5  104 h2w ð1 þ 0:058NÞ (2) The experimental mass flow rate is obtained by calculation from
the establishment of the heat and mass transfer balance between
the inlet and outlet collector hot fluid according to the relation:
hw is the heat-transfer coefficient due to convection at the top of
cover due to wind. hw is calculated by the following relation sug-  
gested by McAdams as given in Duffie and Beckman [23]:
IT ðsaÞhAC ¼ mc
_ p Tf2  Tf 1 (13)

hw ¼ 5:7 þ 3:8v (3) Substituting h, by its expression in Eq. (12), into Eq. (13), we
obtain:
The bottom loss coefficient Ub, which accounts for the conduc-

 
tion heat loss through the back of the solar collector, is calculated I ðsaÞAc T  Ta
_ ¼
m T  FR ðsaÞ  FR UL f 1 (14)
by: IT
cp Tf2  Tf 1
kINS
Ub ¼ (4) The use of a heat exchanger between the storage tank and the
LINS
collector raises the collector fluid temperature, hence lowering the
The overall loss coefficient UL is given by: collector efficiency. The study of the performance of a system of
solar water heater thermosiphon and including a heat exchanger
UL ¼ Ut þ Ub (5) also passes by the study of the performance of this heat exchanger.
Most of those studies are done are made clear sky conditions
The collector heat removal factor FR, which relates the actual
[2,4,11,12,29,30] in order to know the heat exchanger characteris-
useful energy gain of a collector to the useful gain if the whole
tics. The model used is based on a heat exchanger consisting of
collector surface was at the fluid inlet temperature, is estimated by
copper tubes immersed diagonally in a storage tank. The equation
Ref. [24]:
of thermal exchange taking place at the level of the heat exchanger
_ ph
mc 0 _ p
i is expressed by Refs. [9,31]:
FR ¼ 1  eðUL AC F Þ=mc (6)
uL AC  
_ f , Tf3  Tf4 ¼ Ue ,Ae ,½DTm 
m,C (15)
where F0 is the collector efficiency factor and given by:
   
1=U Tf3  Tw  Tf4  Te
0
F ¼ h L i (7) With DTm ¼ Tf3 Tw
(16)
1
w UL ½DþðwDÞF þ c1b þ pD1h ln
I FI
Tf 4 Te

The total heat transfer rate (Q_ ) in an internal heat exchanger can
where hFI is taken as 300 W/m2 K for natural circulation [25] and F
be evaluated from measurements of flow rate and temperature
is the fin efficiency and given by:
change across the heat exchanger, and is expressed as [2]:
tanh mðw  DÞ=2  
F ¼ (8) Q_ ¼ mC
_ P Tf3  Tf4 (17)
mðw  DÞ=2

with m ¼ (UL/kt)0.5 The heat transfer fluid is here transported naturally (passive
The useful energy gain is given by Ref. [26]: systems). But, generally, the heat transfer fluid which passes
through a heat exchanger is transported by forced circulation
     
Qu ¼ Ac S  UL Tpm  Ta ¼ Ac ðsaÞIT  UL Tpm  Ta (9) (active systems).
The heat transfer under full tank recovery operation mode is
where (sa) represents the fraction of the solar radiation absorbed reported in order to characterize the overall heat transfer of the
by the collector and depends mainly on the transmittance of the heat exchanger performance of the heat exchanger. The overall heat
transparent covers and on the absorbance of the absorber transfer of the heat exchanger (UA)hx associated with individual
(sa ¼ 0.836) [27]. thermal resistances can be conventionally defined as in Eq. (18)
The useful heat gain can also be expressed from the fluid flow based on the log-mean temperature difference between the hot
_ through the collector and the inlet (Tf1) and outlet (Tf2)
rate (m) and cold fluids (DTm) in the heat exchanger. The (UA)hx correction
fluid temperatures as: factors for the heat-up tests can be determined from the heat
  exchanger inlet (Tf3) and outlet (Tf4) temperatures and the highest
_ p Tf 2  Tf1
Qu ¼ mc (10) (Tt2) and lowest (Te) tank core temperatures [11]:

By combining Eqs. (8)e(10), the useful heat gain can be Q_

ðUAÞhx ¼ (18)
rewritten as [25]: DTm
h  i
Qu ¼ Ac FR S  UL Tf1  Ta (11) ðUAÞ0hx ¼ FðUAÞhx (19)

The collector instantaneous efficiency (h) is influenced by where F is the correction factor for cross-flow heat exchanger
several factors such as the material used, the design of the absorber, In a conventional heat exchanger, heat is transferred from one
the properties of glass, weather and operating conditions [5]. It is fluid to another. The concept of the heat exchanger effectiveness,
defined as the ratio of the useful heat gain (QU) to the solar radia- εhx, has been introduced by Nusselt to compute directly the rate of
tion intensity (IT (sa)) [28]: heat transfer from the inlet temperatures of the fluids [32]. This
 effectiveness is defined as the ratio of the actual rate of heat
Qu T  Ta
h¼ ¼ FR ðsaÞ  FR UL f1 (12) transfer in a certain heat exchanger to the maximum possible rate
Ac IT IT of heat transfer in an “ideal” heat exchanger and is, therefore, also a
192 P.M.E. Koffi et al. / Energy 64 (2014) 187e199

measure of the thermodynamical quality of the design [33]. The 5.2. Uncertainties on calculated parameters
heat exchanger effectiveness was given by the following relation
[2,31]: The error of the experimental results on the basis of the un-
certainties in the primary measurements is performed using the
Tf3  Tf4 Kline and Mc Clintock relationship [35] as reported by Jia et al. [36]:
εhx ¼ (20)
Tf 3  Tw

