You are here:
Estimating Releases from Chemical Processes for Occupational and Fenceline Community Exposure Assessments
Citation:
Smith, Raymond L., David E. Meyer, Gerardo J. Ruiz-Mercado, William M. Barrett, Michael A. Gonzalez, S. Takkellapati, AND John Chea. Estimating Releases from Chemical Processes for Occupational and Fenceline Community Exposure Assessments. International Society of Exposure Science, 2023 Annual Meeting, Chicago, IL, August 27 - 31, 2023.
Impact/Purpose:
This abstract describes a presentation for the International Society of Exposure Science, for the Annual Meeting to be held in August of 2023. The presentation is on research that supports chemical risk assessments in EPA’s Office of Pollution Prevention and Toxics (OPPT). Chemical processes generate releases of chemicals which can affect occupational workers and fenceline communities. Methods for estimating releases are described in this effort. When fully developed the methods can provide rapid methods for generating release information useful in exposure and risk assessments. The methods can also fill data gaps for assessments. In addition to OPPT, occupational workers, the general public, and fenceline communities could benefit from the information provided.
Description:
New chemicals are introduced into commerce regularly, and existing chemicals are sometimes repurposed for new uses. In both cases the chemicals may present health risks to occupational workers that can be evaluated via exposure and risk assessments. In addition, chemical releases from facilities can represent increased exposures and risks to communities near the facility fenceline. Quantification of releases is an important step in conducting these analyses, along with exposure and hazard assessments. This contribution describes research at the U.S. Environmental Protection Agency’s Office of Research and Development to support risk evaluations, using mechanistic, database, and machine learning models to estimate releases. A type of mechanistic model, known as a Generic Scenario, relates chemical usage, activities, and properties to describe releases. Other mechanistic models, known as bottom-up models, have been built from process simulation and models of unit operations that combine to form a chemical process. Database, or top-down, models are derived directly from inventories available through EPA and others who collect and curate the information. Often collected at a facility level, context-based filter rules are applied to the release data to more accurately represent chemical processes. Finally, machine learning models will be described for work in progress using the predictive power of training and test data sets to estimate releases. These models will be presented for a case study with resulting estimates for releases. The views expressed in this presentation are those of the authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency.