PFAS Cleanup Cost: Why the First Dollar Belongs Upstream
By Antoine Walter · (don't) Waste Water · 5 November 2026 · 43 min read
Ali Ling did the arithmetic nobody in PFAS wanted to see: take just the perfluoroalkyl acids, the best-known forever chemicals, emitted in a year, pay today’s prices to pull them back out of the environment, and you get US$20 to 7,000 trillion a year, against a world economy of about US$106 trillion. So we will not clean our way out of PFAS, and she says that is the point: the number is a prioritisation tool. Because the bill drops by a factor of 1,000 when you move from the environment to the factory outlet, from millions of dollars per kilogram to hundreds or thousands, and it drops again at the drawing board, where the fix is a redesign. Two questions remain: where should the first dollar go, and, since the first dollars are already being spent on drinking water, who ends up paying, the ratepayer or the polluter?
🎧 Built on the podcast
This article is the companion to a (don't) Waste Water episode. Listen on Ausha →
How much does it cost to clean up PFAS?
In 2024, Ali Ling, assistant professor of civil engineering at the University of St. Thomas in Saint Paul, Minnesota, published the only planet-scale estimate we have: removing perfluoroalkyl acids from the environment at the rate we emit them would cost US$20 to 7,000 trillion a year, in a world that earns about US$106 trillion (Science of the Total Environment, 2024).
PFAS, per- and polyfluoroalkyl substances, are a family of several thousand synthetic chemicals built around the carbon-fluorine bond, the strongest bond in organic chemistry, which is why non-stick pans, rain jackets and firefighting foams love them, and why nothing in nature takes them apart. Now, her paper looks at one branch of that family, the perfluoroalkyl acids, and asks a deliberately naive question: what if we simply removed, every year, what we add? The answer runs from 20 to 7,000 trillion dollars a year, and even the low end is a fifth of everything humanity produces.
The ladder, rung by rung
Hard to picture, so let’s climb down the ladder. One continent: the Forever Lobbying Project, coordinated by Le Monde with Ali Ling on the cost methodology, put Europe’s bill, in January 2025, at about €2 trillion over 20 years, roughly €100 billion a year, at today’s emissions.
One US state: in 2023 the Minnesota Pollution Control Agency, in a study her consulting firm co-authored, priced removing and destroying PFAS from the state’s municipal wastewater at US$14 to 28 billion over 20 years. And one kilogram, her back-of-envelope on my microphone: millions out in the environment, hundreds to thousands at the industrial discharge. I covered the European number in my newsletter in January 2025, and the shape of it only landed when she stacked the rungs.
At today’s emission rates, removing perfluoroalkyl acids from the environment would cost US$20 trillion to US$7,000 trillion a year, against a world GDP of about US$106 trillion. Sources: Ling 2024 (Science of the Total Environment); Minnesota Pollution Control Agency 2023; Le Monde / Forever Lobbying Project 2025.
Well, her reading of her own number is not despair. Many people see it and feel there is no hope for anything, she told me, “and that’s not what I’m trying to get at here. What I’m trying to get at here really is that prioritization.”
Why is water the most expensive place to catch PFAS?
The cost of removing PFAS from water scales with the volume you treat, Ali Ling explained in October 2025, so a kilogram caught in municipal effluent costs millions while the same kilogram caught at a factory outlet or in landfill leachate costs hundreds to thousands; the EPA’s stated intent for wastewater permits is to push PFAS upstream to industrial dischargers.
So drinking water and wastewater got regulated first, because they are the most monitored things in the United States, so we know what is in them, while nobody knows what sits in most industrial discharges or consumer products. So the utilities became the easy place to start, and they now sit in what she calls a really difficult place: a permit limit on their effluent, an upstream discharger they may not control, and, if that fix fails, a PFAS plant of their own, with its affordability problem and, per a study she is running, a lot of extra greenhouse gas emissions. Hence her law for regulators:
Whereas PFAS, some of it comes from those non-point sources, but a lot of it comes from industrial discharges and from landfill leachate. I think that’s a case where treating it in those upstream locations is gonna make a lot more sense than treating it in wastewater effluent. The cost to remove PFAS from water in any phase is generally gonna be proportional to the volume of water that you treat.
