Health Stream Literature Summary - Issue 46 - June 2007

Breakpoint chlorination and free-chlorine contact time: Implications for drinking water N -nitrosodimethylamine concentrations.

Charrois JWA , and Hrudey SE. (2007) Water Research 41 :674-682.

Regulation of disinfection byproducts currently focuses on the most abundant halogenated classes trihalomethanes and haloacetic acids, however it is now well established that these DBPs cannot be responsible for the increased risks of bladder cancer observed in a number of epidemiological studies. Recent research has highlighted the presence of N -nitrosodimethylamine (NDMA) and other N-nitrosamine compounds in water supplies and raised interest in the possible role of these chemicals in the causation of bladder cancer. Paradoxically, the practice of switching from chlorination to chloramination to reduce levels of THMs and HAAs, may increase the formation of NDMA and related compounds. This paper describes NDMA levels from two full-scale water treatment plants in Canada , the influence of disinfection conditions on NDMA formation, and possible strategies for water utilities to minimise NDMA formation.

Raw source water, partially treated water (after full-scale coagulation, flocculation, sedimentation and filtration but before disinfection) and finished water (from clear well or distribution system) were collected from two water treatment plants in two different cities. Raw and partially treated water were used to perform breakpoint chlorination experiments.

In city A, NDMA levels were measured at the plant, at a point in the middle of the distribution system and at two locations near the end of the system. The highest concentrations of NDMA were seen at the distal points, with intermediate levels in the middle of the distribution system and lower levels at the water treatment plant. This is consistent with other reports that NDMA concentrations increase with residence time. The maximum concentration observed in City A was 180 ng/L. In city B, sampling was carried out at only one point in the middle of the distribution system, and no samples exceeded 5 ng/L. While variations in NDMA levels in city A may have been influenced to some extent by changing doses of poly-DADMAC to assist coagulation, the presence of the polymer was not the sole influence on NDMA formation as the same polymer was also used in city B, and often at higher doses than city A. Possible reasons for differences in NDMA formation also include differences in free chlorine contact time (less than 30 seconds in city A, compared to 2-4 hours in city B) and oxidation of NDMA precursors by potassium permanganate (used for control of iron and manganese) in city B.

Breakpoint chlorination experiments with raw water showed significant difference in NDMA formation for the two waters with city A levels peaking at 15 ng/L while city B never exceeded 3 ng/L. No reason for the difference was evident as both waters originated from the same watershed only 40 km apart and their dissolved organic carbon levels were similar. Investigations did not reveal any known differences in contaminant inputs (eg industrial waste containing NDMA precursors) which could account for the difference.

Bench-scale experiments demonstrated that a free-chlorine contact time of 2 hours prior to chloramination could significantly reduce NDMA formation using city A partially treated water compared to no free-chlorine contact time. This may be an option to reduce NDMA formation in chloraminated water supplies, although the formation of regulated DBPs would also have to be assessed.