Abstract

This study investigated a novel method for increasing desalinated water mass in solar desalination plants. For this purpose, solar panels and a cylindrical parabolic collector (CPC) were used to raise basin water temperature. The effect of different components of basin solar still on freshwater mass was also investigated. The aluminum basin has been associated with maximum water desalination among the different materials constituting a basin. The effects of different colors (e.g. black, brown, and red) on the basin, as well as different water depths (5, 10, and 15 mm), were also explored. The highest amount of freshwater in the black aluminum basin at a 5-mm water depth was 2.97 kg/day. ANN modeling was employed to validate the experimental data, indicating good compliance of experimental data with ANN prediction. According to the results of the simulation with varying numbers of neurons (n = 2–25), the highest and lowest agreement between experimental data and ANN prediction data were related to 24 and 10 neurons, respectively. Under optimum conditions, R2 and %AAD error were 0.993 and 2.654, respectively.

HIGHLIGHTS

  • Solar energy has been used for seawater desalination.

  • Different materials, water depths, and basin colors were studied for solar desalination.

  • Artificial neural network simulation has been performed.

  • The basin with aluminum material, black color, and saline water depth of 5 mm had the highest water desalination during the experiments.

INTRODUCTION

Drinking water production is of paramount importance. The high percentage of the Earth's surface seawater allows turning seawater into drinking water using various sweetening techniques. Seawater desalination is used in remote locations as well as different places for the sake of simplicity. The efficiency of solar stills can also be increased in a variety of ways.

Two seawater desalination processes were investigated (Bagheri et al. 2019). These experiments found that the higher the power of the solar panels, the higher the temperature of the saline water in the basin, resulting in a higher desalination rate. The effects of the direct use of the thermal element in the basin were examined (Bagheri et al. 2020a). The use of a heating element that receives its energy from solar panels raises the water temperature in the basin and, consequently, the volume of desalinated water. Artificial neural networks (ANNs) and mathematical modeling are used to evaluate and validate experimental data (Bagheri et al. 2020b). This research found that the volume of desalinated water decreased in the last days of summer due to a reduction in solar radiation. ANN simulation was also found to be more accurate than mathematical modeling.

The effect of ambient temperature parameters on the seawater desalination rate was investigated (Refalo et al. 2014). A comparison was made between experimental and theoretical data. The results indicated that the distillation rate increases with increasing ambient temperature. The effects of different absorbers on water desalination in solar still were explored (Hansen & Murugavel 2017). The results suggested that the fin-shaped absorber was more efficient than the flat absorber.

ANN-based solar still optimization has been studied (Mashaly et al. 2015). The theory of ANNs has been fully articulated, and various inputs have been defined for simulation. An ANN has been introduced as a powerful tool to predict solar still performance. The effects of side walls on sweetening has been surveyed (Feilizadeh et al. 2017). This paper also employs mathematical modeling. The modeling results matched well with the experimental data. The results indicated that the yield was higher in summer than in winter.

Solar still thermal performance has been analyzed (Mashaly & Alazba 2017). This paper presents simulation error formulas. This paper points out that using ANN modeling can save time and money. A new solar seawater desalination method was presented (Fathy et al. 2018). This method utilizes a parabolic trough collector (PTC) to increase the basin water temperature and thus increase the seawater desalination rate. Oil flows in the pipes in the PTC to heat water in the basin, increasing the efficiency of the solar seawater desalination process.

The stepped solar-powered desalination system was examined using experimental and mathematical modeling (Muftah et al. 2018). Efficiency, water desalination, heat transfer coefficients, glass coating temperature, and water temperature were examined and compared. The results showed that the post-modification system was more efficient than the pre-modification system. Experimental studies, cost analysis, and simulation of solar-powered water desalination systems were performed (Rashidi et al. 2017). The results indicated that small vortices increase the efficiency of the solar-powered desalination system by increasing both heat and mass transfer.

