Impact of dust and wind on heat rejected from solar panel

Abstract

Keywords: Dust Density; Heat Losses; Response Surface Methodology; Wind Direction; Wind Speed

Introduction

As a result of the exhaustion of natural resources, such as crude oil and gas, the world is switching to renewable energy sources [1-3]. The importance of renewable energy sources, such as solar, wind, biomass, and hydropower, is increasing globally [4-6]. The arid region is an ideal location for constructing large photovoltaic power plants because it has an abundance of sun energy and inexpensive land resources [4]. However, extreme atmospheric conditions in these locations may have a negative effect on the ability of solar cells to generate electricity [6,7]. Rate of energy conversion of the commercially available PV technology typically ranges between five and twenty percent. Regrettably,

this rate is reduced because the PV panel is constantly functioning at high temperatures. Typically, the efficiency degrades by 0.25%–0.5% for every degree Celsius [8,9]. Numerous studies have been conducted over the past decade to determine how different environmental conditions affect the performance of solar modules [9]. Additionally, the electrical competence of photovoltaic modules can be significantly affected by the operating temperature of the solar cells because of the characteristics of the crystalline silicone used in solar panels. The amount of energy that these cells produce decreases at high temperatures [10]. A large portion of solar radiation is converted into heat by photovoltaic cells, which has a negative effect on PV production [11].

*Corresponding author. *E-mail address: soudali2010@mu.edu.iq This paper was recommended for publication in revised form by Editor-in-Chief Ahmet Selim Dalkilic Published by Yıldız Technical University Press, İstanbul, Turkey © Author. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

When the temperature of the surface module rises above a set threshold, the competence of the solar PV array dramatically declines [12]. The surface temperature of the PV panel rises quickly as solar radiation strikes it, which causes a sharp decline in the power output and PV panel efficiency [13,14]. Photovoltaic (PV) module performance decreases due to overheating, which is caused by a substantial amount of incoming solar radiation being transformed to heat [15-18]. Surface temperature and dust affect the effectiveness of conventional solar cells in generating electricity. The competence of PV power generation can be increased by lowering the surface temperature of solar cells and removing surface dust. The removal of dust and cooling have significant effects on the production of solar power, according to previous studies [18]. Xu et al. [4] studied how dust accumulation affected the temperature of the module. According to the results, under identical environmental conditions, the dust-deposited PV module temperature increased by approximately 64.7% in comparison to the clean PV module temperature. Xu et al. [4] noticed that an increasing wind speed causes an exponential drop in the photovoltaic glass’s final temperature. Vasel and Iakovidis [19] found that with all other effective elements remaining constant, the power output of the solar plant increased dramatically in the presence of a southerly breeze. Kaldellis et al. [20] found that when wind speed increases, the difference of temperature between the cell and the ambient temperature diminishes. Jha and Tripathy [21] discovered that a dusty panel’s efficiency reduction on a windy sunny day was 19.5%. Mekhilef et al. [22] found more heat may be evacuated from the PV cell surface by increasing wind speed. Furthermore, wind spreads and lifts dust, which causes shadowing and subpar performance of the PV cells. Goverde et al. [23] investigated how wind speed affects an average decrease of the module’s surface temperature. These procedures resulted in temperature drops of 11, 16, and 21 °C, with wind speeds of 1, 16, and 5 m s¯¹, respectively. Goossens et al. [24] examined how soiling affected the wind’s ability to cool and its implications for solar module performance. On a glass-glass module, cooling and electrical characteristics were assessed for wind velocities of up to four m/s, and soiling grades of up to forty-one g/m². Soiling can help photovoltaic modules cool down and the wind’s impact on the module’s performance.

