Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Contents lists available at ScienceDirect

Journal of King Saud University – Engineering Sciences

journal homepage: www.sciencedirect.com

Original article

Investigating the cooling of solar photovoltaic modules under the

conditions of Riyadh
A. Almuwailhi a,b,⇑, O. Zeitoun a,c
a
Mechanical Engineering Department, College of Engineering, King Saud University, Riyadh, Saudi Arabia
b
Renewable Energy Center, Water and Energy Research Institute, King Abdulaziz City for Science and Technology, Riyadh, Saudi Arabia
c
K.A.CARE Energy Research and Innovation Center at Riyadh, Riyadh, Saudi Arabia

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

Article history: Cooling enhances the energy conversion efficiency and output of photovoltaic (PV) panels. In this work,
Received 10 November 2020 the effects of natural convection, forced convection, and evaporative cooling on the performance of poly-
Accepted 15 March 2021 crystalline PV panels were investigated. The output and efficiency of a cooled PV panel were monitored
Available online xxxx
and compared to those of an uncooled PV panel under the same conditions. The cooling was conducted
using an insulated channel installed below a PV panel. Natural convection cooling was investigated for
Keywords: various channel air gaps (H = 30, 60, 90, and 120 mm). Natural convection currents in the cooling chan-
Photovoltaic
nels were capable of cooling the panel with wide air gaps. In forced convection cooling, the air was intro-
PV cooling
PV efficiency enhancement
duced by fans installed at the bottom opening of the cooling channel with various air velocities (ua = 1, 2,
Passive cooling and 3 m/s). Evaporative natural convection cooling was performed by a wetted fabric along the lower sur-
Active cooling face of the cooling channel, whereas evaporative forced convection cooling by pushing air along the wet-
ted lower surface of the channel. The experimental data showed that the panel efficiency and output
increased due to cooling. The experimental results of natural convection cooling revealed that the use
of an air gap of 120 mm to cool the solar panel contributed to an increase in the panel daily energy pro-
duction and efficiency by 1.7% and 1.2%, respectively. For forced convection cooling, using air at a speed of
3 m/s increased the daily energy production by 4.4% and the efficiency by 4%. Natural convection evap-
orative cooling increased the daily energy production and the efficiency by 3.6% and 2.7%, respectively.
Forced convection evaporative cooling contributed, at a speed of 2 m/s, to an increase in the daily energy
production by 3.8% and an increase in efficiency of 3.8%.
Ó 2021 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction 2013). The absorbed thermal energy can raise the PV module tem-
perature up to 40 °C above the surrounding temperature (Stropnik
The power received from the Sun is about 1.8
1011 MW, and Stritih, 2016; Makki et al., 2015). The increase in the temper-
which equals 10,000 times the current energy demand rate of ature of the PV modules degrades panel efficiency and output
Earth (Parida et al., 2011; Mirzaei et al., 2014). Photovoltaic (PV) power (Schwingshackl et al., 2013). Based on the PV cell type,
panels are used to convert solar energy into electrical energy. An the reduction in output power varies between 0.2 and 0.6% per
ideal PV panel converts 6–20% of the incident solar irradiance into degree rise in the module temperature (Home Energy, 2019). In
electricity (Dubey et al., 2013). However, nearly 50% of the incident addition, a high panel temperature for extended periods can cause
solar energy is absorbed by the panels as heat (Chandrasekar et al., delamination between its layers, glass breakage, bus bar (BB) fail-
ure, ethylene vinyl acetate (EVA) browning, moisture ingress, and
diode failure (Bouaichi et al., 2019; Phinikarides et al., 2014;
⇑ Corresponding author. Rajput et al., 2016; Wohlgemuth et al., 2006) and can lead to a loss
E-mail address: [email protected] (A. Almuwailhi). of 9.5 years from the module lifetime (Ogbomo et al., 2018). Find-
Peer review under responsibility of King Saud University. ing a solution for the overheating of the PV module will give the
confidence to the PV constructors and will accelerate the transfor-
mation process from the conventional to renewable photovoltaic
energy. This transformation is expected to need 40–45 years
Production and hosting by Elsevier (Zekry, 2020). Thus, the negative impact of the temperature

https://doi.org/10.1016/j.jksues.2021.03.007
1018-3639/Ó 2021 The Authors. Production and hosting by Elsevier B.V. on behalf of King Saud University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article as: A. Almuwailhi and O. Zeitoun, Investigating the cooling of solar photovoltaic modules under the conditions of Riyadh, Journal of
King Saud University – Engineering Sciences, https://doi.org/10.1016/j.jksues.2021.03.007
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Nomenclature

E Daily energy (Wh) Greek symbols

ECP Generated energy from the cooled panel (Wh) x Humidity ratios (kg of water vapor/Kg of dry air)
Ef Energy consumed by the cooling fans (Wh) DE Daily energy difference between two panels (Wh)
ENP Generated energy from the uncooled panel (Wh) DTav Two panels average temperature difference (°C)
H Air gap height (mm) DTmax Maximum temperature difference between two panels
I Current (A) (°C)
k Thermal conductivity (W/m∙K) Dg Efficiency difference (%)
P Power (W) mc-Si Microcrystalline silicon
TCP Average temperature of the cooled panel (°C) ṁ Mass flow rate (kg/s)
TNP Average temperature of the uncooled panel (°C) b Temperature coefficient of maximal power (%/°C)
ua Air velocity (m/s) g Module efficiency (%)
V Voltage (V) q Density (kg/m3)
W Cooling channel width (m)
Abbreviations
Subscripts AM Air mass
a Air a-si Amorphous silicon
av Average BB Electrical bus bar
CP Panel with isolated box (Cooled panel) BIPV Building-integrated photovoltaic
ex Exit CdTe Cadmium telluride
f Electrical fans CFD Computational fluid dynamics
in Inlet EVA Ethylene vinyl acetate
max Maximum KACST King Abdulaziz City for Science and Technology
mp Maximum power pint MC4 Multicontact connectors
NP Normal panel (Non-colded panel) m-si Monocrystalline silicon
oc Open circuit NOCT Normal operating cell temperature (°C)
sc Short circuit P-si Polycrystalline silicon
w Water PV Photovoltaic
RTD Resistance temperature detector
STC Standard test conditions
XPS Extruded polystyrene foam