An ANN was modeled using meteorological data (Santos et al. 2012). Meteorological data related to different months were collected. This paper introduces an ANN to predict solar still performance. It was argued that adding a fan could increase the mass of desalination water (Selvaraj & Natarajan 2018). This paper also addresses the effects of wind speed on the process of solar water desalination. Wind speed affects the amount of heat transfer between the environment and the glass coating. Increasing the wind speed increases the heat transfer rate and condensation in the distillation process.

The economic analysis of the stepped solar still was discussed (El-Agouz 2014). This paper specifies that the total dissolved solids (TDS) of seawater is very low after distillation. The formula for calculating system efficiency is presented in this paper. These experiments indicated a significant reduction in conductivity of seawater and saline water after desalination. Multi-stage stacked tray solar still water desalination was investigated (Chen et al. 2017). This paper also examines the effects of water depth. The results showed that the use of shallower basin water increased the mass of desalination water. The solar PV was shown to increase the efficiency of the solar-powered desalination system (Pounraj et al. 2018). The results indicated that the temperature of the ‘conventional solar PV’ was higher than the ‘proposed solar PV’ at different times of the day.

The mechanical vapor compression (MVC) desalination system was investigated, and the mathematical modeling of this system was expressed (Shen et al. 2014). Examining the ‘compressed steam temperature vs. saturation temperature difference’ figure shows that the temperature in the ‘without water injection’ mode is higher than in the ‘with water injection’ mode. Different solar stills were studied (Rufuss et al. 2018). Based on efficiency decomposition diagrams, a comparison between 20 solar still models shows that the transportable hemispherical solar still had the highest efficiency.

A comprehensive study was conducted on the pyramid solar still (Nayi & Modi 2018). This paper discusses the results of water desalination. This study showed that water desalination in the pyramid solar still is less expensive and more efficient than the conventional solar still. The energy and exergy analysis of a solar distillation system was performed (Ibrahim et al. 2017). This paper also introduces system optimization. The volume of desalinated water during the year and in nine different cases was compared. According to the results, the maximum energy efficiency was 27.6%, which occurred in August.

A study surveyed the efficiency of a solar-powered desalination system and the effects of climatic conditions thereupon (Hamed et al. 2016). The results indicated the lowest and highest amounts of thermal energy absorbed by the receiver in January and June, respectively. The cost of freshwater production by the tubular solar still with a semi-circular absorber (TSS-SC) is less than that by the tubular solar still with a flat absorber (TSS-FP) (Elshamy & El-Said 2018). A comparison between thermal efficiency at different test hours shows that the thermal efficiency of TSS-SC has always been higher than that of TSS-FP. A comparison of exergy efficiency measured on different days also shows that the exergy efficiency of TSS-SC has always been higher than that of TSS-FP.

An economic analysis of various solar-powered water desalination systems was performed (Kabeel et al. 2010). This paper also compares the average volume of desalinated water during the day for different types of solar still. This research found that the lowest and highest desalinated water production costs were related to the pyramid-shaped solar still and modified solar still with sun-tracking, respectively. The use of fins in solar stills increases the amount of freshwater (Velmurugan et al. 2008). This paper found that the use of wicks in solar stills can increase the volume of desalinated water.

In Bagheri et al. (2020a), two processes of solar water desalination and economic analysis were examined. This paper comprehensively studied different constituents of the basin during different experiments. Moreover, different basin colors and saline water depths in the basin and their effects on water desalination were explored. To evaluate the experimental results of this paper, ANN simulation was employed.

This paper introduced a novel approach to seawater desalination. Herein, different genera were used to construct the solar still basin, and the effect of varying basin colors and water depths on the amount of freshwater was studied. Several solar panels and a cylindrical parabolic collector (CPC) are used to raise basin water temperature. This paper used a new ANN modeling method to validate the data from different experiments conducted from October 26 to November 30, 2018. During the ANN-based simulation, the accurate results of experimental data were obtained using the prediction data of ANN modeling.

