Defining a Controlled Testing Protocol for Continuous Emission Monitoring Systems Performing Leak Detection at Natural Gas Facilities.

Abstract:

New leak detection and repair (LDAR) programs consider combining continuous emission monitoring systems (CEMS) with traditional survey techniques to improve mitigation and potentially reduce program costs. New programs promote the use of CEMS to detect large sources quickly, independently of scheduled surveys. The resultant shortened times-to-repair thereby reduce the contribution of “fat-tail” sources to total emissions from oil and gas operations.

We share a protocol for evaluating the performance of CEMS under controlled release experiments. The protocol is developed with input from stakeholders including representative from operators, technology developers, and air quality regulatory agencies. Detection data including timing, attribution, localization, and quantification estimates (if applicable) are reported to the test center. Reported detections are classified as true positive or false positive by matching to controlled releases. The classification also identifies false negative detections (controlled releases that were not detected). Metrics including probability-of-detection curves, detection time, localization precision and accuracy, and quantification precision and accuracy are evaluated using the classified detection and controlled release data. This is the first protocol to establish a standard methodology for evaluating the performance metrics needed to model the emissions mitigation potential of LDAR programs incorporating CEMS using simulation tools such as the Fugitive Emissions Abatement Simulation Toolkit (FEAST). Testing according to the protocol is scheduled at the Methane Emission Technology Evaluation Center (METEC) in Fall 2020 and Spring 2021.