Experimental Setup

The trials were conducted at the Material Laboratory of the Engineering Faculty of the Universiti Putra Malaysia (UPM). Apparatus A sun simulator was used to obtain lighting. The simulated lighting system was constructed at the foundry laboratory of the engineering faculty. Solar irradiance obtained

using LED lamps. Recent advancements in light-emitting diode (LED) lamps have made them a viable option for replicating solar radiation [25]. The light source system consists of three units of LED Flood Light lamps holding the trademark name Kontena, with a power output of 50 W each. The dimensions of the holder of the LED lamps were 1040 mm, 25 mm, and 500 mm with an equal distance of 100 mm between each bulb. The lamps were arranged in a single row. Monocrystalline solar module with 36 solar cells. The technical specifications of the module are listed in Table 1. A monocrystalline solar module was positioned directly underneath the sun simulator at a distance of 380 mm, with both LED lamps and a solar panel tilted at 10° [26]. Temperature was measured using an infrared thermometer (RS-820) at one test point (on the centre of surface) of PV module with a resolution and accuracy of 0.1 °C and ± 1.0 °C, respectively. An industrial fan (Mahita SF-650) at three speeds was used to generate wind velocities of 0 – 4 m/s. Additionally, an out-fan regulator was used to obtain accurate speeds. A handheld anemometer (Custom TL-300) was utilised to gauge the wind velocity and environmental temperature with resolution of 0.1 m s¯¹ and 0.1 °C, respectively. Solar charge controller to protect the circuit. Battery to save the electrical power produced and then provide it to the load. The Photovoltaic inverter transforms DC power into AC energy. The load can be powered by DC or AC power. The rig of the experimental system is shown in Figure 1. Preparation of the Dust A sample of dust from the University of Putra Malaysia soil was collected. The concrete fragments and other large solids were manually removed. A sieve with a 0.6 mm opening was used to filter out larger particles. The sample was then heated in a furnace (Memmert UFB 400) at 105 °C for 24 h to ensure that it was dry and humidity-free. Subsequently, the sample was milled to obtain a homogeneous sample. This was accomplished using a Smooth Double Roll Crusher from the Premur Group. Subsequently, the sample was heated in a furnace (Memmert UFB 400) at 105 °C for 24 h to remove moisture.

Table 1. Technical characteristics of the photovoltaic module Maxson solar panel

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

Figure 1. Experimental rig system (1) AC power, (2) Fan regulator, (3) Fan, (4) Solar panel with sun simulator, (5) Solar charge controller, (6) Battery, (7) Photovoltaic inverter, (8) Load. Dust was added at various densities (0 - 8 g/m²). The dust was weighed using a balance (AND GF-300) with a maximum capacity of 310 g and a high precision of 0.0001. Every experiment in this study used a sieve with a 0.6 mm opening to spread the dust evenly. The solar panel was cleaned of dust at the end of each test run. The dust was then collected and disposed of.

Experimental Design

A rotatable central composite design (CCD) with a full factorial method was utilized in order to plan the study’s trials and create a mathematical model. This model describes

the association between heat losses (response) together with the studied factors: dust density, wind speed, and wind direction. Minitab 19 was used as a tool. A set of 26 experiments were conducted with base block one, ten cube centre points, 8 axial points, and 8 cube points. In which α = 1.41421 and a level of significance = 0.05. The independent factors (predictors) employed in this study were dust density (A), wind speed (B), and wind direction (C), while the dependent factor (response) was heat losses. The pace at which dust accumulates varies depending on the site environment. The dust densities at different places, periods, and seasons are shown in Table 2 [27]. Furthermore, the average wind velocity atlas in the

Table 2. Dust densities at different places, periods, and seasons Country

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

Table 3. Minimum and maximum values of input factors Coded factor

* natural (free) stream (wind speed was 0.01 m/sec and was considered zero). ** wind direction: 1 in front of solar panel and 2 rear of solar panel.

Middle East at 10 m above ground is less than 4.5 m /s with the exception of a few places [25, 28,29]. For these reasons, in this study, the rate of dust density was 0 - 8 g/m², and the wind speed was 0 - 4 m/s. Table 3 displays the number of autonomous factors at minimum and maximum values.

