Evaluation of natural fabrics for protection of slow sand filtration.

The advantages of Slow Sand Filtration (SSF) are well established in literature. The major disadvantages of SSF are its inability to treat highly coloured and turbid water for long durations, the frequency and labour intensiveness of the regular cleaning and maintenance. Previous research carried out in Europe and Africa has shown that slow sand filters protected by layers of synthetic fabrics have increased run times compared to un-protected slow sand filters by up to a factor of 8. The high cost of synthetic fabrics has rendered them out of reach of most developing countries. A suitable alternative is the use of natural fabrics, which are plentiful in most developing countries. The main objective of this research was to identify a suitable natural fabric available in Tanzania for protection of SSF beds. As well as extending filter run times, the fabrics protected sand filtration was also expected to eliminate the need for frequent sand scraping and cleaning. A literature and laboratory study of the range and types of tropical fibres available in Tanzania was carried out. The study resulted in the identification of five vegetable fibres for initial consideration. The suitability of these five fibres in the filtration process was further assessed. This later on resulted in the choice of three fibres for application in the pilot plant SSF protection trials. In the pilot plant SSF trials, monitoring of daily head loss and certain key water quality parameters of both the raw water and the filtrates from each of the units was carried out. Furthermore, the degradation rate of the fabrics was also monitored. Through extended pilot plant trials, this research demonstrated that with the use of jute fabrics, a filter run time extension of up to 2.0 can be achieved whilst with the use of sisal and luffa cylindrica, filter run times extension by up to factors of 1.6 and 1.7, respectively can be achieved. With ali the fabric protected filters, solids penetration through the fabrics on to the sand bed was nearly eliminated hence resulting in substantial reduction of operational costs associated with cleaning. The presence of cellulose has been shown to have a detrimental effect on the potability of the filtered water. Luffa cylindrica protected filter produced filtrates with the highest increased organic carbon as compared to sisal and jute. Also the apparent colour of natural fabric protected filtrates, was found to be outside the allowable guideline limits of drinking water. It is recommended that pre-treatment of natural fabrics/fibres with carbon disulphide and an aqueous sodium hydroxide in order to minimise the negative influence on apparent colour and organic carbon is necessary before further application of natural fabrics to protect SSF. It is also recommended that further pilot plant trials are carried out with deeper sand beds in SSF than in this study (approximately 60cm).