MATERIALS AND METHODS

Experiments on a solar desalination plant were carried out in the autumn of 2018 at 1850 meters above sea level at Yasuj city (30°40′06″ N and 51°35′17″ E), Iran. Supplementary Material, Table A1 shows the amount of solar radiation and ambient temperature from October 26 to November 30, 2018, at different times.

Figure 1 shows variations in ambient temperature and solar intensity at different experiment times on the first day of the experiment (i.e. October 26, 2018). As illustrated in this diagram, the maximum solar radiation is at 12 am, and the highest ambient temperature is at 1 pm.

Figure 1

Ambient temperature and solar intensity on October 26, 2018.

Figure 1

Ambient temperature and solar intensity on October 26, 2018.

Experimental setup

This experimental setup is used to produce freshwater from seawater. The basin for this setup is a double-slope type (floor area = 0.5 m2), with wings and the lower part fully insulated with glass wool. The basin glass cover is 5-mm thick with an angle of 30° relative to the horizon, according to its geographic coordinates. The difference between this experimental setup and a simple solar still is that the latter is continuously undergoing desalination after saline water entry. At the same time, in the former, the basin water drains after each hour of testing and enters a separate tank. A thermal element (powered by 300 W solar panels) is used to raise the temperature of the water entering the tank. The experiments used different depths of saline water in the basin. Saline water depths of 5, 10, and 15 mm in the basin equal 2.5, 5, and 7.5 L of saline water, respectively. In this system, once the thermal element is heated, the reservoir water is pumped through a 2 m copper pipe into the center of the CPC device, causing the water to heat further. The resultant hot water enters a fully-insulated tank with no heat transfer to the outdoor environment. While half of the basin saline water is distilled and sweetened, the other half is heated by the thermal element and CPC. This process is repeated every hour. The water desalination process in these experiments lasts from 9 am to 5 pm.

The following is a detailed explanation of the devices used in the process:

  • Solar panel: Electricity is generated from solar energy for use in the heating element and water pump by this device.

  • Pump: Water moves from the water heater tank to the cylindrical parabolic collector (CPC) by a pump.

  • Water heater: The tank has a heating element in which the saline water is heated before entering the CPC and solar still.

  • Cylindrical parabolic collector (CPC): In this device, saline water is heated by a pipe that passes through the focal point before entering the solar still.

  • Battery: This device stores energy output from solar panels.

  • Insulated tank: Before entering the solar still, the saline water is placed inside this device. It is then completely insulated to prevent energy loss.

  • Solar still: The main device in solar water desalination process, the desalination of seawater by distillation occurs in this device.

  • Saline water tank: A reservoir for storing saline water to be desalinated in this process.

  • Charge controller: This device controls the amount of battery charge in the process. It prevents an empty and a fully-charged battery from charging.

  • Tank for recycle: This device stores the saline water removed from the solar still. The saline water then enters the water heater for reheating.

Figures 2 and 3 illustrate the experimental setup with details and a schematic diagram of the process, respectively.

Figure 2

Experimental set-up with details.

Figure 2

Experimental set-up with details.

Figure 3

Schematic diagram of the process.

Figure 3

Schematic diagram of the process.

Furthermore, aluminum, glass, plastic, and wood were used separately to study the effects of basin constituent in the solar still on the amount of freshwater. The effects of black, brown, and red basin colors on the solar still basin have also been examined individually.

Constructional features and specifications of the basin for all materials used (e.g. wood, Al, glass, and plastic) are presented in Tables 1 and 2.

