Experimental investigation of a new design of drum solar still with reflectors under different conditions

This comprehensive study employed a new design of drum solar still which keeps the basin water depth constant using constant head tank. Three solar stills nominated as drum solar still (DSS), modified drum solar still (MDSS), and conventional solar still (CSS) were tested under the same environmental conditions to be able to compare between their performances in fair. Both DSS and MDSS had a free to rotate drum fixed on a rotating shaft run by a DC motor. A PV system was used to eliminate the use of the traditional sources of energy. Besides, the effects of using a flat-sheet glass cover instead of the semi-circular shape, mirrors on the rear high-wall, front small-wall from glass sheet, and nanoparticles on the performance of MDSS were investigated. Various drum speeds were tested to point out the optimum speed of rotation for the MDSS with and without wick. Results revealed that the MDSS obtained better performance than that of the other solar stills. In addition, the drum speeds 0.05 and 0.1 rpm were the optimum speeds for MDSS with and without wick, respectively. Besides, at 0.1 rpm, the productivity of the MDSS with reflectors and nanofluid was increased by 296% more than that of the CSS. While the maximum productivity rise obtained for MDSS with wick at of 0.05 rpm was 280%. Also, the cost of 1 L from the CSS and MDSS was 0.05 and 0.041 $, respectively.

One of the main factors affecting the freshwater production from the solar still is the free surface area of evaporation and condensation [29,46]. Consequently, numerous designs and modifications were proposed to increase the evaporative surface area of the solar distillers with the hope of augmenting the productivity. To achieve this purpose, solar distillers incorporated with moving parts were suggested. The moving parts lead to break the molecules' bonds of the water on the surface (surface tension).
One of the rotating systems introduced to improve the performance of the solar still is the rotating cylinder. A rotating cylinder was proposed inside the solar distiller by Malaeb et al. [47] in an attempt to increase the evaporative surface area and decrease the water thickness. This modification improved the freshwater productivity of the solar still by 250%. In another study, Ayoub et al. [48] tested the different parameters affecting the performance of the drum solar still such as the drum speed, depth of basin water, and cover cooling. They concluded that the drum speed should be as low as possible without creating dry spots on the surface of the drum. Kabeel et al. [49] increased the productivity by 25% via installing a rotating fan inside the solar still. Moreover, Omara et al. [50] installed a rotating fan inside the solar distiller and tested its performance under various water depths. They reported that the depth of basin water should be as small as possible at the low speeds of the fan. The freshwater yield of the solar distiller integrated with a rotating fan was enhanced by 17%. Besides, Essa et al. [25] presented an experimental and empirical study to improve the performance of a tubular distiller by utilizing a rotating drum with two different ends (closed and opened ends). They enhanced the productivity of the tubular distiller by 175% at 0.05 rpm with open ends drum and wick.
Covering the rotating part with wick material is an important modification that remarkably increased the productivity of the solar still. This modification was reported by Haddad et al. [51] and Gad et al. [52] who proposed vertical and horizontal rotating wick belts inside the solar distiller, respectively. Their systems increased the evaporative surface areas without adding more horizontal area to the distiller. In addition, Abdullah et al. [18,41] employed a horizontal and vertical wick belt inside the solar still in an attempt to use the inside space for increasing the evaporative and exposure surface areas without adding more projected area to the distiller. They investigated the distiller performance under various stop times, and they reported that the optimum OFF time of the belt was 30 min with another 5 min as ON time. They achieved an increase of 315% in productivity of the solar still. Furthermore, a rotating shaft was fixed inside the distiller to augment its productivity [53]. The thermal efficiency was enhanced by 5.5%, 5%, and 2.5% in July, June, and May, respectively. Essa et al. [16] investigated the effect of using flat and corrugated rotating discs inside the solar still. They increased the productivity by 124%. Recently, Abdullah et al. [17] introduced a rotating drum inside the solar distiller to enlarge the evaporative surface area. They tested the performance of the drum solar still under different operational conditions such as using solar air heater, condenser, nanofluid, and various drum speeds. They reported that the productivity was 2025 and 6420 L/m 2 . day for the conventional and drum distillers, respectively. So, they increased the productivity by 217% due to using the rotating drum. Also, using the solar air heater, condenser, and nanofluid with the drum still raised the productivity by 350%. Installing a rotating drum inside the solar still has some advantages such as creating a thin water film over the drum surface, which is evaporated quickly. Also, the drum has large evaporative and exposure surface areas, which increase the evaporation rate inside the solar still. Besides, rotating the drum breaks the surface tension of the basin water due to the turbulence in the water and water vapor contents inside the solar still.
Regarding the aforementioned literature and our previously published paper [17], there was a problem appeared during the testing of the drum solar still. That problem was falling a lot of condensed droplets from the highest surface area of the glass to the basin water due to the convex surface of the glass as illustrated from Fig. 1. As a result, we intended to make the glass cover flat sheet not semi-circular. The basic idea of this research was enhancing the evaporation rate of the solar still by minimizing the water film and maximizing the surface area of evaporation and exposure to solar radiation. The basin design was modified to have the smallest amount of water using a constant head tank. Also, the new design kept the basin water depth fixed unlike the traditional drum still. The front small wall of the basin was made from a glass sheet to increase the surface area of exposure to solar radiation and evaporation. Moreover, the drum and modified drum solar stills were investigated under various rotational speeds of the drum (0.02, 0.05, 0.1, 0.2, 0.5, 1.0, 2.0, 3.0, and 4.0 rpm). Consequently, the novel points of this work are: Fig. 1. The condensed droplets inside the previously published drum solar still [17]. 1. A reflector is placed on the rear high-wall. This reflector reflects the incident solar radiation on a part of the drum which increases the evaporation of the drum. Also, the existence of the reflector decreases the heat loss from the wall. 2. The glass cover is used as flat sheet not a semi-circular to prevent the condensed droplets to fall in the basin water. Also, this gives the opportunity to make better condensation. 3. The solar still integrated with a rotating drum is investigated when using water-copper oxide nanofluid. 4. The effect of using wick over the outer body of the rotating cylinder on the solar still performance is studied.

