Report No FR/DW0004
DISINFECTION BY-PRODUCTS FROM DRINKING WATER TREATMENT
The broad objective of the contract was to provide the Department of the Environment/Department of the Environment, Transport and the Regions with information on the identity and concentrations of disinfection by-products (DBPs) generated in water treatment and on concentration changes occurring within distribution systems.
To meet this overall objective, and its associated sub-objectives, a series of research and review studies were carried out.
A review of the modelling of the formation of trihalomethanes (THMs) in distribution showed that there are a number of models which can predict the formation of THMs in distribution to some degree. They appear to perform well under the reported conditions. However, they do have potential weaknesses which include not having been tested over a broad range of water and distribution system types, and/or not taking into account the concentration of certain potentially-important variables, such as bromide ion, ammonia or Total Organic Carbon (TOC). Following on from the review of modelling, a study of the factors influencing changes in THM concentrations in four distribution systems was carried out. A number of conclusions could be drawn from the results.
Unquenched ex-works waters showed the potential to form more THMs within a 24 hour period. If the water was spiked with further chlorine then the formation of THMs was greater. The magnitude of the chlorine residual in the water as it leaves the works, the fraction of chlorine demand satisfied at the works, and any booster chlorination will play an important role in determining the increase in THM concentrations in distribution.
Where works operated to target minimum chlorine residuals leaving the works then a rise in THM precursors in the works did not necessarily lead to higher concentrations of THMs in the distribution. As the precursor level, and hence the chlorine demand, rises then the chlorine residual leaving the works falls. As long as the residual remains above the target minimum then the chlorine dose remains constant. Thus, whilst the THM concentration leaving the works will have risen, the formation in distribution will have been limited by the fall in chlorine residual. Therefore there may not be an appreciable rise in the concentration at the tap. However, if the THM precursor level rises such that the chlorine dose has to be increased to maintain the residual leaving the works then it is likely that the THM concentration at the tap will rise appreciably.
*For works C the difference between the maximum and minimum residuals is 0.45 mg/l. The experiments where 0.5 mg/l chlorine was added to the samples on their return to the laboratory showed increases of 10 - 25 µg/l in THM concentrations over 24 hours. Given that the range of concentrations reported for tap THM concentrations in 1996 was 40 - 118µg/l, the variation allowed by the residual limits appears to cover a small, but appreciable, portion of the annual variation in precursor levels.
*There did not appear to be any evidence of THMs arising due to reactions of chlorine with pipe materials, sediments or biofilms in the distribution system.
An investigation of the influence of UV disinfection on the formation of disinfection by-products was carried out, with emphasis on any operational implications. It was demonstrated that, if UV units are operated at a dose of around 100 mJ cm-2, then at high nitrate levels (i.e. close to the 50 mg/l drinking water limit) the concentration of nitrite may approach or exceed the regulatory limit of 100 µg/l. In this case some monitoring for nitrite should be undertaken, and if necessary, adjustments made to the UV dose applied to allow an operational safety margin. If the UV units are operated in the range 25 - 30 mJ cm-2 then, even for waters with high nitrate levels, the concentration of nitrite formed is unlikely to approach the regulatory limit.
If UV units are operated at a dose of around 100 mJ cm-2, then if tetrachloroethene (PCE) is present in high concentrations, i.e. at, or above, 1 mg/l, then the concentrations of dichloroacetic acid (DCA) produced may approach the WH0 Drinking Water Guideline value of 50 µg/l. Therefore, if PCE is found to be present at high concentrations in the source water then monitoring should be applied to ensure that significant concentrations of DCA are not being formed.
The study found no firm evidence, within the scope of the Gas Chromatography - Nitrogen Phosphorus detection (GC-NPD) and Gas Chromatography-Mass Spectrometry (GC-MS) methods applied, that the UV irradiation of water resulted in the formation of any nitrogen containing organic compounds.
An analysis of the potential for the development of surrogate measurements for DBPs was carried out using data collected for a previous DoE study. The implication of the analysis is that any relationship that can be drawn between an individual by-product and a surrogate varies from works to works. At each works the relationship will depend in some degree on the nature and extent of the treatment applied. Thus it does not appear likely that surrogate relationships can be established covering a wide range of works/water types. However, it may be possible to construct works-specific relationships. For such relationships to be a useful monitoring tool then it would have to be determined whether a relationship which holds at the works also holds throughout the distribution system.
A review was conducted of the state of knowledge with regard to the balance between the formation of bromate and the removal of pesticides by ozonation. There have been a number of studies which have specifically examined this balance. Whilst these studies are not extensive enough to be able to apply the results to different treatment situations, and to a wide range of pesticides, a number of useful pointers can be drawn. *There appears to be a law of diminishing returns for pesticide removal, i.e. the degree of removal tails off as the ozone dose increases.
*Some of the measures which can be taken to limit bromate formation include lowering of pH and the addition of hydrogen peroxide. For most pesticides these measures will tend to increase the degree of removal.
*The addition of hydrogen peroxide to limit bromate formation is most effective when the ozone dose is kept constant. However, in some cases a reduction in bromate has been observed where the final ozone residual has been kept constant. These latter conditions are the better option for pesticide removal.
In principle, it should be feasible to extend the models of bromate formation such as those produced by Von Gunten (e.g. Von Gunten and Hoigne 1992) to take account of the requirements to remove specific contaminants. However, this is likely to take a significant degree of effort to generate sufficient data on the kinetics of reaction of individual pesticides. For some frequently studied pesticides, such as atrazine, sufficient data may already exist.<Copies of this report may be available as an Acrobat pdf download under the 'Post 2000 Reports' heading of the Research Page on the DWI website.