An event-based methodology to estimate emissions from upstream O&G sites

Mozhou Gao, Zahra Ashena, Steve H.L. Liang, Sina Kiaei, Saeedi Sara

Estimating annual site-level emissions in the oil and gas (O&G) sector is a key requirement for regulatory and voluntary initiatives worldwide. This study introduces an Emission Event Data Model (EEDM) that applies Allen’s interval algebra and spatial proximity to group multi-scale emissions observation and O&G operational data into discrete events. The model classifies emissions events into three categories: resolved (known emission rate and duration), partially resolved (known emission rate but unknown or estimated duration), and unresolved (unknown emission rate and duration). To support the application of this data model, we developed a framework integrating numerical and simulation methods to calculate total emissions and associated uncertainties. The framework includes three Monte Carlo-based approaches to enhance emissions estimation: (1) estimating the duration of partially resolved events by incorporating null detects, leak production, and natural repair processes; (2) simulating emissions below the detection limits of deployed measurement technologies; and (3) estimating emissions from unresolved events by combining the probability of emission occurrences with best-fit emission rate and duration distributions. To demonstrate this framework, we estimated site-level emissions for a fictitious site using synthetic emission observations, including OGI inspection records, continuous monitoring systems data, aircraft flyovers, and O&G operational data. The proposed data model and methodologies have important implications for improving annual emissions reporting, reducing uncertainties, and supporting measurement-informed inventories, particularly under the Measuring, Monitoring, Reporting, and Verification (MMRV) framework.