Retire or Retrofit? Why WIN Waste Is Closing Hudson Falls Rather Than Meet the 2026 Combustor Rule

WIN Waste filed to retire Hudson Falls in June. Five hundred tons a day, built in 1991, and it burned roughly 150,000 tons through the grate last year, according to Waste Dive. The company priced out what it would take to keep the plant legal under the new federal limits and walked. Not because it can't burn waste. Because the numbers on whether to retire or retrofit an aging waste-to-energy unit finally tipped the wrong way. It's the first closure I'd point to and say: watch the rest of the fleet run the same arithmetic.

Most of the coverage got this backwards. The EPA rule that triggered the decision didn't ask for anything exotic. A fresh electrostatic precipitator and a NOx control stage are off-the-shelf equipment I've specified more than once. The question was never whether the hardware works. It's whether a 35-year-old plant with a one-off design has room to bolt it on, and whether the local waste market will let you charge enough to pay for it. Hudson Falls failed both tests. The rule was the trigger; the site pulled it.

What the 2026 combustor rule actually forces

EPA finalized the large municipal waste combustor rule on March 10, 2026, the first federal update to these standards in twenty years, per J.J. Keller's read of the final rule. It lives at 40 CFR Part 60 Subpart WWWW for existing units and catches any combustor over 250 tons a day. It tightens particulate matter, metals, hydrogen chloride, and the expensive part for old plants, nitrogen oxides and carbon monoxide. It also strips out the NOx averaging alternatives and the startup/shutdown exemptions, so you comply at all times or you don't comply at all.

But the timeline is where operators get lulled. Existing units don't owe compliance until three years after their state plan clears, or by 2031, whichever lands first. That reads like breathing room. It isn't. I walked through why the states, not EPA, hold your real deadline in a separate piece on the state-plan compliance clock, and the short version is you make the retire-or-retrofit call years ahead of the deadline, because the capital cycle for a boiler retrofit is that long. WIN Waste made theirs in 2026.

The retrofit path: what the equipment costs you

Start with the precipitator. An ESP pulls better than 99% of the particulate out of the flue gas [manufacturer rated], and on a plant this old you aren't tuning the existing one, you're replacing it. The plates and rappers from 1991 won't hold the new limit (and the casing's half-corroded anyway). That's a major structural install into a building that was poured around the old box.

Then NOx. Two ways to skin it. Selective non-catalytic reduction, SNCR, sprays urea or ammonia into the furnace at 850 to 1,100°C and knocks out roughly a third of the NOx [typical range]. It's cheap and it's simple. Selective catalytic reduction buys you 80 to 90% but wants a catalyst bed, more real estate, and a flue-gas temperature window the plant has to be plumbed for. WIN Waste's own filing named SNCR, which tells you they were reaching for the cheapest compliant option and it still didn't pencil. Actually, that one-third figure only holds while the furnace sits in the SNCR temperature band; let the feedstock drift wet and you drop out of the window and get less.

The reason it didn't pencil is what "unique design" really means in a decommissioning notice: nothing's standard. A Wheelabrator or Martin mass-burn line has a documented retrofit path, because B&W and the grate OEMs have run these upgrades on sister units and they know where the ductwork goes. A one-off has no sisters. Every bracket is a custom fab, every residence-time calc starts from a blank sheet, and the OEM thermal margin you're leaning on is 10 to 15% optimistic by default. A bypass damper that's been "temporary" since the Clinton administration is a design decision now, and you inherit it. If you want the shape of a modern build instead, our notes on waste-to-energy technology lay out what a from-scratch thermal train looks like.

And retrofits into old iron bite you on a delay. On the Pinellas County retrofit in 2021 we added a thermal scrubber stage, commissioned clean, felt good about it, and paid for it in year five when chloride attack cracked the superheater and we lost three weeks of availability the OEM swore we wouldn't. That's the retrofit tax nobody lines out in the capital request - a new stage changes the chemistry upstream of metal that's already tired.

