Design factors for improving the efficiency of freeproduct recovery systems in unconfined aquifers

J. J. Kaluarachchi, Utah State University
R. Elliot


Free-product recovery system designs for light hydrocarbon recovery were investigated to evaluate the effects of multiple-stage pumping, delayed startup, and uncertainty of key residual oil saturation input data using a vertically integrated three-phase flow model, ARMOS. The results obtained from a single well recovery system subjected to a given well location and uniform soil properties suggested that multiple-stage water pumping can enhance the recovery and provide optimal design conditions if cost and recovery oil volume are the targets. However, if containment of the spill area with simultaneous recovery is the target, then single-stage pumping still provides the optimal design. Delayed startup of the operation from the initial site characterization can cause substantial plume movement during the lag period making predictions made with zero time lag invalid. Allowance should be made in the design to account for this time delay. Uncertainty of maximum residual oil saturations in both saturated and unsaturated zones may cause considerable differences in predicted results. Careful attention should be paid to obtaining these data for accurate remediation design.