Introduction

Unsafe drinking water supply, inadequate sanitation and insufficient hygiene practice are the factors causing the major share (88%) of all diarrhea cases. Diarrhea is a leading killer of children, accounting for nine percent of all deaths in 2015 (UNICEF 2016). About 6000 children underage five die every day, and a child dies at every 8 s from water-related disease around the globe. This accounts 19% of total child deaths in developing countries (Gomez-Couso et al. 2009). It also causes malnutrition, with the subsequent consequences on physical development and susceptibility to other infections (Gomez-Couso et al. 2009; Byrne et al. 2011; Fontan-Sainz et al. 2012).

Despite the fact that developing countries have large amount of freshwater resources, treatment plant construction and inappropriate treatment cost limit the distribution of the system at a household level. According to Bekele and Leta (2016), 51% of Ethiopian rural residents depend on surface water without treatment. Although conventional water treatment improves water quality, studies have shown that household water treatment techniques could also be used to treat water (Sobsey et al. 2008). WHO estimated improving access to safe water and sanitation ought to forestall at least 9.1% of the international burden of ailments and 6.3% of all deaths (Byrne et al. 2011). Point of use systems refers to the range of water treatment methods including solar treatment, physical treatment, chemical treatment and combined treatment which treat water at the point of use by avoiding contamination during distribution, collection, transportation and storage (Sobsey 2002).

SODIS is a cheap and easy to use, environmental friendly and effective drinking water disinfection technology (Sobsey et al. 2008). The method treats contaminated water in transparent plastic bottles through exposure to sunlight for a minimum of 6 h (Byrne et al. 2011). Following the exposure time, the water is safe to be consumed as the microbial loads can be significantly reduced. SODIS technology enhances bactericidal effect of UVA electromagnetic region (wavelengths in the range of 320–400 nm) of solar radiation with synergism effect of heat (infrared wave) and in the presence of dissolved oxygen species for inactivation of pathogens in the water. UV radiation showed an adverse effect on microbial skill to achieve cellular respiration and generation of adenosine triphosphate (Bosshard et al. 2010).

Drinking water disinfection by natural and/or artificial sunlight is widely studied using plastic or/and glass materials by the aid of photoconcentrator catalysts for the inactivation of a number of bacteria (Helali et al. 2013). Light intensity of wavelength, solar exposure time, availability of dissolved oxygen, turbidity level and water temperature are major study variables that affect the efficiency of SODIS (Byrne et al. 2011).

The study aimed to compare the disinfection potential of (GB)-, (PV)- and (AL)-based SODIS disinfection using some indicator organisms like E. coli, total coliforms and heterotrophic plate counts by using M. oleifera as pre-treatment for turbidity reduction.

Materials and methods

Experimental setup

Laboratory- and field-based experiments were conducted from May 24, 2016, up to August 2, 2016, to investigate the solar disinfection (SODIS) potential of three concentrators (GB, PV and AL).

Quality assurance

During laboratory experiment, all activities were performed in line with laboratory quality standards. All sample bottles and glass wares were sterilized keeping standard time, temperature and pressure. Control sample was kept at room preventing from sunlight and high temperature exposure.

Physical parameters

Turbidity removal test

A turbidity removal test was conducted on natural surface water from Ginjo Gudru by using M. oleifera seed powder. Initially, the raw water turbidity level was 28.6, 30.7 and 45.6 NTU taken at different sampling times. The jar test apparatus was used in a turbidity removal experiment, whereby 10, 30 and 50 mg/L dose of M. oleifera seed powder was added in each 1-L beaker containing the water samples at three turbidity levels, and the change was measured after 2 h with their control (Lea 2010).

UVA irradiation measurement

The solar irradiance was measured at different weather conditions from May 24, 2016, to August 2, 2016, consistently high between 645 and 1200 W/m2 (Gomez-Couso et al. 2009; Giannakis et al. 2014).

UVA was measured within every 30-min intervals (Reed 1997) from 12 up to 31/2 pm by calibrated solar taster and clear sky calculator with a central wavelength of 320–400 nm which provides data in terms of incident WUV/m2. QUV was calculated for comparison of solar test results (Helali et al. 2013).

