Paper No. 7
Presentation Time: 2:35 PM

GROUNDWATER GEOCHEMISTRY: ARE WELLS UNIQUE? A CASE STUDY FROM UPSTATE NEW YORK


HASBARGEN, Leslie1, CASTENDYK, Devin N.2, KEEFE, Leandra3, LOWRY, Fiona4, MOORE, Myles T.5, SPAULDING, Joseph5 and FUESS, Alayna5, (1)Earth & Atmospheric Sciences, SUNY Oneonta, 219 Science 1 Building, Ravine Parkway, Oneonta, NY 13820, (2)Dept. of Earth and Atmospheric Sciences, State University of New York, College at Oneonta, Oneonta, NY 13820, (3)Barton and Loguidice, 10 Airline Drive Suite 200, Albany, NY 12205, (4)School of Public Health, University of Minnesota, 420 Delaware Street Southeast #819, Minneapolis, MN 55455, (5)Earth & Atmospheric Sciences, SUNY Oneonta, Science 1 Building, Ravine Parkway, Oneonta, NY 13820, Leslie.Hasbargen@oneonta.edu

In advance of high volume hydraulic fracturing activities planned for central New York, we have initiated an inventory of groundwater chemistry in 53 separate wells. A few dozen of the wells were sampled repeatedly. Wells penetrate unconsolidated sand and gravel and Devonian age sedimentary strata, including the Marcellus shale. We find that deep wells are more concentrated, exhibit higher pH, and greater alkalinity. We could detect 30 to 45 elements in most wells and 15 major and minor elements were detected in all wells. We present a statistical summary of elemental concentrations for all of the wells we have sampled.

Well chemistry provides a baseline for detecting change in a well. We introduce a simple method for comparing wells against each other. The elemental concentrations in one well are plotted against another well, and a power law is fit to the data. The parameters (that is, the coefficient and the exponent) in the power law, along with a measure of the scatter, provide a powerful tool to characterize similarity and uniqueness. When the power law coefficient, correlation coefficient, and power law exponent approach unity, the samples approach identical concentrations. A high degree of similarity implies uniform dilution or concentration for all species being compared. When the exponent approaches unity, the coefficient indicates which sample is more or less concentrated than the other. When the exponent is greater than unity, major elements are more enriched in one well. The correlation coefficient (R2, in this case) measures the scatter around the power law relation. As the correlation coefficient approaches 0, a wide scatter exists, even if the exponent indicates similarity. We apply this method to our wells and discover that individual wells look far more like themselves than any other well. One implication is that groundwater flow paths have characteristic chemical reactions with rocks along their path to the well, and reach a steady state concentration. Temporal variations amount to uniform changes in concentrations across all elements, such as might occur from mixing with very fresh water. Any mixing with non-identical water will yield either more scatter, or values for the power law parameters other than unity. We show that the elements commonly detected in all wells provide a local fingerprint of groundwater.

Handouts
  • Hasbargen Groundwater Identity GSA 2013 post meeting.pdf (1.4 MB)