Date of Award

Winter 1996

Project Type


Program or Major


Degree Name

Doctor of Philosophy

First Advisor

Michael R Collins


Slow sand filters are an established treatment method for water with low turbidity. They usually are effective for the removal of turbidity, microorganisms (including cysts of Giardia and Cryptosporidium), and particles, but they require significant periods of time for cleaning. In the 1950's, West Hartford, CT developed a harrowing process to reduce the time and labor required for cleaning at that plant. A 1988 study observed those filters had higher removal rates for non-particulate dissolved organic carbon and UV absorbing materials, surrogates for trihalomethane formation, than did filters at two other plants cleaned by the conventional scraping method.

This study was planned to compare the effectiveness of the two cleaning methods and their effects on performance of full-scale filters on a side-by-side basis using a new plant at Gorham, NH. Headlosses through those filters developed very slowly, and the study was transferred to a similar plant at Newport, NH where operations were studied through the initial ripening phase and one cycle of cleaning by each cleaning method. This information was supplemented with data collected from separate plants which had been using the two methods since the 1950's and from pilot scale filters. The effects of filter application rates, source water, and filter media characteristics were studied with laboratory scale columns. Removal performance of the full scale filters were compared for temperature, turbidity, particles, nonpurgeable dissolved organic carbon, and UV absorbing materials. The upper 30 cm of filter media at each of the plants was sampled over the study. Concentrations of volatile solids, protein, carbohydrates, bacteria, iron, manganese, calcium, and aluminum were compared and related to performance. The differences between filter cleaning methods were compared in relation to labor and time required, wastes generated, and resultant media characteristics.

Overall performance of the slow sand filters was influenced by water temperature, sand media age, filter biomass content, source water quality, filtration rate, and empty bed contact time. Some removal trends suggested filter harrowing resulted in higher removals of organic carbon and UV absorbing materials but the conclusion must be qualified because the trend was not consistent and was dependent on other confounding factors, e.g. water source, temperature, and sand age.