There is no single waste to energy business model. There are three, and the boiler does not decide which one you are in. The payer does.
That is the whole argument, and it is worth stating before the revenue tables, because the revenue tables are where most analyses go wrong. A waste to energy plant is a disposal business that happens to spin a turbine. Whether it is bankable depends on whether the entity paying for disposal can still pay in year eleven, and that question has a different answer in Ohio, in Java and on a factory site with no municipality anywhere in the deal.
What a waste to energy business model actually sells
Read the accounting rather than the brochure. Covanta, for years the largest energy-from-waste operator in the United States, told the SEC its projects generate revenue from three primary sources: fees for receiving waste, the sale of electricity or steam, and the sale of ferrous and non-ferrous metals recovered from the waste stream. Waste first, energy second, scrap third. That ordering is not stylistic. It is the order of the cash.
The energy line is smaller than the name suggests. The entire US waste to energy fleet generates around 14,000 gigawatt hours a year, less than 1 percent of American electricity, with about 90 percent delivered to the grid and the remaining 10 percent sold as steam to nearby industrial hosts. A whole national fleet produces less than a rounding error on the grid, and yet the plants keep getting built, because nobody builds them for the electricity. They get built because a city has to do something with its garbage on Monday morning.
So the model question is never "what does the plant produce". It is "who has a legal obligation to pay, and for what".
The revenue lines, ranked by how bankable they are
Five lines show up in almost every pro forma. Ranked by contract quality rather than size, they fall into three tiers.
Tier one: the gate fee
Waste generators pay to deliver material, and in most markets that fee is between $40 and $120 per tonne, accounting for 35 to 45 percent of total revenue and providing the most predictable cash flow when backed by long-term municipal or industrial supply contracts. Planning models in high-cost disposal markets work off a gate fee of $70 to $125 per inbound ton against roughly 450 to 575 kWh of net exported power per ton, which is the clearest way to see the asymmetry. One ton buys you one disposal fee and about half a megawatt hour. Price them both and the fee wins.
Tier two: the electricity
Real revenue, one grade lower. A mass-burn plant runs about $7,000 per kW of capex and needs roughly $0.16/kWh at 60 percent utilization to clear a 10 percent equity IRR on the energy alone. Utility-scale solar clears power purchase agreements at a fraction of that. A waste plant cannot win a merchant power market on price and should never be underwritten as though it could.
Tier three: metals, char, steam and carbon
Useful, volatile, and never the foundation. Landfill diversion typically credits 0.5 to 1.0 tonnes of CO2 equivalent avoided per tonne diverted, at $30 to $80 per credit in the 2026 market range. That is real money on a large plant and it is getting more valuable, but it is priced by a voluntary market, not by a contract a court would enforce. Treat it the way we treat any adjacent carbon removal technology revenue: upside, not underwriting.
Model 1: the gate fee utility
This is the model most of the English-language internet describes, and it works exactly where landfill is expensive.
The numbers set the boundary. US landfill tipping fees averaged $62.28 per tonne in 2024, up 10 percent in a single year, with waste-to-energy states running a 28 percent premium at $71.28 against $55.57 elsewhere, and a regional spread from $44.87 in the South Central region to $80.67 in the Northeast. A plant has to price its gate at or above the nearest landfill to pull tonnage away from it. Where the alternative costs $80, that is a business. Where it costs $45, the energy line has to carry weight it cannot carry, and the project should die at feasibility rather than at financial close.
Scale sets the other boundary. Below roughly 250,000 tonnes a year the capex per tonne of annual capacity turns punishing, running past $1,000 against about $680 at mid scale, per the same analysis. The gate fee model is a big-plant model in a rich-disposal market. Stated that plainly, it excludes most of the world.
Model 2: the single tariff IPP
Indonesia just built the cleanest available example of the opposite structure, and it is the most instructive thing to happen to this sector in a decade.
Under the old framework, developers earned two revenue streams: a feed-in tariff of about USD 0.1335/kWh from the state utility PLN, plus a tipping fee from the local government capped at IDR 500,000 per tonne. That regime produced exactly two operating facilities, and was replaced by Presidential Regulation 109/2025 with a single fixed electricity price of USD 0.20/kWh for the 30-year life of the project, with PLN required to sign the PPA within 10 working days of pre-construction licences.
Read what that reform actually did. No combustion breakthrough arrived. The regulator looked at a two-counterparty structure, concluded that municipal budgets are an annual political negotiation rather than a 20-year security, and deleted the tipping fee from the business model entirely. One creditworthy payer, one contract, one number.
