Tritium as a Model Tracer for Predicting Per- and Polyfluoroalkyl Exposure and Potential Ecosystem and Human Health Impacts

For this thesis, historical tritium and per-and polyfluoroalkyl substances (PFAS) concentrations measured in surface waters of South Carolina at or near the Savannah River Site (SRS) were compared. Statistical linear regression analysis was used to assess the ability of tritium as a model tracer for predicting PFAS concentrations, specifically perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) distributions. Nine years of tritium monitoring data, (2014 – 2022), were analyzed and used in linear regressions to assess their association with PFAS levels of PFOA and PFOS measured during 2023, in and around the SRS and the adjoining Savannah River. Initially, a normality test was performed to determine if these data were normally distributed, followed by linear regression comparing tritium levels measured at the SRS to PFOA and PFOS concentrations measured in surface waters of the adjoining Savannah River. This linear regression analysis yielded significant correlation coefficients for tritium compared to both PFOA (R = 0.53-0.69) and PFOS (R2 = 0.65-79) which were significant (p < 0.05) and suggest a strong association between tritium and both PFOA and PFOS. Similarly, the Coefficients of Determination for both PFOA (R2= 0.28-0.46) and PFOS (R2= 0.42 and 0.61) were significant (p < 0.05) and indicated that tritium levels may explain 28-46% of the variability of PFOA and 42-62% of the variability of PFOS. These results suggest that tritium concentrations may be a potential tracer for prediction of surface water levels of both PFOA and PFOS including surface waters used as drinking water sources. PFAS concentrations in surface waters of the Savannah River were at levels above the proposed drinking water limits and may pose a potential hazard to human health. These finding suggest that routine monitoring of tritium at the SRS has the potential to be used as a tracer to predict levels of PFAS for both PFOA and PFOS. The rational for this association is that tritium emanating from nuclear activities at SRS may co-occur with other industrial chemicals used in plant operations at SRS or in adjoining urban areas near the plant. Studies conducted by the USGS throughout the U.S., including the southeastern U.S., have shown that tritium levels had the highest association (R2 = 0.177) with PFAS levels in both ground water and surface water sources used for drinking water (McMahon et al., 2022). These USGS results indicated that tritium levels along with distance to the nearest fire-training area, % of urban land use, Dissolved Organic Carbon (DOC) and Volatile Organic Carbon (VOC) levels were the top five predictors of PFAS groundwater concentrations, accounting for > 63 % of the variability in the model. In addition, PFOA and PFOS were 2 of the 3 most frequently detected PFAS, with 2.4% of the samples having PFOA and PFOS levels > the 70 ng/L health advisory level. These finding support our conclusion that tritium levels may have the potential to be used to estimate PFAs levels of PFOA and PFOS and could be used to notify consumers of surface water sourced community drinking systems of the potential for excess exposure of PFOS and PFOA that may exceed proposed Environmental Protection Agency’s (EPAs) drinking water guidelines for PFOA and PFOS standards to protect human health. The potential for PFAS impact on human health was also discussed, including potential adverse health effects associated with PFOS and PFOA.