Modelling temporal variability in the surface expression above a methane leakage: The ESCAPE model

Abstract:

Natural gas pipelines are used in distribution and midstream sectors to transport natural gas. These pipes may leak due to corrosion, external stressors, or other factors. While many distribution leaks are detected by odor reports, across all sectors, leaks are identified by walking surveys, where a hydrocarbon gas (CH₄ and light alkanes) sensor is used to detect gas concentration enhancements at the surface. Surveyors often use these concentration observations to assess the severity of the leak, which is a combination of the size of the leak and its proximity to structures or other infrastructure. However, as leakage is a dynamic process, changes in subsurface, surface and atmospheric conditions can mask leak severity. Controlled field experiments indicate surface CH₄ concentration measurements vary with changing meteorology and micrometeorology, increasing by up to four times during low winds and stable atmosphere events. It is possible that during sunny, unstable conditions large leaks may appear smaller or in many cases, go undetected, remaining unrepaired and causing substantial long term emissions. Conversely, in low wind and stable conditions, smaller leaks can appear larger, creating ongoing confusion and uncertainty in the correlation between gas concentrations and leakage rate. This study aims to untangle this uncertainty and confusion by providing guidance to the interwoven contribution of key atmospheric variability on the surface presentation of the leak. To improve leak classification, we estimated leak rates from readily-obtained surface CH₄ concentrations by the model developed in this study, the process-based ESCAPE (Estimating the Surface Concentration Above Pipeline Emissions) model. Controlled field experiments that measured CH₄ concentrations were in good agreement with model output.