The software (DOE) was used, and the results are listed in Table 4. Table 4 displays the factors without a response. Experiments were conducted, and the surface temperatures of the solar panel and air were recorded after 10 s for each run [4]. Each surface temperature represents the average of the five readings recorded. Furthermore, the solar panel

Table 4. The central composite design and heat losses values for the experiment Std

* Wind direction: 1 in front of solar panel and 2 back of solar panel.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

was cooled and cleaned of dust for each run. Subsequently, the heat losses were calculated. Heat losses include radiation and convection losses. It is the sum of the heat radiation loss and heat convection loss. Heat losses were calculated using equations (1-9) [33]. Radiation loss (QR) represents the heat exchanged between the glass layer of the PV panel and the environment, and can be computed in the manner described below: (1) The heat convection loss can be calculated using equation (2). (2) Local Nusselt number was used to find the value of the convective heat transfer coefficient. For forced convection: Nu = f (Re, Pr) while for free convection: Nu = f (Gr, Pr), where Nusselt, Reynold, Prandtl and Grashof numbers are dimensionless. For forced convection: (3) Nux Indicate local values of the Nusselt number based on the fluid’s characteristics and the distance from the leading edge of the plate. (4) (5) Rex represents the local Reynolds number at any x position. For free convection: (6)

where the C and m values that are given for every case. (7) (8) (9)

Results And Discussion

Mathematical Model The values of heat losses were recorded in table 4. Through experimentation and analysis using CCD technique, the quadratic model that predicts the connection between the heat losses and the independent variables utilised in this study was created using the values listed in table 4. The outcomes indicate that the generated quadratic model is well-matched with the empirical data, with R² value 93.20 %, adjusted R² value 90.01 %, and a lack-offit p-value of 0.085. Equation 10 represents general (both directions). Equations 11 and 12 represent direction 1 and direction 2, respectively. Heat losses = - 4.07+3.726 A+6.882 B+1.355 C1-1.355 C2-0.3596A²- 0.880 B²- 0.257A B -0.150 A C1+ 0.150 A C2-1.387 B C1+1.387 B C2 (10) heat losses = -2.72+3.576 A+5.50 B-0.3596 A²-0.880 B²-0.257AB (11) heat losses = -5.43+3.877 A+8.27 B-0.3596 A²-0.880 B²-0.257AB (12) Plotting predicted versus observed heat losses figures allows for additional confirmation of the model’s correctness. The responses’ residual plot values are displayed in Figures 2–6.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

Figure 3. Residual plots for heat losses (a) normal probability, (b) versus fits, (c) histogram, and (d) observation order.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

Figure 5. Main effects plot for heat losses: (a) effect of dust density, (b) effect of wind speed, and (c) effect of wind direction.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

The residual plot’s normal probability is displayed in Figure 3a. In this figure, the points fall near the line. This suggests that the distribution of the data is normal and deviation from normality is small, therefore proving the projected model’s accuracy. The fit of the residual plot is disclosed in Figure 3b. The data shown in this figure is to be randomly distributed with no discernible pattern, suggesting that the variance is constant, and that validates the projected model’s accuracy. The residual vs. order plot is displayed in Figure 3d. It is possible to determine whether or not the variables are associated with one another from this plot. The data dispersion at random verifies that there is no correlation between the variables, indicating the accuracy of the projected model. A Pareto chart can be used to show and compare the significance of each variable’s standardised effect level on the heat losses (Fig. 4). It illustrates that the independent factors have a declining effect on the elimination percentage according to the order of: wind speed, wind direction, and dust density. The influence of the investigated independent factors on the heat losses (response); are displayed in the main effects plot of these factors (Fig. 5). It demonstrated that heat losses increased as dust density rose within a given range. Any additional rise in dust density above that range, however, will have a detrimental effect on heat losses. Heat losses were increased when wind speed was increased to a particular range. Afterward, with each subsequent increase in wind speed, these losses decreased. In the wind direction, the heat losses in direction one is less than in direction two. The links between the experimental variables and how they influence the response are shown in Figure 6. Effect of Dust Density The heat produced by the solar cells passes through the dust particles on the surface of module before exiting. A

fraction of the heat is retained during this process because some of it is transmitted to the dust particles [24]. As shown in Figure 5a, positive effects of the dust density within the range of 0 - 4 g/m² on heat losses, due to the small quantity of dust particles on surface of module. At density above 4 g/m², the heat losses decreased because the dust particles absorbed a portion of this heat. Effect of Wind Speed As shown in Figure 5b, the heat losses increase with rising wind velocity from zero up to 3 m/s, due to the increasing variation in temperature between the solar panel’s surface and the surrounding. At wind speed above 3 m/s the heat losses decreased because the temperature of PV glass dropped significantly as wind speed increased [4], which means reducing the temperature differential between the solar panel’s surface and the surrounding. Effect of Wind Direction Figure 5c shows the heat losses at direction one was less than at direction two, due to the air blowing in front of the solar panel at direction one, which means reducing the variation in temperature between the solar panel’s surface and its surroundings, while the air blowing in the rear of the solar panel at direction two and slight effect on temperature difference.