Experimental setup
The experimental setup included three solar stills (conventional still, drum still, modified drum still), constant head tank, feed water tank, PV system, and DC electric motor as illustrated in Fig. 2. Also, Fig. 3 shows a schematic diagram of the experimental setup. A feeding tank (φ = 500 mm and H = 1000 mm) was used to feed the solar stills by saline water. The conventional solar still (CSS) was utilized as a reference for all measurements and the improvement percentages of productivity of the other solar stills. The three investigated solar stills had the same projected area of 5000 cm 2 . The CSS was made of 1.5 mm thick galvanized steel sheet. The surfaces of the solar stills were painted by matt black paint to maximize the absorption of solar radiation. Moreover, as obtained in Fig. 3, the heights of lower-and higher-walls of solar stills were 160 and 390 mm, respectively. The solar still was covered by a 3 mm thick glass sheet as shown in Fig. 2. There was a tilted trough fixed at the edge connecting between the lower wall of distiller and lower edge of glass. This trough was used to collect the condensed droplets of freshwater and get them out the solar still to be accumulated into the calibrated flasks. Besides, the basin still was insulated well by 50 mm thick glass wool to prevent the heat loss.
On the second hand, both drum still (DSS) and modified drum still (MDSS) had a closed-ends drum each. This drum was installed on a shaft to rotate about its axe using a DC electric motor (3 W) as illustrated in Figs. 2 and 3. The rotating drum (φ = 440 mm and L = 980 mm) was made of 0.5 mm thick aluminum sheet. A PV system (10 W) was utilized to run the motor. The PV system consisted of the photovoltaic panel, battery, and converter. The DC motor was integrated with a speed controller to control the rotation of the drum at the target rotational speed. The commercial matt-black color was used to paint the drum and the solar stills to maximize the absorption of solar radiation. Here, the drum solar still was covered by a semi-circular glass (φ = 500 mm and L = 1000 mm) as shown in Figs. 2 and 3. Furthermore, the distilled freshwater was collected in four troughs to get it out of the drum still into the freshwater bottles. A 30 mm thick wooden box was used as a container for the solar still. The vapor leakage was prevented using silicone rubber as a sealing material.
For the modified drum solar still, it had the same drum as in the drum solar still with a different basin constructure. The difference between the DSS and MDSS is being the glass cover as a sheet not semi-circular as obtained from Fig. 2. Also, the MDSS had mirrors fixed on the rear higher wall of the basin, Fig. 3. Moreover, the smaller wall of the basin was replaced by a glass wall to increase the surface area of the evaporation and solar radiation exposure, Fig. 2c.   Fig. 2. Photo of (a) the experimental setup, (b) drum still, and (c) modified drum still.

