Combined radiogenic (87Sr/86Sr, 234U/238U) and stable (δ88Sr) isotope systematics as tracers of anthropogenic groundwater contamination within the Williston Basin USA

Abstract

The Williston Basin has produced oil and gas from conventional structural and stratigraphic traps for more than 60 years. The advent of horizontal drilling and hydraulic fracturing of shale in the Devonian-Mississippian Bakken Formation has increased the production of oil in this basin from areas that partially overlap the prairie pothole region of North America. Massive scale of oil production in the Williston Basin increases the risk of accidental releases of coproduced formation water (brine) into the environment. The prairie pothole region is named for a multitude of small lakes and wetlands that provide critical habitat for waterfowl and other wildlife. The increased risks raise the importance of developing robust tracers that can identify the source(s) and quantity of contamination of this premier wetland ecosystem. Radiogenic 87 Sr/ 86 Sr and 234 U/ 238 U isotope tracers are widely used in hydrological studies to detect potential sources of groundwater contamination. Here, we used paired 87 Sr/ 86 Sr and δ 88 Sr values in wetland water, which is essentially shallow groundwater at the Goose Lake and Fuller sites in Montana and North Dakota where produced water contamination was previously established. We also analyzed brines from the Bakken Formation and Mississippian Charles Formation to evaluate potential sources of contamination. This is the first attempt to combine these isotope tracers to estimate a magnitude of groundwater contamination by the oil-field produced water. Moreover, because U +6 is soluble in water under subaerial oxidizing conditions and the U +4 content is orders of magnitude lower (pg/g levels) in reducing brines, the 234 U/ 238 U isotope tracer is insensitive to brine contamination. However, it elucidates potential variability of uncontaminated end-members and, therefore, helps to improve the accuracy of estimated degrees of brine contamination based on binary mixing relationships. In addition, U isotopes in groundwater have a potential for detecting some U-rich anthropogenic contaminants (e.g., phosphate fertilizers). Using these isotopic systematics we confirm previous conclusions that surface water and shallow groundwater in two studied sites are variably contaminated by produced water and estimate the degree of the contamination at ∼7% in a sample collected close to a brine tank and <0.5% in most other samples. The U isotope analyses in combination with the Sr isotope results help outline variable groundwater pathways and uncontaminated end-members. Also, U isotopes show no appreciable groundwater contamination from U-enriched P-fertilizers. Our study in the Williston Basin demonstrates the potential of using combined radiogenic ( 87 Sr/ 86 Sr, 234 U/ 238 U) and non-traditional stable (δ 88 Sr) isotopic systematics as tracers of anthropogenic contamination in surface and groundwater systems.