Modeling Detection Probability for Mobile Natural Gas Leak Surveys: Understanding the Factors That Affect Leak Capture

• Activity Number: 340 - SPEED: SPAAC SESSION III
• Type: Topic-Contributed
• Date/Time: Thursday, August 12, 2021 : 10:00 AM to 11:50 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Emily Luetschwager* and Joseph von Fischer and Zachary Weller
• Companies: Colorado State University and Colorado State University and Colorado State University
• Keywords: logistic regression; environmental monitoring; greenhouse gases; methane; mobile sensing

Abstract

Advanced leak detection (ALD) methods for surveying local natural gas distribution systems have developed rapidly in recent years. Commercial providers currently account for most survey efforts and there is little publicly available data for characterizing methane (CH4) plumes, a main feature of ALD surveys, or evaluating the efficacy of these methods for detecting leaks. We analyzed data from ALD surveys in 15 metropolitan areas and an extensive controlled release experiment to investigate the factors that affect leak detection from these surveys. Our analysis examines a breadth of factors influencing leak detection, including infrastructure characteristics, driving protocols, and environmental conditions. Our findings demonstrate how leak detection declines with leak size and distance between the survey vehicle and leak point. We also show how leak detection varies with pipeline material, where leaks attributable to cast iron pipes are most likely to be detected. We also show how wind direction affects leak detection. The results of our analysis can be used to inform driving effort and understand how many leaks may go undetected.