From Permit to Product: Why Execution, Not Technology, Decides Pyrolysis IRR

Plastics pyrolysis is not a science experiment. The chemistry, cracking mixed plastic waste into an oil that can re-enter the petrochemical chain, works in any competent lab. The reason advanced-recycling plants keep destroying capital is that running one at scale, on real-world feedstock, every day, is a different problem entirely. The IRR lives or dies there.

A 100,000-Tonne Plant That Recycled 2,000

Brightmark’s Ashley, Indiana facility is the case study every circular-economy investor should keep close. It was designed to process 100,000 tonnes of plastic a year. By early 2024 it had recycled roughly 2,000 tonnes, about 5% of capacity. In March 2025 the operating subsidiaries missed a $12.9 million payment and entered Chapter 11, carrying around $178 million of secured debt including $172.5 million of green bonds.

Note what did not happen: the pyrolysis technology was not disproven. The plant simply could not run reliably on the feedstock it was fed, at the rate its financing assumed. A facility at 5% of nameplate has no path to covering its debt service, whatever the unit economics look like at full load. This is the single most important fact in advanced recycling, and it is an execution fact, not a technology one. Encina tells the companion story: its planned billion-dollar Pennsylvania plant was cancelled after two years, with permitting and the EPA’s treatment of catalytic pyrolysis as an incineration process, not recycling, among the obstacles.

What Actually Moves the Return

We modelled a 50,000-tonne-per-year commercial pyrolysis plant and stress-tested one variable at a time against a 7.7% unlevered base-case IRR. The base case already assumes the plant runs well. The ranking shows how little room there is for it not to.

Single-variable stress	IRR	Change vs base
Base case (well-run)	7.7%	–
Yield 40% (contaminated feedstock)	(1.2%)	−8.9 pts
Uptime 45% (chronic underperformance)	0.1%	−7.6 pts
Pyoil price −25% (weak premium)	0.5%	−7.2 pts
Feedstock cost +$100/t (must buy clean)	2.6%	−5.1 pts
Capex overrun +30%	4.0%	−3.7 pts

Yield and uptime sit at the top, and each on its own turns a positive return negative. The base case is already thin, around 8%, which means advanced recycling has almost no tolerance for the execution slips that are, in practice, the norm. The reactor design is not a line in this table.

Yield Is a Feedstock Problem, Not a Reactor Problem

The yield assumption is where most models quietly lie. In the laboratory, mixed-plastic pyrolysis can return oil yields anywhere from 48% to 78% by weight. On a real tipping floor, the feedstock arrives contaminated with moisture, food residue, biomass, salts, pigments, and crucially chlorine from PVC and other halogens. Those contaminants do two things: they drag the usable yield well below the lab figure, and they corrode and foul downstream equipment, which is how a yield problem becomes an uptime problem. A project that has not locked down a clean, consistent, contractually specified feedstock is not a 55%-yield plant that occasionally dips. It is a plant whose yield, and therefore whose entire revenue line, is set by whatever shows up at the gate.

Feedstock logistics compound the yield problem in a way models rarely capture. Pyrolysis feed is bulky and low-density, so the economics are acutely sensitive to the catchment radius: every additional 10 km of trucking can take $10 to $20 a tonne off the netback, which means a plant sited a little too far from a steady waste stream can be margin-negative on logistics alone, before the reactor is even in the question.

The Offtake and the Premium Are Real, but Conditional

The demand side is genuinely better than the SAF or hydrogen stories. Consumer-brand recycled-content commitments and, in Europe, the Packaging and Packaging Waste Regulation, create real pull for circular polymers, sold through ISCC PLUS mass-balance certification at a premium to virgin material. That premium is what carries the model; strip 25% off the realised price and the base case collapses to breakeven. The catch is that the premium is paid for consistent, on-spec output. A plant that produces variable, contaminated pyrolysis oil does not sell into the premium circular market; it sells, if at all, into the commodity fuel-oil market at a fraction of the price. The premium and the yield problem are the same problem viewed from the two ends of the plant.

