Hydraulic fracturing has the potential to introduce saline water into the environment due to migration of deep formation water to shallow aquifers and/or discharge of produced water during storage, transport or disposal. High-volume hydraulic fracturing of the Marcellus Shale produces water with a chemical composition similar to deep groundwater found in Appalachian Basin formations, including high concentrations of Ca, Na, and Cl, and metal concentrations consistent with deep Appalachian Basin brines. Fingerprinting water impacted by hydraulic fracturing is challenging given the other potential anthropogenic and natural sources of salinity and metals in areas in which drilling is taking place. We have developed a multivariate statistical modeling framework for quantitatively fingerprinting natural and anthropogenic sources of high salinity (> 20 mg/L Cl) in shallow groundwater pre- or post- hydraulic fracturing. The model was developed using geochemical data for 203 groundwater wells sampled by Project SWIFT (Shale-Water Interaction Forensic Tools) in New York State (NYS), which has a hydraulic fracturing moratorium, along with published data for an additional 1,113 shallow wells in NYS. Chemical compositions of salinity sources were taken from published data for Appalachian formation water, road salt, septic effluent, and animal waste.  The model stochastically simulates the expected geochemistry of high salinity groundwater impacted by different sources of salinity and then employs linear discriminant analysis to classify samples from different populations. Model results indicate formation water is the most probable source of elevated salinity in 28% of the NYS groundwater wells with >20 mg/L Cl sampled by Project SWIFT. The model provides a robust means for differentiating groundwater impacted by formation water versus other contaminants and, using our proposed framework, similar discriminatory tools can be derived for other regions using background water quality data.

Fingerprinting sources of salinity in shallow groundwater prior to hydraulic fracturing: Statistical model development and application

248th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.