Although natural gas (NG) is predominantly composed of methane (CH4), the ratio of other hydrocarbon gases (e.g., ethane, propane, butane) varies substantially between production basins and sections of the supply chain. In the event of a belowground pipeline leak, NG composition impacts its transport in the belowground, surface expression, and emission into the atmosphere, complicating leak detection. Mobile survey methods do not currently account for NG composition variability, as its effect on the probability of detection (POD) remains unclear. This study investigates the impact of NG composition on POD for mobile survey methods, including walking, driving, and simulated unmanned aerial vehicle (SUAV) surveys. Four controlled belowground NG release experiments were conducted at the Methane Emission Technology Evaluation Center at Colorado State University, Fort Collins, CO. Experiments, with a constant leak rate of 5 standard liters per minute, varied in NG composition: standard distribution-grade NG (85% CH4 v/v), a control composition (70% CH4 v/v), and compositions similar to those found in the Permian Basin in Texas and the Denver-Julesburg (DJ) Basin in Colorado. Results indicate that the POD for walking, driving, and SUAV surveys is variably influenced by NG hydrocarbon composition. In simulations for the DJ and Permian Basins, walking surveys showed POD values that were 2.5 and 0.8 times higher, respectively, compared to those for distribution-grade gas. These increases are attributed to larger surface plume areas associated with higher vapor density gases, as heavier hydrocarbons promote enhanced lateral subsurface migration. In contrast, driving and SUAV surveys exhibited negligible differences relative to the distribution-grade composition, with no statistically significant attribution to compositional effects. The observed differences are more likely attributable to environmental variability than to gas-specific behavior. This statistical ambiguity highlights the need for caution when interpreting POD trends based solely on visual inspection of POD curves. To confidently attribute observed effects to gas composition, it is essential to conduct experiments across multiple test facilities, apply rigorous statistical analyses, and collect data under a broad range of environmental and subsurface conditions.
Kolodziej et al. (Thu,) studied this question.