JEM Spotlight: Monitoring the treatment efficiency of a full scale ozonation on a sewage treatment plant with a mode-of-action based test battery

Tertiary treatment of wastewater with ozone is a promising technique for removing residual micropollutants that remain after secondary biological treatment. We monitored the performance of a full-scale ozonation reactor on a sewage treatment plant in Switzerland with a screening battery of bioassays. Six toxicity endpoints were selected that covered non-specific toxicity, as well as selected receptor-mediated modes of action and reactive toxicity. Non-specific toxicity was assessed with two bioassays, the bioluminescence inhibition of the marine luminescent bacterium Vibrio Fischeri and the growth inhibition of the green algae Pseudokirchneriella subcapitata. Treatment efficiency was around 90% for the secondary treatment, but only 65% and 76% for the ozonation step in the two nonspecific endpoints, respectively. This finding is consistent with this type of oxidation reaction because ozone only modifies the organic molecules but does not mineralize them fully leaving residual toxicity of the transformation products. In contrast, the specific receptor-mediated endpoints of inhibition of photosystem II in algae and estrogenicity were largely reduced by ozonation. While compounds inhibiting photosynthesis proved to be rather recalcitrant toward biological treatment with only 47% removal, an additional 86% removal by ozonation yielded an overall treatment efficiency in the entire treatment chain of 89%. The effect on estrogenicity, quantified with the yeast estrogen screen, was even more significant: A treatment efficiency of 95% in the secondary treatment, 86% during ozonation plus a small effect by biological sand filtration yielded an overall treatment efficiency of 99.5%. Insecticides that inhibit acetylcholinesterase were fairly resistant to degradation, but an overall treatment efficiency of 91% was achieved in two steps: 72% in biological treatment and 60% during ozonation. Finally, no significant genotoxicity was observed with the umuC test after ozonation, while the influent showed a genotoxic response when it was enriched by a factor of 15 to 60. Treatment efficiency increased with the ozone dose and remained virtually unchanged over ozone doses above 500 g ozone per kg dissolved organic carbon. The reduction of toxicity can be rationalized by the chemical oxidation processes likely to occur for each group of chemicals that are typical for a given mode of toxic action. For comparison, tertiary treatment with powdered activated carbon was also evaluated, which poses a viable alternative to ozonation with respect to removal of micropollutants.