Repository logo

Using above-ground downwind methane and meteorological measurements to estimate the below-ground leak rate of a natural gas pipeline


Natural gas (NG) leaks from below-ground pipelines present a safety, economic, and environmental hazard, and triaging the severity of leaks remains a significant issue for pipeline operators. Typically, operators conduct walking surveys using hand-held methane (CH4) detectors which output CH4 concentrations to indicate the location of a leak, but quantification often requires excavation of the pipeline. Industry-standard CH4 detectors are lower-cost and have a higher detection threshold and lower precision than optical-cavity CH4 analyzers typically used to quantify emissions. It remains unclear whether coarser CH4 concentration measurements could be used to identify the large leaks that require immediate response. To explore the utility of industry-standard detectors, above-ground downwind CH4 concentration measurements made by the detectors as input to a novel modeling framework, the ESCAPE-1 model were used to estimate the leak rates from below-ground NG pipelines. Controlled below-ground emission experiments were conducted to test this approach over a range of environmental conditions. Using 10-minute averaged CH4 mixing/meteorological data and filtering out low wind/Pasquill Gifford Stability Class (PGSC) A events, the ESCAPE-1 model estimates small distribution leaks (0.2 kg CH4 h-1) to within -31 to +75% (95% CI), and medium distribution leaks (0.8 kg CH4 h-1) to within -73 to +92%(95% CI) of the actual leak rate. When averaged over a longer period (more than 3 hours of data), the average calculated leak rate was an overestimate of 55% for the small (0.2 kg CH4 h-1) leak and an underestimate of 6% for a medium distribution leak (0.8 kg CH4 h-1). Results suggest that as the wind speed increases, or the atmosphere becomes more stable both accuracy and precision of the leak rate calculated by the ESCAPE-1 model decreases. This is likely the result of a trade-off between the high enough wind to move the gas but not high enough that the plume becomes collimated and less homogenous. Optimizing this approach for oil and gas industry applications, this study suggests that CH4 mixing ratios measured by industry-standard CH4 detectors lasting at least 3 hours could be used as a guide to prioritize NG leak repair by estimating the below-ground leak rate from a pipeline within reasonable uncertainty bounds (±55%) in favorable atmospheric conditions.


Rights Access


leak detection
greenhouse gases
pipeline safety
leak quantification


Associated Publications