Multi-scale measurements and temporally resolved modeling of methane emissions at natural gas distribution stations

Coleman Parker Vollrath, Chris Hugenholtz, Thomas Barchyn , Clay Wearmouth

Methane (CH4) emissions from natural gas distribution stations are not well characterized by measurements, contributing to uncertainty in urban emissions inventories and mitigation strategies. We conducted a multi-scale, multi-temporal study of four distribution stations in Calgary, Alberta, combining component- and facility-level measurements with modeling to quantify, attribute, and compare emissions rates. Component-level surveys found low-emitting and readily abatable fugitive emissions, vents, and emissions from catalytic heaters, which accounted for 1.6–40.1% of total emissions. Facility-level measurements using the human-portable flux plane method revealed comparatively higher rates of 3.4 x 10-2 to 7.4 kg CH4/h that were correlated with throughput, with most emissions attributed to incomplete combustion from line heaters that combust process gas prior to pressure regulation. To compare measured with conventional rate estimates, we developed a Monte Carlo-based modeling framework integrating component-level rates, historical leak detection data, gas throughput, and source-specific emission factors (EFs). Measured rates exceeded modeled rates by factors of 1.6–148.3. Across all measurements, facility-level rates totaled 17.4 kg CH4/h (10.8–24.1 kg CH4/h); modeled rates totaled 2.3 kg CH4/h (0.9–4.2 kg CH4/h). Overall, results suggest that conventional approaches underestimate distribution station CH4 emissions because underlying EFs underrepresent emissions from equipment incompletely combusting natural gas.