Proceeding of experiments
The outdoor environmental conditions were used to test the experimental setup from August and September 2019. The whole setup was fixed on the south direction for maximizing the incident solar radiation. To evaluate the performance of the solar still, numerous parameters were measured hourly from 08:00 to 18:00. The measured parameters were the solar intensity, ambient air temperatures, air speed, glass cover temperature, saline water temperature, and amount of distilled freshwater. The tested were ran for 10 h through the daytime, and the distilled water was reported daily. The feeding reservoir was utilized to feed the distillers by saline water. The three solar stills were kept having a constant water depth using the constant head water tank. The depth of saline water inside the distillers was 10 mm. For both DSS and MDSS, the drum was immersed into the basin water by 0.5 cm. With the first movement of the drum at the target speed, a thin film of water was created on the outer surface of the drum. This film was evaporated easily by the effect of solar radiation incident on the drum surface. So, the vapor was generated and raised up to be condensed on the inner surface of the glass cover.
For the steps of experimentations, first, we investigated the performance of the CSS, DSS, MDSS without any additional modifications to compare between the productivity and efficiency of all distillers. This point led to evaluate the performance of the modified drum solar still and compare it with the others. The second step of experimentation was investigating the effect of the rotational drum speed on the performance of both DSS and MDSS. The tested speeds were 0.02, 0.05, 0.1, 0.2, 0.5, 1.0, 2.0, 3.0, and 4.0 rpm. Also, the performance of DSS and MDSS was compared to that of CSS. After that, the influence of using wick with the MDSS was investigated in the third step of tests. Moreover, the effect of using water-copper oxide nanofluid and internal mirror on the performance of the MDSS without wick was studied in the last step of tests. The specifications of the CuO nanoparticles are obtained in Table 1.

Measuring instruments
The various parameters affecting the solar still performance were measured using the suitable devices. The solar radiation was measured hourly via using the Datalogging solar power meter with the range from zero to 5000 W/m 2 . In addition, Calibrated K-type thermocouples were utilized to measure the temperature at different points of the system. Moreover, a G4L-CUEA modular programmable logic control (MPLC) was used to read the signal coming from the thermocouples to obtain the digital values of the temperatures. The air speed was measured using a van type anemometer having the range of 0.4-30 m/s. Furthermore, the freshwater distillate was recorded hourly by the calibrated graded bottles (1.5 L capacity and 5 mL accuracy). Table 2 presents the parameters of the measuring instruments utilized in the experiment.

Uncertainty analyses
The uncertainties of the obtained experimental data were computed regarding the method reported by Holman [54]. Assume that R is a function of independent parameters with "n" numbers. Thus, As well, W 1 , W 2 , W 3 , W 4 …, …., W i are the uncertainties of independent parameters. Thus, the uncertainty in the result (W R ) is found by: (2) Fig. 3. Schematic of the experimental setup. The daily efficiency, η d is evaluated by: A is constant and for the range of measured temperatures assuming C p is constant, so, The uncertainty of efficiency can be deduced as, The resulting error of the ƞ of the solar distiller is around ± 2%. Table 2 indicates the uncertainties of our data.