increase on the module should be minimized to enhance PV mod- mance of the rooftop-installed PV modules using computational
ule performance. For cold regions, PV modules can be cooled using fluid dynamics (CFD) and found that the average module tempera-
natural convection. However, cooling PV modules installed in hot ture decreased with increasing tilt angle and air gap. In addition,
regions is challenging. CFD results indicated that the average panel temperature
PV modules are very sensitive to relatively high ambient tem- decreased with increasing the module length to at least 0.08 m
peratures. The rise in the surrounding temperature reduces the for air gaps. However, the maximum PV temperature increased
output power and efficiency of PV modules. (Mazón-Hernández with increasing panel length. To reduce the possible overheating
et al., 2013) investigated PV cooling by natural and forced convec- of the PV modules, (Gan, 2009) suggested the minimum air gaps
tion using a rectangle steel channel installed beneath the PV mod- for single- and multi-panel installations to be 0.12–0.15 m and
ule. They characterized the electrical behavior of the solar panels to 0.14–0.16 m, respectively.
improve their performance for roof installations. For natural con- Building-integrated photovoltaics (BIPVs) installed on building
vection, they found that the air gap of the cooling channel beneath roofs or walls play a key role in building shading and consequently,
the PV module must be wide enough to accommodate natural con- decrease air conditioning load. The air gap between the PV panel
vection currents needed to cool the PV module. They concluded and roof or wall controls the convection heat transfer from the
that natural and forced convection cooling increased the module module to the roof and wall. The studies showed a 38% saving in
peak power by 7.5% and 15%, respectively. the annual cooling load and a 2.5 °C decrease in the inner roof tem-
(Cuce et al., 2011) experimentally investigated the effects of perature under the PV module (Dominguez et al., 2011).
passive cooling on the performance of silicon solar cells. They used For BIPV systems, (Lau et al., 2012) investigated the effect of the
an aluminum heat sink to dissipate heat from a PV cell. The dimen- air gap and inclination angle on PV module cooling. The buoyancy-
sions of the heat sink were determined based on the steady-state driven flow in the cooling gaps below the PV modules was consid-
heat transfer analysis. Their results showed that the performance ered similar to the flow in a heated open-ended channel. In this
of PV cells enhanced as the ambient temperature fell with and investigation, the channel inclination angle ranged from 15° to
without the heat sink. At a solar radiation of 800 W/m2, using 90°, and a channel height-to-gap ratio of 20 was used. (Lau et al.,
the heat sink achieved an increase of 20% in the PV cell power out- 2012) found that the open-ended channels with small inclination
put. The maximum cooling effect was observed at a solar radiation angles were characterized by a small chimney effect, which
of 600 W/m2. damped the natural currents. Therefore, the heat transfer along
The air gap beneath a PV module plays a primary role in remov- the PV panels decreased for small inclination angles.
ing heat from the panel via natural convection currents. Many (Moshfegh and Sandberg, 1998) investigated natural convection
numerical and experimental studies were conducted to determine along the lower surface of PV panels. In their numerical and exper-
the optimum air gap beneath a PV module to ensure optimal work- imental investigations, they considered heat exchange by radiation
ing conditions. (Gan, 2009) studied the air gap effect on the perfor- using a module with a length of 7.0 m and an air gap of 0.23 m. Dif-

2
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Fig. 1. Ambient temperature, relative humidity, and the global radiation on July 6, 2020, in Riyadh, Saudi Arabia.

Fig. 2. PV module test setup. (a) Natural convection and natural convection with evaporation experiments, (b) forced convection and forced convection with evaporation
experiments. The panels on the left are uncooled, and those on the right are cooled using the cooling channel.

ferent input heat fluxes and emissivity of the bounding surfaces for different module technologies. They found that polycrystalline
were considered. They concluded that the surface under the PV and amorphous silicon modules had the maximum normal operat-
panel received 30% of the heat emitted from the solar panel while ing cell temperature (NOCT). Moreover, they considered the power
the rest was dissipated through convection. degradation due to temperature. They found that the polycrys-
(Schwingshackl et al., 2013) studied the effect of wind on PV talline module had the highest power degradation due to the mod-
module cooling. For several cell types, they investigated the effect ule temperature increase. For this type of panel, the data of
of the wind speed and direction and ambient temperature on cell (Schwingshackl et al., 2013) indicated a power degradation of
temperature and output power. They found that the more wind 0.45% per degree rise in the module temperature.
was present; the power plant became more efficient. Also, they cal- The performances of PV panels in a windy location were com-
culated the maximum normal operating cell temperature (NOCT) pared with those in an unwindy location by (Gökmen et al.,

3
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Table 1 to be evaporated. The results indicated that the output voltage of

Mechanical Characteristics of Panels. the PV modules increased by 19.4% and the power output by 19.1%.
Solar Cells Polycrystalline silicon 156.75
156.75 mm (6 in.), 5 Bus Bar (Haidar et al., 2018) experimentally studied the effect of the
No. of Cells 72 (6
12) evaporative natural convection cooling of a PV panel under
Dimensions 1,956 mm
992 mm
40 mm (77
39.05
1.57 in.) Riyadh’s outdoor conditions. The evaporative cooling was achieved
Weight 22 kg (48 lbs.) by attaching a piece of fabric to the bottom surface of the PV panel.
Front Glass 3.2 mm tempered glass This fabric was kept wet by dripping water using rubber hoses
Frame Anodized aluminum alloy
Junction IP67 rated (3 bypass diodes)
attached to the panel’s bottom surface. The latent evaporation heat
Box was utilized to absorb heat from the body of the tested PV module.
Cables 4.0 mm2 (0.006 in. 2), symmetrical length 1,000 mm (39.4 in.) They achieved a 20 °C reduction in PV panel temperature and a 14%
Connectors MC4 connectors increase in electrical power generation efficiency compared with a
similar PV module operated under the same conditions.
Cooling using air especially with forced convection is a good
solution. For example (Salaudeen, 2019) evaluated the perfor-
Table 2
Electrical Characteristics of Panels. mance of a thermal energy storage system with the use of paraffin
wax as the storage medium, and air as the heat transfer fluid.
STC (Standard test conditions; 1000 W/m2, 25 °C, AM1.5) Model # KS-
Salaudeen (2019) found that the air was the ideal heat transfer
320PC
fluid due to its availability and durability.
Optimum Operating Voltage (Vmp) (V) 36.39
Forced convection with evaporation cooling was investigated by
Optimum Operating Current (Imp) (A) 8.8
Open Circuit Voltage (Voc) (V) 44.95 (Haidar et al., 2016). They developed a theoretical model for con-
Short Circuit Current (Isc) (A) 9.24 current air and water films in a cooling channel installed below
Maximum Power @ STC (Pmax) (W) 320 the PV panel. The model simulated heat and mass transfer that
Module Efficiency (%) 16.5 occurred at the bottom of the PV panel. The results revealed the
Operating Module Temperature (°C)  40 to + 85
Maximum System Voltage (VDC) 1,000
effects of geometrical parameters, air flow rate, temperature, and
Maximum Series Fuse Rating (A) 20 humidity on the cooling process. They concluded that when the
Power Tolerance ±3% air inlet temperature decreased, the temperature of the PV panel
Maximum Power at NOCT (Pmax) 237 slightly decreased. A reduction of 6 °C in the PV panel temperature
Optimum Power Voltage (Vmp) 34.64
was obtained with cooling.
Optimum Operating Current (Imp) (A) 6.84
Open Circuit Voltage (Voc) (V) 43.04 In this study, the performances of PV modules using various
Short Circuit Current (Isc) 7.29 cooling techniques, namely natural convection, forced convection,
Nominal Operating Cell Temperature (NOCT) (°C) 45 ± 2 natural convection with evaporation, and forced convection with
Temperature Coefficient of Pmax (%/°C) 0.400 evaporation were experimentally investigated. The effects of these
Temperature Coefficient of Voc (%/°C) 0.330
Temperature Coefficient of Isc (%/°C) 0.058
cooling methods on the daily energy produced by the panel, peak
efficiency of the panel, and panel temperature were investigated.

