Optimizing Methane Emission Source Localization in Oil and Gas Facilities Using Lagrangian Stochastic Models and Gradient-Based Detection Tools

Abstract. Oil and gas facilities are responsible for approximately 35 % of global emissions of methane (CH₄), a potent greenhouse gas, posing significant environmental and regulatory challenges. While Continuous Emission Monitoring Systems (CEMS) are widely implemented to track real-time emissions, their effectiveness in localizing specific CH₄ emission sources remains limited, particularly under complex environmental conditions. This study integrates CEMS technologies with a Lagrangian Stochastic Back-Trajectory Model, along with an automated Gradient Indicator (GI) tool to improve methane source localization accuracy in oil and gas settings. The model was then applied to a real-world gas distribution site to validate its performance in accurately localizing methane emissions under operational conditions. Using synthetic data simulations, we evaluated the performance of this integrated system under various atmospheric stability conditions, sensor-source height differences, and source proximity. Our results indicate that this combined approach significantly enhances localization performance, achieving a 90 % probability of detection (POD) within a 25–75 meter source-sensor distance under optimal conditions. However, detection performance varied across configurations, with false positive rates (FPF) ranging from 22 % to 86 %, and localization accuracy (LA) ranging from 14 % to 78 %, depending on atmospheric stability, source-sensor geometry, and height differences. The Localization Accuracy (LA) improves when sensor placements are exactly downwind of the emission sources (alignment). The system meets Canadian regulatory requirements for CEMS applications by maintaining localization accuracy above 90 % for unstable and slightly neutral atmospheric conditions, ensuring that emissions are correctly attributed. However, neutral atmospheric conditions and large height differentials between sensors and sources reduce localization accuracy, making optimized sensor configurations important. Findings of the research can be useful for upgrading CEMS systems and help them to overcome some difficulties from regulations associated with methane emission reporting and mitigation efforts.