Rigorous Methane Inventories for Oil and Gas facilities based on Continuous Monitoring Systems

Accurate methane emissions inventories for oil and gas facilities are increasingly required to support regulatory reporting, voluntary frameworks, and international natural gas trade. Onsite continuous monitoring systems (CMS) provide time-resolved methane concentration measurements, making them a promising avenue for inventory development. To infer emissions from the concentration measurements, however, requires a careful inversion framework considering near-field turbulence and short-term wind conditions. Specifically, it is crucial to be aware of time periods when wind conditions and sensor placement are such that potential emissions are not observable, which we refer to as periods of no information. We present a general framework for constructing measurement-derived methane emissions inventories using CMS data alone, without reliance on bottom-up emission factors or operational estimates. The framework explicitly accounts for no-information periods and provides fully transparent rigorous uncertainty quantification that propagates both inference uncertainty and imputation uncertainty into methane emission inventory estimates. We validate the approach using controlled-release experiments and demonstrate case studies from multiple production sites.