ABSTRACT

A glacial aquifer system in Woodbury, Connecticut, was studied to identify factors that affect the groundwater quality in the zone of contribution to a community public-supply well. Water samples were collected during 2002–06 from the public-supply well and from 35 monitoring wells in glacial stratified deposits, glacial till, and fractured bedrock. The glacial aquifer is vulnerable to contamination from a variety of sources due to the short groundwater residence times and the urban land use in the contributing recharge area to the public-supply well. The distribution and concentrations of pH, major and trace elements, stable isotope ratios, recharge temperatures, dissolved organic carbon (DOC), and volatile organic compounds (VOCs), and the oxidation-reduction (redox) conditions, were used to identify recharge source areas, aquifer source material, anthropogenic sources, chemical processes, and groundwater-flow paths from recharge areas to the public-supply well, PSW-1.

The major chemical sources to groundwater and the tracers or conditions used to identify them and their processes throughout the aquifer system include (1) bedrock and glacial stratified deposits and till, characterized by high pH and concentrations of sulfate (SO₄²⁻), bicarbonate, uranium (U), radon-222, and arsenic (As) relative to those of other wells, reducing redox conditions, enriched delta sulfur-34 (δ³⁴S) and delta carbon-13 (δ¹³C) values, depleted delta oxygen-18 (δ¹⁸O) and delta deuterium (δD) values, calcite near saturation, low recharge temperatures, and groundwater ages of more than about 9 years; (2) natural organic matter, either in sediments or in an upgradient riparian zone, characterized by high concentrations of DOC or manganese (Mn), low concentrations of dissolved oxygen (DO) and nitrate (NO₃⁻), enriched δ³⁴S values, and depleted δ¹⁸O and δD values; (3) road salt (halite), characterized by high concentrations of sodium (Na), chloride (Cl⁻), and calcium (Ca), and indicative chloride/bromide (Cl:Br) mass concentration ratios; (4) septic-system leachate, characterized by high concentrations of NO₃⁻, DOC, Na, Cl⁻, Ca, and boron (B), delta nitrogen-15 (δ¹⁵N) and δ¹⁸O values, and indicative Cl:Br ratios; (5) organic solvent spills, characterized by detections of perchloroethene (PCE), trichloroethene (TCE), and 1,1-dichloroethene (1,1-DCE); (6) gasoline station spills, characterized by detections of fuel oxygenates and occasionally benzene; and (7) surface-water leakage, characterized by enriched δ¹⁸O and δD values and sometimes high DOC and Mn-reducing conditions. Evaluation of Cl⁻ concentrations and Cl:Br ratios indicates that most samples were composed of mixtures of groundwater and some component of road salt or septic-system leachate. Leachate from septic-tank drainfields can cause locally anoxic conditions with NO₃⁻ concentrations of as much as 19 milligrams per liter (mg/L as N) and may provide up to 15 percent of the nitrogen in water from well PSW-1, based on mixing calculations with δ¹⁵N of NO₃⁻.

Most of the water that contributes to PSW-1 is young (less than 7 years) and derived from the glacial stratified deposits. Typically, groundwater is oxic, but localized reducing zones that result from abundances of organic matter can affect the mobilization of trace elements and the degradation of VOCs. Groundwater from fractured bedrock beneath the valley bottom, which is old (more than 50 years), and reflects a Mn-reducing to methanic redox environment, constitutes as much as 6 percent of water samples collected from monitoring wells screened at the bottom of the glacial aquifer. Dissolved As and U concentrations generally are near the minimum reporting level (MRL) (0.2 micrograms per liter or μg/L and 0.04 μg/L, respectively), but water from a few wells screened in glacial deposits, likely derived from underlying organic-rich Mesozoic rocks, contain As concentrations up to 7 μg/L. At one location, concentrations of As and U were high in sediment sample extractions, and the concentration of dissolved As also was above background levels in the water sample from the well. The relatively low concentrations of U in water samples from this well are consistent with the low mobility of U under iron- (Fe) or Mn-reducing conditions. Experimental data for adsorption batch reactor experiments were modeled using PHREEQC with the generalized two-layer surface complexation model to determine adsorption reactions. A substantial amount of As and U was adsorbed onto four different glacial stratified deposit samples, whereas much less was adsorbed onto the sample of glacial till. In fact, there was a net release of U(VI) to solution by the glacial till, which may result from uranyl-carbonate complexes that inhibit U adsorption, although it is possible that there was some U(VI) co-precipitated with carbonate minerals in the sediment that was released during the adsorption experiment. The net release of U(VI) from glacial till sediments experiments is consistent with higher concentrations of U in water samples from wells screened in deep glacial deposits and in bedrock than those screened in shallow, carbonate-poor glacial deposits. Radon-222 (²²²Rn) activities in groundwater samples from the Woodbury study area ranged from 490 to 13,000 picocuries per liter (pCi/L), and all exceeded the proposed maximum contaminant level (MCL) of 300 pCi/L. The median ²²²Rn activity of 1,100 pCi/L for glacial wells was greater than the median of 440 pCi/L for glacial aquifers of the northeastern United States, and appears to result from an abundant but relatively immobile mass of U and radium-226 (²²⁶Ra) on aquifer sediment surfaces.

Groundwater age plotted as a function of the number of VOC detections indicates that VOC occurrences are related to short groundwater traveltimes, and therefore, young groundwater is more vulnerable to contamination than older groundwater. VOCs were detected at 32 of the 36 wells sampled for this study. Locally high concentrations of gasoline oxygenates (methyl-tert-butyl ether, or MTBE) and chlorinated solvents (PCE, TCE, and 1,1,1-trichloroethane or TCA), and low concentrations of disinfection byproducts were detected in groundwater samples from several wells. Eleven VOCs were detected in untreated water from PSW-1—PCE, TCA, TCE, 1,1-dichloroethane (DCA), cis-1,2-dichloroethene (cis-1,2-DCE), 1,1-dichloroethene (1,1-DCE), MTBE, tert-amyl methyl ether (TAME), ethyl-tert-butyl ether (ETBE), 1,1,2-trichlorotrifluoroethane (CFC-113), and chloroform. Concentrations of TCE in untreated groundwater sampled from PSW-1 frequently exceeded the maximum contaminant level of 5 μg/L. Most VOCs were detected in water sampled from shallow or intermediate depths of the glacial aquifer and were associated with commercial- and light-industrial development. MTBE was detected in 50 percent of the wells and was the most frequently detected compound. Chloroform was found in 49 percent of wells and was the second most frequently detected compound, but concentrations in samples from all but two of the wells sites were less than the MRL. The sources of chloroform were attributed to the use of chlorinated drinking water and the associated recharge from septic-tank drainfields or lawn irrigation, or for well disinfection. Chlorinated solvents were the third most frequently detected group of compounds; two of them—TCE and PCE—exceeded health-based screening concentrations. PCE was detected in 42 percent of wells, and concentrations were highest (up to 11 μg/L) in samples from several wells downgradient from the commercial development. TCE was found in 33 percent of wells, at concentrations as high as 25 μg/L. Pesticides were detected at low concentrations in few wells and this reflects the small amount of agricultural land use in the study area. Atrazine, prometon, simazine, and fipronil were detected at concentrations near or less than the MRL in monitoring wells downgradient from urban areas.

First posted December 2009

Revised February 10, 2010

Revised February 17, 2010