Optimizing the Remedial Process at a Petroleum Hydrocarbon Contaminated Site Using a Three-Tier Approach

Abstract

Pump and treat (P&T) system is widely used to remove contaminants in groundwater and maintain hydraulic capture of contaminants. A P&T system usually becomes costly in removing contaminant mass after operation for a period of time because the cost of pumping stays high while concentration of contaminants in extracted groundwater decreases significantly over time. However, the P&T system may have to continue operation before contaminants in groundwater reach the cleanup goals. In such cases, the optimization of the existing remedial system and an exit strategy will be necessary and beneficial to all stakeholders. This study examined a P&T system at a remediation site near a coastal wetland. We developed a three-tier approach to optimize the existing remedial system. The remedial process optimization (RPO) was achieved by reducing pump wells and monitoring frequency and developing an exit strategy. In the first tier, we used descriptive trend plot, trend factor analysis, and autocorrelation analysis to capture the temporal trend in the monitoring data and determine optimum sampling frequency; in the second tier, we used multivariate analysis include principal component analysis (PCA) and partial least square regression (PLSR) to identify the spatial and temporal pattern in the monitoring data, and to construct a statistical model to predict future contaminant levels; in the third tier, we used MODFLOW, a USGS developed finite-difference flow module for aquifer simulation, and interpretation of the hydraulic capture of extraction wells and contaminant migration. Based on the tiered analysis, we designed an improved groundwater monitoring program and pumping strategy. The RPO is expected to save 68% groundwater monitoring cost, and 30% P&T operation and maintenance (O&M) cost.