Aircraft-based mass balance flights that estimate emissions in large regions, such as oil and gas production basins encompassing thousands of square kilometers, have been widely used to provide measurement based estimates of total regional emissions. The number and density of sources and the temporal and spatial variability in methane emissions from oil and gas facilities introduce uncertainties into the analyses of mass balance flights. To quantitatively characterize these uncertainties, mass balance flux calculations were conducted for simulated flight scenarios using a site-level emission inventory for >10,000 oil and gas sites and fine spatial scale (∼1 km) modeling of horizontal and vertical transport of emissions. Uncertainties due to varying wind directions, vertical mixing profile assumptions, boundary layer height assumptions, and temporal and spatial variability in emission sources were quantified. Some of the largest uncertainties evaluated in the simulations were due to temporal variations in large emission events. Other uncertainties, specific to the high density of sources in oil and gas production regions, are due to shifting wind directions which blur the boundaries of the domain measured. More precise definition of upwind source locations sampled by mass balance flights would reduce uncertainties in the use of mass balance flight data, and this work demonstrates that upwind domain footprints can be estimated using trajectory tools.
Measurements of ambient ethane concentrations at a regional air quality monitor in the Eagle Ford oil and gas production region are compared to concentrations predicted using site-level hydrocarbon emission inventories coupled with a Gaussian puff dispersion model (CALPUFF). To account for more than half of mean concentrations due to routine emissions, sites at distances 20–50 km from the receptor site were included in the simulations. Nearly all of the highest observed concentrations were observed at night. For each night in the simulation, the location and magnitude of the maximum predicted concentration and maximum observed concentration were compared, and approximately two-thirds of the highest observed nighttime maximum concentrations were accounted for by routine emissions. In contrast, approximately a third of the highest daytime maxima could be accounted for by routine emissions. Most of the large observed maxima that are attributable to routine emissions are predicted to be caused by sources that were within 10 km of the receptor site, but sources up to 20 or more kilometers from the receptor also contributed to the predicted concentrations. A case study is provided demonstrating the potential of coupling site-level inventories of routine emissions with dispersion modeling for attributing sources of elevated hydrocarbon concentrations.
Methane emissions from >100 tank battery sites in the Permian Basin of west Texas were quantified using two different aircraft-based measurement platforms deployed asynchronously, but within a 5-day period. Emissions were detected by aircraft with rates ranging from 0.71 kg/hr to >750 kg/hr. On-site emission measurements were also made using a drone-based technology at a representative sample of 33 sites and were compared to engineering estimates of emission rates. A large fraction of total emissions from the sites sampled by the aircraft systems (90%±6%) were accounted for by sites with emissions greater than 10 kg/hr, when engineering estimates of emissions were assigned to sites with no emissions detected by aircraft. At least half of the sites with emission rates above 10 kg/hr had emissions that did not persist longer than a few days, and consequently, site by site comparisons of asynchronous aircraft measurements showed poor agreement. The two aircraft systems differed in their estimates of total emissions from the ensembles of sites sampled, and in the percentage of sites with emissions greater than 10 kg/hr. The differences in frequencies of detection between the two aircraft based platforms are attributed to a combination of factors, however, both aircraft-based emission measurement systems attribute a large fraction of emissions to sites with emission rates >10 kg/hr.