Globally, children are dying at the rate of nearly one million deaths per year from diarrhoea and dysentery. These diseases are linked to both poor sanitation and contaminated drinking water. In the Pacific Ocean, Kiribati has the second highest rate of under-five mortality (World Health Organization, 2015). This is unlikely to improve with growth rates of 4.5% and an over-stressed water resource (Lal, 2014). One solution may lie in the use of solar disinfection or the SODIS method. The SODIS method has been used effectively around the world to disinfect drinking water since the late 1980s but has only recently been applied in the Pacific. The method involves filling a plastic bottle with water and lying it in the sun all day where upon the water is safe to consume. With its location straddling the equator, Kiribati is ideally placed with a year-round abundance of solar radiation. This free resource can be utilised to improve the quality of drinking water and ultimately reduce child mortality.

The aim of this research was to assess the influence of variables specific to Kiribati on the SODIS method. Experiments were conducted in a controlled environment in the Environmental engineering laboratory at the University of Canterbury to simulate the conditions found in Kiribati. The variables investigated included water conductivity, pH, total water hardness, depth of water i.e. bottle size and age of bottles used.

The method involved filling nine polyethylene terephthalate (PET) bottles with de-ionised water before irradiating the bottles using UVA-340 fluorescent lamps. Seawater was added to the de-ionised water to adjust the conductivity, while sodium bicarbonate and calcium chloride were added to fix the pH and increase the hardness respectively. Primary effluent from the Christchurch wastewater treatment plant was employed as the source of the indicator organisms. These were total coliform and Escherichia coli (E.coli). After every hour, one bottle was removed and sampled. Most Probable Number (MPN) of pathogens was determined using the Colilert-18 system by IDEXX.

The laboratory-based experiments did not have a strong correlation with real world experiments due to the use of constant irradiance lighting, artificial water and locally sourced pathogens. However, a strong correlation between experimental results carried out in the laboratory was found. The maximum temperature reached by the bottles under the lights was 38 °C; this meant the dominant form of inactivation was optical due to the ultraviolet radiation.

E.coli had the highest inactivation rate of 50 cm2/kJ even after beginning with a significant lag. This rate was double the inactivation rate of total coliform of 25 cm2/kJ, which raised concerns about the usefulness of E.coli as an indicator organism. E.coli clearly demonstrated that the SODIS method works, but may lead to unsafe water being consumed as it had a higher inactivation rate than other pathogens.

The pathogen inactivation rate reduced with increasing pH. There was a statistically significant difference (p=0.05) in pathogen inactivation between experiments carried out at pH = 6.5-6.8 and pH = 8.3. The implication of this for Kiribati (pH = 8.3) is that more time may be required to achieve the same inactivation as a location with lower water pH, subject to the same amount of solar radiation.

The conductivity of the ground water in Kiribati is weather dependant with levels of 400 μS/cm (post-rain) and 900 μS/cm (prolonged drought) being recorded, with the average level = 700 μS/cm. All three conductivity were investigated with no statistically significant differences between the inactivation rates being determined (p=0.05). The implication of this is that any SODIS method developed specifically for Kiribati will be independent of weather.

Increasing total water hardness (> 530mg/L CaCO3) appeared to increase the inactivation rate of E.coli. Increasing total water hardness by adding calcium chloride while maintaining specific conductance led to both variables increasing. Therefore, the increase in activation rate could be related to the increase in conductivity or hardness. It is likely the increase in total hardness was responsible however, as they was no statistically significant difference in the pathogen inactivation across the conductivity experiments.

Results from the small bottle experiments to identify the effect of water depth on the SODIS method were at variance with published literature (Dessie et al., 2014; Kehoe et al., 2001). The smaller bottles (Ø = 60mm) performed significantly worse (p = 0.01) than the larger bottles (Ø = 93mm) for inactivation of E.coli. The total inactivation after 9 hours was 1.8 Log and 2.4 Log for small and large bottles respectively. The increased wall-thickness was theorized to be responsible for the poor performance. This highlighted the need for pre-experiment physical examination of bottles to ensure compatible results because SODIS bottles are often recycled from a previous use and/or users may prefer a heavier bottle with the idea that it will last longer. This will result in a significant reduction in the inactivation of the pathogens.

PET bottles were artificially aged under ultraviolet lights for 150h, 1000h, 1500h and 3000h. This resulted in a 15% reduction in light transmitted between the 150h sample and the 3000h sample. Further research is needed to quantify how this reduction in light transmission affects the inactivation of pathogens.