USGS estimates at least 45% of U.S. tap water contains PFAS

Regulatory background

• Per- and polyfluoroalkyl substances (PFAS) are a class of synthetic organic chemicals used in many different industrial and commercial applications. Although no single definition of “PFAS” exists, the chemicals are generally identified by a chain of carbon atoms saturated with fluorine atoms. These carbon-fluorine bonds are some of the strongest in organic chemistry, meaning that most PFAS readily persist in the environment.
• The Environmental Protection Agency (EPA) has taken several steps to reduce and/or study PFAS in the environment and in humans, most recently releasing a framework for new PFAS entering the market [for an analysis of the new framework, see the AgencyIQ piece here]. Under the framework, new PFAS are expected to be considered persistent, bioaccumulative, and toxic (PBT) chemicals, which invokes another level of regulation for these future PFAS.
• As there are thousands of different PFAS chemicals, testing for discrete PFAS in drinking water or other environmental media can be difficult. A water source may have contamination by PFAS not covered by a test method, as the current approved EPA methods for testing only test for 18 (Method 537.1) and 25 PFAS (Method 533), respectively. Testing for total organic fluorine (TOF) content can provide information on the total amount of PFAS in a drinking water source, but EPA has not approved a specific test method for it yet and TOF testing does not provide insight on the specific PFAS present in the water source.
• PFAS regulation has become more and more prevalent over the past decade, with states and the federal government beginning to prohibit or limit the substances in drinking water, consumer products, and more. EPA has published a proposed National Primary Drinking Water Regulation (NPDWR) that would cover six PFAS: perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorononanoic acid (PFNA), hexafluoropropylene oxide dimer acid (HFPO-DA or GenX), perfluorohexane sulfonic acid (PFHxS), and perfluorobutane sulfonic acid (PFBS) [see AgencyIQ analysis]. The Senate’s Environment and Public Works Committee (EPW) also recently introduced a bill to accelerate PFAS research in the country and set enforceable deadlines for the EPA’s PFAS NPDWR. State regulations have also targeted PFAS levels in consumer products and certain states have already set action levels for PFAS in drinking water.

Survey basics

• On July 5, 2023, the U.S. Geological Survey (USGS) released Per- and polyfluoroalkyl substances (PFAS) in United States tapwater: Comparison of underserved private-well and public-supply exposures and associated health implications, a study on PFAS contamination in both private wells and public water systems (PWS). The study tested tap water collected from 716 locations in the U.S. from 2016-2021 in addition to temporal sampling at three separate locations. The survey authors utilized the test results to compare to nearby land uses and potential sources of PFAS to better understand PFAS contamination from sources and form a more useful hypothesis on the estimated number of households in the U.S. whose drinking water may contain PFAS.
• The study authors utilized volunteers to conduct many of the tap water samples. 409 of these samples were collected at point-of-use (e.g., at the tap) from 155 private wells and 252 PWS in all 50 states, the District of Columbia, Puerto Rico, and the U.S. Virgin Islands. Two private wells and one PWS were also sampled repeatedly for a week, and then weekly for up to two months to review PFAS concentrations shifting over time. The samples also included 307 tap water samples conducted as part of a USGS point-of-use tap water research study conducted between 2016 and 2021.
• The samples were analyzed by one of three labs: the EPA’s National Exposure Research Laboratory (NERL), the Colorado School of Mines (CSM) laboratory, and the USGS National Water Quality Laboratory (NWQL). Each lab utilized different methods for determining the number of PFAS in the samples. NWQL reviewed the most samples at 608, followed by NERL providing analysis of 26 samples and CSM providing analysis of 82. NERL’s method only analyzed for 10 total PFAS, whereas CSM analyzed for 28 and 44 PFAS, and NWQL analyzed for 34 total PFAS. Each of these methods does not appear to be one of the traditional PFAS detection methods used by the EPA.
• The method detection limits ranged from .1 to 61.8 nanograms per liter (ng/L or parts per trillion (ppt)). In addition, the study authors attempted to account for known bias associated with variability in the specific detection limits.

