Abstract
Methane emission measurements are important elements in quantifying and mitigating methane emissions from oil and gas systems. Short duration remote sensing measurements of methane emissions are becoming economically feasible and deployable at large scale, but interpretation of these snapshot measurements is complex due to the inherent spatial and temporal variability of methane emissions. Even if measurement technologies have no uncertainty, short duration sampling of these complex temporal emission patterns introduces a sampling uncertainty. This work examines the sampling uncertainties associated with monthly, quarterly, semi-annual and annual short-duration measurements of methane emissions for a group of fifty sample sites selected to replicate conditions in the Barnett Shale oil and gas production region. One of the key characteristics of Barnett Shale methane emissions is the presence of high-emission rate events which contribute to approximately half of total emissions. These high-emission events typically account for only 1-2% of emission rate observations in the Barnett Shale and are of uncertain duration. Analyses were conducted to assess how the duration of these high-emission events impacts the accuracy of remote sensing measurements of methane emissions, conducted with frequencies ranging from monthly to annual. Emissions from pneumatic controllers, pneumatic pumps, tanks, and leaks, which collectively account for approximately half of emissions in the Barnett Shale, are accurately captured with annual sampling frequencies. The addition of high-emission events increases the required number of measurements to maintain a targeted level of uncertainty, even under the assumption of no measurement uncertainty. If emission events have a one-week duration, a monthly sampling frequency has an estimated sampling error >15% for the types of sources in the Barnett Shale, due to the inherent variability of the magnitudes of the high emission rate events. The error associated with short-duration sampling increases as the duration of the high-emission events becomes shorter, suggesting that understanding the temporal persistence of significant emission events is important to consider when designing measurement protocols.