2. Experimental setup
2016). They also studied the effect of changing the tilt angle of the
panel for various durations from 15 min up to a year. Their data The experiments were carried out in a testing facility (KACST PV
indicated that the maximum relative energy difference increased Laboratory) located in Riyadh, Saudi Arabia, which has significant
by about 3.55% for tilt angle moving periods of 15 min, 30 min, high temperatures and dry weather conditions. The tilt angle of
and one hour. the PV panels was set at 23°, which is the optimal fixed tilt angle
(Teo et al., 2012) conducted experiments to examine PV panel for Riyadh City. The experiments were conducted from March to
performance with and without active cooling. They observed that September which covers the hot dry conditions of Riyadh City.
without active cooling, the temperature of the modules was high, Fig. 1 shows a sample of the climate data collected by the weather
and a solar cell efficiency in the range 8–9% was achieved. How- plant of King Abdulaziz City for Science and Technology. This
ever, when the module was cooled by forced convection, the tem- weather station (VSN800 Weather Station with advanced sensor
perature significantly dropped, leading to an increase in the set) located in KACST PV Lab was used to obtain needed weather
efficiency between 12% and 14%. (Teo et al., 2012) developed a conditions. It was installed 2 m away from the experimental setup.
model to predict the module temperature, and the comparison Fig. 1 shows the ambient temperature, relative humidity, and hor-
between the measured and predicted temperature profiles indi- izontal global radiation for typical daytime solar radiation with a
cated good agreement. peak value around noon. The relative humidity and dry-bulb tem-
For natural convection evaporative PV panel cooling, perature reflect the typical Riyadh weather in the summer.
(Chandrasekar et al., 2013) developed a simple passive evaporative Two identical PV panels with 72 cells and 5 BBs were used. The
cooling system with a cotton wick installed on the bottom surface uncooled PV panel had an aluminum metal frame, and the cooled
of the PV panel. They studied the effect of evaporative cooling on panel had no metal frame. The first panel was used to examine
the thermal and electrical performances of the PV module. The PV performance without cooling. The second panel was supported
cooling system consisted of a cotton wick wetted by water, above an isolated channel, as shown in Fig. 2, to examine how cool-
Al2O3/water nanofluid, or CuO/water nanofluid. The temperature ing affects its performance. The two panels were identical and had
of the PV module was reduced to 59 °C and 54 °C. These results cor- typical power outputs. The panels’ data given by the manufacturer
responded to a reduction of 11% and 17%, respectively, in module are listed in Tables 1 and 2.
temperature compared to the module temperature obtained with- Initial power measurements were conducted to ensure that the
out cooling. two panels were properly working and had typical power output
(Alami, 2014) investigated a passive evaporative cooling tech- values. The verification was carried out using a Sun simulator
nique to control the temperature rise of the PV modules due to the under standard test conditions (STC: Irradiation = 1000 W/m2,
absorption of solar irradiance using a synthetic clay element sup- panel temperature = 25 °C AM = 1.5). The results are shown in
ported at the back of the module that allowed a thin film of water Fig. 3. The measured output powers of the panels were 309.2 and
4
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Fig. 3. I–V curve under STC for (a) uncooled panel (b) cooled panel.

304.2 W. This deviation is acceptable since the manufacturer

allows a 3% deviation in power measurement at STC.
The dimensions of the cooling channel installed below the sec-
ond panel are shown in Fig. 4. The open-ended cooling channel was
made of an extruded polystyrene foam (XPS) board with an adjus-
table height (H). A thermal insulation board (Thermal
conductivity = 0.029 W/m∙K (An, et al., 2017)) was used to build
this channel to minimize heat transfer through channel walls. In
natural convection and natural convection with evaporation exper-
iments, the bottom and top ends of the cooling channel were open,
as shown in Fig. 2a. For forced convection and forced convection
with evaporation experiments, the bottom end was equipped with
fans, as shown in Fig. 2b.
Calibrated T-type thermocouples were used to measure panels’
back surface temperature. Nine thermocouples were attached to
the back of each panel, as shown in Fig. 5. In addition, six thermo-
Fig. 4. Panel cooling channel dimensions. couples were used to measure the ambient temperature, air tem-

5
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Fig. 6. Calibration setup for the thermocouples.

Table 3
Experimental instruments and uncertainties.

