Traditional Aquifer Tests: Comparing Apples to Oranges?

Traditional interpretation of aquifer tests rely on fitting the observed hydrograph at one observation well to Theis type curve or Jacob's approximation solution, which assumes aquifer homogeneity. Recent studies based on Jacob's approach show that transmissivity estimates obtain from late time data converge to effective transmissivity value for uniform flow while the corresponding estimates for storativity exhibits strong spatial variability. Using Monte Carlos simulations, we investigate effects of heterogeneity on the analysis of aquifer tests using the Theis approaches. First, we develop a method that is consistent with the homogeneity assumption to estimate the effective transmissivity and storativity values for radial flow. Results using this method indicate that the effective transmissivity and storativity for radial flow are time variables but converge to geometric means at large time. This suggests the existence of REV for radial flow in the cases we examined. However, the estimated transmissivity based on the drawdown from a single observation well reflects the local heterogeneity surrounding the observation well but varies with pumping time. At large time, the estimated transimisivity value converges to the geometric mean of the flow domain while the storativity estimate fluctuates and do not approach the geometric mean. The estimate of storativity eventually approaches the local average values of storativity near the observation well. Finally, we use sensitivity analysis to explain the results of our numerical experiments and to determine the pumping time required to reach the geometric mean of the transmssivity field. Keywords: pumping test, heterogeneity, distance drawdown curve, time drawdown curve.