Ozone decomposes rapidly following application, and for this reason no GV has been proposed for ozone.

By-products of ozonation that have been identified include formaldehyde and other aldehydes, bromate, carboxylic acids, ketones, and brominated acids. When ozonation is followed by chlorination, the concentration of brominated THM may increase (Bull and Kopfler 1991). GVs have been recommended for bromate and formaldehyde.

Based on inhalation studies, IARC has classified formaldehyde in Group 2A, probably carcinogenic to humans. However, in a manner similar as for asbestos, the review group meeting concluded that formaldehyde was not carcinogenic by the oral route, and a TDI approach was adopted. In a two-year study in rats receiving doses of 0, 1, 15, or 80 mg/kg bw/day of formaldehyde in drinking water, the NOAEL was 15 mg/kg bw/day, based on decrease in body weight, pathological changes in the stomach, and effects on the kidney at the higher dose. Using an uncertainty factor of 100 and allocating 20% of the TDI to drinking water results in a GV of 0.9 mg/L. The taste and odor threshold of formaldehyde is about 50 mg/L, well above the health-based GV.

Bromates are used as oxidizing agents in the treatment of flour. In 1995, JECFA evaluated bromate, and because of its carcinogenicity in long-term feeding studies in rats, the committee did not establish an acceptable daily intake (ADI) but recommended that bromate should not be present in foods as consumed. IARC has classified bromate in Group 2B, possibly carcinogenic to humans.

The review group meeting concluded that bromate should be treated as a potential carcinogen. Applying the linearized multistage extrapolation model to the dose-response curve for kidney tumors in rats results in a GV of 0.003 mg/L, for an estimated excess cancer risk of 10-5. This GV is below the limit of quantification, however. A provisional GV of 0.025 mg/L was therefore recommended, corresponding to an excess lifetime cancer risk of 7 x 10-5.