Table 1

Constructional features of the basin for all used materials (wood, Al, glass and plastic)

ParametersValueReference
Basin floor area (aluminum, glass, plastic, wood) 0.5 m2 This experimental work 
Thickness (aluminum, glass, plastic, wood) 0.005 m This experimental work 
Glass cover area 0.55 m2 This experimental work 
Mass of basin (aluminum) 17 kg This experimental work 
Mass of basin (glass) 16 kg This experimental work 
Mass of basin (plastic) 8 kg This experimental work 
Mass of basin (wood) 4 kg This experimental work 
Mass of glass cover 6 kg This experimental work 
Mass of water (water depth = 5 mm) 2.5 L This experimental work 
Mass of water (water depth = 10 mm) 5 L This experimental work 
Mass of water (water depth = 15 mm) 7.5 L This experimental work 
Insulation thickness 0.05 m This experimental work 
Angle of glass cover 30° This experimental work 
ɛglass cover 0.85 Muftah et al. (2018)  
ɛwater 0.96 Muftah et al. (2018)  
αglass cover 0.05 Muftah et al. (2018)  
αwater 0.05 Muftah et al. (2018)  
τglass cover 0.98 Ouar et al. (2017)  
τwater 0.96 Ouar et al. (2017)  
σ 5.67 × 10–8 (W/m2 K4Ibrahim et al. (2017)  
ParametersValueReference
Basin floor area (aluminum, glass, plastic, wood) 0.5 m2 This experimental work 
Thickness (aluminum, glass, plastic, wood) 0.005 m This experimental work 
Glass cover area 0.55 m2 This experimental work 
Mass of basin (aluminum) 17 kg This experimental work 
Mass of basin (glass) 16 kg This experimental work 
Mass of basin (plastic) 8 kg This experimental work 
Mass of basin (wood) 4 kg This experimental work 
Mass of glass cover 6 kg This experimental work 
Mass of water (water depth = 5 mm) 2.5 L This experimental work 
Mass of water (water depth = 10 mm) 5 L This experimental work 
Mass of water (water depth = 15 mm) 7.5 L This experimental work 
Insulation thickness 0.05 m This experimental work 
Angle of glass cover 30° This experimental work 
ɛglass cover 0.85 Muftah et al. (2018)  
ɛwater 0.96 Muftah et al. (2018)  
αglass cover 0.05 Muftah et al. (2018)  
αwater 0.05 Muftah et al. (2018)  
τglass cover 0.98 Ouar et al. (2017)  
τwater 0.96 Ouar et al. (2017)  
σ 5.67 × 10–8 (W/m2 K4Ibrahim et al. (2017)  

ɛ, emissivity; σ, Stefan–Boltzmann constant; α, absorptivity; τ, transitivity.

Table 2

Specifications of the basin for all used materials (wood, Al, glass and plastic)

MaterialDensity (kg m–3)Specific heat capacity (J kg–1 K–1)Thermal conductivity (Wm–1 K–1)References
Aluminum 2,700 900 200 Sharshir et al. (2020)  
Glass 2,500 720 1.2 Hamadou & Abdellatif (2014)  
Plastic (PVC) 1,380 900 1.16 Misra et al. (2013)  
Wood (oak) 720 1,255 0.16 Incropera et al. (2007)  
MaterialDensity (kg m–3)Specific heat capacity (J kg–1 K–1)Thermal conductivity (Wm–1 K–1)References
Aluminum 2,700 900 200 Sharshir et al. (2020)  
Glass 2,500 720 1.2 Hamadou & Abdellatif (2014)  
Plastic (PVC) 1,380 900 1.16 Misra et al. (2013)  
Wood (oak) 720 1,255 0.16 Incropera et al. (2007)  

Supplementary Material, Table A2 presents the different conditions of the process. From October 26 to November 30, process tests were performed under different circumstances. The basin material, the basin color, and the water depth were different on different days of the experiment. The table shows detailed conditions for performing the test on different days.

RESULTS AND DISCUSSION

The tests were conducted from 9 am to 5 pm at Yasuj city in the autumn of 2018. Various experiments were carried out to investigate the effects of constituent materials as well as different colors. The energy was received from solar panels using a thermal element, and the temperature of the saline water in the basin increased as a result of using a CPC device.

Solar still temperature

This section reviews the solar still basin temperature under different scenarios.