It estimates the accumulated UV energy in the solar reactors per unit of treated water volume for given periods of time.

$${\text{QUV}}\; = \;\mathop \sum \limits_{n}^{uv} n - 1 A_{r} /V_{t} \left( {t_{n} - t_{n} - 1} \right)^{ }$$
(1)

where tn is the experimental time for n-sample, UVn−1 is the average solar ultraviolet radiation measured during the period (tn − tn−1), Ar is the illuminated reactor surface, and Vt is the total water volume.

Dissolved oxygen

Dissolved oxygen was one of the critical parameters in this study, and the initial dissolved oxygen availability was measured to compare with the final contents at different solar exposure times. All containers were agitated at 30-min intervals, to maintain oxygen equilibration within the water samples. By using dissolved oxygen measuring apparatus, the variable was measured at 30-min interval similarly to another variables (Burgess et al. 2007).

Microbial indicator organisms

E. coli and total coliforms

Membrane filtration technique was used for both total coliforms and E. coli, 3.81 g of membrane lauryl sulfate broth mixed with 100 mL of distilled water and sterilized at standardized time, temperature and pressure (Oxoid, Basingstoke, England) as ISO 9308-3:1999 and ISO 9308-1:2000 (Almeida et al. 2015). All samples were replicated and prepared at required dilution (0.1 and 0.01) since it is the standard dilution for surface water or rivers. One hundred milliliters of diluted sample discharged to vacuum pumper which contains filter paper with pour size of 0.45 μm and 47 mm diameter. Finally, the filter paper transferred to sterilized petri dishes with 50 × 12 mm diameter that contains absorbent pad already pipetted with 2 mL of culture media. After incubating at optimum temperature and time, colonies were counted to compare with initial load (Myers 2003).

Heterotrophic plate counts

The pour plate method or standard plate count method was used to determine heterotrophic plate count bacteria density. 2.35 g of plate count agar was mixed with 100 mL cool water and sterilized as standard; then, melted medium was kept in a water bath between 44 and 46 °C until used. The appropriate amount (2 mL) of diluted sample pipetted into the sterilized petri dishes for each different volume of diluted sample used 12 mL of liquefied media, and each dilution was replicated. The medium poured into the dish by gently lifting the cover just high enough to pour. Melted medium was mixed thoroughly with the sample in the petri dish by rotating the dish in opposite directions or by rotating and tilting the plates placed on a surface level inside the hood and solidify within 10 min. Finally, dishes placed inverted to prevent condensation and seal in a plastic sheet, followed by incubation at 35 °C for 48 h (Stillings and Herzig 1998).

Chick–Watson model of microbial disinfection kinetics

The inactivation efficiency of solar irradiation and water temperature was evaluated using natural surface water in different exposure times. The level of microbial inactivation was described by first-order Chick’s law reaction number of microorganisms destroyed per unit of time which is proportional to the number of organisms in the form of log inactivation.

$${\text{d}}N/{\text{d}}t\; = \; - k*N \to \, \ln \, \left( {N/ \, No} \right) \, = \, - k*t$$
(2)
$$N/N_{O} = e - k.t \to {\text{reduction factor}}\,\left( R \right)$$
(3)
$${\text{Log inactivation}}\; = \;{\text{Log}}_{10} N_{t} /N_{O}$$
(4)
$${\text{Percent inactivation}}\; = \;\left( {1 - N_{t} /N_{o} } \right)\; \times \;100$$
(5)

Therefore, the relationship between log inactivation and percent inactivation is as follows:

$${\text{Percent inactivation}}\; = \;\left( {1 - \frac{1}{{10\log {\text{inactivation}} }}} \right)\; \times \;100$$

or

$${\text{Log inactivation}}\; = \;\log \left( {\frac{100}{{100 - {\text{percent}}\, {\text{inactivation}}}}} \right)$$

where NO = initial (influent) concentration of viable microorganisms; NT = concentration of surviving microorganisms; Log = Logarithm to base 10, t = time, k = rate constant—this depends on disinfectant concentration, organism and temperature (Mcguigan et al. 1998).

Solar disinfection methods

Glass bottle solar disinfection

Figure 1a presents that half-blackened three glass bottles with size of 8 cm were exposed to UV light and heat at three different turbidity levels. The optical inactivation process was improved by wrapping reflective aluminum foils and coating a dark opaque substance at the rear half of the glass bottle. The glass bottles have 89–90% transmittance efficiency in the UVA (320–400 nm) wavelengths range, whereas transparent polyethylene terephthalate bottles transmittance might be as high as 85–90% in the same wavelengths range (Navntoft et al. 2008).

