Formatted Title

Use of Ricker Method to Evaluate Plume Stability and Support Remedy Refinement in Pesticide Plumes

Background/Objectives

At a former pesticide formulation facility in the Coastal Plain of South Carolina, pesticide plumes in groundwater had been contained using a single, shallow recovery well since the late 1990s pursuant to CERCLA decision documents.  However, a reevaluation of the conceptual site model was required following a five-year review, given suggestion of plume migration in long-term monitoring (LTM) data and erratic exceedances of remedial goals for many constituents of concern (COCs).  Given that pesticides at the site are analyzed by gas chromatography, and analyses of low concentration pesticides are often confounded by the presence of the multi-component pesticide toxaphene, a holistic assessment of plume stability was warranted, to determine if plumes were moving over the course of LTM, or if the site remedy was effective.

Approach/Activities

The Ricker Method (RM) was used to reassess conditions within an 85-foot thick section of the Barnwell Formation, including multiple pesticides (dieldrin, BHC isomers, DDx, etc.) in two discrete intervals.  This technique was used to assess the area, average concentration/mass, and movement (by gauging by spatial changes in the location of the center of mass).  LTM data over the period 2012-2012 were compiled for eight events which were selected because they had the most robust datasets: 2012, 2013, 2014, 2015, 2016, 2017, 2020, and 2021.  Thirteen pesticides were modeled for the shallow interval, and seven pesticides were modeled for the deep interval.  LTM data were used to create an 8,500-node grid file (a node approximately every 15x15 feet).  Concentrations and mass were interpolated only within the 8,500-node data hull.  Spatial data were interpolated using Surfer via Kriging, with concentration estimates generated at each grid node.  All data were used for gridding, with a default linear variogram to interpolate the rate of change between points.  Once values were assigned to each node, plume characteristics (the area, average concentration, mass, and center-of-mass movement) were assessed for each of the eight events being compared.  Plume characteristics were then assessed using the Mann-Kendall test for trend at 80%, 90%, and 95% confidence levels.

The effects of multiple factors (non-detect treatment, interpolation method, grid node density, linear vs. logarithmic interpolation methods, and data hull shape) were assessed via a sensitivity analysis to gauge the impact on Ricker Method outputs and the trend of plume characteristics over time. The assumptions made to perform the analyses were found to have minimal impact on outcomes.  

Results/Lessons Learned

RM findings indicated the shallow aquifer zone had more consistent reduction in plume area, average concentration, and mass than the deep plumes for the 2012-2021 evaluation period, indicating plume containment/plume stability in the shallow zone. The analysis concluded that the deeper plume was mobile, with the downgradient plume edge undefined.  Ricker Method results, including a combined/composite plume, were used to locate both downgradient wells for plume delineation (given adjacent property constraints) as well as evaluate potential locations for supplemental recovery well(s).  The Site is currently undergoing supplemental field activities to support remedy refinement in the deeper zone.