EPA's PFAS Rule in 2026: Bifurcated Compliance Timelines After the 2025 Reset

When EPA finalized the PFAS NPDWR in April 2024, the operational narrative was clean. Six PFAS were regulated. Initial monitoring was required by 2027. Treatment compliance was required by 2029. Utilities had five years to characterize their systems and install treatment where needed.

Eighteen months later, that narrative is no longer accurate. On May 14, 2025 EPA announced its intent to make two structural changes. It would propose to extend the PFOA and PFOS treatment compliance date from 2029 to 2031, and it would propose to rescind the MCLs for perfluorohexane sulfonic acid (PFHxS), perfluorononanoic acid (PFNA), hexafluoropropylene oxide dimer acid (HFPO-DA, commonly called GenX), and the mixture-based Hazard Index. The 2027 monitoring obligations and the PFOA/PFOS MCLs at 4.0 ng/L would stay.

Those proposals are now on the table in concrete form. Both rules published in the Federal Register on May 20, 2026: the compliance extension at 91 FR 29425 and the rescission at 91 FR 29413. The comment period closes July 20, 2026, with a public hearing on July 7, 2026. As of this writing, neither change is final. The 2024 rule, in full, remains the operative law.

That distinction matters. For utilities scoping treatment, and for consultants supporting them, the operative picture is no longer a single-deadline rule. It is a bifurcated picture: monitoring on the original schedule, treatment on a timeline EPA has proposed to extend, and three secondary PFAS whose federal status is genuinely uncertain pending a final rule. This post takes a position on what that bifurcation means for the next five years of utility work, and on the parts of the story that the headline version misses.

The April 2024 NPDWR is the largest expansion of the federal drinking-water regulatory framework in decades. Three pieces of it are operative as of mid-2026 regardless of the 2025 reset, and a fourth is worth naming because it shapes the treatment conversation.

First, PFOA and PFOS are regulated at MCLs of 4.0 ng/L, or 4.0 parts per trillion (ppt). The non-enforceable health goals, the Maximum Contaminant Level Goals (MCLGs), are set at zero; the 4.0 ng/L MCLs are feasibility-based, set at or near the level routine drinking-water laboratories can reliably report. Compliance monitoring runs on two approved analytical methods, EPA Method 533 and EPA Method 537.1. That number is not negotiable in the current regulatory state, and the 2025 proposals do not touch it.

Second, initial monitoring is required by 2027. The schedule allows reduced monitoring frequency for systems with consistently low or non-detect results, but the obligation to characterize each system's PFAS levels is not deferred by the 2025 reset. The proposed extension rule is explicit on this: even systems granted the treatment exemption must still meet all monitoring and reporting deadlines under the 2024 rule.

Third, public notification requirements are tied to MCL exceedances. A system that exceeds the PFOA or PFOS MCL must notify its customers regardless of treatment status.

Fourth, and outside the drinking-water rule itself, EPA designated PFOA and PFOS as hazardous substances under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) in a final rule at 89 FR 39124 (May 8, 2024), effective July 8, 2024. EPA announced in September 2025 that it would retain that designation. This sits in the background of every treatment decision, because it governs how PFAS-laden treatment residuals are handled. More on that below.

These pieces are the floor. The 2025 reset modifies the treatment timeline and proposes to remove three secondary PFAS and the Hazard Index. It does not undo the monitoring or notification structure.

The May 2025 announcement, now embodied in the May 2026 proposed rules, made two distinct changes that should be tracked separately. The popular shorthand flattens both, and the flattening is where planning errors start.

The first is the treatment extension, and it is not a blanket two-year slip. EPA proposes a federal exemption under SDWA 1416(f) and 1450(a)(1) that extends the PFOA and PFOS compliance date from April 26, 2029 to April 26, 2031 only for systems that submit a request and qualify. The MCLs themselves stay at 4.0 ng/L. Two operational details carry real weight for capital planning. Systems must affirmatively apply for the exemption rather than receive it automatically. And systems with PFOA or PFOS at or above 12 ng/L that take the exemption must implement at least two interim control measures during the 2029 through 2031 window, such as certified point-of-use (POU) filters, an alternative water source, or targeted public education. For the highest-concentration systems, which are precisely the ones most likely to want more time, the extension comes with an interim-treatment obligation attached.

