Since the discovery of SODIS, several research and clinical trials have been undertaken by different researchers on its improvement and effectiveness such as Ismail et al. [26] and Swiss Federal Institutes of Science and Technology (EAWAG) Department of Water and Sanitation Developing Countries (SANDEC) [27]. For instance, the effectiveness of using small and large plastic containers, plastic bags, and PET bottles has been tested for purifying contaminated water [28] to meet drinking water requirements in low- and medium-income households [15]. Our design and prototype were based on three main SODIS concepts reported in literature namely the use of simple sunlight bottle exposure, [1] solar reflector incorporated SODIS system [24], and the Dual SODIS and solar still water purifier [26]. [Table 1] shows a matrix evaluation of the three concepts used as a basis in designing this enhanced SODIS prototype. Comparative analysis of the three concepts showed SODIS concept in the dual SODIS and solar was potentially the most effective and would perform better in rural areas of Malawi.

The focus of the design was on improving the SODIS element in the dual water purification system, by incorporating features of the other two concepts to enhance thermal capacity. Furthermore, the third concept was chosen due to the ease of incorporating features of the other concepts for improved performance and its relative advantages over the other two concepts ([Table 1]).

[Figure 1] shows an improved SODIS system developed in this study with four main components, namely solar collector (A), wooden support structure and windshields (B) the transparent plastic sheet, and the heat storage chamber (C). The system was constructed using locally sourced materials including an iron sheet, 2L PET bottles, timber, plywood boards, soft nail wires, black paint, and clear transparent plastic paper. The PET bottles were selected for this system owing to the good transmitting property of light in the ultraviolet A (UV-A) and visible range spectrum, low-cost, durability, reusable and are locally available.

Figure 1:

The SODIS prototype construction showing main components

Figure 2:

The SODIS prototype set-up (prototype) drawn using AutoCAD

[Figure 2] shows the schematic view of the main parts of the SODIS system being constructed and assembled. The solar collector is an aluminium metal sheet shaped into a big parabolic corrugation to allow focus sunlight rays into the water bottles placed in the troughs. The corrugation and the bottles used in the system were half painted black to enhance heat absorption and hence the heating capacity of the system. The windshields were incorporated to reduce heat losses induced by the wind, while the transparent plastic sheet was used to trap the heat generated in the system. The wooden support structure also included a heat storage chamber, which was located underneath the solar collector. The heat storage chamber was incorporated to store heat used to buffer the effects of cloud cover and to minimize heat losses from the system due to changes in atmospheric conditions. The top side of the system was covered by transparent plastic paper to allow sunshine penetration and prevent debris from entering the top chamber. The prototype of the SODIS system was fabricated by a local carpentry artisan in Mzuzu City in Malawi with supervision by the researchers. The size of the collectors’ corrugation and the entire SODIS system were estimated and set using eq. (1):

$T=0.75\left(2\text{π}r\right),$ (1)

where: T is the trough size, r is the radius of the 2L PET bottles used in the SODIS system and π is a constant taken as ²²/₇. Considering the measured radius of the PET bottles (r=0.095 m), the area of the solar collector was therefore found to be 0.495 m².

The optimum angle of tilt or inclination (θ) of the SODIS system was set using eq. (2):

$\theta =\beta \pm Z^{\circ },$ (2)

where β is the latitude of the area, Z is any integer between -10 and +10. Since Malawi lies between latitudes 9°22’S and 17°03’S and longitudes 33°40’E and 35°55’E [29], the 9° was taken as the design latitude parameter β, while +6 was chosen as the integer (Z) to be used and the optimum angle of tilt/inclination (θ) was therefore set at 15°.

The efficacy of the SODIS system is dependent on several variables including time of exposure, [1] solar intensity, the surface area of the collector [24], and the amount of water to be disinfected in the bottles [26]. Thermal analyses were performed by considering several parameters such as temperature measurements and heat energy exchange. The inside system temperatures and those in the ambient were determined by digital thermocouples and recorded the readings every 20 minutes. The temperature of water in the PET bottles was measured using a mercury thermometer. The heat energy that entered the system through the plastic into the water (Q_f) was calculated from eq. (3) [30]:

$Q_{\text{f}}=mc\Delta T,$ (3)

where: m is the mass of water placed in the bottles, c is the specific heat capacity of water (4.187 kJ/kg^oC), and ΔT is the change in water temperature given as $\Delta T=t_{\text{i}}-t_{\text{f}},$, in which (t_i) and (t_f) are initial and final temperature readings, respectively.

The heat transfer rate (Q_{f rate}) was thus estimated by using eq. (4):

$Q_{\text{f}\text{rate}}=\frac{Q_{\text{f}}}{t},$ (4)

where: Q_f is the amount of heat that enters the plastic bottle and absorbed by the water as expressed in Equation 3, and t is the time for obtaining the heat increase.

