Methane Emissions from Oil and Gas Production Sites and Their Storage Tanks in West Virginia
Citation:
Johnson, D., N. Clark, R. Heltzel, M. Darzi, T. Footer, S. Herndon, C. Murral, AND E. Thoma. Methane Emissions from Oil and Gas Production Sites and Their Storage Tanks in West Virginia. Atmospheric Environment: X. Elsevier B.V., Amsterdam, Netherlands, , na, (2022). https://doi.org/10.1016/j.aeaoa.2022.100193
Impact/Purpose:
Advancement in energy extraction technologies has driven unprecedented expansion of oil and natural gas (ONG) production in many regions of the U.S. Methane, ozone precursors, and hazardous air pollutants, such as benzene, are emitted as part of ongoing ONG production operations. A growing body of research indicates that fugitive leaks and equipment/process malfunctions are important ONG emissions categories that may not be properly reflected in emission inventories. These stochastic sources can produce significant impacts to near-source populations, regional air sheds, and greenhouse gas budgets. Environmentally sustainable development of ONG assets requires improved understanding and management of these sources. Development of cost-effective approaches to rapidly detect and control unnecessary emissions, along with a deeper understanding of overall emission levels and atmospheric effects are important research themes in this space. Under its next generation emissions measurement (NGEM) program, ORD is working to develop and use new measurement, modeling, and inventory approaches relevant to EPA partners, industry, and communities. The following research paper describes a field study that used two NGEM technologies, optical gas imaging (OGI) and full flow sampler (FFS), to detect and measure emissions from storage tanks and other sources found in ONG production. The study surveyed 15 ONG production sites in West Virginia documenting storage tanks and other systems supporting 66 working wells. Methane sources were categorized by the following leak and loss groups: pneumatic controllers, tanks, wellheads, compressors, engine crankcases, and other (assorted flanges, valves, fittings). This paper first presents the tank emissions in context of all emissions from a well pad, followed by a more detailed account of the tank emissions. In total, 224 unique sources were quantified yielding a total emission rate of 57.5 ± 2.88 kg/hr for all sites combined. Site specific total emissions rates ranged from 0.35 to 10.51 kg/hr with arithmetic and geometric means of 3.84 and 2.17 kg/hr, respectively. The two largest contributors to total emissions were pneumatic devices (35 kg/hr or ~61 percent of total) and tanks (14.3 kg/hr or ~25 percent of total). The total number of tanks at all sites was 153, which accounted for 42 emissions sources. Tank measurements were compared to various calculations using emission factors (EF) and industry reported data. In addition, 35 tanks components identified with OGI surveys were examined based on make and model data with advertised EF based on their 90 percent pressure threshold. Elevated tank emissions were likely influenced by site engineering design that connected all tanks at a site through a common vent header system increasing the potential impact of a single emission point or process malfunction.
Description:
In April of 2018, a two-week measurement campaign was conducted to characterize methane and other gas emissions from 15 unconventional oil and natural gas production sites (well pads targeting shale) operated by Southwestern Energy in Ohio County, West Virginia. The sites were surveyed using optical gas imaging (OGI) cameras to identify fugitive and vented sources of natural gas emissions, with the mass emission rate of the methane from components subsequently quantified using a full flow sampler. Exhaust gases of natural gas engines driving compressors, and combustor effluent streams were not quantified. Calibrations were completed before, during, and after the campaign, yielding a measurement uncertainty of ±5.5 percent for methane mass emissions rates from 1 to 1000 g/hr. Methane sources were categorized by the following leak and loss groups: pneumatic controllers, tanks, wellheads, compressors, engine crankcases, and other (assorted flanges, valves, fittings). This paper first presents the tank emissions in context of all emissions from a well pad, followed by a more detailed account of the tank emissions. In total, 224 unique sources were quantified yielding a total emission rate of 57.5 ± 2.88 kg/hr for all sites combined. Site specific total emissions rates ranged from 0.35 to 10.51 kg/hr with arithmetic and geometric means of 3.84 and 2.17 kg/hr, respectively. The two largest contributors to total emissions were pneumatic devices (35 kg/hr or ~61 percent of total) and tanks (14.3 kg/hr or ~25 percent of total). Produced water and condensate tanks at all sites employed emissions control devices primarily in the form of enclosed combustion devices to capture and reduce tank emissions (working, breathing, and flashing). Nevertheless, such tanks may still lose gas via leaks at hatches and other pressure relief components as was observed in this study. The total number of tanks at all sites was 153. However, Site 8 experienced a major malfunction at the storage tanks and direct measurements were not conducted due to safety concerns. The remaining 14 sites had 143 permanent tanks, which accounted for 42 emissions sources. Tank measurements were compared to various calculations using emission factors (EF) and industry reported data. In addition, 35 components identified with OGI surveys were examined based on make and model data with advertised EF based on their 90 percent pressure threshold. Major findings or trends were likely obfuscated by variations in production rates and their time periods, by highly variable methane fractions from tank related canister samples, and by the communication of all tanks at a site through a common vent header system.
URLs/Downloads:
DOI: Methane Emissions from Oil and Gas Production Sites and Their Storage Tanks in West Virginia