EPA's PFAS TECH Strategy Funds Four Air Force Pilots but Defines No Destruction Standard

Lee Zeldin announced EPA's first agency-wide PFAS technology strategy on 26 October 2026 at the Klein Water Treatment Facility in Commerce City, Colorado. The venue made an argument the press release didn't. South Adams County Water and Sanitation District had just finished an $86 million ion-exchange upgrade there, nearly $61 million of it Bipartisan Infrastructure Law grant money, according to the contractor's project record. Ion exchange destroys nothing. It concentrates PFAS onto resin, and somebody then has to decide what the resin is. The EPA PFAS TECH strategy is the agency's answer, and it's a research answer rather than a regulatory one.

TECH stands for Technology, Evaluation, Coordination and Harmonization, and per WaterWorld's account it commits EPA's national and regional program offices to two years of funded research, field demonstrations and community technical assistance. Foam fractionation and supercritical water oxidation were named first. Grant money moves through the existing Science to Achieve Results and Small Business Innovation Research programs. "Technology innovation is essential to accelerating our efforts to address these contaminants," Zeldin said, and I don't disagree. But I read every agency announcement looking for the same thing: which document does a state permit writer cite the morning after?

Two years of demonstrations, and still nothing to cite

EPA reports better than 99.99% destruction of PFAS in aqueous film-forming foam using an industrial-scale supercritical water oxidation system, a figure Smart Water Magazine reported from the October announcement. Take it at face value. The number still doesn't tell a permit engineer which analyte list was run, what the detection limits were, whether the fluorine balance closed, or what left the vent. Destruction and removal efficiency means something precise for a hazardous waste incinerator because 40 CFR Part 264 Subpart O attaches a test method and a surrogate compound to the phrase. For PFAS there's no equivalent. So 99.99% is a measurement somebody made under conditions they chose.

The pilots themselves are not a press exercise. At Air Force bases in Utah, Colorado, Wyoming and South Dakota, EPA and the Air Force are testing electrical discharge plasma, high-pressure membranes, micro and nanobubble aerators, supercritical water oxidation and phytoremediation on contaminated groundwater (phytoremediation at a base boundary contains rather than destroys). Hydrothermal alkaline treatment sits in the research column alongside them. That slate covers more of the plausible technology space than EPA has funded.

Guidance that reissues every year can't underwrite a twenty-year asset

The document a destruction decision rests on is the Interim Guidance on the Destruction and Disposal of PFAS, published under a National Defense Authorization Act mandate and refreshed on 20 April 2026 for a 60-day comment period. Interim guidance binds nobody. The 2024 version said in plain terms that it mandates no recommended technology, and the update didn't move off that posture. What changed was the revision cycle, from the legally required three-year schedule to annual updates.

Annual updates sound responsive. Try financing against them. A thermal destruction unit is a fifteen- to twenty-year asset with a permit renewal cycle bolted to it, and the question a lender asks is whether the compliance basis at year eight resembles the one in the credit memo. Nobody can answer that about a document that reissues every twelve months.

A second wrinkle should bother anyone procuring thermal capacity. That April 2026 draft ranks the three established routes by environmental release risk and puts thermal treatment last, behind Class I underground injection and RCRA Subtitle C landfilling, mainly over products of incomplete combustion, per CIRS Group's summary of the ranking. Thermal treatment stays viable above roughly 1,100 degrees C with adequate mixing and residence time. So one EPA workstream funds thermal destruction research while another document tells a state that thermal destruction carries the highest release risk of the options on the table. Both positions are defensible alone. Read together by a reviewer with a queue, they supply a reason to say no and none to say yes.

The Air Force bases are the part that needs no permit

Now the part I'd push back on, which concerns site selection rather than science. Every demonstration venue named in the strategy sits inside a legal shelter that a municipal utility doesn't have. CERCLA Section 121(e)(1) exempts on-site response actions from federal, state and local permitting. The treatability study exclusions at 40 CFR 261.4(e) let a researcher run contaminated material through a novel unit in sample quantities without a treatment, storage and disposal permit. Both exist for good reason. They're also why a plasma pilot can stand up at an Air Force base inside one budget cycle while the identical skid at a publicly owned treatment works waits on a state construction permit, a Title V applicability determination and possibly a pretreatment program revision (the pretreatment piece is the one nobody budgets for). That asymmetry decides whether the data travels: numbers generated under a permit exemption reach a state agency carrying no permitting precedent, which is what a reviewer most wants and what a federal facility pilot can't supply.

