The US PFAS rule: what is enforceable now, and what the 2029 date actually starts
Six numbers, two dates and one misreading that decides real money. The federal drinking water limits for PFAS were finalised in April 2024 and all six remain in force today. Alongside them, two EPA proposals would extend the PFOA and PFOS compliance deadline to 2031 and rescind four of the six limits — and neither proposal is final. The misreading is to plan a treatment train around 2029 as though it were the only date that matters, when 2029 is the end of a monitoring cycle rather than the beginning of the problem.
Three states are kept apart on this page: proposed, in force, and the compliance date. The drinking-water limits shown apply to public water systems; private wells are not covered by them.
The six numbers
The 2024 National Primary Drinking Water Regulation set individual limits for five PFAS compounds, and a mixture index that covers four of them. Five of the six figures are in parts per trillion — a unit worth pausing on: one part per trillion is roughly one drop in twenty Olympic swimming pools. The sixth is a unitless index.
| Compound | Limit | Where it came from |
|---|---|---|
| PFOA | 4 ppt | Legacy fluoropolymer manufacturing; the older of the two most widely known PFAS. |
| PFOS | 4 ppt | Aqueous film-forming foam and older stain-resistant chemistry. |
| PFHxS | 10 ppt | Largely a replacement chemistry for PFOS. |
| PFNA | 10 ppt | Fluoropolymer production by-product. |
| HFPO-DA (GenX) | 10 ppt | A processing aid introduced as a replacement for PFOA. |
| Hazard Index mixture | 1.0 (unitless) | A combined-exposure measure across PFHxS, PFNA, HFPO-DA and PFBS — the four are assessed together rather than one at a time. |
The four compounds in the index act on the same endpoint, so assessing each against its own threshold would understate a supply in which all four are present below their individual limits. The index adds their relative contributions and compares the sum with 1.0. It is a different kind of standard from the others in the table, and it is the one most often left out when the rule is summarised.
The two dates, and the one that moved
The rule has a monitoring date and a compliance date, and in 2026 a third date — a proposed extension — was added to the picture. Only two of the three are in force.
Where this stands in 2026: in force, or proposed
This is the section that has to be read with its date attached, because it is the part of the picture that moves. As at 30 September 2026:
In force
- All six limits from the 2024 rule remain in force. The PFOA and PFOS limits at 4 ppt, the three individual 10 ppt limits, and the Hazard Index of 1.0.
- The monitoring deadline of 26 April 2027 and the compliance deadline of 26 April 2029, both unchanged.
- Reporting obligations — systems report results to the primacy agency and publish detectable PFOA and PFOS levels in their annual consumer confidence report.
Proposed, or in litigation — not settled
- A proposal published in May 2026 would keep the PFOA and PFOS limits but let a system apply for two more years, moving its compliance date to 2031.
- A second proposal from the same period would rescind the limits for PFHxS, PFNA, HFPO-DA and the Hazard Index mixture. The comment period closed in July 2026 and no final rule has been issued.
- Litigation continues in parallel. The D.C. Circuit declined requests to set the challenged limits aside while the proposals proceed, heard argument in September 2026, and has yet to rule. A decision was widely expected in the second half of 2026.
Three things are routinely conflated: what is in force, what has been proposed, and what a court may yet do. A proposal does not change a limit, and a case being heard does not suspend one. Where a document states a compliance date without saying which of the three it belongs to, treat the date as unverified until you have established that. This article is dated for that reason — the position above is stated as at 30 September 2026, and the underlying documents are listed in the sources at the foot of the page.
What the federal rule does not reach
The federal limits apply to public water systems. A private household well is not a public water system, so the rule imposes nothing on it: no monitoring requirement, no compliance date, and no enforcement. That is not a loophole — it is the boundary of the statute the rule is made under — but it has a practical consequence that is easy to overlook.
| Supply | Position |
|---|---|
| Community and non-transient public water systems | Covered. Monitoring by 2027, compliance by 2029, results reported to customers. |
| Private household wells | Not covered. No obligation is placed on a well owner by this rule, and no date in it creates one. Testing and any treatment are the owner's decision. |
| Bottled water | Regulated separately, by a different agency and under different limits. |
| State-level limits | Several states set their own PFAS limits, and some are stricter than the federal values. A state rule is not displaced by anything in the federal proposals. |
For a well owner, the practical upshot is inverted. There is no deadline to wait for and no agency that will tell you your water is affected — which means the decision rests entirely on whether the water has been tested. A test is the only route to a fact here, and its absence is not evidence that the water is clear. The treatment question that follows is a separate one, and it is answered on the well water page.
Why 2029 is the wrong date to plan around
2029 is a compliance date, which makes it an end point. Planning a treatment train backwards from an end point tends to produce the same three errors.
- It treats a deadline as a start date. A treatment train is specified, then procured, then installed, then commissioned, and on a municipal scale it may involve land, power, permitting and a capital programme. Working back from 2029, the decision points sit well before it — which is precisely why the monitoring phase runs from now until 2027. The monitoring results are what the design is based on, so the schedule that matters is the monitoring schedule.
- It overlooks that the deadline may move. The extension proposal, if it is finalised, would not apply automatically: a system would apply for it. A plan that assumes an extension before it exists is a plan built on a draft. A plan that ignores the possibility is a plan with no headroom. Both failures come from the same place — treating one date as fixed.
- It treats the four rescinded limits as one problem. If the four-limit rescission is finalised, PFOA and PFOS remain the two compounds with a settled, maintained limit. A treatment chain chosen for the whole set may be more than the retained limits require; one chosen for PFOA and PFOS alone may need revisiting if the rescission does not go through. The robust position is to design the chain so that stages can be added or bypassed, rather than to optimise it for one version of the rule.