 2  2  2 1=2
vf vf vf
The mean daily efficiency (hd) of a solar water heater is an Dy¼ ðDx1 Þ2 þ ðDx2 Þ2 þ:::::::::::þ ðDx1 Þ2
important parameter which describes the thermal performance of
vx1 vx2 vxn
the system, and a value with comparative constancy. The mean (23)
daily efficiency (hd) can be calculated as follows [12]:
where f is the given function of the independent variables, x is one
 
 MCp Ti  Tf of the variables of the function and Dx is the absolute error asso-
hd ¼ (21) ciated with the variable. The relative error is shown as
Ac ,H

   
Dy vf 2 Dx1 2 vf 2 Dx2 2
¼ þ þ :::::::::::
y vx1 y vx2 y
  
vf 2 Dx1 2 1=2
5. Uncertainties analysis þ (24)
vxn y
Uncertainty analysis is needed to prove the accuracy of the ex- The calculated parameters are the mass flow rate and efficiency
periments. In this study uncertainty analysis focuses on the calculated from the measured parameters. The analysis of the re-
measured and calculated parameters. sults indicates an overall accuracy of the mass flow rate as 6.5%,
5% for the efficiency, 5% for the coefficient of thermal exchange
of the heat exchanger, 2% for the total heat transfer rate, 6% for
5.1. Uncertainties on measured parameters
the overall heat transfer of the heat exchanger and 4% for the heat
exchanger effectiveness.
The measured parameters are temperatures. To determine the
uncertainties and errors in the temperature measurements, one
uses a digital thermometer as described in the previous section. 6. Results and discussions
One then proceeds to a series of several measurements using each
of the 16 probes and the digital thermometer under the same 6.1. Heat fluxesetemperaturesefluid flow rateeeffectiveness
experimental conditions to determine statistically representative
uncertainties that can be calculated. One determines then the Performance and testing of a thermosiphon SWHS was carried
relative uncertainty by taking the differences between the mea- out in Yamoussoukro and throughout a sunny (29/01/2011) and
surement obtained by the digital thermometer and that of the cloudy (05/10/2012) day. The daily irradiations of sunny and cloudy
probe. day chosen are 5575 Wh/m2/day and 1858 Wh/m2/day, respec-
This difference is then reported to the digital one. An average of tively. The experiments were performed at different meteorological
the relative uncertainties is finally made according to the formula conditions in two periods:
[34]:
- 1st period: from September 2010 to February 2011;
 n 
DT 1X Tei  Tmi - 2nd period: from March 2012 to October 2012.
¼ (22)
Tet n i¼1 Tei
During the experimental period, the following measurements
n is the number of measures were carried out on daily basis: total irradiation received by the
This average gives a relative uncertainty of 4% for the tem- collector, total daily irradiation, the ambient air temperature,
peratures measured by the probes. temperatures in various points of the system as shown in Fig. 2.
Measurements relating to the collector characteristic variables and
to those of the storage tank are regularly recorded an interval of

Fig. 3. Instantaneous thermal heat flux versus time of day. Fig. 4. Ambient temperature versus time of day.
 P.M.E. Koffi et al. / Energy 64 (2014) 187e199 193

Fig. 5. Outlet collector temperature versus time of day.

Fig. 8. Fluid mass flow rate versus time of day.

Fig. 6. Inlet exchanger temperature versus time of day.

Fig. 9. Collector efficiency versus time of day.

12 h per day, i.e. from 6:00 am to 6:00 pm and are stored on a
compact disk.
For any thermal system, it is important to know the energy obtained. It is observed that the system behavior is mainly
quantity received and its distribution in time. So, Figs. 3e9 present depending up on the state of sky (sunny or cloudy days).
the heat fluxes, the ambient temperature (Ta), the collector outlet Fig. 3 presents the heat flux received by solar collector according
fluid temperature (Tf2), the exchanger inlet fluid temperature (Tf3), to the time for both days. We notice that the heat flux evolve ac-
the outlet tank temperature of water in the storage tank (Tw), the cording to time. The heat flux reaches its peak at 12:30 am for both
mass flow rate (m)_ and the collector efficiency (h) versus time, for
days and then begins decreasing. The reached maximal values are
the sunny and cloudy day. Table 1 presents the maximum values respectively of 1233 W/m2 for the sunny day and 300 W/m2 for the
cloudy one. These results are in perfect concordance with those of
Nahar [5], Sakhrieh and Al-Ghandoor [8] and Enibe [37].
Fig. 4 shows the variation of the ambient temperature with time.
The temperature increases slowly until reach their maximum
temperature at 1:40 pm, both days. The maximum values recorded
are respectively 33
C for sunny day and 27
C for cloudy day. Then
for both days, the ambient temperatures decrease with the time.
Ambient temperature has a strong effect on water inlet tempera-
ture because inlet temperature increases with the rise of ambient

Table 1
Maximum values of incident heat flux, efficiency, mass flow rate and different
temperatures.

