Hexavalent and total chromium at low reporting concentrations in source-water aquifers and surface waters used for public supply in Illinois, 2013

Abstract

On the basis of their recent review of the human health effects of hexavalent chromium [Cr(VI)] in public drinking water, the U.S. Environmental Protection Agency is considering the need for Federal regulation of Cr(VI). Presently, only total chromium is regulated, at a Maximum Contaminant Level (MCL) of 100 micrograms per liter (µg/L). The occurrence of Cr(VI) in groundwater and surface waters generally is attributed to industrial sources, but can be of natural origin. California’s recently established MCL for Cr(VI) of 10 µg/L illustrates the drinking-water concerns associated with Cr(VI). To improve understanding of the possible impact of a Cr(VI)-specific standard that approximates the California level on the management of Illinois’ public drinking water, the U.S. Geological Survey, in cooperation with the Illinois Environmental Protection Agency, assessed the occurrence and distribution of Cr(VI) in the State’s public-water supplies. During 2013, untreated water samples were collected to be analyzed for Cr(VI) and total chromium [Cr(T)] at 119 water-supply wells and 32 surface-water intakes; also, 32 treated surface-water samples were collected near the point of treatment and 32 near the furthest point of distribution. Public-supply sample sites were selected by a stratified random method. Samples typically were analyzed within 24 hours of collection at reporting limits of 0.02 µg/L for Cr(VI) and 0.1 µg/L for Cr(T). The occurrence of Cr(VI) was compared with selected geophysical, physical, and sampling factors that might more fully explain its distribution and magnitude of concentrations. The maximum concentration of Cr(VI) in groundwater was 2.1 µg/L. Maximum concentrations in untreated and treated surface water were 0.29 µg/L and 2.4 µg/L, respectively. All sample concentrations were below the California MCL; only 35 percent were below that State’s non-enforceable public health goal of 0.02 of µg/L. Cr(VI) was undetected in 43 percent of untreated groundwater samples, with a median of 0.06 µg/L when detected. All but two (94 percent) of untreated surface-water samples had detections. In untreated surface water, the median concentration was 0.09 µg/L, whereas in treated (tap and distributed) water the median was 0.20 µg/L. Surface waters treated with lime for softening typically had the greatest Cr(VI) concentrations (maximum, 2.4 µg/L; median, 1.2 µg/L). The maximum concentration of Cr(T) in groundwater was 1.8 µg/L. Maximum concentrations in untreated and treated surface water were 1.8 µg/L and 2.5 µg/L, respectively. All sample concentrations were below the Federal MCL. Total chromium was detected in 65 percent of untreated groundwater samples, with a median of 0.40 µg/L, when detected. All but one (97 percent) of untreated surface-water samples had detections. In untreated surface water, the median concentration was 0.40 µg/L, whereas in treated (tap and distributed) water the median was 0.30 µg/L. As with Cr(VI), surface waters treated with lime typically had the greatest Cr(T) concentrations. Examination of factors that might account for or be associated with the occurrence of Cr(VI) in public-supply source waters found few clearly evident factors. Associations in frequencies of occurrence and range of concentrations indicate that surface waters and groundwaters of shallow, unconsolidated, unconfined aquifers, particularly alluvial aquifers, are possibly most commonly affected by anthropogenic sources of Cr(VI). Groundwaters of deep (greater than 500 feet) bedrock aquifers, particularly the Cambrian-Ordovician aquifer system, are possibly most commonly affected by geologic sources of Cr(VI). Additional study, with supporting geologic and geochemical data that were not collected in this study, would be necessary to verify these associations. There was a weak positive relation (ρ = 0.23) between concentrations of Cr(VI) and Cr(T) in untreated water samples, with a much stronger positive relation (ρ = 0.86 and ρ = 0.90, respectively) in samples collected soon after treatment and near the endpoint of distribution. The stronger relation and greater similarity between Cr(VI) and Cr(T) concentrations in treated water samples indicate that Cr(VI) represents a greater proportion of the measured concentrations of Cr(T) in treated waters than in untreated waters. The analysis of spikes and other quality-assurance samples indicate uncertainties associated with obtaining or confirming consistently accurate analytical results for Cr(VI) at near the applied reporting limit of 0.02 µg/L.