Fig. 1
figure 1

Solar concentrators exposed to thermal and optical energy. a Half-blackened and aluminum-foil-coated glass bottle. b Photovoltaic box with aluminum foil. c Schematic representation of the mechanism of photocatalysis on titanium dioxide particles. d Aqua lens disinfection

Full size image

Photovoltaic solar disinfection

Figure 1b shows that the bottom surface of the PV box is alone coated with semiconductor photocatalyst TiO2, while the top portion of the box was left; 92–98% transparent acrylic sheet was used to pass UV light through and fall on the TiO2 layer at the bottom surface to determine bacterial inactivation after exposure. Bacterial inactivation was calculated within the interval of 30 min from each turbidity category (Murugan and Ram 2018).

Solar Reactor—Photocatalytic Reactions of TiO2

Figure 1c presents a photon (hv) from the UV spectrum of the sun’s rays fall on the TiO2; it produces an electron–hole pair as shown in the following reaction.

$${\text{TiO}}_{2} \; + \;{\text{ hv}} \to {\text{TiO}}_{2} \cdot \left( {{\text{e}}^{ - } + {\text{ hole}}^{ + } } \right)$$
(6)

The recombination reaction is

$${\text{TiO}}_{2} \cdot \left( {{\text{e}}^{ - } \; + \;{\text{hole}}^{ + } } \right)\; \to \;{\text{TiO}}_{2 + } {\text{heat}}$$
(7)

The overall reaction of the photocatalytic reaction when water flows through it is:

$${\text{TiO}}_{2} \; + \;{\text{ hv}}\; + \;{\text{H}}_{2} {\text{O}}\; \to \;{\text{ TiO}}_{2} \cdot \left( {{\text{e}}^{ - } + {\text{ hole}}^{ + } } \right) + {\text{ OH}}^{ - } + {\text{ H}}^{ + } .$$
(8)
$${\text{hole}}^{ + } \; + \;{\text{hole}}^{ + } \to {\text{OH}}$$
(9)

This highly oxidative OH° degrades the organic molecules like bacteria, thereby killing the bacteria present in the water. After this, OH° reacts with H+ and the electron to reform H2O water back (Murugan and Ram 2018).

$${\text{OH}}^{0} + {\text{e}}^{ - } \; \to \;{\text{OH}}^{ - }$$
(10)

Aqua lens solar disinfection

Aqua lens is a developing technology, with low design expenditure by locally available materials and ease for operation. As indicated in Fig. 1d, the structure contains four each 2.5-m-long stand woods and 75–100 cm in diameter plastic sheet to hold water which used as solar concentrator or lens, at the bottom four small tires connected for each stand to push or pull easily in the light direction to get focal point.

Data analysis

Data were analyzed using SPSS software for Windows Version 20 and Microsoft Excel tool 2013. The linear regression coefficient of determination and descriptive statistics of standard deviation were used to summarize the data. The log inactivation tests were conducted to compare inactivation rate of concentrators for each indicator organism (Dessie et al. 2014).

Results

Moringa oleifera (MO) turbidity removal efficiency at low turbidity level

As indicated in Fig. 2, the removal efficiency of M. oleifera became increased as of the water turbidity increases.

Fig. 2
figure 2

Moringa oleifera turbidity removal efficiency

Full size image

UVA irradiation measurement for solar disinfection experiment

All experiments were exposed for 31/2 h. to follow log inactivation of target indicator organisms. In three disinfection setups, all variables were measured within 30-min interval including UVA irradiation. In different weather conditions and solar exposure time, the UVA intensity was measured in the range of 645–1200 WUV/m2.

Microbial inactivation of aqua lens (AL), photovoltaic (PV) and glass bottle (GB) at different turbidity levels

At 2.81 NTU turbidity level

As shown in Fig. 3a, both AL and PV concentrators completely inactivated total coliforms before 21/2 h. with different log inactivation levels, but glass bottle continued to 3 h for complete inactivation. Figure 3b presents aqua lens concentrator inactivated E. coli before both PV and GB with log inactivation of log 2.5 at one and half hour but completely inactivated before 2 h. Except glass bottle, both concentrators completely inactivated heterotrophic plate counts before 3 h. of solar exposure time (Fig. 3c).