IT (W/m2) Ta (
C) Tf2 (
C) Tf3 (
C) Tw (
C) _ (kg/s)

m h (%)
Sunny day 1233 33 88 74 55 0.010364 68.33
Cloudy day 300 27 58 50 39.4 0.00758 50
Fig. 7. Outlet tank temperature versus time of day.
194 P.M.E. Koffi et al. / Energy 64 (2014) 187e199

temperature. This effect will increase the useful energy gained so

the efficiency of the solar water heater will also increase as shown
in Fig. 9.
Fig. 5 shows the experimental results of the temperature values
at the outlet of the solar collector for a sunny day and for a cloudy
day. Because of the natural circulation of water between the col-
lector and the tank, water temperatures at outlet kept rising during
the test time as long as the amount of energy gained from the sun is
higher than the energy lost from the system. After that, the water
temperatures at the outlet start to decrease. At the morning times,
the water temperatures at the outlet are relatively low due to the
low flux of incident solar radiation. During the day times, water
temperature increases due to the increase in heat flux. We note that
the maximum water temperatures at the outlet of the solar col-
lector reached 88
C in the case of sunny day. Whereas for the
cloudy day, the water temperatures at the outlet reached 58
C. This
maximum outlet water temperature occurs at 1:00 pm for both
days. These results are interesting compared to some values in the Fig. 10. Collector efficiency with (Tf1Ta)/IT.
literature because, for the sunny day, the maximum value of the
outlet collector temperature is 72
C in Khalifa’s study [38], 70
C in
Abd-Al Zahra and Joudi’s study [39], 74
C in Pillai and Banerjee’s in the storage tank and environment with an absorber made of
study [40], 76
C in Sakhrieh and Al-Ghandoor’s study [8]. For the rolled copper tubes painted in black mat, not selective. For the
cloudy day, our result is better than those of Norton et al. [41] and sunny day, with a selective absorber this variation at these same
Abd-Al Zahra and Joudi’s study [39] which are founding that the points is 26
C in the study of Nahar [5] and 12
C, 21
C and 27
C
maximum value of the outlet collector temperature is 42
C and respectively in Enaburekhan and Yakasai’s [46] works when R12,
22
C, respectively. It should be noted, moreover, that the collector ethanol and R-134a respectively are using as fluid.
with a reflector on the back of the insulator used for this study Fig. 8 presents the evolution of the mass flow rate in the primary
makes it possible to obtain a maximum variation between the circuit of both days. As one expects, the mass flow rate increase
temperature of the outlet water temperature of the primary fluid gradually with heat flux to reach its maximum in the middle of the
and that of ambient air of 55
C, when for the same points, this day and then decrease with the falling of the sun to reach their
variation is 42
C in the study of Esen and Esen [42] and 51
C in minima in the night. The mass flow rate is found to reach a
those of Sakhrieh and Al-Ghandoor [8]. maximum respectively at 12.40 am and 12.00 am for sunny day and
Fig. 6 presents the evolution of the temperature of the water in cloudy day. This maximum value is 0.010364 kg/s for sunny day and
the heat exchanger for sunny and cloudy day. We note an increase 0.00758 kg/s for cloudy day. These results are interesting compared
of this temperature for both days reach their maximal values to some values in the literature because, for the sunny day, the mass
respectively at 1:40 pm for both days then decrease with the falling flow rates are 0.0087 kg/s in Khalifa’s [38] study and 0.0065 kg/s in
of the sun. The maximal value of 74
C is reached for the sunny day that of Ong’s [47] and for cloudy day, these values are 0.005 kg/s in
when it is 50
C for the cloudy one. For the sunny day, the difference Abdull Azziz and Mukbel’s [48] study and 0.007 kg/s in that of
between the exchanger inlet water temperature and the ambient Khalifa and Mehdi [49].
temperature is 41
C, when it is 34
C at these same points in the The collector efficiency, which is the ratio of useful energy to
work of Pierson and Javelas [9] and 23
C in those of Hussein [10]. solar radiation, followed the same trend as the heat flux and useful
The storage tank temperature for the selected partly sunny and energy [46]. It increases until noon time and then decreases as
cloudy day is plotted in Fig. 7. shown in Fig. 9. As can be seen, the highest efficiency is satisfied
For the sunny day, the water temperatures increase rapidly with the sunny day where the useful heat is the highest, and the
around solar noon, and then decreased gradually whereas for the lowest efficiency is satisfied with the cloudy day one, where the
cloudy day, it increases slowly with the time. useful heat is the lowest. The maximum efficiencies occur respec-
The maximum storage tank temperature for the sunny and the tively at 12:40 am and 12:00 am true solar time for sunny day and
cloudy day during the test were. cloudy day. The maximum efficiencies values are respectively
55
C at 1:00 pm and 39.4
C at 3:00 pm respectively. Although 68.33% and 50% for sunny day and cloudy day. This trend is due to
solar flow shows some fluctuations, the water temperature showed increasing heat losses with time of day, accompanied with
increases with time. Thus, it can be concluded that the working decreasing net energy absorbed in the afternoon hours. Our results
fluid charged systems operate well even when sunlight is more or are better than those of Khalifa [38] and Abd-Al Zahra and Joudi’s
less continuous, which shows the ability of the system in providing [39] which are 52% and 64%, respectively for the sunny day. For the
hot water suitable for household use under such weather condi- cloudy day, the results obtained by Chen et al. [50] are below the
tions. It is noted, moreover, that for sunny day the outlet temper- maximum value for our performance is only achieved 20%.
ature in the storage tank goes up regularly to reach a maximum Fig. 10 presents the effectiveness versus (Tf1Ta/IT). This figure
value of 55
C, while the results of Hawas and Muneer [43] and permits to determine some physical characteristics of the system
Bargach et al. [44] indicate only one peak at 50
C and 38.5
C,
respectively. Similar results are reported by Hussein [45] and Abd- Table 2
Al Zahra and Joudi’s [39]. For the cloudy day, this temperature goes Efficiency equation and collector principal physical characteristics.
up regularly to reach a maximum value of 39.4
C, while the results
Efficiency equation FR (sa) FR FRUL UL
of Hawas and Muneer [43], Esen and Esen [42] and Norton et al. [41] (W/m2 K)
are 25
C, 30
C and 37
C, respectively.
Cloudy day h ¼ 0.7804.252(Tf1Ta)/IT 0.780 0.933 4.252 4.55
We have recorded, then, a variation of 22
C and 12.4
C (for Sunny day h ¼ 0.7774.689(Tf1Ta)/IT 0.777 0.929 4.689 5.05
sunny day and cloudy day respectively) for the water temperature
 P.M.E. Koffi et al. / Energy 64 (2014) 187e199 195