Ali Ling, Assistant Professor of Civil Engineering, University of St. Thomas, on the (don’t) Waste Water podcast, S13E14, October 2025 · hear her say it (23:17)
Where I have seen this film before
Look, micropollutants were the beginning of my career, and we all reached her conclusion, reduce at the source, before discovering how hard that is when nobody wants to give up the benefits of the molecule. Switzerland then showed where to spend: treat wastewater and you protect the whole water cycle, treat drinking water and you protect the human only until the next flush. For PFAS, given what is already out there, we will probably need both ends anyway.
Yes, once, and quietly: PFOA and PFOS, the two best-known PFAS, were largely phased out in the United States and Europe in the 2010s, and the CDC’s blood survey shows PFOS in Americans’ serum fell by more than 85% between 1999 and 2018, a decline Ali Ling rounds to “a factor of 10”.
She calls it a quiet success story that not a lot of people talk about, and the mechanism is the one she wants more of: companies phasing PFAS out, often voluntarily and often without announcing it, because saying “we used to have PFAS” is an invitation to a lawsuit. Now, IKEA and H&M, in her telling, took PFAS entirely out of their supply chains. And the blood data is the one number in this whole story that goes down.
If you look back at the human PFAS data in blood serum over the past 20 years, and you look at PFOA and PFOS, which were largely phased out in the 20-teens, the concentrations in blood serum have gone down for PFOS by like a factor of 10. I think that’s a quiet success story that not a lot of people talk about where PFAS was mostly phased out.
Ali Ling, Assistant Professor of Civil Engineering, University of St. Thomas, on the (don’t) Waste Water podcast, S13E14, October 2025 · hear her say it (46:20)
The regrettable-substitution trap
Now, regrettable substitution, in her definition, is substituting something and regretting it later because the replacement was not actually better. GenX, the chemical that replaced PFOA (its technical name is HFPO-DA), is the PFAS example, and HFC refrigerants replacing the CFCs that ate the ozone layer is the historical one. That trap is why Europe is trying to restrict the family as a whole rather than one molecule at a time: the European Chemicals Agency’s risk committee adopted its opinion on the universal PFAS restriction in March 2026, its socio-economic committee is expected to finish by the end of the year, and only then does the Commission write the law.
Her own standard is more modest, and I like it better than most: don’t let trying to be perfect get in the way of doing some good. I was slow on PFAS myself, which I admitted in writing in January 2025.
The treatment menu, and where the bill hides
Nothing destroys PFAS at scale without a removal step first, Ali Ling insists: granular activated carbon or ion-exchange resin concentrates the molecules, then something has to break them, and the default in Minnesota’s 2023 study was high-temperature incineration, at US$2.7 to 18 million per pound of PFAS removed and destroyed from municipal wastewater.
Now, granular activated carbon, GAC for the insiders, is charcoal engineered to be porous; ion exchange is a resin that swaps the PFAS molecule for a harmless ion; both end up as a spent medium that somebody must regenerate, landfill or burn. And her latest paper prices exactly that spent-media question, with vendor-sourced energy and operating costs, so that a utility planning a plant can read the menu before ordering. And the menu has fine print: small plants pay more than six times more per pound than large ones, Minnesota found, and the short-chain PFAS cost up to 70% more to remove than the long chains they replaced (regrettable substitution again, this time with an invoice).
The PFAS treatment menu, with Ali Ling’s verdicts (table)
Role
Technology
What it does
What Ali Ling said in the episode
Covered on dww.show
Remove
Granular activated carbon (GAC)
Porous carbon that adsorbs PFAS; the spent carbon must be reactivated or burned.
“There’s nothing available at scale that can do destruction without the removal step.” (19:57) Her latest paper prices GAC regeneration against the alternatives.
A resin that swaps the PFAS molecule for a harmless ion; single-use resin ends in an incinerator.
Same paper: spent-media options with energy and operating costs “such that someone could look at that paper and say, okay, if we wanted to use this option or that option, here’s about how much it would cost”. (33:51)
Air bubbles lift PFAS into a foam you skim off; strong on long chains, weaker on short ones.
“If they can get the chemistry down where they’re removing the short-chain ones more effectively, like that’s gonna be quite a bit more cost-effective than some of the other separation technologies.” (15:29)
Burns the spent media or the concentrate; the default of the Minnesota cost study.
