Bioslurping Model for Assessing Light Hydrocarbon Recovery in Contaminated Unconfined Aquifer. II: Optimization Analysis

Abstract

The bioslurping model was successfully applied to simulate the recovery efficiency of light nonaqueous phase liquids (LNAPLs) in a companion remediation study. This paper carries out an advanced environmental systems analysis for handling the application challenges related to a seamless integration of simulation outputs, regression analyses, and optimization practices to search for the best remediation solution. In view of the inherent complexity of integrating simulation outputs at various scales for building the optimization steps and searching for the ultimate risk-informed and cost-effective solutions, a set of regression submodels describing the system response to water and gas pumping and oil skimming in hydrocarbon recovery are derived and employed to bridge the traditional gap between simulations and optimization studies. The bioslurping model is used to generate regression submodels of the system behavior during remediation operations. To test the robustness of the operational scheme, three planning scenarios based on a common linear programming framework are defined and utilized to assess the optimal control mechanics for LNAPL recovery. These scenarios are designed to shed light on improving the predictive capabilities for environmental decision making to handle a complex multiphase and multimedia remediation program. The research findings indicate that water-pumping volume can be minimized and oil and soil gas recovery can be maximized whether the candidate well locations are fixed beforehand or not. However, flexible choices in well locations in the optimization steps merit substantial examination in the future.