Figure 4 illustrates the temperature of basins made of aluminum, glass, plastic, and black wood at water depths of: (a) 5 mm, (b) 10 mm, and (c) 15 mm. As can be seen, the temperature of the aluminum basin is higher than that of the other basins. This can be attributed to the higher thermal conductivity of aluminum, which allows the basin, and consequently the water therein, to get warm faster. A basin with 5 mm water depth due to less saline water has a temperature higher than a basin with 10 and 15 mm water depths.

Figure 4

The temperature of different materials and black basin at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 4

The temperature of different materials and black basin at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

As shown in Figure 4(a)–4(c), and according to the experiments, the highest and lowest temperatures measured are at 1 pm, and 9 am, respectively. A comparison of Figure 4(a)–4(c) shows that the more saline water is placed in the basin, the lower the basin temperature is.

Figure 5 depicts the effects of black, brown, and red colors on the temperature of the aluminum basin at water depths of: (a) 5 mm, (b) 10 mm, and (c) 15 mm. In this case, the temperature of the black basin is higher than that of brown or red ones, which may be due to the larger amount of light absorbed by the black color. In Figure 5, like Figure 4, the basin temperature with a 5 mm saline water depth is higher than that of the basin with 10 and 15 mm saline water depths.

Figure 5

The temperature of different colors in the aluminum basin at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 5

The temperature of different colors in the aluminum basin at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

The difference in ambient temperature on different days of the experiment can affect seawater solar desalination. As shown in Figure 5, the effect of the basin color is a crucial parameter in conducting experiments.

More light is absorbed by dark colors (black), leading to an increase in the basin temperature and, consequently, saline water temperature in the basin. According to Figure 5, like Figure 4(a)–4(c), the maximum and minimum basin temperatures are measured at 1 pm, and 9 am, respectively. Shallower water depths also increased the measured temperature in a 5 mm-deep basin compared to 10- and 15 mm-deep saline water basins.

Figure 6 shows the temperature difference between the basin and the glass cover/saline water. The former (maximally 11 °C at 1 pm) has always been greater than the latter, which often falls within the range of 0.4–1 °C.

Figure 6

The temperature difference between the basin, the glass cover, and the water in the black wooden basin.

Figure 6

The temperature difference between the basin, the glass cover, and the water in the black wooden basin.

Freshwater production

Figures 79 demonstrate the freshwater mass during desiccation at water depths of: (a) 5 mm, (b) 10 mm, and (c) 15 mm. In Figure 7(a), basins made of different materials were compared with black ones, suggesting the highest amount of freshwater in the aluminum basin and the lowest amount in the wooden basin at 5 mm water depth. Figure 7(b) and 7(c) also shows the effects of using different materials to build basins at 10 and 15 mm water depths. According to the results, the highest freshwater mass is related to the 5 mm saline water depth and black aluminum basin.

Figure 7

Comparison of freshwater produced in different basins (color: black) at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 7

Comparison of freshwater produced in different basins (color: black) at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 8

Comparison of freshwater produced in different basins (color: brown) at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 8

Comparison of freshwater produced in different basins (color: brown) at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 9

Comparison of freshwater produced in different basins (color: red) at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

Figure 9

Comparison of freshwater produced in different basins (color: red) at water depths of: (a) 5 mm, (b) 10 mm, (c) 15 mm.

In Figure 7(a)–7(c), the highest seawater desalination rate always occurred in the aluminum basin. As indicated, there is a significant difference between desalinated water levels in the aluminum basin and other basins. The lower the saline water depth in the basin, the higher the amount of desalinated water, which is less than the saline water depth due to the higher thermal level in the saline water depth.

Figures 8 and 9 also show the comparison of brown and red basins made of different materials at water depths of: (a) 5 mm, (b) 10 mm, and (c) 15 mm. The highest and lowest amounts of freshwater were for the black aluminum basin (5 mm water depth) with 2.97 kg/day and the red wooden basin (15 mm water depth) with 1.87 kg/day, respectively, during the test. Figures 8 and 9 investigate different states of saline water depth. In general, applying 5 mm water depth in the solar still produced the highest mass of freshwater. The lower the saline water content in the basin, the faster the saline water heats up, and the greater the process of evaporation and desalination.

