Abstract

Artificial recharge of the Equus Beds aquifer using runoff from the Little Arkansas River in south-central Kansas was first proposed in 1956 and was one of many options considered by the city of Wichita to preserve its water supply. Declining aquifer water levels of as much as 50 feet exacerbated concerns about future water availability and enhanced migration of saltwater into the aquifer from past oil and gas activities near Burrton and from the Arkansas River. Because Wichita changed water-management strategies and decreased pumping from the Equus Beds aquifer in 1992, water storage in the aquifer recovered by about 50 percent. This recovery is the result of increased reliance on Cheney Reservoir for Wichita water supply, decreased aquifer pumping, and larger than normal precipitation. Accompanying the water-level recovery, the average water-level gradient in the aquifer decreased from about 12 feet per mile in 1992 to about 8 feet per mile in January 2006.

An important component of artificial recharge is the water quality of the receiving aquifer and the water being recharged (source water). Water quality within the Little Arkansas River was defined using data from two real-time surface-water-quality sites and discrete samples. Water quality in the Equus Beds aquifer was defined using sample analyses collected at 38 index sites, each with a well completed in the shallow and deep parts of the Equus Beds aquifer. In addition, data were collected at diversion well sites, recharge sites, background wells, and prototype wells for the aquifer storage and recovery project. Samples were analyzed for major ions, nutrients, trace metals, radionuclides, organic compounds, and bacterial and viral indicators.

Water-quality constituents of concern for artificial recharge are those constituents that frequently (more than 5 percent of samples) may exceed Federal [U.S. Environmental Protection Agency (USEPA)] and State drinking-water criteria in water samples from the receiving aquifer or in samples from the source water. Constituents of concern include major ions (sulfate and chloride), nutrients (nitrite plus nitrate), trace elements (arsenic, iron, and manganese), organic compounds (atrazine), and fecal bacterial indicators. This report describes the water quality in the Equus Beds aquifer and the Little Arkansas River from 1995 through 2005 before implementation of large-scale recharge activities.

Sulfate concentrations in water samples from the Little Arkansas River rarely exceeded Federal secondary drinking water regulation (SDWR) of 250 milligrams per liter (mg/L). Sulfate concentrations in groundwater were exceeded in about 18 percent of the wells in the shallow (less than or equal to 80 feet deep) parts of the aquifer and in about 13 percent of the wells in the deep parts the aquifer. Larger sulfate concentrations were associated with parts of the aquifer with the largest water-level declines. Water-quality changes in the Equus Beds aquifer likely were caused by dewatering and oxidation of aquifer material that subsequently resulted in increased sulfate concentrations as water levels recovered.

The primary sources of chloride to the Equus Beds aquifer are from past oil and gas activities near Burrton and from the Arkansas River. Computed chloride concentrations in the Little Arkansas River near Halstead exceeded the Federal SDWR of 250 mg/L about 27 percent of the time (primarily during low-flow conditions). Chloride concentrations in groundwater exceeded 250 mg/L in about 8 percent or less of the study area, primarily near Burrton and along the Arkansas River. Chloride in groundwater near Burrton has migrated downgradient about 3 miles during the past 40 to 45 years. The downward and horizontal migration of the chloride is controlled by the hydraulic gradient in the aquifer, dispersion of chloride, and discontinuous clay layers that can inhibit further downward migration. Chloride in the shallow parts of the Equus Beds aquifer migrated less than 0.5 mile during the past decade. Migration is slower because of the decrease in the hydraulic gradient since 1992. On the basis of these results, artificial recharge (especially at depths of 100 to 150 feet) could create an effective barrier to saltwater migration.

Nutrients, such as nitrite plus nitrate (hereinafter referred to as nitrate), are a water-quality concern because of the predominantly agricultural land use in the 150-square-mile study area. All nitrate concentrations in water samples collected at the two surface-water monitoring sites on the Little Arkansas River from 1995 through 2005 were less than the Federal maximum contaminant level (MCL) of 10 mg/L for nitrate. Groundwater sampling results indicated that average nitrate concentrations exceeding the MCL were detected in 13 percent of the wells in 9 percent of the shallow parts of the aquifer in the study area. Little nitrate is present in the deeper parts of the aquifer because of chemical reducing conditions.

