Analysis of pumping test data for the identification of the spatial distribution of hydrological parameters

Abstract

The evaluation of groundwater resources holds an important role for the development of sustainable reservoir management plans. An important step in that direction is the detailed characterization of subsurface systems. The use of pumping tests is standard technique for the estimation of subsurface hydraulic properties. The analysis of drawdown data from pumping tests is normally performed using graphical techniques based on the assumption of aquifer homogeneity. However, natural subsurface formations are heterogeneous with complex patterns of spatial variability. This heterogeneity plays an important role in subsurface flow and contaminant transport processes. Therefore, estimating the spatial variability of subsurface flow parameters is essential for the development of models that can accurately predict groundwater flow and contaminant transport. This thesis examines the use of different interpretation methods for the analysis of time-drawdown data derived from pumping tests The main goal of the research was to estimate the flow parameters of the subsurface system, namely: transmissivity, storativity, conductance and leakance, and to infer some information about the spatial variability of these parameters. Both leaky and non-leaky confined aquifer systems were considered. The pumping test interpretation methods evaluated include conventional methods such as the Theis method, the Cooper-Jacob method for confined aquifers, and the Walton and Hantush inflection point method for leaky aquifers. In addition two recently developed methods, the Continuous Derivation method and the double inflection point method were tested. In order to demonstrate the effect of heterogeneity on pumping test interpretations, heterogeneous transmissivity fields were first generated and used to simulate transient drawdown data which was then used to estimate the flow parameters using the different interpolation methods. The interpretation methods were also applied to real field data. The resulting estimated parameters are shown to be space dependent and vary with the interpretation method since each method gives different emphasis to different parts of the time-drawdown data. Also, the heterogeneity in the pumped aquifer influences the estimates and they show different behaviors according to interpretation method. It was also shown that more information about the flow parameters can be obtained when the results of different interpretation methods are combined together.