The amount of heat energy received by the system is dependent on the latitude of an area, which affects the solar constant [26]. This parameter was determined using eq. (5):

$Q_{\text{y}}=I\times A,$ (5)

where: Q_y is the amount of heat energy received, I is the solar constant, and A is the collector surface area.

The amount of radiation from the sun lost (Q_i) in SODIS system depends on properties of the material (in this case the transparent plastic paper) namely transmissivity (τ) and absorptivity (α) product is given by eq. (6) as used by Robaa [31]:

$Q_{\text{i}}=I\times A\times \left(\tau \times \alpha \right),$ (6)

where: τ is the percentage of solar radiation passing through the plastic paper (also called transmissivity), and α is the percentage of solar radiation absorbed by the collector.

Regardless of the property of the material used there are always heat gains and losses. The heat gained or lost by the water was determined using the heat gain or loss coefficient (U_L) as expressed in eq. (7) [32]:

$U_L=\frac{Q_i}{\left(A\times \Delta T\right)},$ (7)

where: Q_i is the amount of radiation received, A is the collector surface area, and ΔT is the change in water temperature.

The efficiency η of the SODIS system was estimated using eq. (8) as used in Kandpur and Sudhir [33]:

$\eta =\frac{\Delta Tmc}{IAt},$ (8)

where: ΔT is the change in water temperature observed in a short time interval t, A is the surface area of the collector, c is the specific heat capacity of the water, I is the solar irradiance constant of the particular area, and t is a time interval the temperature change was observed.

Literature shows that the heating efficiencies of the solar dryers or heaters vary widely depending on mass loaded in the system and the weather conditions of an area. In SODIS systems, typical efficiency values have been reported to fall within the ranges of 10 to 15% [34].

The coliforms are bacteria that originate mainly from the intestinal tract of warm-blooded animals. Their presence in any water source provides an indication of faecal contamination and the disease-causing microorganisms [35]. The effectiveness and efficiency of the developed SODIS system were determined by using contaminated raw water samples collected from rural areas of Chisenga and Chitipa (Malawi). The standard operating procedures and guidelines were followed to collect, preserve and transport the water samples used for SODIS system testing: Malawi Standards (MS) 682-1:2002 [36], MS 682-3:2002 [37], and American Public Health Association (APHA) [38]. Thus, adherence to standard protocol helped to prevent any cross-contamination and alteration of the samples. The samples were analysed for faecal and total coliform in triplicates in accordance with standard procedures [39]. The bacteria were analyzed using the membrane filtration method within a 24 h required period after sampling. Both the raw and treated water samples (100 mL) were filtered through a special membrane (0.20 µm pore size; 47 mm diameter). Petri dishes containing filter membranes and selective nutrient media (M-Lauryl Sulphate Broth) were incubated at 44.5 ± 0.5 °C for 24 h. The number of bacteria colonies obtained were expressed as colony-forming units (cfu) per 100 mL. The choice of using water samples from the two areas followed the results of a baseline study on the assessment of microbial contamination of drinking water sources where gross contamination of the drinking groundwater was reported [40]. The turbidity levels of the water samples were previously evaluated to be below 30 Nephelometric Turbidity Unit (NTU); hence the SODIS System testing with such water samples was practicable and in line with the literature [41]. Further, a comparison of inside system temperature to those in ambient was done to evaluate the efficacy of the enhancements made to the SODIS system.

Considering the market study and costs obtained for the different materials used to build the system, the cost of the SODIS system was evaluated. The social feasibility and product acceptability of the system was not evaluated due to the time and financial constraints of the study.

To ensure quality control and assurance, distilled water samples (<0.5 NTU) were used as controls in the microbiological water quality analyses. One set of the distilled water samples were spiked with coliforms while the other set was not spiked. The control samples were used to check the effects of turbidity on the disinfection efficiency of the system. These control samples were prepared and run concurrently in the SODIS system with the raw and untreated water samples. The raw water samples were collected in triplicate to improve the accuracy and reliability of the results. Prior to their reuse, the PET bottles were disinfected and cleaned with a dilute hypochlorite solution and washed thoroughly with a detergent. Before filling with the water samples, the bottles were further rinsed with distilled thrice followed by the water used in the experiments.

One-way Analysis of Variance (ANOVA) and t test were used to determine the variation and correlation of the bacterial contamination and disinfection efficiency of the water samples. To ensure the accuracy of the thermocouples used in temperature measurements, calibration using a mercury thermometer was done. The data and descriptive statistics such as averages, ranges, and standard deviations were performed using MINITAB and Microsoft Excel (2013). The temperature variation was analyzed by checking trends from graphs drawn using Microsoft Office Excel (2013). The microbiological quality results of the treated and untreated water were compared with the MS678 [42] and World Health Organisation (WHO) drinking water guidelines [43].

