|Over the past century chlorine has been a reliable disinfectant to reduce transmission of waterborne diseases in drinking water. Concerns about chlorination have increased since it was discovered in the 1970s that use of chlorine in drinking water produces trihalomethanes (THMs), when chlorine reacts with natural organic matter (NOM), which has been observed in increased levels in surface water during the past decades. THM and other disinfection by-products (DBPs) such as some of the haloacetic acids (HAAs) and some nitrosamines are considered probable human carcinogens by USEPA.
Since DBPs are still formed even when using alternative disinfectants such as chloramines, treatment processes by which disinfection by-product precursors are removed continue to be studied. Many researchers have demonstrated that the use of pre-ozonation/biological processes in the production of drinking water has the potential to decrease levels of disinfection by-products in finished water more than conventional treatment alone.
Two of the parameters which affect the efficiency of DBP precursor removal in biofilters are filter media and filter flow rate. In this research, the biofiltration process was examined using pilot-scale filters receiving ozonated water to determine the relative effectiveness of these parameters for influencing the removal of natural organic matter. The research presented in this thesis initially focuses on determining the effects of flow rate and filter media including GAC (granular activated carbon) and anthracite on decreasing the levels of THM, HAA and nitrosamine precursors in biologically active filters. In the second part, the performances of full-scale and pilot-scale filters at the Mannheim Water Treatment Plant were compared.
THM and HAA precursor removal was found to decrease when loading rates were increased, likely due to associated shorter contact times in the filters. Also, higher THM and HAA precursor removal was always observed in the GAC filters than in the anthracite filters. However, removal of nitrosamines was not affected by flow rate or the type of filter media.
In general, the pilot-scale filter performance was representative of full-scale filter performance, especially in regards to THM precursor and chlorine demand removal. Statistical evaluation and interpretation of the data for HAA and NDMA precursor removal was more difficult, likely due to the low concentrations of these DBPs which was near their method detection limits (MDLs) and also because of some operational problems with pilot filter #1. Despite these limitations, the results of this study add to the literature concerning the use of different types of media to support biofiltration and reduce DBP precursor concentrations during drinking water treatment.