Quantifying, Assessing, and Mitigating Methane Emissions from Super-emitters in the Oil and Gas Supply Chain

Methane emissions from the oil and gas (O&G) supply chain reduce potential climate benefits of natural gas as a replacement for other fossil fuels that emit more carbon dioxide per energy produced. O&G facilities have skewed emission rate distributions with a small fraction of sites contributing the majority of emissions. Knowledge of the identity and cause of these high emission facilities, referred to as super-emitters or fat-tail sources, is critical for reducing supply chain emissions. This dissertation addresses the quantification of super-emitter emissions, assessment of their prevalence and relationship to site characteristics, and mitigation with continuous leak detection systems. Chapter 1 summarizes the state of the knowledge of O&G methane emissions. Chapter 2 constructs a spatially-resolved emission inventory to estimate total and O&G methane emissions in the Barnett Shale as part of a coordinated research campaign using multiple top-down and bottom-up methods to quantify emissions. The emission inventory accounts for super-emitters with two-phase Monte Carlo simulations that combine site measurements collected with two approaches: unbiased sampling and targeted sampling of super-emitters. More comprehensive activity data and the inclusion of super-emitters, which account for 19% of O&G emissions, produces a emission inventory that is not statistically different than top-down regional emission estimates. Chapter 3 describes a helicopter-based survey of over 8,000 well pads in seven basins with infrared optical gas imaging to assess high emission sources. Four percent of sites are observed to have high emissions with over 90% of observed sources from tanks. The occurrence of high emissions is weakly correlated to site parameters and the best statistical model explains only 14% of variance, which demonstrates that the occurrence of super-emitters is primarily stochastic. Chapter 4 presents a Gaussian dispersion model for optimizing the placement of continuous leak detection systems at three example well pads. The model demonstrates that large leaks can be detected quickly with first generation systems. Continuous leak detection can be used in the near future to cost-effectively mitigate methane emissions from O&G super-emitters.