24To reduce child mortality and improve health in Ghana boreholes and wells are being installed across 25 the country by the private sector, NGOs and the Ghanaian government. Water quality is not generally 26 monitored once a water source has been improved. Water supplies were sampled across Ghana from 27 mostly boreholes, wells and rivers as well as some piped water from the different regions and analysed 28for the chemical quality. Chemical water quality was found to exceed the WHO guidelines in 38% of 29 samples, while pH varied from 3.7 to 8.9. Excess levels of nitrate (NO 3 -) were found in 21% of the 30 samples, manganese (Mn) and fluoride (F -) in 11% and 6.7%, respectively. Heavy metals such as lead 31 (Pb), arsenic (As) and uranium (U) were localised to mining areas. Elements without health based 32 guideline values such as aluminium (Al, 95%) and chloride (Cl, 5.7%) were found above the 33 provisional guideline value. 34 35Economic information was gathered to identify water costs and ability to pay. Capital costs of wells 36and boreholes are about £1200 and £3800 respectively. The majority of installation costs are generally 37 paid by government or NGO, while the maintenance is expected to be covered by the community. At 38 least 58% of the communities had a water payment system in place, either an annual fee/one-off fee or 39 "pay-as-you-fetch". The annual fee was between £0.3-21, while the boreholes had a water collection 40 fee of £0.07-0.7/m 3 , many wells were free. Interestingly, the most expensive water (£2.9-3.5/m 3 ) was 41 brought by truck. Many groundwater sources were not used due to poor chemical water quality. 42Considering the cost of unsuccessful borehole development, the potential for integrating suitable water 43 treatment into the capital and maintenance costs of water sources is discussed. Additionally, many 44 sources were not in use due to lack of water capacity, equipment malfunction or lack of economic 45 resources to repair and maintain equipment. Those issues need to be addressed in combination with 46 water quality, coordinated water supply provision and possible treatment to ensure sustainability of 47 improved water resources. 48 49 50

a b s t r a c tInfluence of organic matter (OM) on uranium removal mechanisms by ultrafiltration (UF) over a pH range of 3-11 was investigated. Humic, alginic and tannic acid were used as OM. It was found that uranium adsorbed strongly to the membrane while retention by size exclusion did not occur. Adsorption was dependent on pH and type of OM used. Speciation predictions performed using Visual Minteq explain some of these results. In the absence of OM, uranium primarily adsorbed to the membrane at pH 5 and 7 where UO 2 OH + and UO 2 CO 3 were the dominant species. In the presence of humic acid (HA), uranium adsorption increased in the acidic range, especially at pH 3 (from 11% to 74%) due to the complexation. The structure of alginic acid (AA) did not favour complexation with uranium and therefore did not have a significant influence on its behaviour in UF. The exception was at pH 3 where adsorption increased from 11% to 52%. At this pH no charge repulsion between the uranium species and the AA occurs and complexation is favoured. The highest effect on uranium adsorption was obtained in the presence of tannic acid (TA) at pH 10 and 11 where adsorption increased from 20% up to 100%. Uranium is most likely forming complexes with the gallic acid fraction of the dissociated TA.

In order to provide safe drinking water in isolated communities where water supply and electrical infrastructure is limited, a system combining solar energy and advanced water treatment technology (a two-staged membrane process with ultrafiltration followed by nanofiltration/reverse osmosis) has been developed. The objective of this study is to model the speciation of commonly occurring and sometimes health-threatening trace contaminants, and then to evaluate the impact of speciation on fluoride, arsenic, and magnesium retention by nanofiltration/reverse osmosis. A series of experiments were performed in central Australia in late 2005 to assess elemental retention with four different membrane modules (BW30, ESPA4, NF90, and TFC-S) at various pH values between 3 and 11 at two different source waters (Pine Hill Station and Ti Tree Farm). The removal of fluoride and arsenic was observed to be independent of pH, while magnesium removal is pH dependent. Compound speciation is considered to explain dependence observations. Membrane type was found to impact retention. The results obtained in this study will provide valuable information on the reliability and the optimization windows of the renewable energy powered desalination system that has been developed.

During an extensive sampling trial in Ghana, a number of physico-chemical water quality problems have been identified. For example, pH values of the collected samples ranged from 3.69 to 8.88, while conductivity ranged from 10 to 45000 µS/cm and turbidity from 0 to > 542 NTU. Many water samples analysed breached the drinking water quality guidelines. High levels of heavy metals were found and also locations high in sulphate and nitrate. In some regions chemical contaminants such as fluoride occur naturally and result in a significant number of boreholes being capped. While such capping results in the protection of the population of exposure to such chemicals, it may increase the risk for microbiological safety if surface water sources are used.

Factors affecting uranium removal from brackish groundwater using a direct solar powered ultrafiltration-nanofiltration/reverse osmosis membrane system were investigated during a field trial in the Australian outback. The key variables were uranium speciation (as a function of pH), groundwater type as well as energy variation over the course of a day. It was found that uranium was retained by the membranes over the pH range 3-11, but strongly adsorbed to membranes at pH 4-7. The speciation of uranium pH 4-7 explained the adsorption to the membrane. The presence of other inorganic species, in particular calcium, was a likely cause of uranium co-precipitation at pH 10-11. During solar energy experiments, it was found that the specific energy consumption increased over the course of the day. This indicated fouling through precipitation on the membranes which caused reduced retention of uranium towards the end of the solar day.