maximum efficiency value is of 65.05% for an optimal mass flow

rate of 0.0096 kg/s for the sunny day and 46.79% for an optimal
mass flow rate 0.007745 kg/s for the cloudy day. This may be
explained by the fact that the increase in mass flow rate is
accompanied by an increase in the convection heat transfer coef-
ficient to the fluid, thus enhancing the rate of heat transfer to fluid.
In previous works of the literature, this effectiveness is 38% in the
experimental study of Karaghouli [28], 50% in the theoretical study
of Kalogirou [53], 51.9% in experimental studies with a black
absorbing surface and 57% with a selective surface collector in
Nahar’s works [5,54].
Fig. 12 shows the evolution of the temperature of the absorber
tubes depending on the position of thermocouples as shown in
Fig. 2. Fig. 12 reveals that the tube temperature increases gradually
as one approach the flat plate collector output.
In order to judge the performance of the system, it is also
necessary to study the behavior of SWHS for rainy season. We are
Fig. 11. Collector efficiency with fluid mass flow rate.
thus going to study the system for two various days which are:

and particularly the total thermal loss coefficient (UL) for both days. - First day (30/09/2010): partly sunny, partly cloudy, rainy and
The two curves obtained are a straight line. By identification of the cloudy weather condition;
equation of these curves, with Eq. (12), from its slope and its y - Second day (04/09/2012): partly cloudy, sunny, rainy and cloudy
intercept, one can, in addition to the total thermal loss coefficient, weather condition.
deduce the system effectiveness factor. These values, presented in
Table 2, are in good agreement with those of Pierson and Javelas [9], Figs. 13 and 14 present the heat fluxes, the ambient temperature
Karaghouli and Alnasser [28] and of Benallou and Bougard [17]. (Ta), the collector outlet fluid temperature (Tf2), the exchanger inlet
Actually, according to Duffie and Beckman [51] and Tiwari et al. fluid temperature (Tf3), the outlet tank temperature of water in the
[52], the quality of a solar thermal collector is determined accord- storage tank (Tw), the mass flow rate (m) _ versus time, for the two
ing to the pair of values of the intercept FR(sa) and the slope FRUL. type of day previously defines. Table 3 presents the maximum
So, for a good collector, this pair of values is 0.8 and 4.5 W/m2
C values obtained. It is observed that the system behavior is mainly
respectively. These parameters are 0.6 and 8.5 W/m2
C, respec- dependent up on the state of sky. The collector outlet fluid tem-
tively for a poor collector. The calculated average values of FR(sa) perature and the exchanger inlet fluid temperature follow the ra-
and FRUL are 0.780 and 4.252 W/m2
C respectively for the cloudy diation pattern. These results are interesting compared to some
day and 0.777 and 4.689 W/m2
C respectively for the sunny day. values in the literature because for the same type of day, the
This result shows the undeniable qualities of the solar collector maximum value of the outlet collector temperature is 35
C in Abd-
realized. Al Zahra and Joudi’s [39] study and the maximum value of the inlet
Variation of collector efficiency with operating parameter for heat exchanger temperature is 50
C in Pierson and Javelas’s [9]
different mass flow rates is illustrated in Fig. 11. This figure shows one.
the strong dependence of the system effectiveness on the hot fluid For both days, the outlet tank temperature of water in the
mass flow rate, which increases with this effectiveness. The storage tank increases steadily, irrespective of the fluctuating

Fig. 12. Temperature variation along the length of the solar collector.
196 P.M.E. Koffi et al. / Energy 64 (2014) 187e199

Fig. 13. Incident heat flux, mass flow and temperatures with time of day.

