Per- and polyfluoroalkyl substances (PFAS) are a class of over 14,000 fluorinated chemicals widely used due to their water and oil repellent properties and high heat tolerance. Their persistence in the environment has led to their designation as 'forever chemicals,' raising significant public, scientific, and regulatory concern. PFOS and PFOA, two prominent PFAS, are included in the Stockholm Convention, restricting their use and production. However, this has led to a shift towards novel PFAS. Regulatory guidance on safe PFAS concentrations varies globally, with Health Canada having one of the most restrictive recommendations (sum of all PFAS < 30 ng l⁻¹). The US EPA has proposed limits for several specific PFAS. Toxicity concerns generally increase with fluorinated chain length (FCL). While PFAS production has been estimated and their presence in various products quantified, their fate and the extent of global surface and groundwater contamination remain largely unknown. This study aims to address this gap by investigating the global extent and distribution of PFAS contamination in surface and groundwater, assessing PFAS concentrations against various regulatory thresholds, and investigating PFAS sources, including their distribution in consumer and industrial products.

Numerous studies have investigated PFAS in various environmental compartments, including one suggesting four select PFAS exceed planetary boundaries. Other studies have assessed aqueous phase PFAS concentrations in specific regions. While PFAS pervasiveness is acknowledged, the extent of global surface and groundwater contamination and the frequency of exceeding drinking water guidelines remain unclear. Previous research has estimated global PFAS production and quantified PFAS in commercial and industrial products, but their environmental fate is not fully understood. The limited existing data focuses on select PFAS and regions, making a global assessment challenging.

This study compiled data from 273 environmental studies (since 2004) encompassing over 12,000 surface water and 33,900 groundwater samples. The dataset was built from 367 published papers and government websites, representing a total of 48,985 samples. PFAS concentrations were assessed against various international regulations and advisories (e.g., US EPA, Health Canada, EU). Data were analyzed and statistically validated using Python scripts and MS Excel. Where PFAS concentrations were below detection limits (BDL), a random value between zero and the detection limit was assigned to minimize bias. Data were mapped using the latitude and longitude of sampling locations; if unspecified, a major city within the country was randomly assigned. PFAS were categorized into classes (PFCAs, PFSAs, precursors such as fluorotelomers, sulfonamides, PAPs, and novel PFAS). The US EPA's draft method 1633 served as a benchmark for assessing the proportion of PFAS captured by current methods. Consumer and industrial products were divided into AFFF and non-AFFF categories to understand their respective contributions to PFAS environmental burden. The data was converted into an Excel file from other formats via open-source converters, and then analyzed using Python for statistical validation and analysis and visualization using MS Excel.

PFAS are pervasive in surface and groundwater worldwide. While Australia, China, Europe, and North America appear as hotspots (Fig 1a), this may reflect higher sampling density rather than actual distribution (Fig 1b). A significant portion of water samples exceeded regulatory thresholds (Fig 2). Exceedances were higher for samples with known contamination sources (AFFF or non-AFFF) than those with unknown sources. Groundwater samples with known AFFF contamination frequently exceeded US EPA criteria (71-72%), while those with unknown sources still showed elevated exceedance (31-50%). The proportion of samples exceeding thresholds varied depending on the specific regulatory guidelines. Groundwater samples without known contamination exceeded Health Canada's criteria in 69% of cases, but only 6% exceeded the EU's sum of all PFAS criteria (500 ng l⁻¹). Surface water samples showed similar exceedance rates for known AFFF sources but lower rates for unknown or non-AFFF sources, as expected due to shorter residence times. Analysis of PFAS in 943 non-AFFF consumer products across 15 categories revealed that fluorotelomers and PFCAs were dominant (Fig 3). However, using only the PFAS quantified by EPA method 1633 significantly underestimates the total PFAS mass, altering the distribution of PFAS subclasses (Fig 4). A significant fraction of PFAS in consumer products are not currently regulated, but many will degrade into regulated PFAS in the environment.Analysis of 148 AFFF samples from various suppliers showed differing PFAS compositions based on manufacturer and production year. PFOS dominated older 3M AFFF (51%), while fluorotelomers and PFCAs were dominant in Angus AFFF. Using only EPA method 1633 underestimates the total PFAS mass in AFFF by a factor of 2.8.Across 33,940 groundwater samples, 57 distinct PFAS were quantified, with an average of 16 PFAS per study. PFCAs, PFSAs, and sulfonamides were routinely quantified, whereas fluorotelomers (primarily FTS) were less frequently measured and FTOH were rarely detected. This is despite the importance of FTOH in consumer products. Two studies examining urban rivers in China and Bangladesh showed FTOH representing a substantial portion of total PFAS (53% and 2%, respectively). Studies of wastewater treatment plants (WWTPs) and landfills showed limitations in PFAS quantification, with FTOH not frequently measured despite atmospheric emissions reported at these sites. The TOP assay, designed to measure PFAS precursors, revealed considerable increases in PFAS after oxidation in some studies, highlighting limitations of current methods in capturing the total PFAS burden.

This study's findings show a substantial and widespread global contamination of surface and groundwater by PFAS, exceeding international regulatory guidelines in a large fraction of samples. The current underestimation of environmental PFAS burden arises from the limited suite of PFAS typically quantified and the existence of numerous PFAS precursors that degrade into regulated PFAS in the environment. The varying regulatory thresholds globally further complicate a comprehensive assessment. The higher exceedance rates in samples with known sources compared to those with unknown sources highlight the importance of identifying and mitigating point sources of PFAS contamination. The dominance of fluorotelomers and PFCAs in consumer products, coupled with their incomplete quantification by current methods, underscores the need for more comprehensive analytical approaches and broader monitoring efforts. The discrepancies in PFAS quantification between different methodologies highlight the critical need for standardization and development of advanced analytical techniques capable of measuring a broader range of PFAS, including FTOH and other under-measured classes.

This research highlights a significant global PFAS contamination problem, exceeding various international regulatory standards and advisories in surface and groundwaters. The underestimation of the environmental burden is substantial due to limitations in current analytical methods that fail to capture a vast number of PFAS, including significant precursors that transform into regulated PFAS over time. Future research should focus on developing comprehensive analytical techniques for a wider range of PFAS, implementing more robust sampling strategies globally, and fully assessing the human and ecological impacts of the diverse PFAS present in the environment. Standardizing the TOP assay and developing robust methods to quantify FTOH are crucial steps.

The study's reliance on existing datasets limits its ability to fully represent the global distribution of PFAS, potentially leading to bias related to sampling location and analytical methods employed in previous studies. The use of random values to replace BDL data introduces uncertainty, though random substitution is a common approach to address this issue in environmental studies where it is not practical to perform extensive statistical analysis to account for censored data. The lack of consistent PFAS quantification across all studies makes direct comparisons challenging. Furthermore, the lack of a globally standardized TOP assay method and its limitations prevent definitive conclusions on the full extent of PFAS precursors. The use of a major city within the country to replace undefined locations creates uncertainty in the precision of the locations, but attempts to mitigate the uncertainty of not having precise coordinates.