Performance of modified distiller with rotating drum
The ambient conditions (solar radiation and air temperature) and temperatures of glass cover and basin water of the solar distillers at 0.1 rpm drum speed are illustrated in Fig. 4. It is revealed that the water temperature of the drum still (DSS) is lower than that of the CSS by 0-10 • C, Fig. 4. This occurs due to that the water is not heated directly by the incident solar radiation and gets warmed and heated by the heat of the drum. While the water of the CSS is heated directly by the incident solar intensity. As a result, the water temperature of the DSS is lower than that of the CSS. On the second hand, the glass temperature of the CSS is lower than that of the DSS by 0-2 • C due to the high rate of evaporation inside the DSS.
Furthermore, the basin water temperature of the modified drum solar still (MDSS) is greater than that of the CSS by 0-4 • C as shown in Fig. 4. This is because the MDSS has lower quantity of saline water (2.2 kg) compared to the DSS (5.0 kg) and CSS (5.0 kg). In addition, the MDSS has a transparent wall (the front wall of the basin) which leads to increase the incident solar radiation due to the high surface area of exposure to solar radiations. Moreover, the MDSS has a glass temperature higher than that of CSS by 0-3 • C due to the high rate of evaporation generated inside the MDSS. Besides, the maximum solar radiation was recorded at 12:00 where it was 1140 W/m 2 . The maximum water temperature was recorded at 13:00 where the water temperatures of MDSS, DSS, and CSS were 70, 56, and 66 • C, respectively. While the glass temperature was reported as 47, 46, and 44 • C for the MDSS, DSS, and CSS at 13:00, respectively. The hourly productivity of the solar distillers at 0.1 rpm drum speed is illustrated in Fig. 5. It was revealed that the productivity begins to increase from zero at the beginning of experiments to reach its maximum value at noon, then it is declined again in the afternoon. The hourly productivity of the solar stills was recorded at 13:00. Where, the maximal hourly productivity of the MDSS, DSS, and CSS was 1250, 1050, and 450 mL/m 2 at 13:00, respectively. As illustrated from Fig. 5, the hourly productivity of the DSS is greater than that of CSS. This is due to the high evaporation rate generated by the effect of thin water film on the drum body as compared to the CSS. The DSS has greater evaporative surface area than the CSS. In addition, the DSS has higher exposure area to solar radiation which raised the generation of vapor compared to the CSS. For example, the exposure area of the DSS equals 0.69 m 2 compared to only 0.5 m 2 for CSS. Consequently, the DSS generates higher amount of evaporation compared to the CSS. Another reason for the increased productivity of the DSS is that the rotation of the drum makes a turbulence in the basin water which leads to break the surface bonds among the molecules of water. Hence, the evaporation and water production are increased.
On the second hand, Fig. 5 illustrates that the hourly productivity of the MDSS is higher than that of the DSS and CSS. This is because the MDSS has a water temperature higher than the other solar stills (DSS and CSS). In addition, the evaporative surface area of the MDSS is higher than that of either the DSS or CSS. This leads to increase the vapor generation inside the MDSS compared to that of the other distillers. In addition, the area of exposure to solar radiation for the MDSS is greater than that of the other distillers due to using the front wall from glass instead of steel. For instance, the exposure area of the MDSS equals 1.02 m 2 compared to 0.69 and 0.5 m 2 for DSS and CSS, respectively. This increases the evaporation rate and freshwater production compared to that of the other distillers. Additionally, using the glass cover as a flat sheet, not a semi-circular, in the MDSS helped to prevent the condensed droplets to be fallen in the basin water. This gives the opportunity to make better condensation as compared to the DSS.
Furthermore, Fig. 6 shows the total accumulated productivity of the solar distillers at 0.1 rpm drum speed. Results revealed that the total productivity of DSS is greater than that of CSS. In addition, the MDSS has total distillate higher than either DSS or CSS. The total distilled freshwater of the MDSS, DSS, and CSS are 8800, 7450, and 2400 mL/m 2 . day, respectively. As a result, the freshwater generation is improved by 266% and 210% when using the MDSS and DSS respectively over the CSS. This improvement is caused by the above explained reasons of increasing the output distillate.

