Effects of oxidants on drinking water biofilters

Abstract

Biological rapid filtration represents an important process step for the production of high quality drinking water. Bacteria attached to the filter media as biofilm use biodegradable organic matter (BOM) present in the filter influent as a source of carbon and energy. The decrease of the BOM levels through biofiltration is important with respect to the prevention of bacterial regrowth and related problems in the distribution system.

The research presented in this thesis examined the effects of different oxidants on the biological performance, i.e. removal of BOM, of biofilters. The investigated oxidants included hydrogen peroxide (H202), free and combined chlorine (Ch) and ozone (03). Experiments were performed at bench scale using several anthracite/sand filter columns, operated in parallel. The filter influent was dechlorinated tap water to which a cocktail containing BOM, N and P was added. The BOM cocktail contained four easily biodegradable components, i.e. acetate, formate, formaldehyde and glyoxal.

The inhibitory effects of the different oxidants on BOM removal in the filters varied over a relatively wide range. When continuously present in the filter influent, free Ch had the strongest negative effect followed by combined Ch and H202. In general, the presence of an oxidant residual in the biofilter effluent was more important than the bactericidal and oxidative strength of a given oxidant/disinfectant per se. The threshold concentration above which the removal of selected easily biodegradable BOM components was found to be inhibited was< 0.2 mg/L in the case of free Ch, 0.1-0.3 mg/L for combined Ch and between 1 and 5 mg/L in the case of H20i. The inhibition of BOM removal in the filters continuously dosed with an oxidant was likely the result of the suppressed growth of biomass in these filters.

The intermittent dosage of free Ch to the influent of an anthracite/sand filter (-0.5 mg/L once per week for 6 hours) was found to have a substantial negative effect on the removal of BOM. The negative impact of free chlorinated backwash water (~1 mg/L Ch) on BOM removal was considerable, whereas the effect of chloraminated backwash water was negligible. The inhibitory effect of the periodic presence of ozone (-0.15 mg/L} in the influent of an anthracite/sand filter was minor, likely because 0 3 rapidly disappeared in the filter as a result of its reaction with the anthracite.

A CT-approach similar to the traditional CT-approach used for disinfection was successfully used to compare the effects of different oxidants, when oxidants were continuously present in the filter influent. For the intermittent presence of the oxidants, the experimental results could not be entirely described with this CT-approach. The comparison between continuous vs. intermittent presence of oxidants in the biofilters showed that the mechanisms of BOM removal inhibition in the biofilters were likely different for these two conditions.

A new method for the measurement of active, i.e. substrate-degrading, biomass in biological drinking water filters was developed during this research. The method (biomass respiration potential method) is simple and requires only standard laboratory equipment. Therefore, the method is of potential interest to the water industry and its further development should be pursued.

The investigation of mass transfer limitations for the removal of BOM in the control filter (no oxidant) indicated that BOM removal was likely not limited by external (to the biofilm) and internal (in the biofilm) mass transfer. The application of the mass transfer calculations to a broader range of biofiltration conditions suggested that BOM removal in such filters is likely generally limited by biodegradation kinetics, rather than mass transfer. These results are relevant in regards to the overall optimization of biological filtration for BOM removal, particularly in terms of future modeling.