Instrument or Parameter Unit Uncertainty

T- type Thermocouples °C ±0.125%
Air Speed Indicator (Pacer DA400 Anemometer) m/s ±12.5%
Panel Efficiency % ±1.3%
PV panel Energy Wh ±0.1%
Water Consumption L/h ±12.6%

peratures along the channels, and wet-bulb temperatures. Fig. 5

shows the location and given numbers of the thermocouples.
The thermocouples used were calibrated using a temperature
calibration unit (ISOTECH Hyperion 936 Basic) equipped with a
high accuracy RTD200 temperature sensor reference with an
uncertainty of ±0.05%. The calibration data indicated an uncer-
tainty of ±1.2 °C in the readings of the thermocouples.
Temperature data were collected using a data logger (Agilent
34972A), which recorded the measurement every 2 min. The tem-
perature data were recorded from sunrise to sunset.
The air velocity indicator (Pacer DA400 anemometer) was used
to measure air velocity at the cooling channel inlet and exit.
Microinverter (Enphase IQ7) was used to generate loads for the
PV panels. It converted DC to AC and delivered it to the grid system.
This module was equipped with a maximum power point tracking
(MPPT) system that allows the extraction of the maximum power
from the PV panels. Enphase IQ7 microinverter had a built-in mon-
itoring system, which indicates the actual produced power and
energy, which was used to monitor daily energy produced by the
panels.
A thermal camera (FLIR-T440) was used to produce the thermal
images of the PV panel surfaces at the peak of the Sun irradiation
(around 1:00 pm), i.e., at the time when the system was stable
and at the highest temperature.
Axial flow fans were used to force the air through the channel.
They were used in forced convection and forced convection with
evaporative experiments.
For the experiment with evaporation, a fabric was spread along
the bottom surface of the cooling channel. This fabric was kept wet
during the experiment. A submerged water pump was used to cir-
culate the water from the bottom water tank to the upper edge of
the cooling channel exit, where 12 nozzles were used to spread
water on the fabric.
To evaluate the accuracy of the main experimental findings, the
Fig. 5. Thermocouple locations and corresponding channel number for a) uncooled
panel and b) cooled panel. uncertainty propagation function built in EES software, Eq. (1), was

6
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Table 4
Summary of natural convection experiments.

Date H (mm) E (Wh) DE

ENP (%) g (%) Dg
gNp (%)
DTmax (°C) DTav (°C) ua (m/s) Mean
ECP ENP gCP gNP DE
ENP (%)
Dg
gNp (%)
DTmax (°C) DTav (°C)

8-Mar-20 30 1848 1842 0.3 13.7 13.9 1.5 3.3 1.2 0.3–0.9 0.6 0.7 1.6 0.8
9-Mar-20 1971 1947 1.2 13.6 13.5 0.4 1.9 0.2
10-Mar-20 1951 1933 0.9 13.4 13.4 0.4 0.7 0.2
8-Jun-20 1692 1695 0.2 14.6 14.8 1.3 0.7 2.4
6-Mar-20 60 1861 1852 0.5 13.6 13.6 0.4 0.5 0.1 0.3–0.4 1.2 0.2 1.4 0.0
7-Mar-20 1624 1608 1.0 16.4 16.7 1.5 3.6 0.8
11-Mar-20 1954 1913 2.1 13.2 13.1 0.8 1.0 1.6
7-Jun-20 1732 1715 1.0 14.7 14.6 0.4 3.5 0.8
3-Mar-20 90 2095 2053 2.0 13.7 13.5 1.1 2.2 2.1 0.5–1.6 1.5 0.6 0.1 1.3
4-Mar-20 1978 1957 1.1 13.3 13.3 0.4 0.5 1.3
5-Mar-20 1941 1916 1.3 13.4 13.4 0.0 1.7 1.3
6-Jun-20 1778 1750 1.6 15.0 14.8 0.9 0.5 0.3
12-Mar-20 120 1816 1771 2.5 13.0 12.8 1.5 3.1 2.1 0.7–1.2 1.7 1.2 2.0 1.2
13-Mar-20 1638 1613 1.5 14.8 14.6 1.2 3.0 1.5
14-Mar-20 785 775 1.3 15.1 14.8 2.0 1.8 0.9
4-Jun-20 1776 1754 1.3 14.8 14.8 0.0 0.1 0.1

Fig. 7. Thermal images of natural convection cooled and uncooled panels for a) H = 30 mm, b) H = 60 mm, c) H = 90 mm, and d) H = 120 mm.

used. The used sensors were calibrated before performing the 3. Results and discussion
experiments to estimate their uncertainties.
Assuming the individual measurements are uncorrelated and 3.1. Natural convection cooling
random, the uncertainty in the calculated quantity can be deter-
mined from, The effect of natural convection cooling on PV panel perfor-
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi mance was investigated for different cooling channel gaps
u
uX  @Y 2 2 (H = 30, 60, 90, and 120 mm). A summary of the experimental data
UY ¼ t UXi ð1Þ is listed in Table 4. This table shows the data collected for the
@X i
i cooled and uncooled panels. For each air gap, the experiment
was collected for a period of 4–7 days to get minimum ambient
where U represents the uncertainty of the variable Y. air velocity as much as possible. From the collected data, the data
The thermocouples used in the experiments were calibrated of the day with the lowest wind speed were considered in the data
using a temperature calibration unit (ISOTECH Hyperion 936 Basic) analysis.
equipped with a high accuracy RTD200 temperature sensor refer- The thermal images of the two panels’ upper surfaces are shown
ence as shown in Fig. 6. Based on the calibration data and for the in Fig. 7 for different air gaps. The thermal images indicated that
temperature range from 20 to 80 °C, the uncertainty in tempera- the natural convection currents (chimney effect) in the cooling
ture measurement was ±0.125%. channel were capable of cooling the panel better than that without
Considering the above procedure for experimental parameters, the cooling channel. Each PV panel had nine thermocouples
the uncertainties are listed in Table 3. attached to its back surface. The temperature of the panel was esti-
7
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

temperature. The horizontal dashed lines display the average panel

temperatures for the cooled panel (red) and uncooled panel
(black). As shown in these figures, the panel temperatures reached
the maximum values between 11:00 am and 2:00 pm, which was
about 30 °C higher than the ambient temperature. As shown in
Table 4 and Fig. 8, the daily average temperature difference (DTav)
between the cooled and uncooled panels ranged from 0.8 to
1.2 °C. The highest daily average temperature drop was observed
for a 120-mm air gap channel with a temperature drop of 1.2 °C.
The drop in the maximum temperature (DTmax) of the cooled
panel compared to the uncooled panel was calculated by

DTmax ¼ TNPmax  TCPmax ð2Þ

where TNPmax is the maximum temperature of the uncooled panel,

and TCPmax is the maximum temperature of the cooled panel.
As shown in Table 4, Figs. 7 and 8, the temperature of the cooled
panel was slightly lower than the uncooled panel, except for the
ones with gaps of 30 and 60 mm. This temperature difference
increased as the air gap increased. However, for the narrow air
gaps (H = 30 and 60 mm), the friction in the cooling channel
resisted natural convection currents. Consequently, the natural
convection around the uncooled panel was more effective than
that inside the cooling channel. As a result, utilizing channels with
30- and 60-mm air gaps reduced cooling efficiency. The data in
Table 4 show a variation ranging from 1.6 to 2 °C in the maxi-
mum temperature difference. The highest temperature difference
(2 °C) occurred for the 120-mm air gap.
The efficiency of the panels calculated during the power peak
time was calculated from

