Regional methane emissions estimates in northern Pennsylvania gas fields using a mesoscale atmospheric inversion system

Shale gas development has been accompanied by concerns regarding the overall greenhouse gas (GHG) footprint of this energy source. Estimates of total methane (CH₄) leakage from shale wells to point of use have ranged from about 1% to more than 7% of production, with the greatest variability due to estimates of leakage in the production phase, which accounts for 1-4% leakage in these estimates. Government accounting of shale gas emissions has changed frequently and by large amounts. We present here the first regional emissions estimates of CH4 from shale gas production activities using a mesoscale inverse system at high resolution. Two CH₄ CRDS analyzers were deployed in the North East of Pennsylvania (Susquehanna county) and near Windsor, New York during June and July of 2013 measuring continuously the CH₄ atmospheric concentrations. Using the PennState WRF-FDDA modeling system coupled to a Lagrangian particle dispersion model, we simulated the atmospheric transport at 1km resolution. The regional CH₄ emissions over northeastern Pennsylvania were quantified using a Bayesian atmospheric inversion system, using alternatively one of the two sites as our background CH₄ atmospheric concentration, depending on the wind direction. We also present an analysis of the temporal variability and the amplitude of signals in the observed concentrations, with several peaks of a few ppm occurring over short periods of time and the systematic enhancement of CH₄ concentrations at the downwind site over longer periods of time. We show here that surface networks of high accuracy CH₄ analyzers are able to detect atmospheric signals due to CH₄ emissions in shale gas production areas and could be used in regional atmospheric inversion systems to quantify CH₄ emissions from gas production activities.