Several trace elements frequently exceeded drinkingwater criteria, including arsenic, iron, and manganese. Computed arsenic concentrations in the Little Arkansas River exceeded the Federal drinking-water MCL of 10 micrograms per liter (μg/L) about 14 percent of the time primarily during low-flow conditions. In shallow groundwater, average arsenic concentrations exceeded the MCL in 10 percent of the wells (6 percent of the study area), whereas at depths of more than 80 feet, average arsenic concentrations exceeded the MCL in 34 percent of the wells (35 percent of the study area). In the Little Arkansas River, dissolved iron concentrations exceeded the Federal SDWR of 300 μg/L in 2 percent of water samples, and manganese concentrations exceeded the SDWR of 50 μg/L in about half of the samples collected. In shallow parts of the aquifer, average iron concentrations exceeded the SDWR of 300 μg/L in 44 percent of the study area, and average manganese concentrations exceeded the SDWR of 50 μg/L in 60 percent of the study area. In deep parts of the aquifer, average iron concentrations exceeded the SDWR in 44 percent of the study area, and manganese concentrations exceeded the SDWR in 97 percent of the area.

The areal distribution of larger dissolved arsenic, iron, and manganese concentrations were similar. Larger naturallyoccurring concentrations of arsenic, iron, and manganese in groundwater are associated with more reducing conditions, areas where more clay is present in the aquifer material, and areas that had large water-level declines and subsequent recovery. Effects of artificial recharge on natural dissolved concentrations of arsenic in the aquifer potentially can be minimized by maintaining the oxidation-reduction potential as near 1995–2005 baseline conditions as possible. However, in many areas of the aquifer, especially the deeper parts, the natural geochemical conditions are conducive to large arsenic concentrations. It may be possible to use artificial recharge of oxygenated water to create a less reducing geochemical environment, decreasing some of the arsenic and iron dissolved in the water, potentially improving the overall water quality in the aquifer.

Atrazine was the most commonly detected organic compound in the study area. The Federal MCL for atrazine in drinking water is 3 μg/L an annual average. Computed concentrations of atrazine in the Little Arkansas River exceeded the Federal MCL value of 3.0 μg/L about 27 percent of the time, mostly during the late spring to early fall. Atrazine was detected in about 55 percent of the samples collected from shallow wells, which indicates infiltration from field applications to the shallow groundwater, but concentrations were much less than the MCL.

Large concentrations of coliform bacterial indicators (total coliform, fecal coliform, and Escherichia coli) were detected in all water samples from the Little Arkansas River. These large bacterial indicator densities are typical in central and eastern Kansas streams, especially during runoff conditions. Total coliform detections exceeded the USEPA Federal Maximum Contaminant Level Goal (MCLG) of 0 colonies in water samples from 95 percent of the shallow index wells and in 87 percent of the deep wells in the Equus Beds aquifer. Many of these detections were in the first samples collected from the wells after they were developed, indicating that at least some of these detections may be related to drilling. Almost all wells sampled for this study had at least 1 sample with a total coliform detection; however, the median densities for most of these wells were less than 1 colony per 100 milliliter (col./100 mL). Viral indicators (Clostridium perfringens and E. coli coliphage) were present in samples from the Little Arkansas River during storm runoff but were not detected in any samples of groundwater. These data indicated that natural infiltration of water through the soil removes most bacterial and viral indicator organisms.

Water quality in surface water and groundwater is controlled by the geology of the underlying bedrock and aquifer materials, the hydraulic permeability (porosity) and geochemical (oxidation and reduction) properties of the aquifer, and the effects of humans related to past oil and gas activities and agriculture. When the proposed full-scale artificial recharge of the Equus Beds aquifer is implemented, changes in concentrations of water-quality constituents are expected. The increased water levels from artificial recharge are expected to slow the saltwater migration from the northwest and south of the study area, potentially limiting further chloride migration and improving the quality of water in the aquifer. Continued monitoring and interpretation of these recharge water-quality data relative to drinking-water criteria will help ensure the usable quality of water in the Equus Beds aquifer.

First posted April 13, 2010