Numerical analysis of step-drawdown tests: Parameter identification and uncertainty

An inverse, numerical model is presented as an interpretation method for step-drawdown tests. The model consists of an axi-symmetric model to simulate radial flow in a horizontally stratified aquifer, and a nonlinear regression analysis to identify hydraulic parameters and quantify parameter uncertainty. Compared to commonly used methods based on analytical solutions, the inverse model allows a more representative aquifer schematisation, usage of all drawdown observations, and a comprehensive analysis of parameter uncertainty. Data sets from two step-drawdown tests performed in a homogeneous, confined aquifer are interpreted and it is concluded that transmissivity and well loss coefficient are identifiable. Additional data sampled from an observation well are recommended to estimate aquifer storativity. Reliability of parameter estimates decreases with a decrease in the number of steps and therefore it is recommended to conduct step-drawdown tests having at least four steps. Analysis of data from a third test performed in an unconfined, layered aquifer reveals that only the well loss coefficient is identifiable and that prior information about hydraulic parameters of extracted and adjacent layers is required. In general, it is concluded that the inverse, numerical model is a viable interpretation method to identify the well loss coefficient and additional aquifer parameters from step-drawdown test data.