The retire path: where the tonnage goes

Retiring isn't free either, and anyone selling you "just close it" is skipping the invoice. There's NYISO deactivation approval to clear, because the plant sits on the grid. There's decommissioning and abatement of the building. And there's the part your county actually feels: the 150,000 tons a year has to go somewhere. When a waste-to-energy plant retires, that stream doesn't evaporate. It rides a truck to a landfill, which is a straight loss on landfill diversion and a step backward on the zero-waste-to-landfill goals the plant was built to serve.

That's the piece that should bother a municipal planner more than the stack numbers did. A mass-burn plant of this size cuts landfill volume by close to nine-tenths [EPA/EIA]. Retire it and you aren't only losing fifteen or so megawatts of baseload; you're re-landfilling a waste stream you'd already diverted, and the methane math on buried waste is worse than the stack ever was.

The criterion that decides it: gate-fee headroom

The number that killed Hudson Falls isn't an emission limit. It's the tip fee. WIN Waste said plainly that with other disposal options and transfer stations nearby, they couldn't raise the gate fee high enough to cover the retrofit [Waste Dive]. That's the whole decision. A retrofit is just capital you amortize into what you charge per ton. If your market clears around $90/ton [Northeast range] and a cheap transfer station sits next to your customers, you can't tack the precipitator's cost onto the ticket. The waste routes elsewhere and your utilization craters, which makes the amortization worse, which is a spiral you don't climb out of.

Standard mass-burn lines, the Wheelabrator and Martin fleets, will mostly retrofit, because their capital path is known and their host contracts often let them pass the cost through. The plants that retire will be the ones like Hudson Falls: one-off designs, soft local pricing, a landfill or transfer station close enough to undercut them. Notice that has almost nothing to do with which plant has the worst emissions. Emissions set the deadline. The waste market decides who clears it.

Which call fits your plant

So which is it for you, retrofit or retire? It depends, and rather than a verdict I'll give you the questions I'd actually ask on a walkdown. Is your design standard enough that an OEM has a proven retrofit on a sister unit? Do your host and offtake contracts let you pass capital through, or are you eating it? Is there spare landfill or transfer capacity inside a cheap haul radius that caps your gate fee? Then look under the insulation at the real condition of the pressure parts, because refractory life, not throughput, is the availability constraint most plants under-budget, and if the boiler's near end of life you're pouring a new ESP into a plant that's a retirement candidate anyway.

Answer those honestly and the decision tends to make itself. WIN Waste's answers ran non-standard, limited pass-through, cheap disposal next door, aging iron. Four for four toward the exit. If your plant is a stock Martin line with a captive waste district and forty years of documented pressure-part history, your answers run the other way and you retrofit without blinking. Most plants sit in the messy middle, which is why this rule won't produce one clean wave of closures. It'll produce a slow sort, plant by plant, over the run-up to 2031. If you're mapping which side of that sort your assets land on, that triage is exactly the work behind our waste-to-energy services.

Hudson Falls didn't fail the emissions test. It failed the tip-fee test, and the emissions rule only forced the day it got graded. Fifty-six other large combustors are sitting for the same exam right now, and the ones with a cheap landfill next door already know their answer.

Sources & Notes

• Plant specifics - 500 tons a day, commissioned 1991, about 150,000 tons in 2024, the SNCR-and-ESP retrofit scope, and the tip-fee reasoning - come from Waste Dive's reporting on the closure.
• Rule scope, the 250-ton-a-day threshold, 40 CFR Part 60 Subpart WWWW, and the 2031 / three-years-after-state-plan timing are drawn from J.J. Keller's rundown of EPA's March 2026 final rule.
• Fleet age and count, 57 plants with roughly 90% of capacity built between 1980 and 1995, are from the EIA waste-to-energy explainer.
• The 2021 Pinellas superheater figure is my own from that commissioning; the SNCR and SCR reduction ranges are standard vendor performance bands, not numbers off any one plant.