Assessment of groundwater contamination risks from shale gas development in the Northern Appalachian Basin, USA

The expansion of unconventional oil and gas development (UOGD), made possible by horizontal drilling and hydraulic fracturing of shale formations, has fostered economic benefits in the United States (US), increasing domestic energy supplies and exports for international markets. At the same time, local concerns about risks posed by this industry on the environment and public health persist, especially regarding the potential contamination of drinking water in communities that depend on aquifers for daily use. Quantifying such risks at large scales has been difficult due to spatiotemporal monitoring constraints and challenges in impact attribution. Here, we develop a physically based framework to assess groundwater contamination risks from UOGD in a 300,000-sq km region encompassing three states in the Northern Appalachian Basin, US (Pennsylvania, Ohio, and West Virginia). The region is home to thousands of unconventional wells drilled into the Marcellus and Utica-Point Pleasant Shale, as well as over 4 million residents served by domestic groundwater wells. Our framework integrates publicly available geospatial data with groundwater flow and solute transport modeling. We employed an ensemble calibration approach to derive multiple realizations of model parameters tuned to minimize residuals between simulation outputs and available hydrologic observations. For each realization, forward particle tracking simulations from UOGD well locations were performed to simulate spills and delineate advective transport pathways towards drinking water receptors. Ensemble simulation results were then translated into quantitative metrics of contamination risk. We illustrate how the framework can be applied both ex post, to establish physics-driven pathways between UOGD sources and observed well-water impairments, and ex ante, to identify priority areas for enhanced monitoring and protection. The assessment framework can be used to evaluate risks associated with multiple contaminant sources distributed across large regions.