Formatted Title
Avoiding over Predicting PFAS Soil Porewater Concentrations: Implications for Hydrogeological Risk Assessment and Soil Remediation
Background/Objectives
The vadose zone has increasingly been identified as a significant reservoir for per- and polyfluoroalkyl substances (PFAS) at sites impacted by different sources, including fire-fighting foam applications, manufacturing processes and land applications of biosolids. PFAS concentrations in unsaturated soils are often orders of magnitude higher than the underlying groundwater at these sites, where the potential impacts of long-term leaching on groundwater quality, and dependent receptors, is a primary concern.
Evaluating the risks posed by PFAS leaching to groundwater requires quantitative understanding of soil porewater concentrations and infiltrating water discharge. Direct measurements of PFAS discharge due to leaching may be impractical at environmentally relevant scales and applying comprehensive vadose zone models for PFAS transport requires considerable information for parameterisation. Partitioning-based estimates for soil porewater concentrations and leaching discharge, of the type frequently used in screening-level transport models, may provide a viable alternative. However, the relationship between soil and soil porewater concentrations is more complex for PFAS than for non‑surfactant solutes, due to the adsorption of PFAS at phase interfaces, which are dependent on fluid contents and soil properties.
Approach/Activities
A PFAS partitioning model from the literature has been applied to predict soil porewater concentrations and develop potential risk-based targets for vadose zone soil remediation. Estimates of soil porewater concentrations of perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) were calculated from total soil concentrations using the PFAS partitioning model, including phase and interface adsorption terms, and a conventional non-surfactant partitioning model. These calculations were based on PFOS and PFOA sorption properties and physical characteristics for 10 different soils of varying particle sizes and organic contents.
Results/Lessons Learned
The soil porewater concentrations predicted for representative water contents using the PFAS partitioning model were up to two orders of magnitude lower than those predicted using the conventional partitioning model, demonstrating the significance of PFOS and PFOA retention on air-water interfaces in unsaturated soils. Soil porewater concentrations decreased by between two and five times for an order of magnitude increase in soil water content. These differences are proportional to the predicted discharge via leaching to groundwater and highlight the likely sensitivity of PFAS risk assessment modeling to temporal variability in source zone water contents. Furthermore, soil porewater concentrations are inversely proportional to calculated potential risk-based soil remediation target concentrations. Remediation targets that do not consider PFAS-specific adsorption processes may be biased low, unrepresentative and unachievable with current treatment technologies, potentially pushing stakeholders towards less sustainable risk-management strategies for PFAS impacted sites.
The role of PFAS partitioning modelling within an integrated strategy to assess risks to groundwater from leaching and discharge through the vadose zone will be presented, including concepts to parameterise such models, which should better inform soil remediation objectives and targeted soil treatment selection.