Extension of Methane Emission Rate Distribution for Permian Basin Oil and Gas Production Infrastructure by Aerial LiDAR

Abstract

Aerial LiDAR measurements at 7474 oil and gas production facilities in the Permian Basin yield a measured methane emission rate distribution extending to the detection sensitivity of the method, 2 kg/h at 90% probability of detection (POD). Emissions are found at 38.3% of facilities scanned, a significantly higher proportion than reported in lower-sensitivity campaigns. LiDAR measurements are analyzed in combination with measurements of the heavy tail portion of the distribution (>600 kg/h) obtained from an airborne solar infrared imaging spectrometry campaign by Carbon Mapper (CM). A joint distribution is found by fitting the aligned LiDAR and CM data. By comparing the aerial samples to the joint distribution, the practical detection sensitivity of the CM 2019 campaign is found to be 280 kg/h [256, 309] (95% confidence) at 50% POD for facility-sized emission sources. With respect to the joint model distribution and its confidence interval, the LiDAR campaign is found to have measured 103. 6% [93.5, 114.2%] of the total emission rate predicted by the model for equipment-sized emission sources (∼2 m diameter) with emission rates above 3 kg/h, whereas the CM 2019 campaign is found to have measured 39. 7% [34.6, 45.1%] of the same quantity for facility-sized sources (150 m diameter) above 10 kg/h. The analysis is repeated with data from CM 2020−21 campaigns with similar results. The combined distributions represent a more comprehensive view of the emission rate distribution in the survey area, revealing the significance of previously underreported emission sources at rates below the detection sensitivity of some emissions monitoring campaigns.

Unaddressed Uncertainties When Scaling Regional Aircraft Emission Surveys to Basin Emission Estimates

“…Various sources provide different estimates of the detection sensitivity for the Carbon Mapper method, ranging from a popularly quoted “10–20 kg/h” to 32 kg/h in a 3 m/s wind by Conrad et al to a “full detection limit” of 280 kg/h by Kunkel et al In this study, the mean estimated plume emission rate was 283 kg/h, with 95% of all estimates above 30 kg/h.…”

“…It is important to note that the detectability of maintenance emissions varies with the sensitivity of the aerial method. Since the Carbon Mapper instrument trades coverage for sensitivity, its detection sensitivity is fairly low. , Therefore, fewer events will be detected than would be expected for a method with higher sensitivity. Additionally, maintenance events vary widely in size, ranging from small equipment blowdowns, such as a pig launch, to large venting operations, such as wellbore liquid unloading …”

“…See Section S3 for more complete descriptions. While many logged and reported events likely produced peak emission rates below the method detection limit of the aircraft method, 36 all events were considered as potential matches to aircraft plumes.…”

The Role of Emission Size Distribution on the Efficacy of New Technologies to Reduce Methane Emissions from the Oil and Gas Sector

“…Non-tank emitters above 10 kg/h are classified as large emitters, similar to EPA modeling. The detection threshold of the aerial survey instrument is approximately 3 kg/h . To incorporate below detection threshold emissions, we supplemented the aerial survey data with bottom-up, component-level methane measurement data for emissions in the 0–3 kg/h range.…”

“…To evaluate the impact of emission size distribution on the equivalence of different LDAR programs, we use data from recent multibasin, equipment-level aerial surveys conducted using Bridger Photonics. Bridger Photonics is a light detection and ranging (LIDAR) based system to detect and quantify methane emissions at a spatial resolution of about 2 m . Details on performance parameters such as the probability of detection obtained from controlled release testing, measurement uncertainty, and intercomparison with other technologies can be found in peer-reviewed literature. , Bridger does not distinguish routine and nonroutine emissions in their data set, and therefore, all measurements from Bridger are used in this analysis.…”