Modeling Ground Water-Surface Water Interactions and Hydraulic Containment Using Influence Functions

The dynamic interactions of ground water, surface water, and recovery-well pumping at an operational industrial facility adjacent to a river were characterized through development of an algebraic influence-function model.  Hourly time-series data collected over a 10-month period from 18 monitoring wells and the river showed that water levels in the two aquifers of concern were primarily determined by antecedent river stages and the pumping rates in four recovery wells.  Noting the frequency dependence of wave propagation in aquifers, the time-series records were filtered into six frequency bands.  Ground water fluctuations were then correlated to river stage fluctuations for each frequency band by de-convolution to establish a polynomial response function defining the transient influence of the river on ground water levels.  Shutdown and restart tests were run on the four recovery wells, and the observed aquifer responses were also characterized as polynomial functions of pumping rate.  The river and pumping response functions were assembled into a spreadsheet model that can predict ground water levels at many spatially-distributed points in both aquifers for any given record of observed river stages and combination of pumping rates.  Water level predictions generated by this model were then contoured to identify the hydraulic capture zones of the recovery wells either for actual pumping conditions in the past or for proposed future pumping rate combinations.  The completed model was used to identify optimal operating rates for the hydraulic containment system that will maintain the required hydraulic capture zone encompassing the facility landfill under various conditions of high, low, rising, and falling river stage.  As an alternative to a calibrated finite-difference or finite-element flow model, the influence-function model is convincingly based on direct measurement of the issues of concern.