Fig. 14. Incident heat flux, mass flow and temperatures with time of day.

radiation pattern. The maximum water temperature occurs Those Figs. 13 and 14 show a large variation in the thermosiphon
respectively around 1:30 pm and 12:20 am for the first day and for flow rate because of variation in incoming solar radiation. This
the second day after which it may drop by 1 or 2
C due to heat loss result in a large variation in the useful energy received from solar
from the storage tank only. We have recorded, then, a variation of radiation and transferred to the water during this period as shown.
8
C and 12.3
C (for the first day and the second day respectively) One of the most important features of thermosiphon solar water
for the water temperature in the storage tank and environment heater according to Abd-Al Zahra and Joudi [39] is that the flow rate
with an absorber made of rolled copper tubes painted in black mat, is automatically regulated according to the incoming solar radia-
not selective. For the same type of day, this variation at these same tion, and that the flow stops when the sky is overcast with a heavy
points are 7
C and 13
C in the study of Abd-Al Zahra and Joudi cloud cover. This condition is clearly observed in Figs. 13 and 14
[39], 10
C in the study of Hawas and Muneer [43], and 3
C, 12
C respectively in which the flow rate dropped to zero during the in-
and 21
C respectively in Enaburekhan and Yakasai’s [46] works terval between 3:00 pm and the sunset for the first day and be-
when R12, ethanol and R-134a respectively are using as fluid. tween 2:00 pm and 4:00 pm for the second day. It was raining

Table 3
Maximum values of incident heat flux, efficiency, mass flow rate and different Table 4
temperatures. Efficiency equation and collector principal physical characteristics.

IT (W/m2) Ta (
C) Tf2 (
C) Tf3 (
C) Tw (
C) _ (kg/s)

m h (%) Efficiency equation FR(sa) FR FRUL UL (W/m2 K)

First day 1063 36.9 76.9 68.9 49.2 0.0091 60.65 First day h ¼ 0.4941.023(Tf1Ta)/IT 0.494 0.591 1.023 1.731
Second day 1137 38.2 73.8 63.4 46.2 0.009833 57.67 Second day h ¼ 0.5811.606(Tf1Ta)/IT 0.581 0.694 1.606 2.31
 P.M.E. Koffi et al. / Energy 64 (2014) 187e199 197

Fig. 17. Heat exchanger effectiveness versus the time of day.

Fig. 15. Total heat transfer rate versus time of day.

transfer rate in the internal heat exchanger increase gradually with

during these periods with an overcast sky. Also, both the outlet
heat flux to reach its maximum in the middle of the day and then
temperature of the collector and in the inlet temperature of the
decrease with the falling of the sun to reach their minima in the
heat exchanger are observed to decrease during this period below
night. The obtained curve follows the same trend as those of Soo
the outlet tank temperature of water in the storage tank as shown
Too et al. [11].
those Figs. 13 and 14. This is because heat is lost to the ambient
Fig. 16 presents the evolution of the heat exchanger overall
atmosphere from the collector with no energy absorbed.
ðUAÞ’hx versus average tank temperature. The overall heat transfer
Maximum efficiency values of 60.65% and 57.67%for the first day
coefficient-area product of the heat exchangerðUAÞ’hx evolves
and the second day. Our results are better than those of Enabure-
gradually with the average temperature of the water of storage.
khan and Yakasai’s [46]. For him, maximum efficiency values of
With an initial value of 7.9 W/K for a temperature of 39.4
C, it
21%, 40.9% and 48.7% for R12, ethanol and R-134a were recorded at
increases until its maximal value to 40 W/K with a temperature of
daytime, respectively. Our results are also similar of studies of Bo-
47.1
C then falls until 11.51 W/K at 51.1
C. The cause of the ðUAÞ’hx
ren chen et al. [55].
variation is due to the change of viscosity of water with tempera-
Table 4 gives the efficiency equation and collector principal
ture on the tank side. The variation of the overall heat transfer
physical characteristics. The results show that for the two days
coefficient-area product ðUAÞ’hx for an internal heat exchanger
chosen, our results are less good than the conventional systems
during a test under controlled indoor conditions was found to be
because in general, performance of solar system in rainy season is
8e40 W/K (Fig. 14) with a correction factor of 0.95 [57]. The evo-
poor comparatively at the typically cloudy and sunny days.
lution is similar to the works made by Soo Too et al. [11] that use
forced circulation in their system (active systems).
6.2. Measurement of heat exchanger characteristics
Fig. 17 shows the variation of the heat exchanger effectiveness
with solar time. It increases until noon time and then decreases
The coefficient of thermal exchange of the heat exchanger, Ue, is
rapidly as the collector water temperature increases and decreases,
determined using Eq. (15). The coefficient of thermal exchange of
respectively. The effectiveness of heat exchanger is found to be
the heat exchange is estimated to be 149.15 W/m2 K. This result is in
between 6% and 97% as shown in Fig. 17. The average daily heat
agreement with those found in the literature. According to Roulet
exchanger effectiveness obtained is 70%. This result is interesting
[56], the transfer of heat taking place by natural convection in
compared to some values in the literature. Pierson [9] who worked
primary circuit and in the secondary circuit, the coefficient of
with an external annular heat exchanger, obtained in the same
transfer is of the order of 150 W/m2 K.
weather conditions, efficiency included between 30.8% and 95.1%
Fig. 15 presents the evolution of the total heat transfer rate (Q_ )
with an average daily heat exchanger effectiveness of 75.8%. Roulet
in the internal heat versus time of day. As one expects the total heat
[56] who made his study with an immersed heat exchanger
showed that a good heat exchanger has to have a heat exchanger
effectiveness superior to 50%. Haltiwanger and Davidson [2], which
also used in their works an immersed heat exchanger, showed after
1 h of measure that the efficiency of the used heat exchanger is
understood between 61% and 68% with an average of 65.1%. As for
Syed et al. [29], they showed in their works during the sunny day
that, heat exchanger effectiveness was found to be between 60%
and 70% with an average of 63%.
The mean daily efficiency (hd) of a solar water studied is 49.80%
for the sunny day. Table 5 compare this value to those obtained in
the literature by other authors concerning the systems with col-
lectors made of a selective surface absorber and have similar di-
mensions as ours, for a sunny day. It is wise to indicate that the
absorber of the collector of the studied system is made of copper
tubes painted black mat but not selective. Experimental results
Fig. 16. Evolution of the heat exchanger overall ðUAÞ’hx versus average tank
show that the mean daily efficiency of a thermosiphon flat-plate
temperature. solar water heater with an internal heat exchanger in our study is
198 P.M.E. Koffi et al. / Energy 64 (2014) 187e199

