THE FORMATION OF BROMATE DURING ELECTROLYTIC GENERATION OF CHLORINE
(DWE 9005) Final Report to the Department of the Environment DoE
Report No DWI0136
The use of ozone in drinking water treatment has led to concern about the possible formation of harmful by-products. Of particular concern is bromate, which can be formed from the ozonation of raw waters containing bromide. Bromate is a suspected carcinogen and is included in the revised WHO Guidelines for drinking water. The WHO have set a provisional guideline value of 25 µg/l, taking account of perceived analytical problems associated with the determination of bromate in drinking water. However, in the absence of such problems the low dose extrapolation models commonly used when deriving guideline values give a value of 3 µg/l.
Another potential source of bromate is the use of sodium hypochlorite (either generated on-site or commercially available). There is a growing interest in the use of on-site electrolytically-generated chlorine for water treatment but brine used in on-site generators will contain bromide, and it is possible that it will be oxidised, leading to bromate in the treated water. There are few data on the levels of bromate either in hypochlorite solutions generated on-site or in the associated final waters.
The objectives of this study were threefold:
The work showed that bromate is formed during on-site generation of hypochlorite. The concentrations found in hypochlorite samples ranged from 2.8 to 21.8 mg/l. However, bromate was only found in two out of twelve final waters analysed. These corresponded to the two hypochlorite solutions containing the highest concentrations of bromate. It appears unlikely that the provisional WHO Guideline for bromate of 25µg/l would be exceeded through the use of on-site generated hypochlorite alone. If a lower value (i.e. less than 10 µg/l) were adopted then it is feasible that exceedances could occur at some sites. A major factor determining the level of bromate present in hypochlorite will be the bromide concentration in the brine. This will depend on the bromide content of the salt and the strength of the brine. The brine strength is likely to vary significantly in different systems. In some cases it is increased to maintain a pre-determined hypochlorite concentration in the outlet stream (normally 0.7-0.9% chlorine).
Other operating conditions could affect the amount of bromide converted to bromate. These include cell voltage, current density, operating temperature and residence time in the generator cell. To a certain extent all these parameters are inter-related, and whilst it is difficult to predict the effect of changes in any one of them on bromate production, it is possible that any increase in one or other parameter is likely to favour bromate production.
Storage of on-site generated hypochlorite does not seem to result in significant increases in bromate concentrations.
The type of on-site generating system does not appear to be important in determining the level of bromate formed.
The significance of the findings clearly depends on the numerical value of any standard set for drinking water. However, some further investigations would seem prudent.
Wider monitoring of bromate concentrations in hypochlorite solutions generated by on-site systems, and of bromide concentrations in brines should be carried out.
Since it is not possible to explain the reason for the two detected levels of bromate in treated water, due to insufficient data on the brine feedstocks and detailed operating conditions, a detailed study of the factors that control bromate formation should be carried out. In practice, this would require a test rig to allow bromate monitoring under the full range of operating conditions.Copies of this report may be available as an Acrobat pdf download under the 'Pre 2000 Reports' heading on the DWI website.