P=A
g¼
100 ð3Þ
Irr

whereg is the efficiency of the panel (%), P is the power (W), A is the
area of the panel (m2), and Irr is the solar irradiance (W/m2).
The efficiency improvement due to the cooling was estimated
from

Dg gCP  gNp
%¼
100 ð4Þ
gNp gNp

where Dg is the efficiency difference (%); gCP is the efficiency of the

cooled panel (%), andgNp is the efficiency of the uncooled panel (%).
It was found that the panel efficiency increased as the air gap
increased. The maximum efficiency occurred at an air gap of
120 mm with an improvement of 1.2%. For H = 30 and 60 mm,
the efficiency of the cooled panel was lower than that of the
uncooled panel since the cooling process was unable to cool the
panel.
The microinverters (Enphase IQ7) were used to generate loads
for the PV panels and monitor the daily generated energy. Table 4
shows the difference in the daily produced energy (DE) between
the cooled and uncooled panels. The energy enhancement percent-
age caused by the cooling process was calculated from

DE ECP  ENP
Fig. 8. Temperature histories of natural convection cooled and uncooled panels for %¼
100 ð5Þ
a) H = 30 mm, b) H = 60 mm, c) H = 90 mm, and d) H = 120 mm.
ENP ENP

where DE/ENP% is the energy enhancement percentage, ECP is the

daily energy generated from the cooled panel, and ENP is the daily
mated by averaging the readings of these nine thermocouples. The energy generated from the uncooled panel.
data collected for each air gap in four days are listed in Table 4. As shown in Table 4, the highest energy gain percentage was
Fig. 8 shows panel temperatures, the blue curve represents the obtained for H = 120 mm, where the cooled panel produced 1.7%
ambient temperature, the red curve represents the cooled panel more energy compared to the uncooled panel. On the other hand,
temperature, and the black curve represents the uncooled panel the lowest was 0.6%, which occurred for H = 30 mm.
8
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Table 5
Summary of 12 days of forced convection experiments (H = 120 mm).

Date ua (m/s) E (Wh) DE

ENP (%) g (%) Dg
gNp (%)
DTmax (°C) DTav (°C) Mean
ECP ENP g cp g Np DE
ENP (%)
Dg
gNp (%)
DTmax (°C) DTav (°C)

20-Mar-20 1 1656 1614 2.6 14.3 14.0 1.9 5.0 2.0 1.9 0.6 2.2 1.2
27-Mar-20 1764 1740 1.4 16.9 17.0 0.4 0.3 0.2
28-Mar-20 1504 1478 1.8 13.9 13.9 0.0 2.7 1.3
1-Jun-20 1825 1790 2.0 15.3 15.2 0.8 1.2 1.1
18-Mar-20 2 1622 1580 2.7 15.8 15.4 2.4 5.6 2.6 3.0 2.4 7.6 3.8
19-Mar-20 1967 1898 3.6 14.6 14.1 3.4 11.1 5.6
29-Mar-20 1916 1858 3.1 14.7 14.4 2.7 8.3 4.7
2-Jun-20 1817 1770 2.7 15.3 15.1 1.3 5.2 2.4
16-Mar-20 3 2099 1982 5.9 15.4 14.4 6.3 16.2 9.1 4.4 4.0 11.3 6.0
17-Mar-20 1917 1857 3.2 16.0 15.6 2.6 6.7 4.3
30-Mar-20 1990 1903 4.6 14.7 14.1 4.2 12.4 6.7
3-Jun-20 1848 1782 3.7 15.2 14.8 2.9 10.0 3.8

Fig. 9. Thermal images of panel surfaces for forced convection cooled and uncooled
panels at air velocities of a) 1 m/s, b) 2 m/s, and c) 3 m/s.
Fig. 10. Temperature histories of forced convection cooled and uncooled panels at
air velocities of a) 1, b) 2, and c) 3 m/s.
3.2. Forced convection cooling
tom opening was equipped with eight axial flow fans to push the
The effect of forced convection on the performance of the PV air from the bottom to the top of the channel. These fans were
panels was studied. The cooling channel width was fixed at operated using a monitored DC power source to supply the neces-
H = 120 mm during these experiments. The cooling channel bot- sary power and measure the consumed energy. The air fans con-

9
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Table 6
Summary of natural convection with evaporation experiments.

Date H E (Wh) DE
ENP
g (%) Dg
gNp (%)
DTmax DTav ṁw ua (m/s) Mean
(mm) (°C) (°C) (l/h)
ECP ENP (%) g cp gNp DE
ENP
Dg
gNp (%)
DTmax DTav ṁw
(%) (°C) (°C) (l/h)

30-Jun-20 30 1783 1742 2.4 15.1 14.9 1.3 3.0 3.2 0.8 0.38–0.42 2.5 1.6 3.2 3.5 0.8
1-Jul-20 1773 1734 2.2 15.0 14.8 1.3 2.5 3.3 1.0
2-Jul-20 1771 1722 2.8 15.3 15.0 2.2 3.9 4.0 0.7
3-Jul-20 1799 1756 2.4 15.1 14.9 1.7 3.5 3.4 0.6
26-Jun-20 60 1842 1786 3.1 15.0 14.6 2.5 5.1 4.9 1.6 0.93–1.61 3.2 2.1 5.2 4.7 1.3
27-Jun-20 1859 1800 3.3 15.3 14.9 2.5 5.7 4.9 0.3
28-Jun-20 1835 1781 3.0 15.2 14.9 1.7 4.8 4.3 1.7
29-Jun-20 1836 1777 3.3 15.3 15.0 1.7 5.1 4.5 1.7
22-Jun-20 90 1843 1780 3.5 15.5 15.2 2.1 4.3 4.0 0.8 0.62–1.39 3.4 2.5 5.0 4.2 0.8
23-Jun-20 1825 1768 3.2 15.5 15.1 2.9 5.3 4.1 0.8
24-Jun-20 1854 1787 3.7 15.6 15.2 2.5 5.0 4.5 –
25-Jun-20 1781 1728 3.1 15.7 15.3 2.6 5.6 4.3 –
18-Jun-20 120 1868 1810 3.2 15.3 15.0 2.5 4.9 4.3 1.4 0.24–0.3 3.6 2.7 5.3 4.5 1.2
19-Jun-20 1881 1811 3.9 15.3 14.8 3.3 4.6 4.6 1.4
20-Jun-20 1834 1765 3.9 15.7 15.2 3.0 5.7 4.9 1.0
21-Jun-20 1696 1640 3.4 15.4 15.1 2.1 5.9 4.3 –