Survey results

• The survey found at least one PFAS in 30% (237 of 716) samples. Seventeen PFAS were detected at least once, with PFBS being most often detected (16% of samples) along with PFHxS (15%) and PFOA (14%). The study authors reviewed test results of between one and nine total PFAS detected in a specific sample (with a median of two detected PFAS), and a median concentration of 2.88 ng/L.
• At least one PFAS was detected in 20% of private wells and 40% of PWS. In the eastern U.S., this number rose to 60% for PWS detections. The study authors found that median cumulative PFAS concentrations stayed relatively the same between private wells and PWS.
• The temporal sampling at a private South Carolina well resulted in no PFAS detections over the course of three months of sampling, but PFAS were detected at the other two temporal testing locations at a private well and PWS in New Jersey. The study authors found that the cumulative detected concentrations were relatively stable, with potential changes in the detected concentrations likely leaning towards the fact that some of the PFAS detected were in concentrations close to the lower bound of the detection limit. However, none of the tests exceeded New Jersey’s maximum contaminant limits (MCLs) it had set for three PFAS (PFOA, 14 ng/L; PFNA 13 ng/L; and PFOS 13 ng/L).
• The proposed PFAS NPDWR’s MCLs for PFOA and PFOS (both 4 ng/L) were below the reporting limit for two of the three laboratories (including the NWQL). The proposed MCLs were exceeded 6.7% of the time for PFOA and 4.2% of the time for PFOS across the tests but exceeded that 4 ng/L limit 48% of the time for PFOA and 70% of the time for PFOS when detected.
• The authors also focused on land uses nearby to samples that detected PFAS to better understand PFAS contamination from certain potential sources (such as airports, military bases and installations, oil and gas development, and waste management sites). For individual PFAS, the distance to a probable PFAS source was not a strong predictor of concentration (except for PFBS, which exhibited a positive relation with development and pasture agriculture). In addition, concentrations of PFBS, PFHxA, PFHxS, PFOA, and PFOS all decreased with surrounding cultivated cropland. The authors theorized that point-of-use tap water exposure to PFAS is more closely associated with the type of PFAS source, rather than the number of potential PFAS sources nearby.

Implications of the survey

• The survey portrays a stark picture of potential PFAS contamination at the tap around the U.S. About 40 million Americans rely on private wells, and 52 million rely on small water supplies that serve less than 10,000 people. Sampling practices under the Unregulated Contaminant Monitoring Rule (UCMR) have previously not focused on wells and small systems, but the latest UCMR 5 contains a range of systems under 10,000 and will test for 29 separate PFAS in those systems. However, UCMR 5 does not cover well testing, meaning that the EPA will not have data on incidence of PFAS in water supplies for about 12% of the U.S.
• The modeled results expect that at least one PFAS will be detected in about 45% of U.S. drinking water supplies, with the potential for urban water supplies to detect at least one PFAS upwards of 70% of the time. With additional PFAS regulation in drinking water likely on the way in the form of other NPDWRs that target additional PFAS contaminants, water systems around the country will likely need to install additional infrastructure upgrades to meet PFAS concentration limits.
• The study may prove useful to plaintiffs in the multitude of PFAS-related litigation around the country. Although some of the largest defendants in the South Carolina aqueous film-forming foams (AFFFs) multi-district litigation recently reached preliminary settlements with a group of public water systems, the settlement would not cover personal injury claims as well as claims by water systems that are not currently required to test for PFAS under EPA monitoring rules. The settlements would also not resolve claims by nearly half of the state attorneys general against PFAS manufacturers. As the total settlement amount has crossed $10 billion, it is more and more likely that a global settlement of PFAS claims, whether for personal injury or by the state attorneys general, may be on the horizon, like the Tobacco Master Settlement of 1998.

To contact the author of this analysis, please email Walker Livingston.
To contact the editor of this analysis, please email Patricia Iscaro.

Key Documents and Dates

• Per- and polyfluoroalkyl substances (PFAS) in United States tapwater: Comparison of underserved private-well and public-supply exposures and associated health implications (July 5, 2023)
• USGS press release (July 5, 2023)
• Framework for TSCA New Chemicals Review of PFAS Premanufacture Notices (PMNs) and Significant New Use Notices (SNUNs)
• Proposed National Primary Drinking Water Regulation