The second change is the rescission of the MCLs for PFHxS, PFNA, HFPO-DA, and the Hazard Index. EPA's stated basis is legal and procedural, not scientific: the Agency argues the prior administration acted unlawfully by proposing and finalizing the regulatory determinations and the regulations for these compounds simultaneously, which the Agency now reads the SDWA to prohibit. EPA has been explicit that this is not a reassessment of the underlying health or occurrence data. The Agency has also noted that, because the matter is being reopened, the eventual outcome could be more stringent rather than less. Nothing here is settled.

Two clarifications on scope. PFBS never carried an individual MCL; perfluorobutane sulfonic acid (PFBS) appeared only as one component of the Hazard Index mixture, so the rescission affects three individual secondary PFAS plus the mixture standard, not four individual chemicals. And until a final rescission rule issues, all four standards remain legally in force. Describing the secondary PFAS as already deregulated, as much of the trade coverage does, is premature.

The reset reads as a clean policy pivot. It is better understood as the product of a litigation dead-end, and that context tells a utility how durable to expect the outcome to be.

The 2024 rule was challenged in the U.S. Court of Appeals for the D.C. Circuit by the American Water Works Association (AWWA) and the Association of Metropolitan Water Agencies (AMWA), in a petition consolidated under No. 24-1188. The challenge did not dispute the goal of regulating PFAS; it argued that EPA had not followed the SDWA process and had not used the best available science and cost data. In September 2025 EPA told the court it would continue to defend the PFOA and PFOS MCLs but would seek to vacate the standards for the four other compounds. The court declined to take that shortcut: it denied EPA's motion for partial vacatur on January 21, 2026 and set a merits briefing schedule.

That denial is the reason the rescission is now proceeding through notice-and-comment rulemaking rather than a court order. It also exposes the structural constraint that shapes everything EPA can do here. The SDWA's anti-backsliding provision, Section 1412(b)(9), bars EPA from revising an MCL in a way that reduces health protection. That is why the rescission is framed as correcting an unlawful procedure rather than reweighing the science: a straightforward "we decided these limits are too strict" would run directly into anti-backsliding. The procedural framing is the only viable path, and it is also the more legally contestable one. For a utility, the practical read is that the secondary-PFAS rescission is more exposed to reversal, on appeal or under a future administration, than the PFOA/PFOS extension, which EPA is defending rather than unwinding.

The picture across the country in 2026 is two-tiered.

Large surface-water utilities, especially those with prior PFAS detections under the third Unregulated Contaminant Monitoring Rule (UCMR 3) cycle, have generally moved. Initial monitoring is underway or complete. Treatment alternatives have been evaluated, often against multiple PFAS rather than just PFOA and PFOS. Granular activated carbon (GAC), anion exchange (AIX), and reverse osmosis or nanofiltration (RO/NF) are the three Best Available Technologies the analysis centers on, with cost and operational tradeoffs that vary by source water and existing treatment train. Several of these utilities began treatment construction before the 2025 reset and are continuing on the original schedule.

Smaller systems are a different story. Many community water systems (CWSs) below 10,000 connections have not completed initial monitoring. Cost, capacity, and competing priorities mean sampling is often outsourced, slow, and uncoordinated with treatment planning. For these systems the proposed exemption could create a planning buffer, but the buffer only exists for systems that apply for it, and the planning has not yet been done. The gap between the two tiers is best stated as a practitioner observation rather than a measured figure; the supporting numbers, if quantified, should be drawn from the fifth UCMR (UCMR 5) occurrence dataset segmented by system size.