Influence of drum speed on performance of modified drum distillers
The influence of the different drum speeds on the productivity improvement of the drum-and modified drum stills over the CSS is drawn in Fig. 7. The figure illustrates that the MDSS obtained higher productivity rise over the CSS than the DSS under at tested drum speeds. Also, all productivity rise values above the zero point on y-axe means positive effect (CSS is of lower distillate), and that below the zero point means negative effect (CSS is of higher distillate). Besides, the maximum rise in productivity for both DSS and MDSS was achieved at the drum speed of 0.1 rpm, where the rise in productivity was 210% and 266%, respectively. Moreover, increasing the drum speed more than 0.1 rpm leads to decrease the productivity rise percentage. This maybe because the high speed of rotation did not give enough time to evaporate the whole water film on the drum surface. Then, the vapor content was less than that of 0.1 rpm. In addition, this decrease in productivity at high speeds was due to the higher heat dissipation by stronger forced heat convection, which A.S. Abdullah et al. was achieved at high speeds compared to that occurred at 0.1 rpm. The low drum speed (0.02 rpm) had a bad effect on the distillate (CSS is of higher distillate). This is maybe due to the dry spots created on the drum surface because of the slow motion.
For evaluating the investigated solar stills based on the thermal efficiency, the formula used to calculate the daily efficiency is written as following [49]; where η d , ṁ, h fg , I(t), A, and MP are the thermal daily efficiency, hourly distillate productivity, vaporization latent heat, daily average solar radiation, system area (solar still projected area + PV projected area), and required motor power, respectively. Following the above relation, the thermal efficiency of the tested solar stills under different drum speeds is plotted in Fig. 8. As observed from the figure, it has the same behavior of the distillate rise of the distillers. In addition, the thermal efficiency of the CSS is almost constant at 34-35%. While the efficiency of the DSS and MDSS depends strongly on the drum speed as illustrated from Fig. 8. The maximum thermal efficiency for both DSS and MDSS was obtained at 0.1 rpm drum speed, where the maximum productivity rise was achieved. At 0.1 rpm, the thermal efficiency of DSS and MDSS is 72% and 77%, respectively. In addition, working at the drum speeds of 0.02 and 4.0 rpm obtains lower efficiencies for DSS than that of CSS, where the efficiency of DSS was 28.5% and 32% compared to 34.5% and 34% for CSS at 0.02 and 4.0 rpm, respectively. This decline in efficiency is due to the decrease of distillate of  the DSS at these speeds as shown in Fig. 7. While the efficiency of MDSS is observed to be lower than that of the CSS at 0.02 rpm only (32% for MDSS vs 34.5% for CSS).

Effect of adding wick material on performance of modified drum solar still
When operating the system on the above experiments at the low drum speeds of 0.02 rpm (one loop every 50 min) and 0.05 rpm (one loop every 20 min), it was observed that the drum has dry spots formed on its outer surface due to the small motion. So, we tested the performance of the MDSS with a wick material covering the drum outer surface. Fig. 9 shows the total productivity improvement of the modified drum solar still with and without wick at different drum speeds. It is observed from Fig. 9 that using the wick leads to improve the performance of the MDSS at the low drum speeds (0.02 and 0.05 rpm). In addition, the maximum productivity rise achieved is obtained at the drum speed of 0.05 rpm, where the increase in productivity records its highest value of 280%. This improvement in distillate of MDSS with wick is higher than the highest value of productivity increase for MDSS without wick at 0.1 rpm. This is because the slow motion at 0.05 rpm gives more opportunity for better evaporation without causing the dry spots phenomenon because of using the wick.
Moreover, the productivity rise of MDSS with wick is declined with increasing the drum speed more than 0.05 rpm as shown in Fig. 9. This is because the wick carries an amount of water that needs more time to be evaporated at the speeds more than 0.05 rpm due to the high rotation of the drum. Furthermore, with increasing the drum speed more than 0.05 rpm, the productivity rise of MDSS with wick is lower than that of MDSS without wick. Additionally, at 3.0 rpm, the CSS and MDSS with wick have the same productivity, Fig. 9.  Fig. 9. The total productivity improvement of the modified drum solar still with and without wick at different drum speeds.
A.S. Abdullah et al. A comparison between the efficiency of the CSS and MDSS with and without wick under different drum speeds is plotted in Fig. 10. The efficiency behavior obtained in this figure looks like that illustrated in Fig. 8. The CSS has a constant thermal efficiency of 34-35%. In addition, the highest efficiency for MDSS with and without wick was revealed at 0.05 and 0.1 rpm drum speed, where the thermal efficiency of MDSS with and without wick is 79% and 77%, respectively. Furthermore, with increasing the drum speed more than 0.05 rpm, the thermal efficiency of MDSS with wick is lower than that of MDSS without wick, Fig. 10. Additionally, at 3.0 rpm, the CSS and MDSS with wick have the same thermal efficiency as illustrated in Fig. 10.

