Report No DWI0134

Aug 1989

SUMMARY

I OBJECTIVES

1. To study the effects of disinfection of water by chlorine and ozone on a range of selected substance likely to occur as contaminants in water.
2. To identify products of these reactions.

With a few exceptions, the chemistry of disinfection with respect to organic chemicals occurring in water sources is not well understood. Consequently, it is difficult to predict the reactions of disinfectants, such as chlorine or ozone, with such chemicals and the nature of any byproducts generated. Of particular concern are reactions with chemicals arising from specific pollution incidents and with any chemicals that commonly occur in surface water, eg hydrocarbons, surfactants and fatty acids. These chemicals may arise from waste water discharges, but although the chemicals themselves may not be of concern they may give rise to byproducts during disinfection which are undesirable owing to the production of tastes/odours or risks to health.

To date, few of the studies in this area have been conducted under conditions likely to arise during drinking water disinfection (eg concentration, pH, contact time). In contrast, WRc research has set out to investigate the effects of chlorine and ozone in a systematic manner under conditions as close as possible to those of treatment. This report provides some of the findings of the research at WRc.

The data in the scientific literature on the reactivity of chlorine and ozone with organic substances and the likely reaction products must be treated with caution when trying to assess the effects of water treatment disinfection. Some of our findings in the laboratory produced under conditions related as far as possible to water treatment practice, disagreed with data in the literature.

Some types of potential or known water pollutants were shown to be unreactive and would not be expected to be of concern with respect to production of byproducts. However, a variety of potential or known water pollutants did react. The reactivities of the substances studied are summarised in Table 1.

In some cases where the organic chemical reacted, the products of the reaction were identified.

The presence of bromide during chlorination was an important factor. In certain cases bromide led either to more extensive consumption of organic chemical during chlorination or even reaction when no reaction could be found in the presence of chlorine alone

The data produced in this work should enable better prediction of the likely effects of chlorine and ozone, when applied during water treatment, on organic chemicals occurring in raw water. Such chemicals may occur as common constituents or specific pollutants from pollution incidents.

The work on using chlorine demand as a means of quickly assessing which chemicals react markedly with chlorine, and thus avoiding unnecessary analytical development work, gave promising results and indicated several chemicals likely to be found in water sources and some water treatment chemicals that need to be studied in more detail.

Ozone is currently proposed as an alternative to chlorine, the latter being disliked in some quarters since it produces byproducts during disinfection. However, ozone also produces byproducts (of different type) and in some instances more readily.

Studies on the reaction of chlorine and ozone with further organic chemicals that are either common constituents of raw water or frequently involved in pollution incidents are needed.

Work is needed to determine whether or not products of chlorine and ozone disinfection mentioned in this report can be found in treated waters. Studies are needed to explain why double bonds in some organic chemicals (eg permethrin) do not react as expected with chlorine. The influence of important inorganic species, such as bromide and bicarbonate, on the effects of disinfectants on organic substances needs further investigation. However, more study of the effects of pH (between a range of 6-9) is necessary.

Further work on rapid means of assessing general reactivity of chemicals (particularly those difficult to measure in water, such as treatment chemicals) towards disinfectants has been justified. The chemicals shown by the chlorine demand work to have marked reactivity to chlorine, need to be studied further to confirm the reactivity and assess the significance of the byproducts generated.

Many of the recommendations are being followed in on-going research, funded by the Department of the Environment, and the findings will be disseminated at a later stage.

Comparison of the effect of chlorine and ozone on saturated and unsaturated compounds has shown that both oxidants will react with unsaturated compounds such as alkenes, unsaturated fatty acids, unsaturated esters and unsaturated surfactants. After chlorination the products were found to be chlorohydrin derivatives of the starting material. Ozonation was shown to result in the formation of oxygenated cleavage products, such as aldehydes and carboxylic acids. Neither chlorine nor ozone could be shown to react with saturated substrates such as alkanes and saturated fatty acids.

Aromatic hydrocarbons containing more than one ring, such as naphthalene, phenanthrene, fluoranthene, fluorene and pyrene reacted readily with ozone, whereas toluene and xylenes (one ring aromatic hydrocarbons) were unreactive. Few of the products of these reactions have been characterised with the exception of phthalic acid (from naphthalene) and fluoren-9-one (from fluorene). In contrast only pyrene was found to react with chlorine; it yielded a mono- and a di-chloropyrene as products.

The chlorination of nonylphenyl ethoxylates (a nonionic surfactant) in the presence of bromide was shown to result in the formation of brominated products. Ozone completely consumed nonylphenyl heptakis ethoxylate; however, no products were detected.

From this work it is evident that some types of potential or known water pollutants do not react with either chlorine or ozone and are therefore of little or no concern with respect to the formation of byproducts. Some potential or known water pollutants were reactive. The reactivities of the substances studied and the byproducts formed (where known) are summarised in Table 1. The presence of bromide during chlorination was shown to be an important factor. In some cases bromide led either to more extensive consumption of the substrate or even to reaction when no reaction was found in its absence (ie with chlorine alone).

The data presented in this report should enable better prediction of the likely effects of chlorine and ozone on organic chemicals in water when applied during water treatment. However, more substances need to be studied and the apparently anomalous behaviour of some compounds needs to be understood. Many compounds which need to be evaluated are, because of their physico-chemical properties, extremely difficult and tedious to analyse and consequently require considerable time to study. Consequently, some method of rapidly assessing the reactivity of these substances needs to be developed; with such a method only chemicals shown to have reactivity would need to be studied in greater detail. Investigations have shown that in the case of chlorine such an approach, based on chlorine demand, can give useful indications on the reactivity of a wide range of chemicals.

The studies were funded by the Department of the Environment. On-going work covers the reactivity of ozone, chlorine and chloramine with additional organic chemicals including pesticides of current concern.

Copies of this report may be available as an Acrobat pdf download under the 'Find Completed Research' heading on the DWI website.