The quiet story under WEFTEC 2026

I will be on the floor at the Morial this week. Here is what I expect to hear, what I think it means, and the one thing almost nobody on the exhibit floor will say out loud.

1. PFAS is not slowing down, it is fragmenting

The federal picture looks like a retreat if you read headlines. EPA proposed in May to rescind four of the six drinking water limits and push compliance from 2029 to 2031. Neither proposal was finalised this year. The 2024 rule, six MCLs and an April 2029 deadline, is still the law today.

The part that matters more got settled the other way. On 18 August the D.C. Circuit affirmed the CERCLA hazardous substance designation for PFOA and PFOS. That designation has now survived both a change of administration and judicial review. Liability, not the MCL, is the durable driver, and wastewater utilities are not covered by the AFFF settlements that resolved things for drinking water systems.

Meanwhile Maine and Minnesota are running bans to 2032, New Mexico came in this March, and eleven or more states have their own AFFF rules. So the honest read is not that PFAS demand is going away. It is that one national procurement wave is turning into fifty state-level ones, on fifty timetables.

If you make equipment: your pipeline gets lumpier and more regional, and the value of knowing which state is next goes up sharply.

2. The SRF cliff lands this month, and municipal has already gone quiet

About ten billion dollars of IIJA water money expires on 30 September. The regular FY2026 appropriation leaves roughly $746m for Clean Water SRF capitalisation once earmarks take their cut, which is more than a seventy percent reduction against recent capitalisation levels. Reauthorisation exists in the Senate WRDA bill. The House bill that passed 415 to 9 on 16 September has no SRF provision in it. Nobody knows what comes out of conference.

You can already see it in results. ANDRITZ set a group order record while its Environment and Energy segment fell 12.7 percent. Xylem's Water Infrastructure line was down 4 percent while its services business grew 147 percent organically. Municipal capital is soft. Municipal service and retrofit is not.

If you make equipment: anything that upgrades performance inside existing concrete is the right product to be selling into a constrained municipal budget. New basins are a hard sell right now.

3. Industrial and power are carrying the sector

Manufacturing construction is down 22.2 percent year to date, and computer and electronics plants account for essentially the entire decline, down 45.4 percent. Data centres are up 34.8 percent, and up 57.2 percent in July alone. CECO's orders are up 191 percent on a backlog of $1.8bn, mostly on gas power.

So the phrase to retire is "reshoring." The phrase that describes 2026 is power and data centres. If you are selling cooling water, process water, wastewater pretreatment or air emissions control into that build-out, this has been a good year. If you are selling to POTWs, it has not been.

There is a second wave behind it. The air NESHAP deadlines were deferred, not repealed. Iron and steel moves to April 2027. Coke ovens to 2027 and 2029. Several sectors took two-year presidential exemptions that are already under legal challenge. Those obligations come back in the 2027 to 2029 window, and compliance equipment gets bought twelve to twenty-four months ahead of a date.

4. Water and AI, minus the hype

The Water-AI Nexus program is bigger this year and the interesting sessions are not the ones about optimisation. They are about utility cybersecurity, about the workforce cliff as operators retire, and about how much water a data centre actually uses. That last one is a genuinely new political problem for utilities and it will shape siting fights for the next decade.

And the thing nobody says out loud

Here is what I would tell anyone walking the floor with a badge that says owner or president.

The number of independent manufacturers in this industry is falling fast, and faster than most people in it realise.

Walk the aisles and you will see most of the same logos you saw five years ago. A lot of them no longer belong to the people whose names are on the sign. Well over a third of the booths at a show like this sit inside a strategic, a public parent, a private equity platform or a foreign group, and a further slice are North American sales arms for factories in China, Korea, Italy, Spain or Germany rather than manufacturers at all. Some of the companies standing next to you as peers are quietly buyers.

The consolidation at the top is not subtle either. CECO closed Thermon on 1 June and roughly doubled itself. Parker closed Filtration Group on 13 August. Veralto agreed Cleanwater1 in August at around seventeen times, and the thing worth noticing there is what it paid for: a business with roughly a quarter of its revenue in aftermarket.

That ratio is the whole game. Kadant runs 45 percent gross margin and 20.6 percent EBITDA with 71 percent of revenue from parts. A project-heavy engineered-to-order fabricator of similar size runs 25 to 33 percent gross margin and single-digit to low-teens EBITDA. Nederman's own segment disclosure makes the point inside one company in one year: consumables at 19.6 percent EBITA, engineered projects at 11.0 percent.

So the question I would ask yourself, before someone else asks it for you, is a simple one. What share of your revenue is new-build project work, what share is packaged product, what share is parts and consumables, and what share is service and rebuild?

Almost no owner-operated business tracks revenue that way. But that is the number that decides what the business is worth, and it is the first thing a buyer will build for themselves whether or not you hand it to them. Working it out on your own terms, a year or two before you need it, is worth several turns.

Happy to talk to anyone on the floor this week, or after.

Alex Christodoulou is a Partner at Aveon Partners, which advises owners of water and air pollution control equipment businesses.