Abstract

In 2001, the National Water-Quality Assessment (NAWQA) Program of the U.S. Geological Survey initiated a series of studies on the transport of anthropogenic and natural contaminants (TANC) to public-supply wells (PSWs). The main goal of the TANC project was to better understand the source, transport, and receptor factors that control contaminant movement to PSWs in representative aquifers of the United States. Regional- and local-scale study areas were selected from within existing NAWQA study units, including the south-central Texas Edwards aquifer. The local-scale TANC study area, nested within the regional-scale NAWQA study area, is representative of the regional Edwards aquifer. The PSW selected for study is within a well field of six production wells. Although a single PSW was initially selected, because of constraints of well-field operation, samples were collected from different wells within the well field for different components of the study. Data collected from all of the well-field wells were considered comparable because of similar well construction, hydrogeology, and geochemistry. An additional 38 PSWs (mostly completed in the confined part of the aquifer) were sampled throughout the regional aquifer to characterize water quality. Two monitoring well clusters, with wells completed at different depths, were installed to the east and west of the well field (the Zarzamora and Timberhill monitoring well clusters, respectively). One of the monitoring wells was completed in the overburden to evaluate potential hydrologic connectivity with the Edwards aquifer. Geophysical and flowmeter logs were collected from one of the well-field PSWs to determine zones of contribution to the wellbore. These contributing zones, associated with different hydrogeologic units, were used to select monitoring well completion depths and groundwater sample collection depths for depth-dependent sampling. Depth-dependent samples were collected from the PSW from three different depths and under three different pumping conditions. Additionally, selected monitoring wells and one of the well-field PSWs were sampled several times in response to a rainfall and recharge event to assess short-term (event-scale) temporal variations in water quality. For comparison purposes, groundwater samples were categorized as being from regional aquifer PSWs, from the well field (wellhead samples), from the monitoring wells (excluding the overburden well), from the overburden well, from the PSW depth-dependent sampling, and from temporal sampling. Groundwater samples were analyzed for inorganic, organic, isotopic, and age-dating tracers to characterize geochemical conditions in the aquifer and provide understanding of the mechanisms of mobilization and movement of selected constituents from source areas to a PSW. Sources, tracers, and conditions used to assess water quality and processes affecting the PSW and the aquifer system included (1) carbonate host rock composition; (2) physicochemical constituents; (3) major and trace element concentrations; (4) saturation indices with respect to minerals in aquifer rocks; (5) elemental ratios, such as magnesium to calcium ratios, that are indicative of water-rock interaction processes; (6) oxidation-reduction conditions; (7) nutrient concentrations, in particular nitrate concentrations; (8) the isotopic composition of nitrate, which can point to specific nitrate sources; (9) strontium isotopes; (10) stable isotopes of hydrogen and oxygen; (11) organic contaminant concentrations, including pesticides and volatile organic compounds; (12) age tracers, apparent-age distribution, and dissolved gas data used in age interpretations; (13) depth-dependent water chemistry collected from the PSW under different pumping conditions to assess zones of contribution; and (14) temporal variability in groundwater composition from the PSW and selected monitoring wells in response to an aquifer recharge event.

Geochemical results indicate that the well-field and monitoring well samples were largely representative of groundwater in the regional confined aquifer. Constituents of concern in the Edwards aquifer for the long-term sustainability of the groundwater resource include the nutrient nitrate and anthropogenic organic contaminants. Nitrate concentrations (as nitrogen) for regional aquifer PSWs had a median value of 1.9 milligrams per liter, which is similar to previously reported values for the regional aquifer. Nitrate-isotope compositions for groundwater samples collected from the well-field PSWs and monitoring wells had a narrow range, with values indicative of natural soil organic values. A comparison with historical nitrate-isotope values, however, suggests that a component of nitrate in groundwater from biogenic sources might have increased over the last 30 years. Several organic contaminants (the pesticide atrazine, its degradate deethylatrazine, trichloromethane (chloroform; a drinking-water disinfection byproduct), and the solvent tetrachloroethene (PCE)) were widely distributed throughout the regional aquifer and in the local-scale TANC study area at low concentrations (less than 1 microgram per liter). Higher concentrations of PCE were detected in samples from the well-field PSWs and Zarzamora monitoring wells relative to the regional aquifer PSWs. The urban environment is a likely source of contaminants to the aquifer, and these results indicate that one or more local urban sources might be supplying PCE to the Zarzamora monitoring wells and the well-field wells. Samples from the well field also had high concentrations of chloroform relative to the monitoring wells and regional aquifer PSWs. For samples from the regional aquifer PSWs, the most frequently detected organic contaminants generally decreased in concentration with increasing well depth. Deeper wells might intercept longer regional flow paths with higher fractions of older water or water recharged in rural recharge areas in the western part of the aquifer that have been less affected by anthropogenic contaminants. A scenario of hypothetical contaminant loading was evaluated by using results from groundwater-flow-model particle tracking to assess the response of the aquifer to potential contamination. Results indicate that the aquifer responds quickly (less than 1 year to several years) to contaminant loading; however, it takes a relatively long time (decades) for concentrations to reach peak values. The aquifer also responds quickly (less than 1 year to several years) to the removal of contaminant loading; however, it also takes a relatively long time (decades) to reach near background concentrations.

Interpretation of geochemical age tracers in this well-mixed karst system was complicated by contamination of a majority of measured tracers and complexities of extensive mixing. Age-tracer results generally indicated that groundwater samples were composed of young, recently recharged water with piston-flow model ages ranging from less than 1 to 41 years, with a median of 17 years. Although a piston-flow model is typically not valid for karst aquifers, the model ages provide a basis for comparing relative ages of different samples and a reference point for more complex hydrogeologic models for apparent-age interpretations. Young groundwater ages are consistent with particle-tracking results from hydrogeologic modeling for the local-scale TANC study area. Age-tracer results compared poorly with other geochemical indicators of groundwater residence time and anthropogenic effects on water quality, indicating that hydrogeologic conceptual models used in groundwater age interpretations might not adequately account for mixing in this karst system. Groundwater samples collected from the well field under a variety of pumping conditions were relatively homogeneous and well mixed for numerous geochemical constituents (with the notable exception of age tracers). Groundwater contributions to the PSW were dominated by well-mixed, relatively homogeneous groundwater, typical of the regional confined aquifer. Zones of preferential flow were determined for the PSW, but groundwater samples from different stratigraphic units were not geochemically distinct. Variations in chemical constituents in response to a rainfall and aquifer recharge event occurred but were relatively minor in the PSW and monitoring wells. This observation is consistent with the hypothesis that the response to individual recharge events in the confined aquifer, unless intersecting conduit flow paths, might be attenuated by mixing processes along regional flow paths. Results of this study are consistent with the existing conceptual understanding of aquifer processes in this karst system and are useful for water-resource development and management practices.

First posted November 10, 2011