Publications : 2026

Killius A, Covell LT. Characterization of PFAS in Vermont public water systems. Abstract 3892, Society of Toxicology 65th Annual Meeting, San Diego, CA, March 2026.

Abstract

OPEN ACCESS

Background and Purpose: Per- and poly-fluorinated alkyl substances (PFAS) have gained considerable attention due to their persistence, widespread environmental detection, and frequent presence in drinking water systems. While certain regions are impacted due to local PFAS manufacturing facilities, the State of Vermont makes for an interesting case study because it has no known primary manufacturing sources of PFAS, yet many PFAS congeners are present in Vermont drinking water. The Vermont Water Supply Rule requires public water systems to routinely test for PFAS, including 18 individual congeners and the sum of PFHpA, PFNA, PFOS, PFOA, and PFHxS. Vermont regulations state that the sum of these 5 congeners must be below 20 ppt. Although the testing results are publicly available, the database format is not easily accessible for analysis. Here, we systematically transformed Vermont’s public PFAS drinking water database into a usable format and conducted a quantitative analysis characterizing PFAS occurrence and temporal and spatial patterns across the state. This analysis provides insight into “background” PFAS drinking water concentrations in regions with no known primary PFAS sources. Methods: PFAS drinking water results were obtained from the Vermont Drinking Water Database on August 19, 2025. The PDF file was imported into RStudio to parse and extract data, and the resulting dataset was exported into Microsoft Excel for further manual data tidying and quality assurance due to the inconsistent formatting of the PDF output. Because the limit of detection (LOD) was not provided for results reported as non-detects, a value of 0.1 ppt was used for imputation to enable further quantitative analysis. Each town was mapped to its county for broader geographical analysis. Summary statistics were calculated for each congener, and the sum of the five Vermont-regulated PFAS was used to characterize temporal and geographical trends. Results: A total of 631 unique public water systems across 231 towns and all 14 counties were sampled between July 28, 2019, and August 4, 2025 (n=2,656). Most (94.1%) systems were sampled on at least 3 separate occasions. 503 systems (79.7%) and 148 towns (64.1%) exhibited no detectable PFAS in any sample. Six of the 18 congeners assessed were not detected in any sample. The long chain legacy PFOA and PFOS congeners were the most frequently detected analytes, observed in 16.3% and 13.6% of samples, respectively, and had the highest average concentrations among all congeners (1.15 and 1.59 ppt, respectively). They were primarily detected simultaneously, wherein 306 samples had detectable levels of both PFOA and PFOS, and only 186 samples resulted in concentrations of either compound alone. The detection frequencies of the shorter-chain replacement compounds were lower: PFHxA (12.2%), PFBS (8.6%), PFHxS (4.9%), and HFPO-DA (0.2%), indicating dominance of legacy long-chain PFAS. Years with the highest number of samples had the lowest average annual PFAS concentrations. The average sum of the 5 Vermont-regulated PFAS was 3.63 ppt, which is lower than the regulated sum of 20 ppt. Rutland County exhibited the highest average sum value (9.83 ppt). The ten highest sum concentrations were all observed in a single water system in Mount Holly, suggesting a potential localized secondary source of PFAS. Other towns with relatively elevated sum concentrations compared to the rest of the state included Barre Town, Arlington, and Killington. PFOS and PFHxS were the primary contributors to elevated sum concentrations. Conclusions: The majority of public water systems in Vermont do not have detectable levels of PFAS, which is consistent with the absence of primary point sources within the state. Although long-chain congeners were phased out more than a decade ago, they continue to persist in higher concentrations and detection frequency than the short-chain replacements, indicating historical legacy contamination rather than ongoing inputs. The annual average trend with sample size shows expanded monitoring found smaller overall statewide impacts. PFAS concentrations were elevated in several specific areas, possibly due to secondary sources that historically used PFAS in their production. Overall, the data show that PFAS detections in Vermont drinking water supplies are low and the congener profile is indicative of historical widespread use, rather than continued ongoing contamination.