Effect of using internal mirrors and nanoparticles on performance of modified drum solar still
The effect of using internal mirrors and nanoparticles on performance of modified drum solar still without wick is investigated in this sub-section. Results revealed that the best performance of MDSS with internal mirrors and nanoparticles was achieved at the drum speed of 0.1 rpm. In addition, using the internal mirrors on the back wall of the MDSS increases the freshwater productivity by 14% over that of MDSS without mirrors. As a result, the total productivity of the MDSS with mirrors was increased by 280% compared to that of the CSS. This is due to that the mirror has two features: first, it helps to prevent the heat loss from the back wall of the basin, and second, it reflects the solar rays on the back part of the drum. As a result, the evaporation rate is improved.
Finally, the productivity of the MDSS with mirrors and without wick was investigated when using CuO nanoparticles. The saline water was used as a solution for the CuO nanoparticles to be used as a nanofluid instead of the saline water. It was revealed that the productivity of the MDSS with mirrors and nanofluid was increased by 296% more than that of the CSS. This means that using the nanofluid raised the productivity of MDSS by 16% only over that of MDSS without nanofluid. This is because the nanofluid is of heat transfer characteristics better than that the pure water. In addition, the copper oxide-water mixture has a thermal conductivity higher than the pure water [38]. Moreover, the heat storing of the nanoparticles is better than that of the water. For these reasons, the evaporation rate of the MDSS with CuO nanoparticles is augmented, which leads to improve the freshwater production.

Cost analysis of distillers
The cost analysis of the solar stills depends upon the distiller type and its individual components as tabulated in Table 3. So, the cost estimation is conducted for both fixed and variable costs. Based on the average daily accumulated distillate and the distiller lifetime, the cost of 1 L from the solar still can be evaluated as following.

For the CSS
The fixed cost of CSS is F = 102 $ per 1 m 2 . The annual total cost C is assumed as; According to [34,38], let V equals 0.3 F per year, and it includes the cost of maintenance. Let the expected distiller lifetime is 10 years, then C = 102 + (0.3 × 102 × 10) = 408 $. The average daily productivity is about at 2.45 L/m 2 . day, and that the distiller operates 340 days in a year. The production during the life of distiller is 2.45 × 340 × 10 = 8160 L. The cost of 1 L from a reference distiller = 408/8160 = 0.05 $.
As a result, the cost of 1 L from the conventional distiller equals to 0.05 $. Moreover, the cost of 1 L from the modified drum distiller equals 0.041 $.

Conclusions
This paper investigates the performance of drum solar still and modified drum solar still under different operating conditions. Regarding the above analyses, it was concluded that the thermal performance of the modified drum solar still was better than that of either the drum solar still or conventional distiller under all testing conditions. Besides, the conducted modifications can actually benefit the solar still, where the cost of 1 L from the CSS and MDSS was 0.05 and 0.041 $, respectively. Also, the following points are concluded.
1. Without any modifications, the maximum rise in productivity for both DSS and MDSS was achieved at the drum speed of 0.1 rpm.
At this speed, the productivity of MDSS, DSS, and CSS was 8800, 7450, and 2400 mL/m 2 , respectively. As a result, the output yield was improved by 266% and 210% when using the MDSS and DSS respectively over that of CSS. 2. Using the wick improved the performance of the MDSS at the low drum speeds (0.02 and 0.05 rpm) and had a bad effect on the performance of MDSS at high speeds. So, the maximum productivity rise of MDSS with wick was obtained at 0.05 rpm, where the increase in productivity recorded its highest value of 280% over the CSS.
3. In addition, using the internal reflectors on the back wall of the MDSS increased the freshwater productivity by 14% over that of MDSS without reflectors. Besides, the total productivity of the MDSS with mirrors was increased by 280% compared to that of the CSS. 4. Moreover, using nanofluid with the MDSS increased the freshwater productivity by 16% over that of MDSS without nanofluid. The productivity of the MDSS with mirrors and nanofluid is increased by 296% more than that of the CSS.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.