USGS Publications Warehouse

Abstract

In 1994, the U.S. Geological Survey, in cooperation with the U.S. Environmental Protection Agency and the City of Elkhart, developed a ground-water model of the Elkhart, Indiana, area to determine the avail-ability and source of water at potential new well fields. The modeled area covered 190 square miles of northwestern Elkhart County and a small part of southern Michigan. Three Superfund sites and several other sites in this area are undergoing environmental cleanup. The model would be used to guide the location of well fields so that Superfund sites and environmental cleanup areas would not be within recharge areas for the well fields. The City of Elkhart obtains its water supply from two aquifers separated by a generally continuous confining unit. The upper aquifer is composed primarily of sand and gravel of glacial origin. Thickness of the upper aquifer ranges from 0 to 116 feet and averages 47 feet. The lower aquifer is composed of sand and gravel with interbedded lenses of silt and clay. Thickness of the lower aquifer ranges from 1 to 335 feet and averages 35 feet. The intervening confining unit is composed of silt and clay with interbedded sand and gravel; the confining unit ranges from 0 to 177 feet, with an average thickness of 27 feet. Flow through the aquifers is generally horizontal vertically downward from the upper aquifer, through the confining unit, and into the lower aquifer, except where flow is vertically upward at the St. Joseph River and other large streams. The hydraulic characteristics of the aquifers and confining unit were estimated by analyzing aquifer-test data from well drillers? logs and by calibration of the model. The horizontal hydraulic conductivity of the upper aquifer is 170 feet per day within about 1 mile of the St. Joseph and Elkhart Rivers and 370 feet per day at distances greater than about 1 mile. The horizontal hydraulic conductivity of the lower aquifer is 370 feet per day throughout the modeled area, with the exception of an area near the center of the modeled area where the horizontal hydraulic conductivity is 170 feet per day. Transmissivity of the lower aquifer increases generally from southwest to northeast; transmissivity values range from near 0 where the lower aquifer is absent to 57,000 square feet per day and average about 8,100 square feet per day. The vertical hydraulic conductivity of the confining unit is 0.07 feet per day; the vertical conductivity of the streambeds commonly is 1.0 foot per day and ranges from 0.05 foot per day to 50 feet per day. The areal recharge rate to the outwash deposits was determined by a base-flow separation technique to be 16 inches per year, and the areal recharge rate to the till was assumed to be 4 inches per year. A two-layer digital model was used to simulate flow in the ground-water system. The model was calibrated on the basis of historical water-use data, water-level records, and gain/loss data for streams during May and June 1979. The model was recalibrated with water-use data and water-level records from 1988. For 1979 data, 49 percent of the inflow to the model area is from precipitation and 46 percent is ground-water inflow across the model boundaries. Most of the ground-water inflow across the model boundary is from the north and east, which corresponds to high values of transmissivity?as high as 57,000 feet squared per day?in the model layers in the northern and eastern areas. Eighty-two percent of the ground-water discharge is to the streams; 5 percent of the ground-water discharge is to wells. Source areas and flow paths to the City of Elkhart public well fields are affected by the location of streams and the geology in the area. Flow to the North Well Field originates north-west of the well field, forms relatively straight flow paths, and moves southeast toward the well field and the St. Joseph River. Flow to the South Well Field begins mostly in the out-wash along Yellow Creek south of the well field, moves northward, and turns to the north-west because of the influence of the St. Joseph River. Although pumpage at the Main Street Well Field is greater than pumpage at either of the two other well fields, the flow paths at the Main Street Well Field are much shorter than those at the two other well fields, indicating that the source of water to the wells at Main Street is from the nearby recharge ponds and from sections of Christiana Creek. The computer model was used to calculate locations of recharge for each well field; delineation of these recharge points roughly identifies the source area for each well field. Almost all of the recharge points for the South Well Field are greater than 5 years of travel time from the well field. The recharge points for the Main Street Well Field are sufficiently close to reported contamination sites to be potentially within the 5-year recharge area of the well field. Almost all of the flow from reported contamination sites discharges to one of the streams in the study area, primarily to the St. Joseph River. Longer flow paths tend to begin in the upper aquifer, usually moving downward through the confining unit to the lower aquifer, traveling horizontally until near the St. Joseph River, then flowing upward through the confining unit into the upper aquifer and into the river. Water in this type of flow path has twice been retarded in velocity by its flow through the confining unit before discharging into the river. Shorter flow paths tend to remain in the upper aquifer. To determine ground-water availability, the model was used to estimate the effects of potential future increases in pumpage at the three public-supply well fields. A 50-percent increase in pumpage above rates in 1993 at each of the well fields was simulated, and the resulting maximum drawdown is 5.4 feet. The areas affected by drawdown are small relative to the entire model area, indicating that the ground-water system has the capacity to provide additional amounts of water at the well fields without causing large, areally extensive drawdowns. Although the area affected by drawdown is small, the areas contributing flow to the North and South Well Fields extend well beyond the area of noticeable drawdown. Under the simulated increased pumpage conditions, the source area for the South Well Field is slightly wider but not noticeably longer than the source area for the 1993 pumpage.