The Permitting Trap

Pyrolysis carries a regulatory ambiguity that does not exist for most energy assets. Whether a given process is classified as recycling or as incineration is contested, varies by jurisdiction, and directly affects both permitting timelines and the marketability of the output as recycled content. Encina’s cancellation showed how that ambiguity, combined with local opposition, can kill a project before it pours concrete. That ambiguity is now in flux at the federal level: in March 2026 the EPA proposed removing pyrolysis from its “other solid waste incinerator” definition under the Clean Air Act, a change that would ease permitting but which drew tens of thousands of opposing comments and left the classification, and the timeline it dictates, unresolved as of mid-2026. For an investor, permitting and classification are not a box to tick at the end. They are an early, binary gate that determines whether the asset can exist at all.

De-Risking a Pyrolysis Project

If an advanced-recycling project crossed our desk, this is the diligence that maps to how the return is actually won or lost:

1. Contracted, specified feedstock. Volume, composition, contamination limits, and price, under contract. Without it, the yield and uptime assumptions are fiction.

2. Demonstrated uptime at scale, not pilot yield. Has this exact configuration run continuously on real feedstock? Underwrite 45 to 60% uptime in year three, not the brochure’s 85%.

3. A bankable offtake into the premium market. An ISCC-certified, on-spec product contracted to a circular-polymer buyer, not a hope of selling fuel oil.

4. Permitting and classification resolved early. Confirmed recycling classification in the relevant jurisdiction before capital is committed.

5. Capex certainty and a thin-margin stress test. The base case is ~8%. Model a 30% overrun and a 25% price drop together and ask whether any equity survives.

6. The first plant as a learning asset. As in every first-of-a-kind, the return comes from the second and third unit. Size the first accordingly.

How We’d Read the Opportunity

Advanced recycling is not a fraud and it is not a slam dunk. It is a thin-margin industrial operation wearing a clean-tech narrative, and it is priced by how well it is run. The winners will be the operators who treat it as a feedstock-and-uptime business, with the permitting and offtake locked before the reactor is even the question. The losers, and there have been many, are the ones who believed that proving the chemistry was the hard part. For disciplined capital, the edge is to underwrite the plant as the operating asset it is, and to walk from any deal where the feedstock, the offtake, or the permit is still a hope.

Pyrolysis Economics: Investor FAQ

Why do plastics pyrolysis plants fail if the technology works?

Because the return is decided by uptime, yield and feedstock quality, not the chemistry. Brightmark’s Indiana plant was designed for 100,000 t/yr but recycled about 2,000 t, roughly 5% of capacity, and defaulted on $172.5 million of green bonds. The pyrolysis worked; the plant could not run reliably at scale.

What is the IRR of an advanced recycling project?

On a well-run base case (50,000 t/yr, 57% usable yield, ~$950/t circular-oil price, 85% uptime) we model around a 7.7% unlevered IRR. It is thin, and a drop in yield to 40% or uptime to 45% turns it negative.

What yield does plastic pyrolysis actually achieve?

Lab yields of mixed-plastic pyrolysis oil run 48 to 78% by weight, but real contaminated feedstock (moisture, food residue, chlorine, metals) pushes usable yield lower and damages equipment. Yield is set by feedstock quality, not the reactor.

What moves pyrolysis IRR the most?

Yield and uptime, each of which can turn a positive IRR negative, followed by the pyrolysis-oil price premium and feedstock cost. Capex overruns matter less, and the reactor technology is not a top driver.

How should investors de-risk a pyrolysis project?

Demand contracted specified feedstock, demonstrated uptime at scale, a bankable offtake into the ISCC premium circular market, resolved permitting and recycling classification, and capex certainty. Treat the first plant as a learning asset.

Methodology: 50,000 t/yr pyrolysis plant modelled on an unlevered, pre-tax basis; base IRR ~7.7%, single-variable stresses. Plant data (Brightmark, Encina) from Waste Dive, Resource Recycling, Inside Climate News and C&EN; the EPA’s March 2026 proposal to reclassify pyrolysis per Waste Dive / C&EN; yield ranges from ACS Energy & Fuels. Scenario weights are Trident’s framework. Analysis, not investment advice.