Fig. 3
figure 3

Comparison of microbial inactivation between concentrators at 2.81 NTU turbidity level. a Comparison between concentrators for total coliform counts inactivation. b Comparison between concentrators for E. coli inactivation. c Comparison between concentrators for heterotrophic plate count at 35 °C for 48-h inactivation

Full size image

At 3.41 NTU turbidity level

Glass bottles showed the least inactivation solar exposure time than others (Fig. 4a). On the other hand, both AL and PV concentrators completely inactivated before 21/2 h. exposure time even if PV shows high log inactivation than AL for TCF counts at 3.41 NTU. As indicated in Fig. 4b, all concentrators completely inactivated E. coli at similar solar exposure time with different log inactivation potential progresses at 3.41 NTU.

Fig. 4
figure 4

Comparison of microbial inactivation between concentrators at 3.41 NTU turbidity level. a Comparison between concentrators for total coliform counts inactivation; b comparison between concentrators for E. coli inactivation; c comparison between concentrators for heterotrophic plate count at 35 °C for 48-h inactivation

Full size image

Figure 4c presents that photovoltaic and glass bottle concentrators completely inactivated after log 3 and log 1.63, respectively, but aqua lens inactivated totally before 3 h. with log 2.14.

At 6.9 NTU turbidity level

As shown in Fig. 5a, total coliforms extended to 31/2 h. in GB concentrator to inactivate completely, but both AL and PV inactivated entirely before 21/2 h. As shown in Fig. 5b, E. coli inactivated before 21/2 h. solar exposure time in both aqua lens and photovoltaic concentrators, while glass bottle inactivated completely at 3 h. of solar exposure time. Heterotrophic plate counts were completely inactivated after 3 h. in both PV and GB concentrators, but entirely inactivated before 3 h. in AL concentrator as shown in Fig. 5c.

Fig. 5
figure 5

Comparison of microbial inactivation between concentrators at 6.9 NTU turbidity level. a Comparison between concentrators for total coliform counts inactivation. b Comparison between concentrators for E. coli inactivation. c Comparison between concentrators for heterotrophic plate count at 35 °C for 48-h inactivation

Full size image

With regard to temperature concentrating potential of aqua lens, photovoltaic and glass bottle in similar solar exposure time showed different concentrating potentials. In aqua lens, high temperature was recorded at 6.9 NTU turbidity level as depicted in Fig. 6a–c.

Fig. 6
figure 6

Comparison between aqua lens, photovoltaic and glass bottle temperature concentrating potential at different turbidity levels. a Temperature concentrating potential of concentrators at 2.81 NTU. b Temperature concentrating potential of concentrators at 3.41 NTU. c Temperature concentrating potential of concentrators at 6.9 NTU

Full size image

Discussion

Moringa oleifera as pre-treatment for solar disinfection

Moringa oleifera seed powder was used as a primary natural coagulant for low-cost household water treatment (Yahaya et al. 2011), at a range of turbidity levels. The removal efficiency of M. oleifera was increased when the raw water samples were highly turbid. For water sample of relatively high initial turbidity of 45.6 NTU, M. oleifera produced the best results with an average turbidity reduction of 94%. Similar studies have been reported on turbidity removal efficiency and initial turbidity level (Abatneh et al. 2014).

In all three experimental setups, the water samples that contain relatively high turbidity of 6.9 NTU were achieved low inactivation rate of indicator organisms than the others. A different study that has been conducted in this area explains that inactivation rate clearly decreased when turbidity increased. Naturally dissolved organic matter may act as a photosensitizer and hence advance the inactivation process (Burgess et al. 2007). In aqua lens, E. coli inactivation showed negatively strong association (R2 = 0.98), and both PV and GB were associated negatively (R2 = 0.95) at different turbidity levels. High turbidity increased inactivation times and enhanced bacterial re-growth (Kehoe et al. 2001).

Solar irradiance measurement

Under UVA exposure, the biocidal action of UVA has also been attributed to the production of reactive oxygen species which are generated from dissolved oxygen in water.

The photosensitization of molecules in the cell with any naturally occurring dissolved organic matter can absorb photons of wavelengths between 320 and 400 nm, to induce photochemical reactions (Burgess et al. 2007; Byrne et al. 2011).

