Re-Evaluation of a Conceptual Site Model to Accelerate Remediation of Metals in Groundwater

Abstract

Monitored natural attenuation (MNA) was selected ten years ago by previous operators of a Superfund site in Florida as a remedy for metals-affected groundwater impacted from the historical operation of wastewater evaporation ponds at a metal galvanizing plant. The Record of Decision (ROD) stipulated that MNA should result in a decrease of dissolved zinc and nickel concentrations in a shallow sandy aquifer to the Performance Standards within 7.6 years of pond closures, or else the implementation of a groundwater-recovery-and-treatment contingency remedy would be required. The groundwater model, used by others to estimate the time to achieve the Performance Standards via natural attenuation, assumed no contaminant retardation, and no other potential sources, including releases from the current operator. This model predicted that zinc concentrations would be reduced from a range of 140,000 to 190,000 μg/L, to 10,000 μg/l by natural groundwater dilution or flushing. Historical results indicated substantial decreases of zinc and nickel concentrations, but not to levels below Performance Standards. Therefore, the U.S. EPA concluded that MNA was not a viable remedy and would not result in cleanup within a reasonable time frame. Before moving forward with the contingency remedy, LFR recommended a re-evaluation of the conceptual site model (CSM) to better understand site hydrogeology and to optimize full-scale remediation. It was suspected that the same conditions that limited the effectiveness of MNA would also limit the effectiveness of the groundwater-recovery-and-treatment contingency. Therefore, additional characterization data were collected to develop an appropriate CSM for the Site that would assist in determining the appropriate remedial approach and also shed light on why MNA was not effective. Additional characterization included horizontal and vertical groundwater data collection using direct push technology and nested wells; detailed geochemical analysis; and 3-dimensional imaging of hydrostratigraphic units. The new data helped determine the distribution of a clayey-sand transition zone between the high-conductivity shallow sand aquifer and the underlying low-conductivity clay formation. The 3-D imaging showed the presence of paleostratigraphic depressions in the underlying clay formation, probably a result of karst-related sinkholes in the deeper limestone. Geochemical data indicated the presence of higher concentrations of metals and total dissolved solids in the clayey-sand transition zone relative to the overlying sand aquifer. Based on the revised CSM, it appears that high-density wastewater from the former ponds moved both vertically and horizontally into the surficial aquifer, and penetrated down to the clayey-sand transition zone. Although MNA appears to have progressed as modeled in the sandy surficial aquifer, higher concentrations of metals were retained in the clayey-sand transition zone due to the higher sorption capacity of this unit. Furthermore, the paleostratigraphic depressions in the clayey sediments further slowed down the horizontal movement of metals in the lower part of the aquifer. Based on the additional characterization data and the revised CSM, groundwater-recovery-and-treatment was no longer considered viable due to the majority of the metal mass residing in low-conductivity soils. The U.S. EPA concurred with this conclusion. LFR has proposed an in situ treatment remedy, which is currently being bench- and pilot-tested.