Fig. 11. Thermal images for evaporative natural convection cooled and uncooled panels: a) H = 30 mm, b) H = 60 mm, c) H = 90 mm, and d) H = 120 mm.

sumed approximately 0.56 kWh per day. This value was excluded velocities are shown in Fig. 10. The forced convection cooling effect
from energy calculations. The performance of the PV panel was improved as the air velocity increased from 1 to 3 m/s. As indicated
examined for three air velocities (1, 2, and 3 m/s). Table 5 contains in Table 5 and Figs. 9 and 10, the temperature of the cooled panel
a summary of 12 days forced convection experiments. was lower than the uncooled panel in the examined air velocities.
The thermal images for the PV panel surfaces are shown in The maximum temperature difference (DTmax) between the
Fig. 9. These images were captured at the peak of the power pro- cooled and uncooled panel was about 2.2, 7.6, and 11.3 °C for air
duction at 1:00 pm. The air velocity in the cooling channel affects velocities of 1, 2, and 3 m/s, respectively. For the daily average tem-
the panel temperature, as can be seen in Figs. 9 and 10. A signifi- perature difference (DTav) between the cooled and uncooled pan-
cant drop in the temperature of the cooled panel can be observed els, the cooling by forced convection reduced the temperature to
on the right of the thermal images as the air velocity increased 1.2, 3.8, and 6 °C at air velocities of 1, 2, and 3 m/s, respectively,
to 3 m/s and the panel temperature dropped to 11 °C. as shown in Fig. 10.
As discussed before, the temperatures of panels were estimated The panel temperature drop was reflected in the efficiency
based on the readings of the thermocouples attached to their back improvement (gDg Þ of the cooled panel compared to the uncooled
Np
surface. The panel temperature histories for the examined air
panel. As shown in Table 5, the efficiency of the cooled panel

10
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

was higher than the uncooled panel by 0.6, 2.4, and 4.0% for air
velocities of 1, 2, and 3 m/s, respectively.
Table 5 shows the difference between the cooled and uncooled
panels in daily produced energy (EDNPE ). As indicated in Table 5, the
panel cooled at an air velocity of 3 m/s has the highest energy effi-
ciency compared to the uncooled panel, where a 4.4% enhance-
ment in the daily energy output was achieved.

3.3. Natural convection cooling with evaporation

Natural convection with evaporation was performed for differ-

ent air gaps by keeping the bottom surface of the cooling channel
wet. For each air gap (H = 30, 60, 90, and 120 mm), the experiments
were conducted for at least four days. The experiment procedure
was similar to that performed in the natural convection experi-
ments, except that the lower surface of the channel was wet. A
layer of fabric sheet was placed on the lower surface of the channel
to increase the evaporation effect. A water spray system was used
to keep the fabric surface wet using a circulation water pump. The
water pump was consuming nearly 0.14 kWh per day. This value
was excluded from energy calculations. The wet-bulb tempera-
tures at the channel inlet and exit were measured using the wet-
bulb temperature sensors. The humidity ratio was found using
EES software at measured dry- and wet-bulb temperatures. The
water evaporation rate was estimated from
m _ a ðxex  xin Þ
_w¼m ð6Þ
where ṁw is the water evaporation flow rate (kg/s), xin and xex are
the humidity ratios at the cooling channel inlet and exit, respec-
tively, and ṁa is the air mass flow rate in the channel (kg /s). The
air mass flow rate (ṁa) was estimated from
_ a ¼ q a H W ua
m ð7Þ
where qa is the air density (kg/m3) at channel exit conditions, ua is
the air velocity (m/s) measured in the channel, H is the channel
height (H = 0.03, 0.06, 0.09, or 0.12 m), and W is the channel width
(1.96 m).
Table 6 shows a summary of evaporative natural convection
experiments. The thermal images in Fig. 11 and the temperature
histories of panels shown in Fig. 12 indicated the effectiveness of
evaporation on PV cooling compared to natural convection without
evaporation. The heat transfer from the cooled panel was enhanced
because of the air humidification process. As shown in Table 6 and
Fig. 12, the cooling using evaporative natural convection reduced
the maximum temperature from 3.2 to 5.3 °C compared to the
uncooled panel. For the daily average temperature difference
(DTav) between the cooled and uncooled panels, the cooling by nat-
ural convection with evaporation helped to increase DTav to 3.5–
4.7 °C. The difference in panel temperatures was reflected in the
efficiency improvement (gDg Þ. As shown in Table 6, the efficiency
Np

of the cooled panel increased from about 1.6 to 3.6%. Table 6 indi-
cates that the daily energy produced from the cooled panel was
enhanced by 2.5–3.6% for air gaps of 30 and 120 mm, respectively.

3.4. Forced convection cooling with evaporation

Fig. 12. Temperature histories of evaporative natural convection cooled and

Panel cooling using forced convection with evaporation effect uncooled panels at air gaps of a) 30, b) 60, c) 90, and d) 120 mm.
was studied for a channel gap of 120 mm for air velocities of 1,
2, and 3 m/s. In this experiment, axial fans were used to push air
in the cooling channel while its bottom surface was maintained The thermal images in Fig. 13 show the effect of air velocity on
wet, as discussed in the previous section. the temperatures of cooled and uncooled panels. Moreover, Fig. 14
Table 7 indicates a summary of the results of the forced convec- shows the temperature history for different air velocities. A signif-
tion with evaporation experiments. The consumed water (ṁw) icant drop in the maximum temperature (DTmax) was observed
during the experiments is presented in Table 7. The water con- from 7.3 to 12.5 °C for air velocities of 1 and 3 m/s, respectively.
sumption varied from 1.2 to 4.9 l/h. Compared to the data of forced convection without evaporation,

11
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Table 7
Summary of forced convection with evaporation experiments (H = 120 mm).