The following sub-sections give a discussion on the disinfection system performance. The results and discussion on ambient and the disinfection system temperature variations for the two days of testing are also provided. Furthermore, the subsequent sections give the results and discussion on the application to real water samples and the effectiveness of the SODIS system in water disinfection.

A. Ambient and the disinfection system temperature variations. The system was tested for two days – the first day on 13^th September 2014 (10:33 am- 14: 33 pm) and the second day on 25^th November 2014 (9:33 am - 13:33 pm). [Figure 3] and [Figure 4] show the line graph of temperature recording for the first and second day of SODIS system testing, respectively.

On the first day, the temperature inside the system rose from --25.3 to a maximum of 64.3 °C ([Figure 3]). The maximum temperature difference between the temperature inside and those in the ambient were- 36.4 °C. The mean water temperature inside the bottle rose from 24.3 °C to 66 °C. The temperature rise inside the system was attributed to the thermal enhancement in the system. A 50 °C temperature is recommended for disinfection and such the system surpassed this threshold, meaning it was effective [27] . On the second day, the inside system temperature rose from an initial minimum value of 26.5 °C to a maximum temperature of 69.7 °C ([Figure 4]). The maximum temperature difference between the inside SODIS system and the surrounding was- 36.4 °C. The mean bottle water temperature rose from an initial temperature of 24.8 °C to 68.7 °C. In this SODIS design, the corrugated aluminium sheets and the PET bottles were painted black to enhance heat absorption. This promoted the heating capacity of the entire system, and this is consistent with previous research studies conducted elsewhere notably Ubomba-Jaswa et al. [44], Berney et al. [45], and Atikul [46]. The transmission properties and characteristics of the PET plastic bottles used in this study are similar to those reported by Dejung et al. [47].

Figure 3:

Line graph of temperature recording on the first testing day

Figure 4:

Line graph of temperature recordings on the second testing day

B. Application to real water samples. To evaluate the effectiveness of the SODIS system in water disinfection, microbiological water quality analyses were performed using shallow well water samples obtained from Chitipa. The specific areas where water samples were collected were Chisenga (C1a, C1b, C1c), Magomero (M2a, M2b, M2c) and Namkuntha (N3a, N3b, N3c). [Figure 5] shows the microbiological characteristics of the water samples used for SODIS system testing.

The system efficacy and effectiveness were evaluated after 3, 4 and 5 hours of exposure regimes in the SODIS prototype. [Table 2] and [Table 3] show a summary of the disinfection time for coliforms in the water samples. It is evident that almost all the coliform organisms (total coliform and feacal coliform) were killed within a period of 3 and 5 hours of exposure as indicated in both the results. After exposure to sunlight for 3 hours, the SODIS treatment of the samples generally resulted in nil total coliform count (0 cfu/ 100 mL) except for one from Magomero (1 cfu/100 mL) ([Table 2]; [Table 3]). This was attributed to poor handling of the sample resulted from human errors. Generally, the SODIS system successfully reduced the number of bacteria under study to the levels required for drinking water. Similar results where SODIS treatment was used to kill and inactivate microorganisms in water have been reported extensively in literature including in Nepal by Rainey and Harding [48], and in rural coastal Bangladesh by Islam et al. [46].

According to literature the thermal inactivation of E. coli has been reported to normally occur at temperatures above 40 °C [45],[49]. In this study, this minimum temperature of 40 °C was reached after about 80 minutes of water exposure during the first and second days of testing, respectively ([Figure 3] and [Figure 4]). The system reached the required temperature to kill all the microorganisms present. The results in the current study are consistent with a study by Ubomba-Jaswa et al. [44] who reported on complete bacteria inactivation temperatures between 45 and 55 °C after 5 hours of exposure of the water samples in SODIS.

Figure 5 (A-B):

Microbiological characteristics of the water samples for DAY 1 and DAY 2. Samples from Chisenga (C1a, C1b, C1c), Magomero (M2a, M2b, M2c) and Namkuntha (N3a, N3b, N3c). For DAY 1, other sets of distilled water samples were spiked with coliforms

A t-test was run for only the number of total coliforms after 3 hours because all the other values did not have a variance with the standards set by the WHO and Malawi Bureau of Standards (MBS). The t-test was set to assess if the water treated using the prototype met the WHO and MBS Total coliforms set standards in which a t of 1.00 was obtained, which with 8 degrees of freedom was significant at the 0.347 level. This, therefore, showed that the water was safe for drinking and that the prototype was effective in water treatment.