Generally, regarding different constituent materials of the basin and different colors, red has the least saline water desalination rate since it absorbs less solar energy than black and brown. Thus, saline water in the basin has a lower temperature than black and brown basins, leading to lower desalinated water production. However, stills with aluminum basins have a higher water desalination rate than other basins made of glass, plastic, and wood.

A comparison between different colors indicated that the black color had the highest level of desalination due to the higher basin temperature and thus the higher basin water temperature. In contrast, the red color yielded the least amount of freshwater in these experiments.

Supplementary Material, Table A3 shows the amount of desalinated water on different experiment days under the influence of various parameters, including the basin material, the basin color, the depth of saline water in the basin, and the amount of desalinated water.

According to Table A3, the highest volume of desalinated water during 1 day of testing is related to the black aluminum basin and water depth of 5 mm. Due to the high thermal conductivity of aluminum, the temperature of saline water in the basin increases, and as a result, the evaporation rate of water increases. This increase in the evaporation rate eventually leads to an increase in water desalination. A comparison between different colors shows that black raises the water temperature in the basin compared to brown and red as it absorbs more solar energy, leading to an increase in the volume of desalinated water. The use of shallower water in the basin also causes this water to heat faster due to solar energy absorption, consequently the volume of desalinated water increases.

ANN modeling

This paper used ANN modeling to validate the experimental data (with MATLAB software). To start the ANN-based simulation, data from October 26 to November 30, 2018, were used. The input parameters of the ANN employed in this simulation are water temperature (Twater), basin temperature (Tbasin), and glass cover temperature (Tglass). Training, testing, and validation accounted for 70, 15, and 15% of the input data to the ANN. The number of layers in this simulation is three. The input layer includes Twater, Tbasin, and Tglass inputs (on multi-layer perceptron (MLP) method).

Figure 10 illustrates the structure of the ANN employed in this paper. As shown in this structure, three inputs were used to start the simulation. Several different neurons were exploited in the simulation process, i.e. 2–25 neurons in the hidden layer and the output responses of the ANN were compared in different situations. In other words, to compare simulation results, a particular number of neurons were employed at each time point. The highest and lowest regression rates are for 24 and 10 neurons, respectively.

Figure 10

The structure of the artificial neural network (ANN).

Figure 10

The structure of the artificial neural network (ANN).

Figure 11(a)–11(c) shows diagrams for the volume of desalinated water during different hours of testing and compared with ANN predictions. These diagrams are associated with three different days from October 28 to November 14 and 27, 2018. From these diagrams, the comparison between the simulation results and the experimental data revealed that the simulation results with 24 neurons were much closer to the experimental data. According to the diagram, simulations with 10 neurons are less accurate than those with 24 neurons.

Figure 11

Mass of desalinated water and the ANN-based prediction modeling with 10 and 24 neurons: (a) October 28th, (b) November 14th, (c) November 27th.

Figure 11

Mass of desalinated water and the ANN-based prediction modeling with 10 and 24 neurons: (a) October 28th, (b) November 14th, (c) November 27th.

According to the experimental data, the highest water desalination rate occurs in a black aluminum basin with a saline water depth of 5 mm. Hence, the basin's optimum conditions are aluminum for the material, black for the color, and 5 mm for the saline water depth.

The simulation of an ANN involves determining the Tbasin, Twater, and Tglass inputs. From the first to the last day of the experiment, the experimental data matches the ANN's prediction data. The least error occurs when the number of neurons in the ANN simulation is 24, meaning that the optimal state for the simulation occurred when the number of neurons was 24. In other words, the simulation results from October 26 to November 30 for the number of neurons (n = 24) show the highest correlation between the experimental data and neural network prediction data and the lowest correlation between all ANN prediction data for the number of neurons (n = 10).

