(48a) Data-Driven Estimation of Chemical Releases from End-of-Use Management Scenarios

Ruiz-Mercado, G., U.S. Environmental Protection Agency
Hernandez-Betancur, J. D., U.S. Environmental Protection Agency
Abraham, J. P., U.S. Environmental Protection Agency
Ingwersen, W., U.S. Environmental Protection Agency
Smith, R., US Environmental Protection Agency
Meyer, D. E., U.S. Environmental Protection Agency
Gonzalez, M. A., U.S. Environmental Protection Agency
The Toxic Substances Control Act (TSCA) requires the U.S. Environmental Protection Agency (USEPA) to establish a risk process evaluation to determine whether a chemical poses a considerable risk to the human health or the environment. Consequently, as a part of the procedure for chemical risk evaluation, the USEPA has developed models to understand the relationship between a chemical and a potential receptor, either human or the environment. As part of the effort to meet the TSCA requirements, USEPA published in 2016 an initial list of the first 10 chemicals substances to address risk evaluation. In the conceptual models developed for the 10 substances, the pathways considered can be expanded to releases of chemicals when these are transferred to an off-site facility for their recycling, recovery, and/or reuse, as well as waste treatments regulated by any USEPA statute different to TSCA. Therefore, for dealing with the current needs and considering that the current TSCA Chemical Substance Inventory lists around 85,000 chemicals, this work proposes an end-of-use data-reconciliation and learning-from-data framework. This approach is based on existing USEPA databases for predicting releases that occur following the chemical end-of-use management at off-site facilities. The data-reconciliation framework will be useful to support TSCA risk evaluation for rapid estimation of industrial chemical releases during end-of-use management scenarios, which are not currently analyzed. Therefore, a chemical substance in waste streams can be tracked in its composed end-of-use stage and a learning-from-data generalization would be applied for existing and new chemical substances in the U.S. market to estimate environmental releases and worker exposure.