Abstract
This thesis addresses the major problem of high turbidity water in drinking water treatment to small rural and urban populations. This problem is attributed to polluted surface water sources and untreated piped water supply systems. In this research, the problem is addressed in the context of the Gilgit-Baltistan located in the extreme north of Pakistan, however, the solutions proposed may be applicable in other parts of the world with similar conditions of glacial melt water as a source for gravity fed drinking water delivery systems. Glaciers, lakes and seasonal snow deposits are the principal sources of all flowing water in the Gilgit-Baltistan (GB) region. High river water flows in summer, due to snow and glacial melt, which result in flooding and high turbidity water in almost all surface water sources in GB. Main surface drinking water sources in the region include rivers, lakes, springs and traditional drinking water channels and shallow water pits. Slow sand filter, due to the simple operational, maintenance requirements and high biological treatment efficiency is an attractive technology. However, due to its vulnerability to high suspended solids loadings this could not be applied in Gilgit-Baltistan, where more than 70% of the surface water sources has turbid water ranging from 500-3000 NTUs for 6 months (IUCN 2003). In order to cope with such water quality issues, dosing of a coagulant at upstream of the pre-filtration stage is commonly used worldwide. However, due to high costs and skilled personnel requirements, this is not suitable for small rural drinking water supply systems, such as in Gilgit-Baltistan. The successful use of gravel filters as pre-treatment have been investigated worldwide and in Gilgit-Baltistan, however, in few glacial water sources having finer particles conventional up-flow gravel filters failed to reduce turbidity levels up-to-the range suitable for slow sand filters. The objectives of this study were to: assess dirking water quality in rural and urban settlements, critically review of the physical characteristics of glacial water and to investigate efficiency as well as the effectiveness of up-flow gravel filters in series (UGFS) as pre-treatment in the context of Gilgit-Baltistan. In order to achieve the research aims an experimental plant was built at Mominabad, Hunza. The experimental plant was designed on the bases of the literature review, studies, past experience of up-flow roughing filters in GB and typical design guideline values for up-flow roughing filters. The experimental plant consisted of three series connected with the existing sedimentation tank. Each series was further divided into 3 sub-stages of up-flow filters, with different types, gradings, and depth of filter media. Different types of filter media used in the experimental plant consisted upon commercial charcoal, natural granite gravel, burnt bricks, natural limestone gravel and river bed gravel. The depth of filter media in each stage of the series was kept 0.9 m, with a different type of filter medium of 0.3 m depth. The filter compartment having burnt bricks, commercial charcoal, and limestone media were placed in different positions in each series. Water quality analysis was carried out for three months for turbidity levels revealed that 80 to 99 percent reduction in turbidity levels were achieved at the end of all URF-series in the experimental plant (p=0.000) Performance of burnt bricks, charcoal, and lime stone filter medium at first stage of the URF series coupled with granite filter medium in other sub-stages was found best in reducing turbidity levels of 340 NTUs and 939 NTUs in winter and summer seasons where 94 and 88 percent reduction was observed at outlet of the URF Stage-3 respectively A baseline water quality survey was carried out in the initial phase of the research to assess the bacteriological and physio-chemical quality of improved and traditional drinking water delivery systems, spring sources in rural and urban areas and proposed water source for the experimental plant in Mominabad Hunza. Grab samples were collected from selected representative points of the improved and traditional drinking water systems. Membrane filtration technique was used to assess bacteriological contamination in water samples. Conductivity, TDS, temperature, and pH was measured by a conductivity meter (Eutech Instruments, Cyberscan Con 11). Turbidity was measured by nephelometric turbidity tubes provided with Del-aqua water testing kit. Chemical analysis of urban drinking water supply systems and experimental plant site was analyzed on a spectrophotometer. The rural baseline results showed that with the exception of spring water, almost all the rural water supply systems and sources were biologically contaminated. Out of 284 samples, only 68 (24 %) samples had zero counts for E.coli in 100 mL sample and were fit for human consumption as per WHO Guideline Values (GVs) and National Standards Drinking Water Quality set by Government of Pakistan. (WHO, 2011; PEPA, 2008). Urban baseline water quality results showed that out of the 89 water samples, collected from different points of the urban water supply networks, 20 (22 %) samples were found to be fit for human consumption. However, chemical parameter in most of the cases were below the recommended guideline values of WHO (WHO, 2011) for drinking water in developing countries and National Drinking Water Quality Standards (PEPA, 2008), except in water samples taken from Gilgit town water supply networks, where the concentration of chromium, iron and nitrate was found on higher side as per the recommended values. The research revealed that almost all surface drinking water sources and piped water supply systems were not fit for human consumption with different degree of feacal contamination. High concentration of chromium, nitrite, and iron were observed in some urban drinking water sources. Experimental plant monitoring results indicated that up-flow multi-layer roughing filters in series was a suitable option for pretreatment under low flow rates for high turbidity glacier waters for community managed drinking water supply systems. Experimental plant results showed the effective removal of bacteriological (E.coli) and chemical (Nitrate, Nitrite, Phosphate, and Ammonia. Overall research results indicate that up-flow multi-layer roughing filter in series is a suitable option for pretreatment of highly turbidity glacial waters and is effective to remove bacterial and chemical contamination present in source water. URF-series-3, where burnt bricks, limestone and charcoal were placed in first stage of the series, reduced turbidity level higher than the other URF series. Research conclude that URF could be an appropriate and environment friendly option for rural water supply systems for glacial water sources, such as in Gilgit-Baltistan.