You are here:
Towards Automated Processing of Fenceline Sensor Data
Citation:
MacDonald, M., E. Thoma, I. George, AND R. Duvall. Towards Automated Processing of Fenceline Sensor Data. American Geophysical Union Fall Meeting, Chicago, IL, December 12 - 16, 2022.
Impact/Purpose:
This is an abstract submitted to the American Geophysical Union (AGU) Fall Meeting in Chicago, IL, December, 12-16, 2022 As part of the U.S. Environmental Protection Agency (EPA) Next Generation Emission Measurement (NGEM) program, research on lower cost fenceline sensors is enabling new forms of air pollution source monitoring. Easy to site, solar-powered fenceline and near-source sensors can help industry, regulators, and communities improve understanding of certain types of air pollution sources, such as fugitive emissions and process malfunctions. Current research focuses on non-speciated volatile organic compound (VOC) measurements using passively ventilated and temperature-stabilized 10.6 eV photoionization detectors (PIDs). These sensor pods (SPods) record time-synchronized wind and VOC concentration data at 1 Hz frequency and can trigger canister samples for speciated laboratory analysis. In this presentation, recent progress on analysis of multi-node SPod data is reviewed through discussion of two field studies with increasingly advanced sensor forms. We first describe a 19-month deployment of prototype EPA SPods sited near a chemical facility in Louisville, KY. In this study, we compared PID sensor elements from two different manufacturers and developed an SPod-triggered canister grab sampling approach. Periodically elevated VOC levels were observed originating from the direction of the nearby facility. Canisters acquired during these emission events indicated elevated concentrations of 1,3-butadiene and cyclohexane, compounds known to be emitted by this facility. In a current field study, spatially separated deployments of a commercial version of the EPA SPod and canister acquisition system are helping to characterize VOC emissions from fuel and asphalt storage terminals in Greensboro, NC.
Description:
As part of the U.S. Environmental Protection Agency (EPA) Next Generation Emission Measurement (NGEM) program, research on lower cost fenceline sensors is enabling new forms of air pollution source monitoring. Easy to site, solar-powered fenceline and near-source sensors can help industry, regulators, and communities improve understanding of certain types of air pollution sources, such as fugitive emissions and process malfunctions. Current research focuses on non-speciated volatile organic compound (VOC) measurements using passively ventilated and temperature-stabilized 10.6 eV photoionization detectors (PIDs). These sensor pods (SPods) record time-synchronized wind and VOC concentration data at 1 Hz frequency and can trigger canister samples for speciated laboratory analysis. In this presentation, recent progress on analysis of multi-node SPod data is reviewed through discussion of two field studies with increasingly advanced sensor forms. We first describe a 19-month deployment of prototype EPA SPods sited near a chemical facility in Louisville, KY. In this study, we compared PID sensor elements from two different manufacturers and developed an SPod-triggered canister grab sampling approach. Periodically elevated VOC levels were observed originating from the direction of the nearby facility. Canisters acquired during these emission events indicated elevated concentrations of 1,3-butadiene and cyclohexane, compounds known to be emitted by this facility. In a current field study, spatially separated deployments of a commercial version of the EPA SPod and canister acquisition system are helping to characterize VOC emissions from fuel and asphalt storage terminals in Greensboro, NC. Progress on the development of an open-source, automated data analysis tool that transforms these high time-resolved data into usable information is described. The data analysis system includes a baseline correction approach that assists with isolation of emission plume detection signatures. We discuss the aggregation and automated quality assurance (QA) processing of high frequency data to generate spatially resolved source location data, automated QA filtering, and statistically based detection limit calculations.