Modeling temporal variability in the surface expression above a methane leak: The ESCAPE model

Leaks in natural gas distribution networks are often initially detected by odor and then localized during operators' walking surveys. The size of the enhancement detected and the leak's proximity to infrastructure are used to determine leak severity of the leak and how quickly it is repaired. Methane (CH 4 ) enhancements on the surface above the leak change with atmospheric conditions, but currently the leak severity assessment is made without considering the environmental conditions and could result in misdiagnosis of the leak. This study has developed the “ESCAPE” model, a tool that can be used to estimate CH 4 enhancements on the surface above a leak using meteorological and leak data as input. Surface CH 4 concentrations calculated using the ESCAPE model agrees well with measurements made during controlled experiments (m = 0.95 and R 2 = 0.95). This study highlights the effect of micrometeorology on gas transport from the surface to the atmosphere, where the surface and the atmospheric conditions have the largest effect on the flux. Here, recommendations are made to improve industry best practices, including recording the meteorological conditions at the time of the leak detection, avoiding walking surveys on days where there are strong winds or strong solar irradiance and that known leak locations should be revisited and measured in different conditions. Mitigating CH 4 emissions from the gas distribution network is a cost-effective and economically realistic target. Distribution operators can be directed by the output of ESCAPE model to correctly identify and repair the largest distribution leaks and could reduce annual CH 4 emissions by 0.69 Tg, this may go some way to reducing energy sector emissions in accordance with the Paris Agreement. • Surface methane above pipeline leakage changes with atmospheric conditions. • Model developed to estimate surface methane using weather and leak data as input. • Model agrees with measurements made during controlled experiments, m = 0.95 & R2 = 0.95 • Shows need of using wind speed & irradiance to assess severity of distribution leaks. • Model helps identify/prioritize largest leaks for repair, reduces risk & GHG emission.