Energy Institute, Colorado State University, Fort Collins, Colorado 80524, United States.
Department of Computer Science, Colorado State University, Fort Collins, Colorado 80524, United States.
Environ Sci Technol. 2024 Apr 16;58(15):6575-6585. doi: 10.1021/acs.est.3c08972. Epub 2024 Apr 2.
Wide-area aerial methods provide comprehensive screening of methane emissions from oil and gas (O & G) facilities in production basins. Emission detections ("plumes") from these studies are also frequently scaled to the basin level, but little is known regarding the uncertainties during scaling. This study analyzed an aircraft field study in the Denver-Julesburg basin to quantify how often plumes identified maintenance events, using a geospatial inventory of 12,629 O & G facilities. Study partners (7 midstream and production operators) provided the timing and location of 5910 maintenance events during the 6 week study period. Results indicated three substantial uncertainties with potential bias that were unaddressed in prior studies. First, plumes often detect maintenance events, which are large, short-duration, and poorly estimated by aircraft methods: 9.2 to 46% (38 to 52%) of plumes on production were likely known maintenance events. Second, plumes on midstream facilities were both infrequent and unpredictable, calling into question whether these estimates were representative of midstream emissions. Finally, 4 plumes attributed to O & G (19% of emissions detected by aircraft) were not aligned with any O & G location, indicating that the emissions had drifted downwind of some source. It is unclear how accurately aircraft methods estimate this type of plume; in this study, it had material impact on emission estimates. While aircraft surveys remain a powerful tool for identifying methane emissions on O & G facilities, this study indicates that additional data inputs, e.g., detailed GIS data, a more nuanced analysis of emission persistence and frequency, and improved sampling strategies are required to accurately scale plume estimates to basin emissions.
广域航空方法为生产盆地中的石油和天然气（O&G）设施的甲烷排放提供了全面的筛查。这些研究中的排放检测（“羽流”）也经常被扩展到流域水平，但对扩展过程中的不确定性知之甚少。本研究分析了丹佛-朱尔斯堡盆地的飞机现场研究，以量化羽流识别维护事件的频率，使用了 12629 个 O&G 设施的地理空间清单。在 6 周的研究期间，研究合作伙伴（7 家中游和生产运营商）提供了 5910 次维护事件的时间和位置。结果表明，在先前的研究中，有三个存在潜在偏差的重要不确定性因素未得到解决。首先，羽流经常检测到维护事件，这些事件规模大、持续时间短且难以通过飞机方法进行估算：在生产过程中，9.2%至 46%（38%至 52%）的羽流可能是已知的维护事件。其次，中游设施的羽流既不频繁也不可预测，这使得中游排放的估计是否具有代表性受到质疑。最后，4 个归因于 O&G 的羽流（飞机检测到的排放的 19%）与任何 O&G 位置都不匹配，这表明这些排放已经在某个源的下风处漂移。目前尚不清楚飞机方法对这种羽流的估算有多准确；在本研究中，它对排放估算产生了重大影响。虽然飞机调查仍然是识别 O&G 设施甲烷排放的有力工具，但本研究表明，需要额外的数据输入，例如详细的 GIS 数据、更细致的排放持久性和频率分析以及改进的采样策略，才能准确地将羽流估算扩展到流域排放。