Date ua (m/s) E (Wh) DE

ENP (%) g (%) Dg
gNp (%)
DTmax (°C) DTav (°C) ṁw (l/h) Mean
ECP ENP g cp gNp DE
ENP (%) Dg
gNp (%)
DTmax (°C) DTav (°C) ṁw (l/h)

5-Jul-20 1 1811 1747 3.7 15.4 15.0 2.6 9.1 5.1 0.8 3.1 2.5 7.3 4.6 1.2
6-Jul-20 1810 1733 4.4 15.3 14.7 3.9 9.3 6.1 0.6
7-Jul-20 1798 1732 3.8 15.1 14.7 2.6 9.2 5.5 0.7
8-Jul-20 1831 1758 4.2 15.2 14.7 3.0 7.9 5.1 0.4
19-Sep-20 1831 1788 2.4 12.5 12.3 1.6 6.6 3.9 2.5
20-Sep-20 1868 1828 2.2 12.4 12.2 2.0 5.6 3.8 1.6
21-Sep-20 1731 1692 2.3 12.4 12.1 1.9 4.7 3.9 0.8
22-Sep-20 1702 1667 2.1 12.4 12.1 2.0 5.8 3.5 2.4
9-Jul-20 2 1847 1762 4.8 15.4 14.7 4.8 11.7 6.5 0.5 3.7 3.5 10.3 6.0 2.1
10-Jul-20 1810 1730 4.6 15.4 14.7 4.6 10.0 5.9 0.9
11-Jul-20 1712 1640 4.4 15.5 14.9 4.0 9.1 5.4 1.3
12-Jul-20 1761 1696 3.8 15.2 14.8 2.6 7.4 4.9 1.9
12-Sep-20 1871 1823 2.6 12.3 12.0 2.4 10.2 5.4 2.3
14-Sep-20 1864 1800 3.6 12.4 11.9 4.0 12.6 7.0 3.5
15-Sep-20 1873 1822 2.8 12.5 12.1 2.8 9.4 6.5 3.8
16-Sep-20 1870 1813 3.1 12.4 12.1 2.8 11.9 6.4 2.7
14-Jul-20 3 1616 1555 3.9 14.9 14.4 3.3 12.3 6.2 3.0 3.8 3.8 12.5 6.4 4.9
15-Jul-20 1741 1667 4.4 15.2 14.7 3.5 12.3 6.2 3.9
16-Jul-20 1797 1737 3.5 15.5 15.0 3.0 11.6 5.3 5.1
17-Jul-20 1872 1786 4.8 15.2 14.6 4.2 12.6 6.8 4.9
8-Sep-20 1908 1850 3.1 12.9 12.5 3.5 11.5 6.6 8.6
9-Sep-20 1889 1817 4.0 12.4 11.9 4.0 14.1 7.3 7.7
10-Sep-20 1647 1590 3.6 10.5 10.0 5.0 14.4 5.8 1.3
11-Sep-20 1643 1588 3.5 12.3 11.8 3.8 11.5 6.6 4.4

evaporation facilitates panel cooling. The average daily panel tem- ment in panel efficiency is shown. The data in the figure indicate an
perature drop (DTav) observed was 4.6 °C at an air velocity of 1 m/s improvement of up to 4% in panel efficiency using forced convec-
and increased to 6.4 °C at an air velocity of 3 m/s. The average tem- tion at an air velocity of 3 m/s.
perature values are presented in Table 7 and Fig. 14. The comparison of the daily average energy is shown in Fig. 18.
The drop in panel temperatures was reflected on the efficiency In general, the examined cooling techniques assisted in increasing
improvement (gDg Þ of the cooled panel. As shown in Table 7, the the produced energy. The enhancement in the daily energy varied
Np
from 0.5 to 4.5% for the cooled panel compared to the uncooled
efficiency of the cooled panel was higher than that of the uncooled
panel.
panel by 2.5, 3.5, and 3.8% for 1, 2, and 3 m/s, respectively. The
summarized results in Table 7 show enhancement of 3.1, 3.7, and
3.8% in the daily produced energy at 1, 2, and 3 m/s, respectively. 5. Conclusions

The performances of polycrystalline PV panels were investi-

4. Comparison of cooling techniques gated under natural convection, forced convection, and evapora-
tive natural and evaporative forced convection cooling
The comparison of experimental data obtained on the investi- techniques. In natural convection cooling, the cooling was
gated four cooling techniques revealed that the cooled panel enhanced as the cooling channel width increased. For the 120-
temperature decreased as the air gap and velocity increased. In mm air gap cooling channel, a maximum daily energy enhance-
addition, the evaporative cooling assisted in reducing the panel ment of 1.7%, an improvement in the panel efficiency of 1.2%, a
temperature. maximum temperature drop of 2 °C, and an average daily temper-
In Fig. 15, the maximum temperature drops of the cooled panel ature drop of 1.2 °C were observed.
compared to the uncooled panel is presented. The data in the figure Forced convection cooling was improved as the air velocity
shows how the peak temperature of the cooled panel drops due to increased. A maximum daily energy enhancement of 4.4%, a panel
cooling. In forced convection cooling with evaporation (green efficiency improvement of 4.0%, a maximum temperature drop of
bars), the cooled panel temperature decreased by 12 °C compared 11.3 °C, and an average daily temperature drop of 6 °C were
to the uncooled panel. The daily average panel temperature observed at an air velocity of 3 m/s.
dropped with the proposed cooling methods, as shown in Fig. 16, Evaporative natural convection enhanced the cooling process
except for those with natural convection with air gaps of 30 mm compared to natural convection cooling. An energy output
and 60 mm. enhancement of 3.6%, a panel efficiency improvement of 2.7%, a
The critical parameter that should be addressed in these exper- maximum temperature drop of 5.3 °C, and an average daily tem-
iments was the efficiency of the PV panel. In Fig. 17, the improve- perature drop of 4.5 °C were observed for an air gap of 120 mm.

12
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

Fig. 13. Thermal images of evaporative forced convection cooled and uncooled
panels at air velocities of a) 1 m/s, b) 2 m/s, c) 3 m/s.

Fig. 14. Temperature histories of evaporative forced convection cooled and

uncooled panels at air velocities of a) 1, b) 2, and c) 3 m/s.

Evaporative forced convection cooling improved panel cooling

compared to forced convection cooling. At an air velocity of 3 m/
s, a maximum temperature drop of 12.5 °C, a daily average panel
temperature drop of 6.4 °C, an energy increase of 3.8%, and a panel
efficiency improvement of 3.8% were observed.
Among all, the forced convection cooling at an air velocity of
3 m/s showed the highest improvement in daily energy production
(4.4%) and panel efficiency (4.0%).
Further investigations may be needed for better PV module
cooling. Further investigations may include enhancing evaporation
technique, installing wetted fabrics on the back of the PV module
and investigating effect of air velocity on water droplet carry over.

