You are here:
Incorporating Metadata from US EPA’s Inventories to Generate More Reliable Life Cycle Emissions Data
Citation:
Cashman, S., D. Meyer, W. Ingwersen, AND J. Abraham. Incorporating Metadata from US EPA’s Inventories to Generate More Reliable Life Cycle Emissions Data. Presented at American Center for Life Cycle Assessment Annual Meeting, Ft. Collins, Colorodo, September 25 - 27, 2018.
Impact/Purpose:
The purpose of this presentation is to inform the audience at the American Center for Life Cycle Assessment Annual Meeting about the research from CSS exploring the use of data mining with EPA chemical information data sources (Chemical Data Reporting database, National Emissions Inventory, Toxics Release Inventory, Discharge Monitoring Report, RCRAinfo Biennial Report, Electronic Greenhouse Gas Reporting Tool) to model life cycle releases of chemicals.
Description:
The United States Environmental Protection Agency (EPA) previously presented a method to mine its publicly available emissions inventories for development of quick and accurate chemical manufacturing life cycle inventories (LCI). The original method integrates data at a facility-level from the EPA’s Facility Registry Services (FRS), Chemical Data Reports (CDR), Toxic Release Inventory (TRI), National Emissions Inventory (NEI), electronic Greenhouse Gas Reporting Tool (eGGRT), Resource Conservation and Recovery Act Information database (RCRAInfo), and Discharge Monitoring Reports (DMR) to create U.S. average chemical LCIs. This presentation provides an update to the method based on application of metadata within the existing EPA emissions inventories for building process-specific chemical LCIs. Source Classification Codes (SCC) in combination with NEI or eGGRT emission unit types are used to allocate air emissions to specific chemical processes within a facility and to filter out emissions not associated with the process of interest. Reporting requirements for each inventory data source are investigated to develop recommendations for assigning flow data quality scores, determining maximum, minimum and true zero flow values, creating methods to avoid data quality issues due to reporting exemptions, handling speciation issues to avoid flow overlap within databases, and applying prioritization to flow values when multiple databases report the same flow. Results of incorporating the metadata to produce more reliable chemical LCIs are presented within the context of three chemical case studies.