Regional Differences in Methane Emissions from Oil and Natural Gas Production Sites in the Denver-Julesburg, Uintah, and Marcellus Shale Basins

Several recent studies have indicated that methane (CH₄) emissions at oil and natural gas (O&G) production sites are dominated by fugitive emissions from pneumatic controllers and venting/off-gassing from condensate and produced water tanks. The cumulative totals of these CH₄ emissions in a given production region are determined, in part, by the total number of well pad sites and may vary, within the region, based on quantifiable facility-level operational characteristics (e.g., number of wells per site, oil/natural gas production rates, facility age, and natural gas composition). Understanding regional variabilities in such operational characteristics and their influence on total CH₄ emissions could help inform targeted CH₄ emissions mitigation strategies at the regional scale.

In this study, we examined regional differences in CH₄ emissions from O&G production sites in the Denver-Julesburg (DJB, Weld County, CO), Uintah (Uintah and Duchesne Counties, UT), and Marcellus (northeastern (NEPA)) basins. These three O&G production regions differ significantly in several key operational parameters. First, there are large variabilities in the total number and type of active O&G well pads (e.g., 1260 unconventional gas well pads (UNG) in NEPA versus 12000 mixed O&G well pads in the DJB). Second, regional O&G production and facility-specific O&G production rates vary widely among the three production regions. For example, in May 2015, approximately 1260 NEPA UNG well pad sites produced 6 times more natural gas than all 7700 O&G well pads in Uintah. Third, at the regional scale, average facility age (estimated from well spud dates) are highest in the Uintah (10.6 years old) and DJB (13 years old) and lowest among NEPA UNG sites (3.5 years old). Finally, measured average ethane/methane (C₂H₆/CH₄) molar ratios in natural gas varied from a low 2% (NEPA) to a high 30% in the DJB.

Measurements of facility-level CH₄ emission rates (CH₄ FLER) in these three regions were performed at 103 active O&G production well pad sites in April to May 2015 (Uintah, n = 31, and DJB, n = 23) and May 2016 (NEPA, n = 49). Selection of sites were quasi-random, and were determined by downwind road access, local terrain and meteorology. Nevertheless, efforts were made to ensure that sites were selected from a distribution that adequately matched regional diversity in site production, facility age, and site operators. In all three regions, three top-down ground-based measurement techniques—i.e., dual tracer flux utilizing acetylene and nitrous oxide as atmospheric tracers, drive-by plume intercepts utilizing Gaussian plume inverse modeling, and stationary OTM (Other Test Methods)-33A approaches—were utilized. Measured CH₄ FLER were scaled up to total populations of well pads in each region using a probabilistic modeling scheme that accounts for the contributions of high emitting sites and large episodic emissions based on empirical observations.

Measured mean CH₄ FLER varied by four orders of magnitude, ranging from 0.01 ± 0.01 kg/h (Uintah) to 293 kg/h (NEPA), and from 0.001% to 42% proportional CH₄ loss rates at the facility level. Overall, 72% to 83% of sampled sites in each region emitted less than 3 kg/h CH₄ (average: 0.81 ± 0.89 kg/h), equivalent to 0.13% to 4.27% CH₄ loss rates. However, the top 10% of measured well pad sites in each region contributed approximately 60% (Uintah, DJB) to 80% (NEPA) of respective cumulative CH₄ emissions, underscoring the importance of “super emitters”. For all sites in the three regions, multilinear regression models incorporating oil/gas production rates, well count, facility age, and C₂H₆/CH₄ molar ratios explained only 11% (p = 0.027) of the observed variability in mean CH₄ FLER. However, preliminary results suggest that high mean CH₄ FLER are likely to be observed at sites with high gas production rates in Marcellus (r²_adj = 0.30, p < 0.05) and low C₂H₆/CH₄ content in Uintah (C₂H₆/CH₄ < 8%, r²_adj = 0.25, p < 0.05). Regionally, preliminary analyses indicate total (basin-wide) CH₄ emissions from active O&G production well pads are highest in the DJB and lowest in NEPA corresponding to regional proportional CH₄ loss rates that are lowest in NEPA (<0.5%) and highest in the Uintah and DJB (<3%) during the measurement period. Detailed discussions on factors influencing estimated regional CH₄ emissions, contributions of high emitting sites in each region, comparisons with recent top-down and bottom-up inventory estimates, and uncertainty assessments will be presented.