Declaration of Competing Interest

The authors declare that they have no known competing finan-

cial interests or personal relationships that could have appeared Fig. 15. Comparison of the maximum temperature drops for four cooling
to influence the work reported in this paper. techniques.

13
A. Almuwailhi and O. Zeitoun Journal of King Saud University – Engineering Sciences xxx (xxxx) xxx

authors thank for King Abdulaziz City for Science and Technology
for access to facilities.

References

Alami, A.H., 2014. Effects of evaporative cooling on efficiency of photovoltaic

modules. Energy Convers. Manage. 77, 668–679.
An, W., Sun, J., Liew, K.M., Zhu, G., 2017. Effects of building concave structure on
flame spread over extruded polystyrene thermal insulation material. Appl.
Therm. Eng. 121, 802–809.
Bouaichi, A., Alami Merrouni, A., Hajjaj, C., Messaoudi, C., Ghennioui, A., Benlarabi,
A., Ikken, B., El Amrani, A., Zitouni, H., 2019. In-situ evaluation of the early PV
module degradation of various technologies under harsh climatic conditions:
The case of Morocco. Renewable Energy 143, 1500–1518.
Chandrasekar, M., Suresh, S., Senthilkumar, T., Ganesh karthikeyan, M., 2013.
Passive cooling of standalone flat PV module with cotton wick structures.
Energy Convers. Manage. 71, 43–50.
Cuce, E., Bali, T., Sekucoglu, S.A., 2011. Effects of passive cooling on performance of
Fig. 16. Comparison of average temperature drops for four cooling techniques.
silicon photovoltaic cells. Int. J. Low-Carbon Technol 6 (4), 299–308.
Dominguez, A., Kleissl, J., Luvall, J.C., 2011. Effects of solar photovoltaic panels on
roof heat transfer. Sol. Energy 85 (9), 2244–2255.
Dubey, S., Sarvaiya, J.N., Seshadri, B., 2013. Temperature Dependent Photovoltaic
(PV) Efficiency and Its Effect on PV Production in the World – A Review. Energy
Procedia 33, 311–321.
Gan, G., 2009. Effect of air gap on the performance of building-integrated
photovoltaics. Energy 34 (7), 913–921.
Gökmen, N., Hu, W., Hou, P., Chen, Z., Sera, D., Spataru, S., 2016. Investigation of
wind speed cooling effect on PV panels in windy locations. Renewable Energy
90, 283–290.
Haidar, Z.A., Orfi, J., Kaneesamkandi, Z., 2018. Experimental investigation of
evaporative cooling for enhancing photovoltaic panels efficiency. Results
Phys. 11, 690–697.
Haidar, Z.A., Orfi, J., Oztop, H.F., Kaneesamkandi, Z., 2016. Cooling of Solar PV Panels
Using Evaporative Cooling. J. Thermal Eng. 2, 928–933.
Home Energy, Glossary, PV Temperature Coefficient of Power Accessed: 15/12/
2019, www.homerenergy.com/products/pro/docs/latest/pv_
temperature_coefficient_of_power.html.
Lau, G.E., Sanvicente, E., Yeoh, G.H., Timchenko, V., Fossa, M., Ménézo, C., Giroux-
Julien, S., 2012. Modelling of natural convection in vertical and tilted
photovoltaic applications. Energy Build. 55, 810–822.
Makki, A., Omer, S., Sabir, H., 2015. Advancements in hybrid photovoltaic systems
for enhanced solar cells performance. Renew. Sustain. Energy Rev. 41, 658–684.
Mazón-Hernández, R., García-Cascales, J.R., Vera-García, F., Káiser, A.S., Zamora, B.,
2013. Improving the Electrical Parameters of a Photovoltaic Panel by Means of
Fig. 17. Comparison of panel efficiency improvement for four cooling techniques. an Induced or Forced Air Stream. Int. J. Photoenergy 2013, 1–10.
Mirzaei, P.A., Paterna, E., Carmeliet, J., 2014. Investigation of the role of cavity
airflow on the performance of building-integrated photovoltaic panels. Sol.
Energy 107, 510–522.
Moshfegh, B., Sandberg, M., 1998. Flow and heat transfer in the air gap behind
photovoltaic panels. Renew. Sustain. Energy Rev. 2 (3), 287–301.
Ogbomo, O.O., Amalu, E.H., Ekere, N.N., Olagbegi, P.O., 2018. Effect of operating
temperature on degradation of solder joints in crystalline silicon photovoltaic
modules for improved reliability in hot climates. Sol. Energy 170, 682–693.
Parida, B., Iniyan, S., Goic, R., 2011. A review of solar photovoltaic technologies.
Renew. Sustain. Energy Rev. 15 (3), 1625–1636.
Phinikarides, A., Kindyni, N., Makrides, G., Georghiou, G.E., 2014. Review of
photovoltaic degradation rate methodologies. Renew. Sustain. Energy Rev. 40,
143–152.
Rajput, P., Tiwari, G.N., Sastry, O.S., Bora, B., Sharma, V., 2016. Degradation of mono-
crystalline photovoltaic modules after 22 years of outdoor exposure in the
composite climate of India. Sol. Energy 135, 786–795.
Salaudeen, S.A., 2019. Investigation on the performance and environmental impact
of a latent heat thermal energy storage system. J. King Saud Univ. Eng. Sci. 31
(4), 368–374.
Schwingshackl, C., Petitta, M., Wagner, J.E., Belluardo, G., Moser, D., Castelli, M.,
Zebisch, M., Tetzlaff, A., 2013. Wind effect on PV module temperature: Analysis
of different techniques for an accurate estimation. Energy Procedia 40, 77–86.
Stropnik, R., Stritih, U., 2016. Increasing the efficiency of PV panel with the use of
PCM. Renewable Energy 97, 671–679.
Teo, H.G., Lee, P.S., Hawlader, M.N.A., 2012. An active cooling system for
Fig. 18. Comparison of enhancement in produced energy for four cooling photovoltaic modules. Appl. Energy 90 (1), 309–315.
techniques. Wohlgemuth, J.H., Cunningham, D., Nguyen, A., Miller, J., 2006. Long term reliability
of PV modules. The 4th IEEE World Conference on Photovoltaic Energy
Conversion 4968, 57-58.
Zekry, A., 2020. A road map for transformation from conventional to photovoltaic
Acknowledgments
energy generation and its challenges. J. King Saud Univ. Eng. Sci. 32 (7), 407–
410.
This work was supported by King Saud University, Deanship of
Scientific Research and Research Center College of Engineering. The

14