And permitting timelines are where first-of-kind projects die. A Title V renewal I worked in 2023 tripped PSD review over a modification the whole room had called minor, and it cost the client eight months nobody had budgeted. The technology hadn't changed. The applicability analysis had. Where there's no federal performance standard to point at, a state reviewer builds that record from scratch, and such a record is slow to write and easy to challenge.

Biosolids gasification stalls on fluorine accounting, not heat

The EPA PFAS TECH strategy also puts research money toward whether gasification can help address PFAS in biosolids at wastewater treatment plants. That research line is the one I'd watch hardest, because it's a measurement problem wearing a technology costume. According to work published in Water Environment Research in April 2026, pyrolysis does remove PFAS from the solid phase at concentrations typical of biosolids without industrial inputs, with removal depending on the starting concentration and how far the pyrolysis goes. The same work tracked partitioning into the liquid and gas phases at 500 and 800 degrees C. Removal from the cake isn't destruction of the molecule.

Mineralization wants temperatures above the 1,100 degree threshold the disposal guidance uses, and sludge pyrolysis systems don't run there. So the fluorine is somewhere, and a permit condition can only be written against what a plant can measure at the stack. The analyzers in a sludge dryer cabinet are specified for criteria pollutants; no Siemens or Thermo train in that enclosure reports total organic fluorine. That gap is where monitoring conditions get added at renewal, and retrofitting an analyzer into a permitted stack is a capital conversation.

Where does this stop holding? A drinking water utility with a removal problem and a disposal contract gets the good news without the permitting exposure; the regulated party downstream inherits the destruction question. In states that moved first, an unsettled federal standard matters less than a settled state prohibition: Maine's 2022 ban on land-applying PFAS-containing biosolids already closed the cheapest route. It's the sequencing I wrote about when New York imposed a leachate removal requirement while the federal rule waited. And none of it applies below a few million gallons a day, where capital access settles the technology before any permit question arrives.

What goes in the procurement file before the skid ships

If you're specifying destruction capacity inside the EPA PFAS TECH strategy's two-year window, the federal record won't protect you, so the contract has to.

1. Buy the performance test, not the claim. Require the vendor to run its unit on your feedstock with your analyte list, and make fluorine mass balance a deliverable with a written closure tolerance.
2. Get the applicability determination in writing before the purchase order. A determination in hand survives a change of reviewer; a phone call doesn't.
3. Specify the monitoring you'll be asked for at renewal, not what sits in the current permit. Space, power and stack penetrations for an analyzer you don't need yet cost little at design and a great deal at year six.
4. Comment in the interim guidance docket asking for a destruction metric with a test method attached. Trade groups reliably ask for flexibility; flexibility is already the problem.
5. Price the residual before you price the plant. Spent resin, spent carbon and reject concentrate are the products of every removal step, and a destruction route that isn't under contract is an open liability whatever your zero-waste-to-landfill reporting claims for landfill diversion.

The open question is which way the causation runs, and the strategy document doesn't answer it. A demonstration program can generate the evidentiary record a performance standard needs, in which case two years of pilots is right and a rule follows in 2028 or 2029 with real test data behind it. Or the demonstrations become the answer themselves, with a searchable technology navigator standing in for a binding metric and fifty states deciding independently what 99.99% destruction of an unspecified analyte list is worth. Research dollars and rulemaking dollars come from different offices on different clocks, so the funding doesn't signal which. I'd want to hear whether anyone at the agency has committed to a destruction performance standard as the deliverable at the end of the two years, and nobody has said so on the record.

Sources & Notes

• The strategy's full name, the two-year commitment, the Zeldin quote and the STAR and SBIR funding routes: WaterWorld, 27 October 2026.
• The technology slate at the four Air Force bases, the hydrothermal alkaline treatment work and EPA's destruction figure for AFFF are in Smart Water Magazine's account. The figure is EPA's, from an industrial-scale system; the test conditions weren't published with it.
• For the 20 April 2026 refresh of the Interim Guidance on the Destruction and Disposal of PFAS, including the release-risk ranking, the thermal temperature threshold and the move to annual updates, see CIRS Group's summary.
• ArentFox Schiff's read of the 2024 version is the plainest statement that the guidance binds nobody.
• Pyrolysis partitioning across phases, and the dependence of removal on starting concentration, come from the April 2026 study in Water Environment Research. It measures the solid phase well; gas-phase accounting is where the open questions sit.

Disclosure: I advise waste-to-energy and water-sector developers on permitting and regulatory strategy, and write here for Renewable Waste Energy. The 2023 Title V matter is unnamed because that permit record remains contested.