The effectiveness of a SODIS technology is measured by its thermal conversion of solar energy to sufficiently raise the water temperature to kill the pathogens [28] , which constitute one of the acute health risks of using water that is contaminated in low and middle-income households [50]. The optical properties of the materials used in the construction of the SODIS play a significant role in the thermal performance of the SODIS [15] . This section provides the values obtained from the calculation done using the equations outlined earlier and summarized in [Table 4] and [Table 5].

The heat energies that entered the system through the plastic into the water (Q_f) during the first and second day of testing were 2,095.18 and 2,205.71 kJ, respectively. Although these values differed slightly attributed to changes in sunlight intensity, the differences were not significant (p > 0.05). The rates of heat transfer obtained in this study were 0.20 and 0.21 kW for the first and second day of testing, respectively. The heat gains or loss rates, on the other hand, were found to be 12.91 and 12.26 W ([Table 4]).

The efficiency of the system (η) was determined using eq. (8) as the mass m was 12 kg, c was 4.187 kJ/kg^oC, I was 823 W/m²(J/m²s), t was 5 minutes (300 seconds) and there was a minute temperature change 2.5 ^oC.

$\eta =\frac{2.\text{5}{}^{\text{o}}\text{C}\times 12\text{kg}\times 4187\frac{\text{J}}{{\text{kg}}^{\text{o}}\text{C}}}{823\text{J}/{\text{m}}^{\text{2}}\text{s}\times 0.0495{\text{m}}^{\text{2}}\times 300\text{s}}=10.2\%$

The SODIS system performed fairly well on a sunny day as it was on the second day of testing. Its efficiency was 10.2% which was fairly good compared to the very best achieved efficiency of 15% reported in the literature [44].

The SODIS are designed to be effective (in providing safe water), available, affordable and acceptable to the low and medium-income households, which are mostly affected by outbreaks of waterborne diseases [10] . The SODIS system developed in this study has several benefits over other points of use water treatment options such as Water Guard, chlorination and boiling water treatment commonly used in rural areas of Malawi. The cost of the SODIS system was US$ 32.45 compared to the retail price of Water Guard, which was at US$ 0.32/L during the time of this study. This implies that a household has to buy at least 80 bottles of Water Guard to reach the system’s cost of water treatment achieved by the SODIS developed in this study. Furthermore, the use of Water Guard may put the household at risk if the household has no reliable source of income, unlike incurring the cost of purchasing the SODIS system once and for all. In addition, the supply chain of water guard in rural areas in Malawi is challenging, requiring households to travel to urban centres to source it, which in turn increases the cost per unit of disinfected water. The SODIS system is also better than boiling using biomass, either in wood or crop residue form as people spent time for biomass collection. The distance for collecting biomass for thermal applications at the household level in Malawi has increased over time. The supply chains of agricultural residues that can be used for boiling water are seasonal and have many competing uses [51], which exacerbate the sustainability challenge of the agricultural residues supply chain for water heating. In addition, if burnt in a room, biomass produces a lot of soot or smoke [52], which is a risk factor for lower respiratory functions causing a majority of deaths among children [53]. Furthermore, the use of biomass contributes to environmental degradation as it is estimated that a kilogram of wood is required to boil 1 litre of water. People also dislike the taste that boiling treatment method of water imparts on the water [24] . A study done in Chikwawa – Malawi on the use of Water Guard to treat water showed several residents did not use it due to cost implications [54]. Several studies in the country showed that the majority of people in the rural areas use untreated contaminated water fetched from shallow wells, rivers [55] and groundwater (boreholes) [56]. Use of the SODIS in treating the contaminated water can have positive health benefits among users in the country. This has been demonstrated in other countries such as in Bangladesh [46] , Cameroon [14] and Bolivia [57], where use of the solar disinfection system is being promoted.

The efficacy and effectiveness of the system, auger well with the goals and targets of the SDG 6, the SADC Water Policy (2005) and Malawi National Water Policy (2005). The developed SODIS system, if properly put to use can help promote and improve access to safe and affordable drinking water for all – especially in the rural areas. Since the system does not require use of any form of fuel or electricity, it can significantly help communities treat water for home use. Furthermore, use of the system can help conserve the forests and other forms of energy. The previous SODIS studies were performed in Yaounde, Cameroon [14], Nepal [48], Bolivia . [47] and many other places worldwide, have demonstrated the significance of solar disinfection. The results of the current study has also demonstrated the need for use of SODIS in rural areas where access to clean and potable water is a challenge. The Malawi National Energy Policy (2018) aims at “increasing access to affordable, reliable, sustainable, efficient and modern energy for every person in the country”. This policy is also in tandem with SDG 7, “promoting the use of affordable and clean energy” [7]. The use and promotion of clean and sustainable energy (solar) in the disinfection system to treat water are, therefore, in line with the aims and goals of the Malawi national energy policy and water policy as well the SDG 7.