The Effect of Saprolite on Long-Term Performance of Wells in the Piedmont

Saprolite overlies fractured crystalline rock aquifers throughout the Piedmont Physiographic Province in the eastern U.S., but the effect of saprolite storage on short-term well tests used to predict the long-term performance of water supply wells has been unclear. We used analytical solutions to show that the steady- state specific capacity of wells in uniform aquifers is controlled by transmissivity and interaction with streams. The effect of water released from storage in saprolite was ignored in the analytical solutions, however, so we have expanded the work to include numerical analyses representing saprolite as an unconfined layer over a rock aquifer. Narrow bands of constant head spaced 1 km apart atop the saprolite layer were used to represent streams (typical spacing of first- order streams in this area is 1 km). Values of T = 1 to 30 m2/day and S = 0.0001 to 0.001 were used for the rock aquifer, whereas the properties used to represent saprolite were T = 1 to 30 m2/day and Sy = 0.0001 to 0.5. Several hundred analyses were conducted with various combinations of properties in these ranges. Preliminary results show that when the T and S of the saprolite is the same as the rock (baseline condition), drawdown fits a semi-log straight line until interaction with streams flattens the slope to produce steady-state conditions. This response is consistent with analytical solutions by Hunt (1999) & Butler et al (2001), and with our previous results. As the specific yield of the saprolite is increased, drawdown in the rock aquifer flattens at early time and then steepens, an effect resembling dual porosity behavior. Ultimately, the slope flattens again at steady state. Increasing the magnitude of specific yield prolongs the saprolite effect, with typical durations from as short as a day to as long as a year (high Sy and low K in saprolite lengthen the duration of this effect). Drawdown curves can flatten and appear to go to a steady state, where values of representative simulations are roughly 3/4 of the true steady state drawdown. Accordingly, specific capacity determined while the saprolite effect is occurring during a short-term well test can over-estimate the actual specific capacity of the well by a factor of up to 1.5 or more. It is important to recognize that the effect of release from storage in saprolite is always temporary and will not sustain long-term pumping conditions. Eventually this effect dies out and the long-term capacity of a well is dominated by interaction with nearby streams.