Concentrations of boron greater than the U.S. Environmental Protection Agency (USEPA) 900 μg/L removal action level (RAL) standard were detected in water sampled by the USEPA in 2004 from three domestic wells near Beverly Shores, Indiana. The RAL regulates only human-affected concentrations of a constituent. A lack of well logs and screened depth information precluded identification of whether water from sampled wells, and their boron sources, were from human-affected or natural sources in the surficial aquifer, or associated with a previously defined natural, confined aquifer source of boron from the subtill or basal sand aquifers. A geochemically-based classification of the source of boron in ground water could potentially determine the similarity of boron to known sources or mixtures between known sources, or classify whether the relative age of the ground water predated the potential sources of contamination. The U.S. Geological Survey (USGS), in cooperation with the USEPA, investigated the use of a geochemical method that applied boron stable isotopes, and concentrations of boron, tritium, and other constituents to distinguish between natural and human-affected sources of boron in ground water and thereby determine if the RAL was applicable to the situation.

Boron stable-isotope ratios and concentrations of boron in 17 ground-water samples and tritium concentrations in 9 ground-water samples collected in 2004 were used to identify geochemical differences between potential sources of boron in ground water near Beverly Shores, Indiana. Boron and δ¹¹B analyses for this investigation were made on unacidified samples to assure consistency of the result with unacidified analyses of δ¹¹B values from other investigations. Potential sources of boron included surficial-aquifer water affected by coal-combustion products (CCP) or domestic-wastewater, upward discharge of ground water from confined aquifers, and unaffected water from the surficial aquifer that was distant from human-affected boron sources.

Boron concentrations in potential ground-water sources of boron were largest (15,700 to 24,400 μg/L) in samples of CCP-affected surficial aquifer water from four wells at a CCP landfill and smallest (27 to 63 μg/L) in three wells in the surficial aquifer that were distant from human-affected boron sources. Boron concentrations in water from the basal sand aquifer ranged from 656 μg/L to 1,800 μg/L. Boron concentrations in water from three domestic-wastewater-affected surficial aquifer wells ranged from 84 to 387 μg/L. Among the representative ground-water samples, boron concentrations from all four samples of CCP-affected surficial aquifer water and four of five samples of water from the basal sand aquifer had concentrations greater than the RAL. A comparison of boron concentrations in acid-preserved and unacidified samples indicated that boron concentrations reported for this investigation may be from about 11 to 16 percent less than would be reported in a standard analysis of an acidified sample.

The stable isotope boron-11 was most enriched in comparison to boron-10 in ground water from a confined aquifer, the basal sand aquifer (δ¹¹B, 24.6 to 34.0 per mil, five samples); it was most depleted in CCP-affected water from the surficial aquifer (δ¹¹B, 0.1 to 6.6 per mil, four samples). Domestic-wastewater-affected water from the surficial aquifer (δ¹¹B, 8.7 to 11.7 per mil, four samples) was enriched in boron-11, in comparison to individual samples of a borax detergent additive and a detergent with perborate bleach; it was intermediate in composition between basal sand aquifer water and CCP-affected water from the surficial aquifer. The similarity between a ground-water sample from the surficial aquifer and a hypothetical mixture of unaffected surficial aquifer and basal sand aquifer waters indicates the potential for long-term upward discharge of ground water into the surficial aquifer from one or more confined aquifers. Estimated δ¹¹B values for acidified samples were depleted by 1.9 to 2.8 per mil in comparison to unacidified samples from the four wells sampled; those differences were small in comparison to the differences between δ¹¹B values of representative sources of boron in ground water.

Tritium concentrations ranged from 7.0 to 10.3 tritium units in six samples from the surficial aquifer and were less than 0.8 tritium units in three samples from the basal sand aquifer. Water from wells in the surficial aquifer represents predominantly modern, post-1972 recharge and sources of boron and other constituents. Water from the basal sand aquifer is associated with pre-1952 recharge from sources not affected by local boron inputs.

Ground water from six wells (five domestic wells and one public-supply well) where the ground-water source was unknown had boron concentrations, boron isotope ratios, and tritium concentrations similar to water from the basal sand aquifer. Boron concentrations greater than the RAL were found in water from four of these six wells. The boron isotope and tritium data from these four wells indicate a natural source of boron in ground water; therefore, the RAL does not apply to boron concentrations in water from these wells. Water samples from two domestic wells where the ground-water source was unknown had boron concentrations less than the RAL and boron isotope ratios and tritium concentrations that were similar to domestic-wastewater-affected water from the surficial aquifer. The boron isotope ratio for a sample from one domestic well was similar to that of CCP-affected water from the surficial aquifer and detergent compositions; the boron concentration of that sample was less than the RAL. The classifications of differences among representative sources of boron in ground water and water samples from wells where the ground-water source was unknown generally agreed with distinctions based on strontium-87/strontium-86 ratios and concentrations of strontium, chloride, nitrate, and ammonia. This application of boron concentrations, boron isotope ratios, and tritium concentrations to classify differences in relation to potential sources of boron in ground water was able to distinguish between boron from natural sources and from human-affected sources that are subject to regulation.