Optimisation Process

The response surface methodology is a statistical design that works effectively for data analysis and mathematical model prediction, whereas the CCD is a useful tool for optimising independent factors and getting the best possible response. The relationships between variables and their influences on heat losses are demonstrated through surface and

Figure 7. Surface plot of heat losses vs utilised variables for direction one.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

contour plots of heat losses, as shown in Figures 7 and 8 for direction one, Figures 9 and 10 for direction two. For direction one, Figure 8 illustrates that the region with 4.0 - 6.0 w/m² is the one with the best (lowest) heat losses. The zone limited by 0 - 5 w/m² for direction two is where the optimal (lowest) heat losses are found, as Figure 10 displays. These figures were used to estimate the optimal conditions, and Table 5 ‘s parameters were used to carry out the

optimisation method. Table 6 displays the optimal circumstances necessary to get the optimal response (minimum heat losses of 4.08). In Figure 9, the optimisation plot of the minimum heat losses with the highest, lowest and optimal values for independent factors were presented. In Figure 11, represented the minimal heat loss optimisation plot.

Figure 8. Contour plot of heat losses vs utilised variables for direction one.

Figure 9. Surface plot of heat losses vs utilised variables for direction two.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

Figure 10. Contour plot of heat losses vs utilised variables for direction two.

Sigma J Eng Nat Sci, Vol. 43, No. 4, pp. 1075−1087, August, 2025

Conclusion

This study examined heat losses that were rejected from solar panels in two wind directions, and heat losses connected to the factors via full quadratic regression correlation. In addition, the optimal conditions were established and the impact of each individual variable on the response was examined. These were the conclusions reached: v The mathematical pattern between heat losses and the factors was anticipated and its accuracy was tested using response surface methods and the central composite design for the experiments. v Dust density positively impacts on heat losses within the range of 0 - 4 g/m², and negatively impacts on heat losses when it exceeds 4 g/m². v The wind speed positively affects on heat losses within the range of 0 - 3 m/s, and when it exceeds 3 m/s there is a slight descent in heat losses. v In general, the heat losses at direction one was lower than at direction two, due to the air blowing in front of the solar panel at direction one, while it was blowing in the rear for solar panel at direction two. v The optimal conditions for the minimum heat losses were 1.145 g/m² of dust density and a wind velocity of 0.757 m s¯¹, and wind direction two.

Nomenclature

cp g Gr h Imp Isc k Nu Nux Pmax Pr QC QR Re Rex Ta Ts u Voc X

Specific heat (KJ/Kg °C) acceleration of gravity (m/s²) Grashof number Convective heat transfer coefficient (W/m² °C) Current at Pmax (A) Short circuit current (A) Thermal conductivity (W/m °C) Nusselt number Local Nusselt number Maximum power (W) Prandtl number Heat convection loss (W/m²) Heat radiation loss (W/m²) Reynold number Local Reynold number Ambient temperature (°C) Surface temperature (°C) Velocity (m/s) Open circuit voltage (V) distance from ledge (m)

Greek symbols Β Temperature coefficient (k¯1) µ Viscosity (Kg/s m) σ Stefan Boltzmann constant 5.68x10¯⁸ W/m²K⁴ ε Emissivity of the glass ρ Density (kg/m³)

Subscripts a Ambient mp Maximum power oc Open circuit sc Short circuit s Surface

Abbreviations

AC Alternative current C1 Direction 1 C2 Direction 2 CCD Central composite design DC Direct current DOE Design of experiment LED Light emitting diode PV Photovoltaic module UHI Urban heat islands UPM Universiti Putra Malaysia

Acknowledgements

We would like to express our gratitude to the reviewers for their precise and succinct recommendations that improved the presentation of the results obtained.

Data Availability Statement

The authors confirm that the data that supports the findings of this study are available within the article. Raw data that support the finding of this study are available from the corresponding author, upon reasonable request.

Conflict Of Interest

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethics

There are no ethical issues with the publication of this manuscript.

Share and Cite