Measurement-Informed Methane Emission Inventories Must Account for the Impact of Repair Interventions

Accurate methane emission estimation from the oil and gas sector has become a cornerstone of regulatory actions and voluntary initiatives. Methane emissions in the oil and gas sector follow a heavy-tailed distribution, where underestimating the frequency of short-duration large emitters leads to inaccurate emission inventories. However, even high-frequency surveys may miss short-lived, high-rate-emission events. Furthermore, operator-led surveys are routinely followed by repairs of addressable emissions, affecting inventory estimates and introducing uncertainty into measurement-informed inventories (MII). In this study, we developed a stochastic uncertainty estimation model to quantify uncertainties in MII resulting from variations in emitter duration, survey frequency, sample size, technology characteristics, and the effect of repair interventions. We find that quantification accuracy is mainly driven by survey frequency, while precision is driven by the sample size. Increasing survey frequency substantially improves accuracy initially, but additional surveys beyond a certain point yield diminishing improvements, indicating that an optimal combination of survey frequency and sample size exists to balance operational costs with the development of an accurate MII. For applications that require an unbiased estimate of emissions to conduct reconciliation (e.g., OGMP 2.0), a higher survey frequency is preferable to a larger sample size. We also find that implementing repair programs will strongly influence the accuracy of MII and highlight the importance of properly accounting for repair actions when developing the MII. We conclude with key recommendations for planning measurement campaigns to comply with requirements in regulatory programs and voluntary initiatives.