The effects of highway-deicer application on ground-water quality were studied at a site in northwestern Indiana using a variety of geochemical indicators. Site characteristics such as high snowfall rates; large quantities of applied deicers; presence of a high-traffic highway; a homogeneous, permeable, and unconfined aquifer; a shallow water table; a known ground-water-flow direction; and minimal potential for other sources of chloride and sodium to complicate source interpretation were used to select a study area where ground water was likely to be affected by deicer application. Forty-three monitoring wells were installed in an unconfined sand aquifer (the Calumet aquifer) near Beverly Shores in northwestern Indiana. Wells were installed along two transects that approximately paralleled groundwater flow in the Calumet aquifer and crossed US–12. US–12 is a highway that receives Indiana’s highest level of maintenance to maintain safe driving conditions. Ground-water quality and water-level data were collected from the monitoring wells, and precipitation and salt-application data were compiled from 1994 through 1997.

The water-quality data indicated that chloride was the most easily traced indicator of highway deicers in ground water. Concentration ratios of chloride to iodide and chloride to bromide and Stiff diagrams of major element concentrations indicated that the principal source of chloride and sodium in ground water from the uppermost one-third to one-half of the Calumet aquifer and downgradient from US–12 was from a halite highway-deicer source. Borehole logs of relative electromagnetic conductivity defined a distinct plume of deicer-affected water in the uppermost 8 feet of aquifer at about 9 feet horizontally from the paved roadway edge and a zone of higher conductivity than background in the lower one-third of the aquifer. Chloride and sodium in the deep parts of the aquifer originated from natural sources.

Chloride and sodium from highway deicers were present in the aquifer throughout the year. The highest concentrations of chloride and sodium in ground water were determined in samples collected during the spring and summer from wells open to the water table within about 9 feet of the highway. Chloride concentrations in ground water that were attributable to highway deicers also were found in tested wells about 400 feet downgradient from US–12 during the fall and winter and at greater depths than in wells closer to US–12.

Chloride concentrations exceeded the U.S. Environmental Protection Agency’s (USEPA) secondary maximum contaminant level of 250 milligrams per liter for drinking water at seven wells downgradient from the highway during late winter, spring, and summer samplings. The chloride standard was exceeded only in water from wells with total depths that are less than about 10 feet below land surface. Sodium concentrations in water periodically exceeded the USEPA drinking-water equivalency level of 20 milligrams per liter in both the uppermost (deicer 2 Effects of Highway-Deicer Application on Ground-Water Quality in a Part of the Calumet Aquifer, Northwestern Indiana affected) and lower one-thirds of the aquifer. Sodium concentrations in ground water downgradient from US–12 and in the upper 5 feet of the aquifer also occasionally exceeded drinking-water standards for sodium (160 milligrams per liter) as set by the State of Florida and a standard for taste (200 milligrams per liter) as set by the World Health Organization.

Dispersion was identified by analysis of aquifer-test data, isotopic dating of ground water, and water-quality data to be the process most responsible for reducing concentrations of highway deicers in the aquifer. Chemical analyses of the sand composing the aquifer indicated that cation exchange decreased the mass of deicerrelated sodium in ground water, although the sand has a limited capacity to sustain the process.

Automated daily measurements of specific conductance, correlated to chloride concentrations, indicated that some deicer is retained in the aquifer near the highway throughout the entire year and acts as a continuous chloride source for ground water. Peak concentrations of deicerrelated constituents occasionally were detected by the daily, automated measurements of specific conductance that were made between the monthly samplings of ground water. Data analysis indicated that more frequent sampling than monthly intervals would be necessary if maximum chloride concentrations were to be measured.

Some deicer may be retained in the aquifer and unsaturated zone between annual salt-application periods. Chloride concentrations at wells 1-DG-WT and 2-DG-WT remained greater than background through much or all of the year. The estimated masses of chloride transported in ground water past 2-DG in 1995 and 1996 were either slightly greater than (1995) or less than (1996) the masses of chloride applied to US–12 during the study.