2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 9
Presentation Time: 1:30 PM-5:30 PM


PAYLOR, Randall L. and CURRENS, James C., Kentucky Geological Survey, Univ of Kentucky, 228 Mining and Minerals Building, Lexington, KY 40506-0107, rpaylor@uky.edu

Royal Spring, the largest karst spring in the Inner Bluegrass Region of Kentucky, is the principal water supply for the city of Georgetown. Over 8000 customers of Georgetown Municipal Water and Sewer Service (GMWSS) rely on water produced from the spring. The spring recharge basin encompasses 25 mi2 in Scott and Fayette Counties, including parts of Lexington. A large portion of the basin is developed with suburbs, commercial, and light industrial land uses. Interstates I-65 and I-75, the CSX railroad, and U.S. and State roadways cross the watershed.

Past contamination incidents in the Royal Spring watershed have interrupted the supply. The risk of a catastrophic spill along the transportation corridors is high. GMWSS switches to finished water from Frankfort as an emergency response to a spill, but purchase of the Frankfort water is expensive and needs to be minimized to reduce cost. To facilitate prediction of arrival time and duration of soluble contaminants, the Kentucky Geological Survey (KGS) conducted a quantitative dye tracing study to determine watershed travel times from key swallow holes in the basin.

KGS conducted 11 sets of simultaneous traces for discharges at Royal Spring from 5 ft3/sec to over 100 ft3/sec. An existing USGS gage monitored hourly discharge, and KGS modified the rating curve to include water withdrawal. Water samples were collected and analyzed for tracers using a Varian spectrofluorometer. Tracer first arrival, last detection, center of mass and total mass were determined. Average recovery of tracers was 91 percent. A simple watershed runoff model was used to approximate surface travel times from the watershed boundary to nearest swallow holes, and was calibrated with surface velocity measurements.

Travel times from swallow holes to the spring ranged from 2 hours to 3 days. Using the groundwater tracing and surface modeling results, 5 travel time maps were constructed that indicate the minimum time of arrival for a spill on the surface. Each of the maps is keyed to a specific spring discharge. In the event of a spill, its location, time, and the current spring discharge level are used with the appropriate map to estimate the first arrival time at the spring. A decision can then be made on how long to safely wait after the spill before switching to an emergency supply.