Flow parameters in a shallow conduit-flow carbonate aquifer, Inner Bluegrass Karst Region, Kentucky, USA


In the carbonate aquifers which underlie most karst terrains, groundwater flow is through a dendritic system of solution conduits. In such aquifers, termed shallow conduit-flow aquifers, the methods used to model granular and fracture aquifers are not generally applicable. Investigations were conducted in the Inner Bluegrass Karst Region of central Kentucky with the objective of developing methods of modeling shallow conduit-flow aquifers as well as obtaining quantitative information on a specific portion of the aquifer to assist in its management for water supply purposes. In the Inner Bluegrass Karst Region, groundwater basins are developed, in each of which there is an integrated system of solution conduits which conducts recharge to a major spring. One of the largest of these groundwater basins feeds Royal Spring, which serves as the principal water supply for the town of Georgetown. The basin extends over 15 km to the southeast and most of its flow is furnished by underground diversions of Cane Run, a surface stream with headwaters near the center of the City of Lexington. The principal objectives of the field investigation were to determine discharges at the spring and travel times to the spring from discrete recharge points within the basin, termed swallets. The spring is ungaged, and an attempt was made to obtain a continuous discharge record by the dilution of dye introduced at a swallet. Comparison of the dye-dilution discharge record with stage discharges at the spring revealed substantial discrepancies which are believed to be caused by as much as five-sixths of the low-flow discharge from the upper portion of the basin bypassing the spring. The dye-dilution method, therefore, provided significant insights into the geometry of the conduit system of the groundwater basin although it proved unsatisfactory as a method of determining discharges at the spring. Analysis of the travel times and stage discharges provided information on the conduit geometry by modeling the flow as open-channel flow in a rectangular channel. Flow in the system is rapid, ranging from 140 to 590 m h-1. Although the flow rate increases with discharge, the relationship is not simple owing to substantial increases in conduit cross-sectional area at higher discharges. Flow is turbulent and subcritical under all conditions. The most surprising result was the very low depth of flow calculated; less than 17 cm at even the highest discharge. Although this must be considered an 'equivalent' depth, it is believed to indicate that active flow in shallow conduit-flow aquifers is generally in a thin zone just beneath the water table.

Document Type




Publication Date


Journal Title

Journal of Hydrology