Groundwater quality in relation to drinking water health standards and geochemical characteristics for 54 domestic wells in Clinton County, Pennsylvania, 2017

Despite the reliance on groundwater by approximately 2.4 million rural Pennsylvania residents, publicly available data to characterize the quality of private well water are limited. As part of a regional effort to characterize groundwater in rural areas of Pennsylvania, samples from 54 domestic wells in Clinton County were collected and analyzed in 2017. The samples were evaluated for a wide range of constituents and compared to drinking-water health standards and geochemical characteristics. The sampled wells were completed to depths ranging from 46 to 500 feet in bedrock that was of predominantly sandstone, shale, or carbonate lithology. Results of this study show that the sampled groundwater quality in Clinton County generally met most drinking-water standards that apply to public water supplies. However, a percentage of samples exceeded drinking-water maximum contaminant levels (MCLs) for total coliform bacteria (57.4 percent), Escherichia coli (E. coli) (25.9 percent), nitrate (1.9 percent), and arsenic (1.9 percent); and secondary maximum contaminant levels (SMCLs) for pH (31.5 percent), manganese (29.6 percent), iron (13 percent), total dissolved solids (7.4 percent), aluminum (1.9 percent), and chloride (1.9 percent). Sodium concentrations exceeded the U.S. Environmental Protection Agency drinking-water advisory recommendation in 16.7 percent of the samples. Radon-222 activities exceeded the proposed drinking-water standard of 300 picocuries per liter (pCi/L) in 59.3 percent of the samples. The only volatile organic compounds (VOCs) detected were acetone and methyl ethyl ketone in two separate samples; neither constituent exceeded drinking-water standards.

Higher median nitrate concentrations were found in the carbonate (3.26 milligrams per liter [mg/L]) versus shale (less than 0.04 mg/L) and sandstone (0.27 mg/L) aquifer subsets. Most of the elevated nitrate concentrations were associated with E. coli detections in the carbonate aquifers, where transmissive bedrock can facilitate groundwater contamination by human activities at the land surface.

The median pH of groundwater from the sandstone aquifers (6.53) was less than those for the shale aquifers (7.31) and carbonate aquifers (7.43). Generally, the lower pH samples had greater potential for elevated concentrations of dissolved metals, including beryllium, copper, lead, nickel, and zinc, whereas the higher pH samples had greater potential for elevated concentrations of total dissolved solids, sodium, fluoride, boron, and uranium. Near-neutral samples (pH 6.5 to 7.5) had greater hardness and alkalinity concentrations than other samples with pH outside this range. Many samples from the shale or sandstone aquifers, particularly those with pH less than 6.5, were identified as having serious potential corrosivity based on the combination of the calcite saturation index and the chloride to sulfate mass ratio; however, none of the samples from the carbonate aquifers was identified as seriously corrosive.

Groundwater from 3.7 percent of the wells had concentrations of methane greater than the Pennsylvania action level of 7 mg/L, and 48 of the 54 wells (88.9 percent) had detectable concentrations of methane greater than the 0.0002 mg/L detection limit. Greater methane concentrations were found more frequently in groundwater sampled from the shale aquifers than the carbonate or sandstone aquifers in the study area. Most of the samples containing elevated methane (greater than 0.2 mg/L) were located outside the area of the Appalachian Plateaus. The elevated concentrations of methane generally were associated with suboxic groundwater (dissolved oxygen less than 0.5 mg/L) that had near-neutral to alkaline pH and were correlated with concentrations of iron, manganese, ammonia, sodium, lithium, barium, fluoride, and boron. The stable carbon and hydrogen isotopic compositions of methane in two of four samples analyzed for isotopes were consistent with compositions reported for mud-gas logging samples from gas-bearing geologic units (thermogenic gas) in the Appalachian Plateaus region, whereas two others were consistent with methane of microbial origin or a mixture of microbial and thermogenic gas.

Forty-two percent of samples had chloride concentrations greater than 20 mg/L with variable bromide concentrations. Corresponding chloride/bromide ratios are consistent with low-bromide sources such as road-deicing salt and septic effluent or animal waste, or, in a few cases, high-bromide brine. Brines characterized by relatively high bromide are naturally present in deeper parts of the regional groundwater system and, in some cases, may be mobilized by gas drilling. The chloride, bromide, and other constituents in road-deicing salt or brine solutions tend to be diluted by mixing with fresh groundwater in shallow aquifers used for water supply. One of the four groundwater samples with methane concentrations greater than 4 mg/L had chloride and bromide concentrations and a chloride/bromide ratio that indicates mixing with a salinity source such as road-deicing salt, whereas the chloride and bromide concentrations and ratios for the other three high-methane samples indicate mixing with a small amount of brine (0.03 percent or less). In two other eastern Pennsylvania county studies where gas drilling is absent, groundwater with comparable chloride/bromide ratios, bromide, and chloride concentrations plus other element associations have been reported. Additional sampling and analysis, such as isotopic analysis of the dissolved gas, fracture analysis, and more detailed evaluation of surrounding land uses, may be warranted to better understand the origin of the methane and brine constituents in groundwater at specific locations.