Variability observed over time in methane emissions from abandoned oil and gas wells

Abstract

Recent studies have reported methane (CH4) emissions from abandoned oil and gas wells across the United States and the United Kingdom. These emissions can reach hundreds of kg CH4 per year per well and are important to include in greenhouse gas emission inventories and mitigation strategies. Emission estimates are generally based on single, short-term measurements that assume constant emission rates over both short (hours) and longer (months/years) time periods. To investigate this assumption, we measure CH4 emissions from 18 abandoned oil and gas wells in the USA and the UK continuously over 24 hours and then make repeat 24-hour measurements at a single site over 12 months. While the lack of historical records for these wells makes it impossible to determine the underlying leakage-pathways, we observed that CH4 emissions at all wells varied over 24 hours (range 0.2-81,000 mg CH4 hr-1) with average emissions varying by a factor of 18 and ranging from factors of 1.1 to 142. We did not find a statistically significant relationship between the magnitude of emissions and variability or that variability is correlated with temperature, relative humidity or atmospheric pressure. The results presented here suggest high CH4 emission events tend to be short-lived, so short-term (< 1 hour) sampling is likely to miss them. Our findings present the dynamic nature of CH4 emissions from abandoned oil and gas wells which should be considered when planning measurement methodologies and developing greenhouse gas inventories/mitigation strategies. Incorporation of these temporal dynamics could improve national greenhouse gas emissions inventories.