“High-temperature incineration, which was our default for that study.” (15:29) The US$2.7 to 18 million per pound rests on it.
MPCA 2023 (Sources, 2)
Destroy
Electrochemical oxidation
A current through the water, an electrode breaks the carbon-fluorine bond; 18 vendors with PFAS claims.
Destruction technologies “coming online and getting demonstrated at scale that would be more cost-effective than high-temperature incineration”. (15:29)
“I have not heard anything super convincing from anyone doing plasma. It’s possible they just haven’t shown their data to me, but it’s not super high on my list.” (56:41)
Destroy
Hydrated electrons and other reduction routes
Electron donors strip the fluorine instead of oxidising the molecule.
The sludge after treatment; drying, destruction or landfill, each with its own logistics.
“There’s a lot of other logistics that people haven’t thought through.” (25:30) Maine’s scramble for landfill space is her example.
Every route still ends in a destruction step: separate first, then break the bond. Sources: the episode (timestamps link to YouTube); Minnesota Pollution Control Agency 2023; Leviathan, my water database.
Biosolids, the bill nobody has budgeted
Then come biosolids, the sludge left after treatment that farmers have spread on fields for decades. Maine stopped land application and had to scramble for landfill space, and landfilling only moves the PFAS into the leachate and the landfill gas, where it comes back out. Nobody has priced the regional drying and destruction plants that would replace the field, she notes, or the trucks to feed them.
The destruction side is getting crowded, which I will come back to, and her verdicts there are short: most technologies have some sort of Achilles heel, foam fractionation gets cheaper the day it learns short chains, and plasma is “not super high on my list”. I had Oxyle’s destruction route on the podcast in 2024 and I reviewed H2Plus’s hydrated-electron pitch in July 2025, so I can confirm that every vendor’s heel is somewhere else.
Is drinking water even the right place to look?
The EPA’s own exposure model assumes drinking water carries less than 20% of a person’s PFAS intake, Ali Ling reminded me, so even a perfect drinking-water programme caps the health benefit at about a fifth; the rest comes mostly from diet, which is falling as PFAS leaves food packaging, and from indoor dust.
Well, the exception is the hotspot: Cape Fear in North Carolina, Hoosick Falls in New York, Dordrecht in the Netherlands, places where a contamination event put very high concentrations into the tap and treatment was the right call. But outside them, she says, the risk is diffuse and relatively small, and the honest question for each utility is where its risk sits: PFAS, or arsenic, manganese, disinfection by-products, or plain pipes. Parts per trillion, the unit all of this is measured in, is one drop in twenty Olympic pools, which says how far down the plants are asked to go. Her own tap, she told me, is “almost zero”; her dust and her products are not.
There’s a lot of talk about PFAS in drinking water, but the EPA acknowledges that drinking water is less than 20% of PFAS intake […]. So even if we were to remove all PFAS in drinking water, the best we could do is 20% reduction in people’s PFAS exposure.
Ali Ling, Assistant Professor of Civil Engineering, University of St. Thomas, on the (don’t) Waste Water podcast, S13E14, October 2025 · hear her say it (21:22)
Who pays for PFAS cleanup?
Treat at the end of the pipe and the bill lands on ratepayers and taxpayers, Ali Ling told me; treat at the factory and the companies that emit pay, “where they should be routed”. The largest settlements, 3M’s US$10.3 to 12.5 billion (approved in 2024) and US$1.185 billion from DuPont, Chemours and Corteva, roughly cover one state’s 20-year wastewater bill.
So let’s do the maths: put the three settlements approved in 2024 together, 3M, the DuPont group and Tyco’s US$750 million, and you get US$12 to 14 billion, paid to drinking-water systems across the whole United States over more than a decade. And Minnesota alone, for its wastewater alone, needs US$14 to 28 billion over 20 years. The money the polluters have agreed to pay covers one state’s sludge and effluent, give or take. There is some federal money for drinking water, she notes, “but it’s not nearly enough. It’s not nearly as much as what’s been estimated.”
