Predicting accidental release of engineered nanomaterials to the environment

Challenges in distinguishing between natural and engineered nanomaterials (ENMs) and the lack of historical records on ENM accidents have hampered attempts to estimate the accidental release and associated environmental impacts of ENMs. Building on knowledge from the nuclear power industry, we provide an assessment of the likelihood of accidental release rates of ENMs within the next 10 and 30 years. We evaluate risk predictive methodology and compare the results with empirical evidence, which enables us to propose modelling approaches to estimate accidental release risk probabilities. Results from two independent modelling approaches based on either assigning 0.5% of reported accidents to ENM-releasing accidents (M1) or based on an evaluation of expert opinions (M2) correlate well and predict severe accidental release of 7% (M1) in the next 10 years and of 10% and 20% for M2 and M1, respectively, in the next 30 years. We discuss the relevance of these results in a regulatory context. Quantification of the accidental release rates of engineered nanomaterials (ENMs) would inform risk management strategies and their implementation, but a comprehensive assessment of ENM accidental release is not currently available. Here the authors present a predictive study of the release of ENMs following accidents during their fabrication, transport and end-of-life processes in the next 10 and 30 years, based on conceptual models inspired by the nuclear power sector.

Automated summary

The challenges in distinguishing between natural and engineered nanomaterials and the lack of historical records on accidents involving engineered nanomaterials have made it difficult to estimate the accidental release and environmental impacts of such materials. Drawing on knowledge from the nuclear power industry, this study assesses the likelihood of accidental release rates of engineered nanomaterials in the next 10 and 30 years. By comparing risk predictive methodology with empirical evidence, the authors propose modelling approaches to estimate the probabilities of accidental release. The results from two independent modelling approaches correlate well and predict severe accidental release rates of 7% (M1) in the next 10 years and 10% and 20% (M2 and M1, respectively) in the next 30 years.