Estimating methane emission durations using continuous monitoring systems

A small number of large emissions contribute a significant portion of total emitted methane from the oil and gas sector, making them a critical pathway toward emission reduction. These large emissions are often detected by snapshot measurements from aerial or satellite platforms, which have limited ability to characterize emission duration due to the relatively low frequency at which they observe a given source. Duration estimates are necessary for computing the total methane emitted by a given release and will be required for all emissions >100 kg/hr under proposed updates to the EPA's Greenhouse Gas Reporting Program (GHGRP). Continuous monitoring systems (CMS) are a monitoring technology that measure methane concentrations in near-real time and hence can complement snapshot measurements by bounding the duration of detected emissions. However, CMS will not record concentration enhancements during an emission if wind blows emitted methane away from the sensors. We propose a method for estimating emission durations using CMS that accounts for these non-detect times. The method has an average error of 6.3% when evaluated on controlled releases, with 88.5% of estimates within a factor of 2x error from the true duration (i.e., percent error within [-50%, 100%]). We apply the method to a typical production site in the Appalachian Basin and use it to bound the duration of snapshot measurements. We find that failing to account for CMS non-detect times on this site results in underestimated emission durations of up to a factor of 71x (7,000%).