Water temperature measurement

As laboratory result indicated next to aqua lens, photovoltaic box disinfection showed relatively high inactivation rate than glass bottle, almost for all indicator organisms in short solar exposure time. Total coliforms were inactivated in AL (SD = 15.8 °C, R2 = 0.92) followed by PV inactivation temperature association (SD = 11.6 °C, R2 = 0.90), and the GB concentrator was inactivated (SD = 10.9 °C, R2 = 0.70) at turbidity level of 3.41 NTU. The role of semiconductor TiO2 with natural organic matter to form highly disinfectants reactive oxygen species and the high light transmittance potential of acrylic glass sheet supports to attain high inactivation rate than GB (Mcguigan et al. 2012; Murugan and Ram 2018).

Water temperature was raised through progress when solar exposure time becomes increased. All disinfection setups were measured slightly different from water temperature concentrating potentials. The aqua lens measured highest water temperature (74 °C) followed by photovoltaic (60 °C) and the glass bottle (52.5 °C) which was the lowest water temperature recorded in this study at different turbidity levels. Similar studies have been investigated that glass bottle is able to concentrate 45–50 °C water temperature at different weather conditions with global irradiance of 800 W/m2 (Navntoft et al. 2008; Mcguigan et al. 2012).

Aluminum foil was used at back side of half-blackened GB, PV and AL disinfection setups. High reflective power of aluminum foil to concentrators was elevated water temperature at short exposure time. Experimental studies confirmed that aluminum foil back-coated solar concentrators increased disinfection rate constants by a factor of twofold (Navntoft et al. 2008; Kehoe et al. 2001; Rengifo-herrera et al. 2011; Mcguigan et al. 2012).

In this study, dissolved oxygen in the water container ranged between 6.35 and 8.74 mg/L, and as investigated in laboratory experiment dissolved oxygen was inversely related to water temperature (R2 = 0.78). A similar study explained that the amount of dissolved oxygen quickly utilized in some stages of the reaction due to increment of temperature (Byrne et al. 2011; Giannakis et al. 2014).

All experiments were exposed to both optical (UVA light) and thermal (infrared heat) electromagnetic spectrum inactivation system. The inactivation result which is shown above is direct synergism effect of light and heat. However, bactericidal action was evidenced at temperatures above 40–45 °C with a synergistic SODIS process (Byrne et al. 2011; Mcguigan et al. 2012; Helali et al. 2013; Giannakis et al. 2014). The E. coli, total coliforms and heterotrophic plate counts inactivation experiments were conducted at natural waters source in three separate turbidity levels (2.81, 3.41 and 6.9 NTU). The relation between three indicator organisms like E. coli is assumed to be a subset of total coliforms, which is a subset of heterotrophic plate counts bacteria (Wilson and Andrews 2011).

In three and half hour solar disinfection exposure time almost in all experiments, E. coli were inactivated prior than total coliforms and heterotrophic plate counts with complete inactivation range of 2 h (2.81 NTU) in aqua lens and 3 h (6.9 NTU) in glass bottle disinfection system with the association (SD = 10.8 °C, R2 = 0.81). However, E. coli were inactivated in aqua lens concentrator (SD = 14.1 °C, R2 = 0.91), which shows high association between E. coli inactivation and water temperature. Different studies support this complete inactivation rate with indicated exposure time range, keeping other variables similar to this result (Mcguigan et al. 2012).

Mesophilic characteristics of E. coli thrive between 20 and 45 °C, but for the temperature beyond the range there is a thermal tension exerted to the cells, affecting the cell wall in addition to destroying the proteins and nucleic acids, leading to death of bacteria (Giannakis et al. 2014). Total coliforms and HPC completely inactivated at 21/2 and 3 h. in aqua lens solar concentrator with the association of (SD = 15.3 °C, R2 = 0.90) and (SD = 16.0 °C, R2 = 0.77) at 2.81 NTU turbidity level, respectively. Total coliforms were recorded similar inactivation time in PV disinfection system; however, 30 additional minutes were required to inactive heterotrophic plate counts completely, considering at different turbidity levels (Wilson and Andrews 2011).

Conclusions

The disinfection potential of concentrators was increased at 2.81 NTU. Aqua lens solar disinfection was shown to be an effective household water treatment system. Heterotrophic plate counts were more resistant than the other coliforms in SODIS treatment within similar inactivation exposure time.