The three largest PFAS settlements ever approved, US$12.2 billion together and up to US$14.4 billion, add up to about one state’s 20-year wastewater bill: Minnesota alone expects US$14 billion to US$28 billion. Sources: 3M 2024; settlement administrator 2024; Minnesota Pollution Control Agency 2023; Le Monde / Forever Lobbying Project 2025.
The counter-argument, from my own archive
Henry Charrabe, who led Seven Seas Water to its billion-dollar exit, gave me the other side in the episode that followed Ali’s: “ultimately, ratepayers are the polluters. We all use Teflon.” But Henrik Hagemann of Puraffinity had made the opposite case back in 2022: the utilities never earned a cent from Scotchgard or Gore-Tex. Both are right, which is exactly why the upstream route matters, because it is the only place where the bill and the benefit sit inside the same company.
And a confession: I once calculated what a human life is worth to a regulator, about US$10 million in the United States, and I still think we skipped that cost-benefit step entirely on PFAS. Ali’s answer was gentler than mine: the societal and climate costs are real, they are just very hard to put on the same scale.
Will the 4 ppt limits survive?
In October 2025 Ali Ling predicted that some form of the US limits would stick, and that dropping the four other PFAS would barely change how many systems must treat; in May 2026 the EPA proposed that: keep PFOA and PFOS at 4 parts per trillion, rescind the other four, and let opt-in systems slip from 2029 to 2031.
Now, her reasoning was mechanical rather than political: the systems above 10 ppt of PFHxS, one of the four, are the same systems above 4 ppt of PFOA and PFOS, so keeping two limits makes monitoring easier and changes very little on the ground. And an MCL, a maximum contaminant level, is the legal ceiling for a substance in drinking water, and the American Water Works Association sued over the 2024 rule; she found its arguments “very reasonable”, because asking thousands of small utilities to build PFAS treatment with limited funding, when their bigger health risk may sit elsewhere, is a problem.
Ali Ling called it in October 2025: PFOA and PFOS stay at 4 parts per trillion, the other four limits are set to go, and the compliance clock moves from 2029 to 2031 if the May 2026 proposal holds. Sources: EPA; Harvard Law EELP tracker; AMWA; the episode.
The court, the agency and the 2031 clock
The D.C. Circuit refused in January 2026 to vacate parts of the rule and in March refused to shelve the case, so the court and the agency are now racing each other to the same finish line. Her October reading sits on tape seven months before the formal proposal and before both court rulings: the agency had announced its intention in May 2025, and what she called was which parts of it would survive the lawyers, and why the rescissions would barely move the number of plants that must treat.
The next responsible step, in her view, is effluent limit guidelines on specific industries, upstream, which she doubts this administration will write. As of early November 2026 the May proposals are still proposals; this page gets updated the day they are final.
Where the money is going: 32 companies, 24 installations
The day before this episode went out, I published my own count: 32 companies market a proprietary electrochemical-oxidation process, 18 of them with PFAS-destruction claims, chasing at most 24 installations worldwide, on a little under US$120 million raised between the start-ups. As you know, I’m fun at parties.
Because electrochemical oxidation runs a current through the water and lets the electrode break the carbon-fluorine bond, it is the destruction route with the most vendors, and most of the 18 sit at TRL 7 to 9, TRL being NASA’s one-to-nine scale of how baked a technology is, with 9 meaning fully commercial. So two are at 9, Axine and AECOM, nine at 8, and only four publish a third-party-verified report of more than 90% defluorination. Investors read that crowd as a bloodbath. I read it as validation, because zero competitors would mean zero market, and by my estimate about 250 industrial facilities have acted on PFAS out of roughly 10 million (don’t quote me on the last digit).
Eighteen vendors claim to destroy PFAS with electrochemical oxidation, and at most 24 industrial sites have ever installed it. Source: my own count of electrochemical-oxidation vendors, October 2025.
What my Leviathan database adds
My Leviathan database widens the lens beyond one technology. Sixteen companies whose main business is removing or destroying PFAS have raised about US$196 million since 2019, 84% of it for destruction claims, and only 11 acquisitions of a PFAS treatment business are on record, four of them priced, US$111 million together, the largest being Montrose’s US$53.8 million purchase of ECT2 in 2019. So let’s do the maths one last time: all the venture money and all the deal money in the PFAS-core vertical, about US$0.3 billion over seven years, is one to two percent of Minnesota’s wastewater bill.
