Northeastern Section - 42nd Annual Meeting (12–14 March 2007)

Paper No. 8
Presentation Time: 3:40 PM

NITRATE CONTAMINATION OF A BEDROCK AQUIFER IN CENTRAL VERMONT


KIM, Jonathan, Vermont Geological Survey, 103 South Main Street, Logue Cottage, Waterbury, VT 05671-2420 and COMSTOCK, Jeff, Vermont Agency of Agriculture, Food, and Markets, 116 State Street, Montpelier, VT 05620, jon.kim@state.vt.us

Since ~2000, elevated nitrate levels in groundwater from domestic bedrock wells in the vicinity of a large dairy farm in central Vermont have been a problem. A comprehensive geological, hydrogeological, and geochemical study was conducted in the farm vicinity by the Vermont Geological Survey and Vermont Agency of Agriculture with funding from EPA-Section 319 during 2005. This six month+ study integrated bedrock mapping and structural analysis, borehole camera surveys, monitor well installation, analysis of bedrock aquifer flow directions, EMI geophysical surveys, and sampling of groundwater from wells to analyze for nitrates, herbicides, major and trace elements, nitrogen and oxygen isotopes, fluorescent dye tracers, and CFC ages.

By analyzing nitrate levels in the groundwater from the bedrock wells over time, we delineated groups of wells with similar nitrate patterns. Groundwater flow directions calculated from static water levels indicate multiple flow domains in the bedrock aquifer that are controlled by brittle and ductile bedrock structures. Borehole camera surveys show that a single foliation-parallel zone contributes groundwater to each domestic well. In combination with EMI surveys, the configuration of flow domains can explain the distribution of well groups with different nitrate patterns. Herbicide patterns are similar to nitrate patterns for some wells.

Nitrogen and oxygen isotope data show that all wells fall in the soil nitrogen field on nitrate source discrimination diagrams. Combined CFC113 and Tritium/Helium groundwater dating techniques yielded ages ranging from 13-22 years. CFC11 and CFC12 ages could not be determined because of an excess local source(s) of these compounds. Three distinct fluorescent dyes were introduced at widely-spaced locations, but only one of these dyes reached a well during the study. We were able to fingerprint the groundwater from two wells by integrating major and trace element geochemical trends with physical parameters such as pH, Eh, DO, Temp., and Conductivity.