Formatted Title
A Comparison of PFAS Concentrations Determined Using Modified Method 537.1 and Method 1633
Background/Objectives
The Rocky Flats Plant was a nuclear weapons component manufacturing facility managed by the U.S. Department of Energy (DOE) that was cleaned up and closed in 2005. Since then, the central portion of the property—the Central Operable Unit (COU)—has been managed by the DOE Office of Legacy Management, while the surrounding security zone has been converted to a national wildlife refuge managed by the U.S. Fish and Wildlife Service. In 2018, the Colorado Department of Public Health and Environment (CDPHE) requested that DOE collect environmental water samples for analysis of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). Results confirmed the presence of these two PFAS. The analytical method used was the modified U.S. Environmental Protection Agency (EPA) drinking water Method 537.1. The scope was then expanded, and results continued to show the presence of various PFAS including those two compounds. The highest concentrations reported as of mid-2023 for PFOA and PFOS were 160 and 370 nanograms per liter, respectively. In late 2023, the scope was revised to support comparison of analytical methods to assess whether the earlier data could be confirmed as representative, given that EPA Method 1633 for environmental matrices was then being validated and nearing finalization. This determination is important to the assessment of the presence and distribution of PFAS in the COU. The preference is to be able to use all available PFAS data as desired, irrespective of the analytical method used to generate the data.
Approach/Activities
The initial samples were collected during the second and fourth calendar quarters of 2019 from eight locations, including three monitoring wells, one hillside seep at the base of a former landfill, influent to two groundwater treatment systems (one for chlorinated solvents, one for seepage from a second former landfill), and two creeks. Because there was no approved PFAS analytical method with low detection limits for complex matrices such as groundwater and surface water, analytical laboratories developed modifications to a drinking water method to analyze these matrices. The result, referred to as modified Method 537.1, meets the requirements of the U.S. Department of Defense and DOE Consolidated Quality Systems Manual (QSM) for Environmental Laboratories, but was not EPA-validated for use with environmental matrices. This method was selected for use. The same method was used when the scope was expanded in 2021 to include 12 locations, an expanded analyte list, and quarterly sampling. Starting in the second quarter of 2023, split samples were collected at six of the 12 locations. At these six locations, one sample set was analyzed using the same modified Method 537.1 and the second set was analyzed using draft Method 1633. Starting in the fourth quarter of 2023, the monitoring scope was reduced to just the six locations at which split samples were collected. The six locations included the two monitoring wells presenting the highest concentrations of PFAS in groundwater; effluent from the treatment system at a former landfill, which treats a mix of groundwater and landfill seepage; the seep from the base of the other former landfill; and the two creeks. These locations were selected because they represent the full range of PFAS concentrations that had been observed at the 12 sample locations in the COU, they represent a mix of different environmental water types, and except for the wells, they represent water that flows off the DOE-managed property relatively quickly and, therefore, may be parts of an exposure route to offsite entities, both human and nonhuman.
A minimum of four rounds of split samples were required to support a confident statistical comparison of results to assess for systematic variation. After samples were collected and analyzed, the analytical results were validated for use following approved procedures. They were then grouped by analytical method and compared using the analysis of variance and Mann-Whitney statistical approaches. Subsets of the data were also evaluated in this manner. For example, water issuing from a landfill seep may present greater analytical challenges for one of the methods than water from a creek.
Results/Lessons Learned
Results of these comparisons will be available for the presentation. If systematic variation is suggested, corresponding lessons learned will also be presented.