Establishing a Link Between Hydraulic Fracturing Activities and Stream Water Impacts in North-Central Pennsylvania
The impacts of hydraulic fracturing related natural gas extraction activities on water contamination are poorly understood. Furthermore, while the majority of preliminary studies have focused on the impacts of hydraulic fracturing activities to groundwater little data exists with regards to impacts on overall streamwater quality. This study used a multi‐faceted approach to determine whether there is a link between hydraulic fracturing related well pad density within a given watershed and associated stream and shallow groundwater quality. A total of 77 streamwater samples, 19 groundwater samples, and one spring sample were collected from a total of 38 small watersheds (<20 km2) in Bradford, Susquehanna, and Wyoming Counties in north‐central Pennsylvania during summer 2013. Three samples from three watersheds exceeded their UEPA National Recommended Water Quality Criteria (NRWQC) Criteria Maximum Concentration (CMC) or acute concentration for both Selenium and Cadmium. In addition, one watershed (#45) exceeded its USEPA NRWQC CMC for Zinc. Several additional samples exceed their USEPA NRWQC Criterion Continuous Concentration (CCC) or chronic concentration for Cadmium and Aluminum. Interestingly, while several watersheds exhibited metal concentrations in excess of their associated criterion values during both the early and late summer sampling events, other watersheds exhibited a surprising degree of temporal variation in concentration. Groundwater sampling revealed barium concentrations in excess of the United State Environmental Protection Agency Maximum Contaminant Level (USPEA MCL) in two drinking water well samples (G25C and G29) and arsenic concentrations in excess of the USEPA MCl of in one of the groundwater samples (G46). Several of the groundwater samples also exhibited aluminum, iron, and manganese concentrations in excess of their respective USEPA List of National Secondary Drinking Water Regulations for aluminum, iron, and manganese. Detectable concentrations of methane were identified in twelve of the groundwater samples and ranged from 0 to 1.1 mg/l. None of the methane values were in excess of the US Department of Interior Action Level for Hazard Mitigation. Finally, a principal component analysis (PCA) was performed to determine correlations between well pad density and land use practices within a given watershed and their associated water quality parameters and/or elemental yields. Temperature positively correlated with percent forest and grass land while potassium, nickel, and streamwater conductivity were positively correlated with percent agricultural land. Developed land exhibited the most impact on the elemental yields, positively correlating to conductivity, nickel, vanadium, iron, cobalt, arsenic, rubidium, strontium, and lithium. Lithium, cesium, lead, and uranium negatively correlated with the area of the watershed and the amount of water within each watershed. No correlations were observed between well pad density and any of the elemental yields. In addition, no correlations were observed between any of the principal components and pH, DO, sediment, chromium, manganese, copper, molybdenum, silver, cadmium, and barium. The heterogeneous nature of the observed stream and groundwater impacts, coupled with the lack of correlation of elemental yields with well pad density within a given watershed, likely points to hydraulic fracturing related spills and/or localized releases from compromised wells as playing an important role on regional stream and groundwater quality. The data suggest that more extensive spatial and temporal sampling is required to determine the long term imprint on stream and groundwater impacts to this region. The results also suggest the data should be compared with Pennsylvania Department of Environmental Protection (PADEP) records documenting known hydraulic fracturing well pad violations to determine whether there are any associated links.