Effect of Heterogeneity Correlation Scale on Pump-and-Treat Remediation Design

Pump-and-treat (P&T) remediation is a commonly used technology for the cleanup of contaminated groundwater. The long remediation time, high capital and operational costs are the most important problems associated with these systems. One of the determinants that affect the remediation design and its outcome is the spatial heterogeneity in the hydraulic conductivity (K) field. In this study, the effect of correlation scale (length) (lambda) of the heterogeneous K-field on P&T remediation design and costs was investigated. For this purpose, several synthetic K-fields of similar means and variances (sigma²) but different lambdas were generated using a random field generator. Simulation-optimization approach was used in order to conduct the study. For this purpose, BIO2D-KE linked with a genetic algorithm library was used. Runs were conducted in order to determine the best pumping policies and costs for the heterogeneous fields. According to the initial results, lambda of the spatial variation in K impacted P&T remediation design and costs. The number of wells, pumping rates, and costs of remediation varied for different lambdas. This outcome was more pronounced for higher sigma² of the spatially variable K-field.