A controlled field evaluation of continuous vs. pulsed pump-and-treat remediation of a VOC-contaminated aquifer: site characterization, experimental setup, and overview of results

Sheet pile test cells were used to isolate two adjacent segments of a long-extant groundwater plume containing volatile organic compounds (VOCs), including tetrachloroethylene (PCE) and trichloroethylene (TCE) and their degradation by-products cis-1,2-dichloroethylene (c-DCE) and chloroethene (a.k.a. vinyl chloride or VC). One cell was subjected to continuous flushing and the other to intermittent (pulsed) flushing for over 200 days. Pre-pumping coring and sampling revealed maximum contaminant concentrations immediately adjacent to the basal aquitard for most analytes in both cells. The single exception to this trend was PCE, which showed maximum concentrations 0.5 to 2.5 m above the aquitard in the continuously pumped cell. For contaminants other than PCE, the contaminant mass within the aquitard was similar to that in the aquifer. Substantial “cleanup” was achieved in the aquifer region of both test cells, but levels of contamination within the aquitard region of the two cells were comparatively unaffected. Regarding overall removal efficiency in terms of mass removed per volume of water pumped, differences in pre-pumping contaminant distributions and total masses among the two cells complicated the comparison. Nevertheless, pulsed-pumping was observed to be slightly more efficient than continuous pumping at the same rate. The slight efficiency difference is attributed to diffusive limitations on contaminant removal from low-permeability zones. In this regard, diffusion from the thick underlying aquitard was too slow to contribute strongly to contaminant “rebound” within the time scale of the pumping interruption. We postulate that larger differences in efficiency between pulsed and continuous pumping would be observed for longer interruption periods, or with similar pumping conditions at sites where more of the initial contamination is in small scale low-permeability regions (e.g., sorbing silty or clayey layers within the aquifer). Overall, this work illustrates that contaminant transport and subsurface remediation may be influenced not only by (1) spatial variability of the transport medium (aquifer), a matter reasonably well appreciated, but also by (2) spatial variability in pre-remediation contaminant distribution, and (3) the nature and extent of contamination in adjacent less permeable regions.