Whatever happens to the four limits, one question does not change: what is in the water after the last treatment stage, and before the tap? Every stage in a PFAS-driven chain — carbon, ion exchange, reverse osmosis — is followed by a storage tank, a length of tubing or a pressure vessel, and none of those is sterile. That part of the problem sits downstream of the rule and will still be there when the litigation ends.
Where ultraviolet fits — and where it does not
Ultraviolet treatment does not remove PFAS. Ultraviolet light does not destroy the carbon-fluorine bond at the wavelengths and contact times used in a point-of-use device. Any page or product that presents UV as a PFAS treatment is presenting something that is not the case. The limits in this article are met by adsorption and separation stages — granular activated carbon, anion exchange and reverse osmosis — not by ultraviolet.
That said, a PFAS-driven installation is one of the situations in which a UV stage most often earns its place — and the reason is the paragraph above. The chain that removes PFAS is a chain of removal stages, and removal stages are followed by storage. Water that has just left a membrane or a carbon bed is clean; water that has been sitting in a tank on the permeate side is water that has had time, surface area and no residual disinfectant.
The same logic applies to a household well, where the federal rule imposes nothing at all. If a test shows PFAS above a level the owner is not willing to accept, the response is a removal stage — and the same tank sits downstream of it. The rule decides nothing there; the water decides everything.
The US PFAS rule in practice
Has the federal PFAS rule been rolled back?
Not as at 30 September 2026. Two EPA proposals are in progress — one to extend the PFOA and PFOS compliance deadline to 2031 on application, one to rescind the limits for PFHxS, PFNA, HFPO-DA and the Hazard Index mixture — and neither has been finalised. All six limits from the 2024 rule remain in force, and the 2027 and 2029 dates are unchanged. Litigation over the limits is running in parallel and had not been decided when this page was published.
Does the 2029 date apply to my house?
Only if your water comes from a public water system. The rule applies to public water systems, so it creates no obligation and no deadline for a private well. It also does not tell a household what to do — where a supply is a well, the decision to test and the decision to treat are the owner's, and the first of those is the one that produces a fact.
Is a result above 4 ppt a violation today?
For a public water system in 2026, generally no — the treatment obligation runs to the 2029 compliance date, so a detection above the limit is a signal that treatment will be needed, not a present violation. That distinction matters for how a monitoring result should be read: it is the start of a design process rather than a finding of breach. Formal violations arise against the compliance and monitoring requirements, which are separate obligations.
Do these limits apply outside the United States?
No. This is a United States federal rule made under United States drinking water legislation. Other jurisdictions set their own limits or none, and several United States states set their own on top of the federal values. Where a product is sold into a market, the requirements of that market govern — and this page should be read as a description of the United States position only.
Does a UV unit help with PFAS at all?
Not with the PFAS itself: ultraviolet treatment does not remove or destroy PFAS at the wavelengths and contact times used in point-of-use equipment. It can, however, be part of a PFAS-driven installation for a different reason — the removal stages that meet the limit are followed by storage, and water held in a tank can support bacterial regrowth. That is the problem a UV stage placed after the tank addresses, and it is why the two appear together without UV contributing to the PFAS number. See the PFAS water-problem page for the stage-by-stage position.
A PFAS-driven installation, or a well with a result?
Send the test result, what is already on the line, and the destination market. Where the question is compliance, the answer depends on which supply you are on and which date applies to it — and where it is a private well, the rule imposes nothing and the result is the whole of the input.
- Models: ZXSJ-UVC-1GPM (1 GPM / 4 LPM, 1/4") ｜ ZXSJ-UVC-3GPM (3 GPM / 11.4 LPM, 3/8")
- Position in a removal chain: after any storage tank, immediately before the point of use
- Stated plainly: the UV stage does not contribute to a PFAS limit — this article sets out what it does address
Send an enquiry
Typical response within one business day.
This form is not yet connected to a mail service — the enquiry endpoint will be wired up at launch. Until then, enquiries reach the manufacturer through the contact page.
Sources and scope
5 sources
- The maximum contaminant levels for PFOA, PFOS, PFHxS, PFNA and HFPO-DA, the Hazard Index of 1.0, the initial monitoring deadline of 26 April 2027 and the compliance deadline of 26 April 2029: the 2024 National Primary Drinking Water Regulation for PFAS, as described in the United States Environmental Protection Agency's rulemaking record and in the Federal Register notice published on 20 May 2026 (Vol. 91, No. 97) concerning the proposed compliance extension.
- The two proposals described in this article: the EPA's proposed PFOA and PFOS compliance extension rule, announced on 18 May 2026, for which the comment period closed on 20 July 2026; and the separately proposed rescission of the regulatory determinations and standards for PFHxS, PFNA, HFPO-DA and the Hazard Index mixtures, for which a public hearing and comment period ran sequentially with the first. Both are proposals. Neither is a final rule.
- The litigation position: proceedings in the United States Court of Appeals for the District of Columbia Circuit concerning the 2024 rule, in which the court declined to set the challenged limits aside and heard oral argument in September 2026, with a decision pending at the research cut-off for this article.
- Deliberately not stated: any treatment performance figure for any technology, any removal percentage, and any prediction of the outcome of the pending proposals or the litigation. No UV dose, log-reduction or inactivation figure is asserted for our product anywhere on this site, and none is implied by the use of a UV stage in the chain shown above.
- Research cut-off for this article: 30 September 2026. Regulatory positions change; the date at the top of the page is the date the position was last checked.