(109d) A Data-Engineering Approach for Estimating Chemical Releases from Industrial Pollution Abatement Technologies | AIChE

(109d) A Data-Engineering Approach for Estimating Chemical Releases from Industrial Pollution Abatement Technologies


Hernandez-Betancur, J. D. - Presenter, U.S. Environmental Protection Agency
Ruiz-Mercado, G., U.S. Environmental Protection Agency
Li, S., Pacific Northwest National Laboratory
Martin, M., University of Salamanca
Lima, F. V., West Virginia University
Chemical substances available for industrial and commercial activities or uses may present an unreasonable risk to human health and the environment. In the U.S., the chemical risks should be evaluated according to the Toxic Substances Control Act, while in Europe, this is under the Registration, Evaluation, Authorization, and Restriction of Chemicals (REACH). Therefore, knowing what occurs with a chemical at each stage of its life cycle is essential for its evaluation, selection, use, and regulatory decision-making. However, gathering information to evaluate chemical risk is a time-consuming task, and especially at the end-of-use (EoU) stage because of the high uncertainty about chemical fate and exposure pathway. The contribution of this work is a data engineering approach for gathering information about pollution control units (PCUs) from publicly-available regulatory databases and filling data gaps based on technical information. The data collected include the type of waste stream having the chemical of interest (e.g., liquid waste), composition, and PCUs (e.g., absorber) and their efficiencies, which the framework transforms into a machine-readable structure for future automation. Thus, the developed approach can rapidly streamline chemical release estimations from PCUs and allocate potential effluents. Besides, it could be applied to identify and recommend the application of industrial PCUs for managing chemicals at generator facilities. The developed approach will be applied for case studies with the resulting identification of PCUs and release estimates. This approach may support the risk evaluation process by developing learning-from-data models to select PCUs and estimate releases.

The views expressed in this abstract are those of the authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency.