EPA's PFAS Rule Costs $1.5 Billion a Year. Most Utilities Hadn't Even Started.
Summary
The 2024 rule setting the first drinking-water limits on PFAS runs $1.548 billion a year, by EPA's own accounting — and when GAO checked, 77% of water systems already over the limit hadn't finished installing treatment. In May 2026, before that gap closed, EPA proposed two more years on the deadline and moving to drop four of the six regulated chemicals entirely.
What the rule actually costs
Almost none of the $1.548 billion is paperwork. Ninety-seven percent of it — $1.506 billion — is treatment and disposal: the granular activated carbon, ion exchange, or reverse-osmosis systems utilities install at the tap, plus safely handling the -laden waste those systems generate. Monitoring, state oversight, and utility administration together account for the remaining $42 million.
View data as table
| Treatment & disposal | $1,506M | 97% of total |
|---|---|---|
| Monitoring | $36M | |
| State/primacy agency admin | $5M | |
| Water system admin | $1M |
's own analysis found the rule's quantified health benefits — fewer cancers, heart attacks, strokes, and birth complications — run almost exactly as high: $1.549 billion a year, preventing an estimated 9,600 deaths and 30,000 serious illnesses over time. On paper, costs and benefits are close to a wash. The 66,000 systems bearing the cost and the roughly 100 million people meant to see the benefit are not the same population, which is where the second number comes in.
The compliance gap
Before the federal rule even took effect, six states — Illinois, Massachusetts, New Hampshire, New Jersey, Ohio, and Vermont — already regulated PFOA or PFOS on their own. The Government Accountability Office surveyed public water systems in those states that had PFOA or PFOS at or above 's future limits, and asked whether they'd treated for it. As of 's October 2023–January 2024 survey, an estimated 77% had not yet fully implemented a treatment method — and at least 3.6 million people in 's sample were served by a system that hadn't started. Utilities cited funding as the most common obstacle: an estimated 65% of systems that had begun treating and 68% of those that hadn't said they faced challenges getting federal or state money for it. Among systems that had installed treatment, 86% of large ones said simply explaining health risks to their own customers was difficult.
That gap is the backdrop for what is proposing to do about the deadline it originally set.
Two years, four chemicals
The first of the two May 2026 proposals would let systems request a two-year extension on the PFOA and PFOS limits — the deadline moving from April 26, 2029 to April 26, 2031 — citing construction-materials inflation the agency says ran over 20% between 2024 and 2025 alone, per the Bureau of Labor Statistics, plus a shortage of certified operators qualified to run advanced treatment. 's own cost-benefit table for the extension puts a number on that tradeoff directly: utilities save $90.2 million to $188.2 million (3% and 7% discount rates), against $79.3 million to $105.7 million in health benefits delayed by the same two years.
View data as table
| Utility savings, 3% discount rate | $90.2M |
|---|---|
| Utility savings, 7% discount rate | $188.2M |
| Forgone health benefits, 3% discount rate | $79.3M |
| Forgone health benefits, 7% discount rate | $105.7M |
The second proposal is a harder stop, not a delay. It would rescind federal drinking-water limits entirely for PFHxS, PFNA, HFPO-DA ("GenX"), and any mixture of those three plus PFBS — four of the six chemicals the 2024 rule covers — on the grounds that the Biden didn't follow required procedure in regulating them. The dollar case for doing so is thin: in the 2024 rule's own national cost analysis, the portion attributable to PFHxS alone was $11.6 million, and 's broader sensitivity estimate — folding in fuller costs for PFNA and HFPO-DA too — puts the total cost of regulating all four at up to roughly $93.6 million, against a $1.631 billion fully-loaded rule total. That's 6% of the cost, for chemicals that aren't rare: 's own occurrence data project PFNA alone detectable in water serving 21.2 million people, 227 systems serving 711,000 of them above the level had flagged for health concern.
The takeaway
- The rollback's savings are a rounding error. Rescinding four of six regulated removes at most about $93.6 million of a $1.631 billion cost estimate — roughly 6%. The other 94% is tied to PFOA and PFOS, which aren't being dropped, just given two more years.
- Most systems weren't close to finished anyway. Before proposed any extension, 77% of surveyed systems already over the limit hadn't installed treatment, and funding and workforce shortages — the same reasons now cites for the delay — were already utilities' top-named obstacles.
- "Rare chemical" isn't the right description. PFNA, one of the four proposes to stop regulating, is projected to be detectable in water serving over 21 million people.
Cost, system-count, and population figures are 's and 's own national-level estimates, not independently audited totals; 's 77% figure is drawn from a six-state sample, not a national survey. Both 2026 proposals were open for public comment as of this writing — comments close July 20, 2026 — and neither is a final rule.
Sources
- , National Primary Drinking Water Regulation, final rule, 89 Fed. Reg. 32532 (Apr. 26, 2024) — the $1,549 million national cost estimate, the 4,100–6,700 systems expected to need treatment, and PFNA occurrence/population figures. federalregister.gov
- , Benefits and Costs of Reducing in Drinking Water fact sheet (April 2024) — the $1.548 billion annual cost broken into monitoring, treatment/disposal, and administrative categories, and the $1.549 billion quantified benefit estimate. epa.gov
- U.S. Government Accountability Office, Persistent Chemicals: Additional Actions Could Help Public Water Systems Address in Drinking Water, -24-106523 (Sept. 24, 2024) — the 77% treatment-implementation gap, the 3.6-million-person undertreated population, and utility-reported funding and communication challenges, from a six-state survey (Illinois, Massachusetts, New Hampshire, New Jersey, Ohio, Vermont). gao.gov
- , Extending the Compliance Deadline for the PFOA and PFOS Maximum Contaminant Levels, proposed rule, 91 Fed. Reg. (May 20, 2026), Docket -HQ-OW-2025-0654 — the proposed 2029-to-2031 deadline extension, the construction-cost-inflation and operator-workforce rationale, and the cost-savings/forgone-benefits estimates. federalregister.gov
- , Rescission of Regulatory Determinations and Removal of Related Provisions for Four Substances (PFHxS, PFNA, HFPO-DA, and the Mixture of These Three Plus PFBS), proposed rule, 91 Fed. Reg. 29413 (May 20, 2026) — the proposal to stop regulating four of the six covered by the 2024 rule. federalregister.gov
Comments
Always open. Logged-in readers can annotate paragraphs in place.
In April 2024, set the first legally enforceable drinking-water limits on — the "forever chemicals" linked to cancer, thyroid disease, and low birth weight — covering six compounds and an estimated 66,000 public water systems. put the national price tag at $1.548 billion a year and gave systems until April 2029 to comply. Before most of them finished, the agency started walking part of the rule back. Two proposals published in the Federal Register on May 20, 2026 would give systems two more years on the two chemicals that matter most to the price tag, and drop federal limits on the other four entirely.