Table 5
Different results of mean daily efficiency for a sunny day.

Systems Flat-plate system without Flat-plate system with a mantle Our study All-glass evacuated
a heat exchanger [5] heat exchanger [2] tubular system [2]

Mean daily efficiency (%) 57 50.28 49.80 48.16

near 50%, which is lower than that of a thermosiphon flat-plate References

solar water heater without heat exchanger, but higher than that
of an all-glass evacuated tubular solar water heater. From these [1] Zhai XQ, Wang RZ, Dai YJ, Wu JY, Xu YX, Ma Q. Solar integrated energy system
for a green building. Energy Build 2007;39(8):985e93.
results, one can conclude that our SWHS containing an internal [2] Haltiwanger Julia F, Davidson Jane H. Discharge of a thermal storage tank
heat exchanger is a good one. using an immersed heat exchanger with an annular baffle. Sol Energy
2009;83:193e201.
[3] Zerrouki A, Boumedien A, Bouhadef K. The natural circulation solar water
7. Conclusion heater model with linear temperature distribution. Renew Energy 2002;26:
549e59.
[4] Belessiotis V, Mathioulakis E, Papanicolaou E. Theoretical formulation and
A thermosiphon SWHS with an internal heat exchanger has experimental validation of the inputeoutput modeling approach for large
been fabricated from locally available materials and tested under solar thermal systems. Sol Energy 2010;84:245e55.
[5] Nahar NM. Year round performance and potential of a natural circulation type
real climatic conditions. Studies were carried out by considering of solar water heater in India. Energy Build 2003;35:239e47.
two types of days: a sunny day and the other cloud. The daily ir- [6] Roberts DE, Forbes A. An analytical expression for the instantaneous efficiency
radiations of sunny and cloudy day chosen are 5575 Wh/m2/day of a flat plate solar water heater and the influence of absorber plate absorp-
tance and emittance. Sol Energy 2012;86:1416e27.
and 1858 Wh/m2/day, respectively. It is observed that the system [7] Tang Runsheng, Cheng Yanbin, Maogang Wu, Li Zhimin, Yu Yamei. Experi-
behavior is mainly depend up on the state of sky (sunny or cloudy mental and modeling studies on thermosiphon domestic solar water heaters
days). with flat-plate collectors at clear nights. Energy Convers Manag 2010;51:
2548e56.
The heat flux reaches its peak in the middle of the day for both [8] Sakhrieh A, Al-Ghandoor A. Experimental investigation of the performance of
days and then begins decreasing. The reached maximal values are five types of solar collectors. Energy Convers Manag 2013;65:715e20.
respectively of 1233 W/m2 for the sunny day and 300 W/m2 for the [9] Pierson P, Javelas R. Theoretical and experimental study of a natural circula-
tion type solar water heater with exchanger. Rév Gén Therm 1983;258e259:
cloudy one when for ambient temperature, the maximum values
467e82.
recorded are respectively 33
C for sunny day and 27
C for cloudy [10] Hussein HMS. Transient investigation of a two phase closed thermosiphon flat
day. plate solar water heater. Energy Convers Manag 2002;43:2479e92.
The temperature of the water in the heat exchanger, the tem- [11] Soo Too YC, Morrison GL, Behnia M. Performance of solar water heaters with
narrow mantle heat exchangers. Sol Energy 2009;83:350e62.
perature at the outlet of the solar collector, the mass flow rate and [12] Huang Jinbao, Shaoxuan Pu, Gao Wenfeng, Que Yi. Experimental investigation
the collector efficiency, follow the same trend as the heat flux. As on thermal performance of thermosyphon flat-plate solar water heater with a
one expects their increase gradually with solar intensity to reach mantle heat exchanger. Energy 2010;35:3563e8.
[13] Wong LT, Mui KW, Guan Y. Shower water heat recovery in high-rise resi-
their maximum in the middle of the day and then decrease with the dential buildings of Hong Kong. Appl Energy 2010;87:703e9.
falling of the sun to reach their minima in the night. [14] Pluta Zbyslaw, Pomierny Wlodzimierz. The theoretical and experimental
For the sunny day, the water temperature in the tank increases investigation of the phase-change solar thermosiphon. Renew Energy
1995;6(3):317e21.