Figures 12, A1 and A2 depict the R2, MSE error chart, and %AAD error for training, test, validation, and all data. These figures display the simulation responses based on a varying number of neurons. As shown in these graphs, the highest and lowest regression rates are associated with 24 and 10 neurons. The highest and lowest MSE and %AAD error among a varying number of neurons are related to 10 and 24 neurons.

Figure 12

R2 between experimental and predicted data on a varying number of neurons: (a) train data, (b) test data, (c) validation data, (d) all data.

Figure 12

R2 between experimental and predicted data on a varying number of neurons: (a) train data, (b) test data, (c) validation data, (d) all data.

As shown in Figures 12, A1 and A2, a diagram can be drawn to simulate an ANN in train, test, validation, and all data and several different neurons.

The Appendix and Tables A4 and A5 present the hidden layer, weight, and bias of the output layer and the output layer with 10 and 24 neurons.

CONCLUSIONS

Innovation has been performed in selecting the basin material and examining the differences between different materials. Temperature and radiation characteristics are provided on different days of the experiment. The tests were performed from October 26 to November 30, 2018. The basin temperature can be raised if appropriate materials are used, increasing the water temperature and desalination rate.

This innovative method has taken advantage of solar panels and a CPC in solar desalination plants to increase the basin water temperature during the process. Based on the application of different components of solar still basins and their comparison, aluminum converts more saline water into freshwater thanks to its higher thermal conductivity. Black basins, in general, have sweetened more water than brown and red basins. In these experiments, three different water depths (i.e. 5, 10, and 15 mm) were investigated. At 5 mm water depth, saline water was heated faster and, consequently, distillation occurs to a greater extent. The highest and lowest amounts of freshwater produced in the experiments were related to the black aluminum basin (5 mm water depth) with 2.97 kg/day and the red wooden basin (15 mm water depth) with 1.87 kg/day, respectively.

An ANN was employed to validate the experimental data. The results were examined in a varying number of neurons. The best response rate is yielded with 24 neurons (R2 = 0.993). With the same number of neurons, the least amount of error is predicted by the ANN. The ANN-based prediction of a varying number of neurons according to the ANN inputs (Twater, Tbasin, and Tglass) generally indicates a high agreement between the ANN-based prediction and the experimental data. ANNs have been employed as one of the most reliable data validation and prediction methods. The ANN simulation has been performed to investigate the complete similarity of ANNs (from 2 to 25). In each simulation, R2, MSE, and% AAD errors were reported. During the simulations, the ANN's response, when the number of neurons is 24, was found to have the highest R2 and the lowest error.

Future scope

Given the importance of water desalination, various methods are used. Solar water desalination has attracted the attention of many people in remote areas due to its cost-effectiveness and availability of facilities. The efficiency of this water desalination method can be increased in different ways. For example, we can take advantage of the effects of different basin insulations on solar sweetening efficiency.

DATA AVAILABILITY STATEMENT

All relevant data are included in the paper or its Supplementary Information.

