Environmental risk assessment of biocidal products: identification of relevant components and reliability of a component-based mixture assessment

BACKGROUND: Biocidal products are mixtures of one or more active substances (a.s.) and a broad range of formulation additives. There is regulatory guidance currently under development that will specify how the combined effects of the a.s. and any relevant formulation additives shall be considered in the environmental risk assessment of biocidal products. The default option is a component-based approach (CBA) by which the toxicity of the product is predicted from the toxicity of ‘relevant’ components using concentration addition. Hence, unequivocal and practicable criteria are required for identifying the ‘relevant’ components to ensure protectiveness of the CBA, while avoiding unnecessary workload resulting from including by default components that do not significantly contribute to the product toxicity. The present study evaluated a set of different criteria for identifying ‘relevant’ components using confidential information on the composition of 21 wood preservative products. Theoretical approaches were complemented by experimentally testing the aquatic toxicity of seven selected products. RESULTS: For three of the seven tested products, the toxicity was underestimated for the most sensitive endpoint (green algae) by more than factor 2 if only the a.s. were considered in the CBA. This illustrated the necessity of including at least some additives along with the a.s. Considering additives that were deemed ‘relevant’ by the tentatively established criteria reduced the underestimation of toxicity for two of the three products. A lack of data for one specific additive was identified as the most likely reason for the remaining toxicity underestimation of the third product. In three other products, toxicity was overestimated by more than factor 2, while prediction and observation fitted well for the seventh product. Considering all additives in the prediction increased only the degree of overestimation. CONCLUSIONS: Supported by theoretical calculations and experimental verifications, the present study developed criteria for the identification of CBA-relevant components in a biocidal product. These criteria are based on existing criteria stated in the regulation for classification, labelling and packaging of substances. The CBA was found sufficiently protective and reliable for the tested products when applying the here recommended criteria. The lack of available aquatic toxicity data for some of the identified relevant components was the main reason for underestimation of product toxicity.