rapidly around solar noon, and then decreased gradually whereas [15] Ayompe LM, Duffy A. Thermal performance analysis of a solar water heating
for the cloudy day, it increases slowly with the time. system with heat pipe evacuated tube collector using data from a field trial.
The experimental results of the performance test presented Sol Energy 2013;90:17e28.
[16] IEA. CO2 emissions from fuel combustion highlights. ed.; 2012.
above show that the system reaches an efficiency of 68.33% for the [17] Benallou A, Bougard J. Guide of solar energy: the solar thermics to the service
sunny day and 50% for the cloudy day when one obtain maximum of durable development. Québec: IEPF; 1990.
water temperatures at the outlet of the solar collector of 88
C in [18] Tiéné M. Problématique du développement de l’énergie solaire en Côte
d’Ivoire: Etat des lieux, analyse et recommandations. Mémoire de fin de cycle
the case of sunny day and of 58
C in the cloudy one. The maximum
Ingénieur-INP-HB Yamoussoukro(Côte d’Ivoire); 2004 [Solar energy devel-
water temperature in the storage tank reached 55
C and 39.4
C opment questions in Côte d’Ivoire: Studies, analysis and recomandations].
respectively for sunny and cloudy days, with a thermosyphonic [19] Seka JC. SODEXAM, Société d’Exploitation et de Développement Aéroportuaire
mass flow rate of 0.010364 kg/s for sunny day and 0.00758 kg/s for aéronautique et Météorologique. Service Banque de Données Météo, (Airport
meteorological data exploitation and développent society). Meteo Data Ser-
cloudy day. The values of the thermal performances parameters vice; 2001. p. 10.
FR(sa) and FRUL were 0.780 and 4.252 W/m2
C respectively for the [20] Gupta MK, Kaushik SC. Performance evaluation of solar air heater having
cloudy day and 0.777 and 4.689 W/m2
C respectively for the sunny expanded metal mesh as artificial roughness on absorber plate. Int J Therm Sci
2009;48:1007e16.
day. [21] Badescu V. Optimal control of flow in solar collector systems with fully mixed
The coefficient of exchange thermal of the heat exchanger Ue water storage tanks. Energy Convers Manag 2008;49:169e84.
found is 149.15 W/(m2 K) when, the variation of the overall [22] Malhotra A, Garg HP, Palit A. Heat loss calculation of flat plat solar collectors.
J Therm Energy 1981;2:2.
exchanger UA’hx during a test under controlled indoor conditions [23] Duffie JA, Beckman WA. Solar engineering of thermal processes. New York:
was found to be 8e40 W/K with a correction factor of 0.95. The Wiley; 1991. p. 197e249.
average heat exchanger effectiveness obtained is 70%. The experi- [24] Duffie JA, Beckman WA. Solar engineering of thermal processes. 3rd ed. New
York, USA: Wiley; 2006.
mental results show that mean daily efficiency of the thermosiphon [25] Hottel HC, Whillier A. Evaluation of flat plate solar collector performance. In:
flat-plate solar water heater with an internal heat exchanger is near Trans. of conference on the use of solar energy, University of Arizona 1958.
50%, which is lower than that of thermosiphon flat-plate solar p. 74e104.
[26] Duffie JA, Beckman WA. Solar energy thermal processes. New York: Wiley; 1974.
water heaters without heat exchanger, but higher than that of all
[27] Cardonnel C. Solaire actif et passif, conception, bilan thermique, habitat
glass evacuated tubular solar water heaters. (active and passive solar design, heat balance, habitation). Editions Par-
These results prove that the thermosiphon SWHS with an in- isiennes, CFP; 1983 chaud, froid, plomberie.
ternal heat exchanger is suitable under the weather conditions in [28] Karaghouli AA, Alnaser WE. Experimental study on thermosiphon solar water
heater in Bahrain. Renew Energy 2001;24:389e96.
Côte d’Ivoire. These performances added to the relative simplicity [29] Syed A, Izquierdo M, Rodrıguez P, Maidment G, Missenden J, Lecuona A, et al.
of the system manufacturing and the absence of moving parts make A novel experimental investigation of a solar cooling system in Madrid. Inter J
it an interesting technological solution. Refrig 2005;28:859e71.
 P.M.E. Koffi et al. / Energy 64 (2014) 187e199 199