The consolidators are circling anyway (Axine and Veralto, Aclarity and Xylem), the way SKion did in water and the way the 2025 funding map predicted. I dissected the 32-company census in my newsletter, which is where this whole thread started; if you want the next one in your inbox, subscribe here.
Sixteen companies built only to remove or destroy PFAS have raised about US$196 million since 2019, and PFAS treatment companies have changed hands eleven times, a vertical still small enough to count by hand. Source: Leviathan, my water database, verified 2026-11-04.
What Ali Ling asks for next
Her two asks are both forward-looking. To the vendors who used to call her as a consultant with a great thing and no data: show me your data, because the consultants who put you in front of a utility trust what they understand.
To the rest of us: zoom out, stop treating PFAS as the only risk and municipal effluent as the place to spend. The 2031 clock now runs on the two limits everyone agrees will stick, which makes the upstream permit the next fight, and her thesis the one to test it against:
… we can’t afford to clean up PFAS from the environment. That needs to be like a point-specific problem type of solving thing. It can’t be just cleaning it all up. I think talking about source control and industrial discharges treatment, that’s really where we need to be talking about PFAS treatment. I think treating PFAS in wastewater, municipal wastewater effluent doesn’t really make sense.
Ali Ling, Assistant Professor of Civil Engineering, University of St. Thomas, on the (don’t) Waste Water podcast, S13E14, October 2025 · hear her say it (56:54)
The whole conversation, the regulatory rollercoaster included, is in the episode below, and the comments are open for anyone who thinks the ratepayer is the polluter after all.
Frequently asked questions
Can PFAS ever go away?
Not by cleanup at planetary scale: Ali Ling’s 2024 estimate puts removal at current emission rates at US$20 to 7,000 trillion a year, more than the world economy. They stop accumulating only when emissions stop, and the PFOS phase-out shows what that does: a drop of more than 85% in American blood levels in two decades.
What is the biggest source of PFAS in humans?
Outside the contamination hotspots, it is not drinking water. The EPA’s exposure model assumes tap water carries less than 20% of intake; diet, which is falling as PFAS leaves food packaging, and indoor dust carry most of the rest. That is why Ling questions drinking water as the best benefit per dollar.
How much does it cost to remove PFAS from water?
Minnesota’s 2023 study is the reference: US$2.7 to 18 million per pound of PFAS removed and destroyed from municipal wastewater, small plants paying more than six times more per pound than large ones, and short-chain PFAS up to 70% more than long chains. Per gallon, the AqueoUS Vets and H2Plus pieces linked above carry the vendor numbers.
Who is liable for PFAS cleanup?
The manufacturers, for drinking-water systems, through the settlements approved in 2024: 3M’s US$10.3 to 12.5 billion, US$1.185 billion from DuPont, Chemours and Corteva, US$750 million from Tyco. Everything downstream of that, wastewater and biosolids, falls on ratepayers and taxpayers unless a permit reaches the industrial discharger who emitted it.
Is source reduction really cheaper than treatment?
By about a thousand times per kilogram, in Ling’s back-of-envelope: millions of dollars once PFAS is in the environment, hundreds to thousands at the industrial discharge, and the Minnesota agency reaches the same conclusion, prevention more than 1,000 times more cost-effective than cleanup. The honest caveat is that the cost of redesigning products is not yet known; the European Chemicals Agency is counting.
What is regrettable substitution?
Substituting a chemical and regretting it later because the replacement was not actually better: GenX replacing PFOA in the PFAS family, HFC refrigerants replacing ozone-eating CFCs before that. It is the reason Europe is trying to restrict the whole PFAS family at once rather than one molecule at a time.
How long do PFAS last?
Long enough to earn the nickname: the carbon-fluorine bond does not break in nature on any human timescale, so “forever” describes the molecule’s persistence, not a cleanup horizon. Which is the whole point of putting the first dollar where the molecule is made rather than where it ends up.
Sources
Ali Ling, “Estimated scale of costs to remove PFAS from the environment at current emission rates”, Science of the Total Environment, 2024. https://pubmed.ncbi.nlm.nih.gov/38325453/ (retrieved 2026-11-04)