This compliance gap will be the operative story of 2027 and 2028. The 2027 monitoring deadline will produce a wave of detections that small systems must act on. Even with a granted exemption, the treatment clock runs to 2031, and treatment selection, pilot testing, design, permitting, procurement, and construction can easily absorb three of those years. The long window is narrower than it looks.

The headline treatment conversation stops at technology selection. The engineering reality does not, and this is where the CERCLA designation re-enters.

Every PFAS treatment train generates a concentrated PFAS waste stream. GAC produces spent carbon. AIX produces spent resin. RO and NF produce a reject concentrate that can carry the removed PFAS at elevated levels. None of these residuals is destroyed by the treatment process; the contaminant is moved, not eliminated. Disposal and destruction pathways remain unsettled: thermal destruction efficacy is still under study, landfill disposal raises leachate-recirculation questions, and concentrate management for membrane systems is a site-specific problem with limited clean options. EPA's own guidance on the subject, the Interim Guidance on the Destruction and Disposal of PFAS (Version 2, 2024), reads as provisional rather than settled.

The CERCLA hazardous-substance designation for PFOA and PFOS sharpens the stakes. Handling, transporting, and disposing of PFAS-bearing residuals now carries Superfund liability exposure that did not exist before July 2024. A treatment evaluation that selects a technology without costing and siting its residual-management pathway is incomplete, and for a consulting deliverable it is a defensible-scope gap. This is the part of PFAS compliance least visible in the rule text and most likely to surprise a utility late in design.

The shift from a single-deadline rule to a bifurcated timeline changes the shape of consulting demand through 2031.

The 2027 monitoring obligation produces near-term work in 2026 and 2027: sampling plans, laboratory procurement, quality-assurance review, and the regulatory reporting that follows initial detection. The 2031 treatment obligation produces a longer wave across 2027 through 2030: treatment-alternative evaluation, residual-management and disposal planning, pilot testing, capital planning, and detailed design. The proposed extension makes this wave broader and less compressed than the original 2029 deadline would have, but only for systems that elect the exemption, and the exemption carries the interim-measures condition for high-concentration systems.

The proposed rescission of the secondary-PFAS MCLs does not remove those compounds from project work. State PFAS programs, National Pollutant Discharge Elimination System (NPDES) discharge limits, site assessments at suspected source areas, and biosolids management all continue to involve PFHxS, PFNA, and HFPO-DA. Several states maintain their own enforceable PFAS standards independent of the federal rule, so in those jurisdictions a federal rescission may change little in practice. The federal drinking-water layer may narrow; the broader PFAS regulatory landscape has not.

For utilities, the practical implication is that 2026 is the year to align monitoring, treatment evaluation, residual-management planning, and capital programming into a single integrated schedule rather than running them in sequence. The recalibrated timeline is generous enough that a poorly sequenced compliance program now looks like a choice, not a constraint.

The window for input is open now and short. Both proposed rules published May 20, 2026 and the comment period closes July 20, 2026. EPA will hold a virtual public hearing on July 7, 2026; pre-registration to give verbal comment closes July 1, 2026, and comments on the extension rule's information-collection provisions are due to the Office of Management and Budget (OMB) by June 22, 2026. Utilities and consultants with a view on the implementation details, particularly the exemption-application mechanics and the rescission of the Hazard Index that some state programs incorporate by reference, should be filing now rather than planning to.

The 2027 monitoring deadline remains unforgiving for systems that have not begun. It is not subject to the proposed extension, and enforcement discretion will not move it. Initial sampling should be scoped no later than fall 2026 to leave room for laboratory turnaround, quality-assurance review, and the reporting that follows a confirmed detection.

The PFAS rule has not collapsed. It has split into a binding monitoring track and a treatment track that EPA has proposed to recalibrate, against a backdrop of active litigation and an unresolved residuals problem. The firms that plan to the actual shape of the rule, rather than to either the 2024 headline or the 2026 one, will produce better work than the firms still reading from the press release.