[30] Mertol A, Place W, Webster T. Detailed loop model analysis of liquid solar [44] Bargach MN, Tadili R, Boukallouch M. Survey of thermal performances of a
thermosiphons with heat exchangers. Sol Energy 1981;27(5):367e86. solar system used for the heating of agricultural greenhouses in Morocco.
[31] Sacadura JF. Initiation aux transferts thermiques, technique et documentation. Renew Energy 2000;20:415e33.
CAST (Centre d’Actualisation Scientifique et Technique), INSA de Lyon; 1993. [45] Hussein HMS. Optimization of a natural circulation two phase closed ther-
4ième tirage. mosiphon flat plate solar water heater. Energy Convers Manag 2003;44:
[32] Kreith F. Principles of heat transfer. 3rd ed. New York: Harper and Row; 1976. 2341e52.
[33] FRANCKEN JC. On the effectiveness of a flat plate collector. Sol Energy [46] Enaburekhan Joseph, Yakasai Usman Tanko. Performance evaluation of a
1984;33(3/4):363e6. refrigerant-charged integrated solar water heater in northern Nigeria. Desa-
[34] Koffi PME. Conception et réalisation d’un chauffe eau solaire thermosiphon. lination 2009;243:208e17.
Etude comparative d’un isolant classique et d’un isolant à fibre végétale [47] Ong KS. A finite difference method to evaluate the thermal performance of a
[Design and realization of a thermosyphon solar water heater. Comparative solar water heater. Sol Energy 1974;6:137e48.
study of a traditional insulation with a vegetable fiber insulation]. Thesis. [48] Abdull Aziz GM, Mukbel MA. Thermal performance of solar water heater
Institut National Polytechnique Houphouët Boigny Yamoussoukro-Côte system in Yemen. Renew Energy 1994;4(2):241e7.
d’Ivoire; 2008. [49] Khalifa AJN, Mehdi MM. On the verification of a hydrogen bubble flow meter
[35] Kline J, Clintock Mc FA. Describing uncertainties in single sample experiments. used for monitoring flow rates in thermosiphon solar water heaters. Energy
Mech Eng 1953;1:3e9. Convers Manag 1998;39(12):1295e302.
[36] Jia CX, Dai YJ, Wu JY, Wang RZ. Experimental comparison of two honey- [50] Chen BR, Chang YW, Lee WS, Chen SL. Long-term thermal performance of a
combed desiccant wheels fabricated with silica gel and composite desiccant two phase thermosiphon solar water heater. Sol Energy 2009;83:1048e55.
material. Energy Convers Manag 2006;47:2523e34. [51] Duffie JA, Beckman WA. Solar energy thermal processes. New York: Wiley;
[37] Enibe SO. Performance of a natural circulation solar air heating system with 1980.
phase change material energy storage. Renew Energy 2002;27:69e86. [52] Tiwari RC, Kumar A, Gupta SK, Sootha GD. Thermal performance of flat plate
[38] Khalifa AJN. Thermal performance of locally made flat plate solar collectors solar collectors manufactured in India. Energy Convers Manag 1991;31:309e
used as part of a domestic hot water system. Energy Convers Manag 1999;40: 13.
1825e33. [53] Kalogirou Soteris A. Papamarcou Christos. Modeling of a thermosiphon solar
[39] Abd-Al Zahra HAA, Joudi KA. An experimental investigation into the perfor- water heating system and simple model validation. Renew Energy 2000;21:
mance of a domestic thermosiphon solar water heater under varying oper- 471e93.
ating conditions. Energy Convers Manag 1984;24(3):205e14. [54] Nahar NM. Capital cost and economic viability of thermosyphonic solar water
[40] Pillai IR, Banerjee R. Methodology for estimation of potential for solar water heaters manufactured from alternate materials in India. Renew Energy
heating in a target area. Sol Energy 2007;81:162e72. 2002;26:623e35.
[41] Norton B, Eames PC, Lo SNG. Alternative approaches to thermosiphon solar [55] Chen Bo-Ren, Chang Yu-Wei, Lee Wen-Shing, Chen Sih-Li. Long-term thermal
energy water heater performance analysis and characterization. Renew Sus- performance of a two-phase thermosyphon solar water heater. Sol Energy
tain Energy Rev 2001;5:79e96. 2009;83:1048e55.
[42] Esen M, Esen H. Experimental investigation of a two-phase closed thermosi- [56] Roulet CA. Chauffe-eau solaires. 3003 Berne: Office fédéral des questions
phon solar water heater. Sol Energy 2005;79:459e68. conjoncturelles; 1988.
[43] Hawas MM, Muneer T. Year round performance of thermosiphon solar water [57] Incropera FP, DeWitt DP. Introduction to heat transfer. 3rd ed. John Wiley and
heater in Benghazi. Energy Convers Manag 1984;40(3):237e42. Sons; 1996.