REFERENCES

REFERENCES
Bagheri
A.
,
Esfandiari
N.
&
Honarvar
B.
2019
Experimental investigation of the effect of using the cylindrical parabolic collector, different solar panels and their cooling on seawater desalination in double-slope solar still
.
Energy Sources, Part A Recover. Util. Environ. Eff
.
https://doi.org/10.1080/15567036.2019.1623947
.
Bagheri
A.
,
Esfandiari
N.
,
Honarvar
B.
&
Azdarpour
A.
2020a
An experimental study on the effects of direct and indirect use of solar energy on solar seawater desalination
.
Water Sci. Technol. Water Supply
20
(
1
),
259
268
.
Bagheri
A.
,
Esfandiari
N.
,
Honarvar
B.
&
Azdarpour
A.
2020b
First principles versus artificial neural network modelling of a solar desalination system with experimental validation
.
Math. Comput. Modell. Dyn. Syst.
26
(
5
),
453
480
.
Feilizadeh
M.
,
Soltanieh
M.
,
Estahbanati
M. R. K.
,
Jafarpur
K.
&
Ashrafmansouri
S. S.
2017
Optimization of geometrical dimensions of single-slope basin-type solar stills
.
Desalination
424
,
159
168
.
Hamed
O. A.
,
Kosaka
H.
,
Bamardouf
K. H.
,
Al-Shail
K.
&
Al-Ghamdi
A. S.
2016
Concentrating solar power for seawater thermal desalination
.
Desalination
396
,
70
78
.
Ibrahim
A. G. M.
,
Rashad
A. M.
&
Dincer
I.
2017
Exergoeconomic analysis for cost optimization of a solar distillation system
.
Sol. Energy
151
,
22
32
.
Incropera
F. P.
,
DeWitt
D. P.
,
Bergman
T. L.
&
Lavine
A. S.
2007
Fundamentals of Heat and Mass Transfer
, 6th edn.
John Wiley & Sons
,
Hoboken, NJ
.
Kabeel
A. E.
,
Hamed
A. M.
&
El-Agouz
S. A.
2010
Cost analysis of different solar still configurations
.
Energy
35
(
7
),
2901
2908
.
Misra
R.
,
Bansal
V.
,
Agrawal
G. D.
,
Mathur
J.
&
Aseri
T.
2013
Transient analysis based determination of derating factor for earth air tunnel heat exchanger in summer
.
Energy Build.
58
,
103
110
.
Muftah
A. F.
,
Sopian
K.
&
Alghoul
M. A.
2018
Performance of basin type stepped solar still enhanced with superior design concepts
.
Desalination
435
,
198
209
.
Nayi
K. H.
&
Modi
K. V.
2018
Pyramid solar still: a comprehensive review
.
Renewable Sustainable Energy Rev.
81
,
136
148
.
Ouar
M. L. A.
,
Sellami
M. H.
,
Meddour
S. E.
,
Touahir
R.
,
Guemari
S.
&
Loudiyi
K.
2017
Experimental yield analysis of groundwater solar desalination system using absorbent materials
.
Groundwater Sustainable Dev.
5
,
261
267
.
Pounraj
P.
,
Winston
D. P.
,
Kabeel
A. E.
,
Kumar
B. P.
,
Manokar
A. M.
,
Sathyamurthy
R.
&
Christabel
S. C.
2018
Experimental investigation on Peltier based hybrid PV/T active solar still for enhancing the overall performance
.
Energy Convers. Manage.
168
,
371
381
.
Refalo
P.
,
Ghirlando
R.
&
Abela
S.
2014
The effect of climatic parameters on the heat transfer mechanisms in a solar distillation still
.
Heat Transfer Eng.
35
(
16–17
),
1473
1481
.
Rufuss
D. D. W.
,
Kumar
V. R.
,
Suganthi
L.
,
Iniyan
S.
&
Davies
P. A.
2018
Techno-economic analysis of solar stills using integrated fuzzy analytical hierarchy process and data envelopment analysis
.
Sol. Energy
159
,
820
833
.
Santos
N. I.
,
Said
A. M.
,
James
D. E.
&
Venkatesh
N. H.
2012
Modeling solar still production using local weather data and artificial neural networks
.
Renewable Energy
40
(
1
),
71
79
.
Sharshir
S. W.
,
Peng
G.
,
Elsheikh
A. H.
,
Eltawil
M. A.
,
Elkadeem
M. R.
,
Dai
H.
,
Zang
J.
&
Yang
N.
2020
Influence of basin metals and novel wick-metal chips pad on the thermal performance of solar desalination process
.
J. Cleaner Prod.
248
,
119224
.
Velmurugan
V.
,
Gopalakrishnan
M.
,
Raghu
R.
&
Srithar
K.
2008
Single basin solar still with fin for enhancing productivity
.
Energy Convers. Manage.
49
(
10
),
2602
2608
.

Supplementary data