Simulation-Optimization Approach for the Consideration of Well Clogging during Cost Estimation of In Situ Bioremediation System

Abstract

In situ bioremediation of groundwater has become one of the most widely used technologies for contaminated site treatment because of its relatively low cost, adaptability to site-specific conditions, and efficacy when properly implemented. According to many studies, enhanced bioremediation techniques can change the hydrogeological properties of the polluted aquifers, and the most notable change is biological clogging, resulting in reduction of porosity and hydraulic conductivity of the porous media. Because biodegradation kinetics in previous studies do not simulate microbial growth rate explicitly in the aquifer system, in this study biological clogging is accounted for during the cost optimization of in situ bioremediation system. A simulation-optimization approach based on extreme learning machine and particle swarm optimization (ELM–PSO) techniques is used to design an optimal in situ bioremediation system for a characteristic site. A two-dimensional finite-difference model is used to get the data for training and testing of ELM. Further, a single-objective function is considered to optimize pumping cost, facility capital cost, and well cleaning cost for a clogged well. The application of the ELM-PSO method to problems where clogging of wells occurs provides a more practical and realistic cost for a typical in situ bioremediation system.