The terms confirm it is a genuine independent power producer structure rather than a subsidy dressed up as one. The build-own-operate PPA carries a take-and-pay mechanism on annual contracted energy, lender step-in rights, liquidated damages for construction delay, and purchase of excess electricity at 20 percent of the base tariff, with no escalation across the full 30-year term. On the supply side, local governments must commit an agreed 1,000 tonnes of waste per day for the life of operations, and the flagship Legok Nangka project in West Java processes around 2,131 tonnes per day for roughly 40.79 MW with a Sumitomo and Hitachi Zosen led consortium.
The trade is explicit: the model now lives or dies on one regulated number. That number can move. Vietnam is proving it in the other direction, where the proposed 2026 ceiling price of 2,378.39 VND/kWh is 7.64 percent below the previous year and developers argue it stretches payback on some projects to 15 to 17 years, partly because the pricing model assumes 33 percent generating efficiency while real plants achieve 20 to 24 percent. A single-tariff model is only as durable as the regulator setting the tariff.
Model 3: on site processing, where no municipality is in the deal
The third model is missing from almost every ranking page, and it is the one we spend most of our time deploying.
Here there is no gate fee and no PPA. The payer is the waste producer: a factory, a plantation, a resort, a hospital. Revenue is avoided disposal cost plus displaced fuel, and the comparison price is not the wholesale grid. It is diesel.
That changes the arithmetic completely. A techno-economic study of municipal solid waste incineration in Kaduna, Nigeria put the levelized cost of electricity at USD 0.15 to 0.25/kWh, above the local grid tariff but well below diesel captive generation that often exceeds USD 0.30/kWh, with a discounted payback of 8 to 12 years. The same study found the project moves from viable to unviable on a 20 percent swing in collection efficiency or feedstock heating value, which was measured at roughly 4.12 MJ/kg with 53.5 percent food waste. Wet garbage is the binding constraint, not the turbine. Anyone underwriting this model should start from what electricity actually costs per kWh in Nigeria rather than from a European tariff sheet.
This is where our own track record sits. EX EPIC has designed, financed and installed waste-to-energy systems across 11 countries, with more than 200 units deployed through Zero-X, alongside a EUR 160M capital track record. Small distributed units do not clear the mass-burn scale test, and they were never supposed to. They clear a different test, because the customer is not selling power into a grid, they are buying their way out of a disposal bill and a diesel bill at the same time. It is the same logic that governs how a deep tech venture builder finances and deploys hardware in markets where the infrastructure finance chain does not exist. If you are scoping a project on these terms, there is a project guide to waste to energy for emerging markets worth reading alongside this.
Choosing a model: five questions that settle it
- Is the disposal alternative expensive? If landfill nearby costs $45 a tonne, Model 1 is not available to you at any boiler efficiency.
- Who signs the offtake, and can they still pay after a downgrade? A municipal budget line and a sovereign-backed utility are different credits. Indonesia's reform was, in substance, an answer to this one question.
- Does the tariff escalate? A flat price across 30 years is bankable on day one and eroded by year fifteen.
- Is there a put-or-pay tonnage commitment? Without it, tonnage risk sits with equity while fixed costs, debt service and emissions monitoring carry on regardless.
- Will the feedstock hold its calorific value? Contract for waste acceptance criteria, not for tonnes.
The Casablanca award shows what it looks like when the answers line up. The consortium holds a 33.5-year exclusive treatment concession worth around USD 1.5 billion, with electricity offtake agreements with national utility ONEE and the regional operator SRM Casablanca-Settat, and waste treatment and generation beginning six to ten months before final completion. Two contracted pillars, plus a phased commissioning that shortens the window in which the asset carries construction risk without earning anything.
That last detail is the one spreadsheets consistently underweight. Schedule risk lands on equity, not on lenders. Every month of permitting delay is capital sitting idle behind a fixed-cost build, and none of it shows up in the tariff. Before you choose a technology, settle who pays, get it in writing, and read how these projects actually get funded.
FAQ
Who owns a waste to energy plant under a BOO or BOT concession? Under build-own-operate the developer owns the asset through the concession term and is paid per tonne, per kilowatt hour, or both. Under build-operate-transfer, ownership reverts to the public authority at the end of the term. Casablanca is structured as a 33.5-year concession combining municipal treatment rights with utility offtake; Indonesia's PR 109/2025 PPA also adopts a build-own-operate model. A third variant exists where a public authority owns the facility and pays a private operator a fixed service fee to run it, which shifts commodity risk back to the public side.
What payback period do waste to energy projects target? Emerging-market feasibility work puts discounted payback at 8 to 12 years in optimistic cases. Where the tariff gets squeezed it stretches: Vietnamese developers argue the proposed 2026 ceiling pushes some projects to 15 to 17 years, which is long enough to complicate long-term financing and reduce investor appetite regardless of how well the plant runs.
What is a put or pay waste supply agreement? A commitment by a municipality or hauler to deliver a minimum tonnage or pay as though it had. It converts variable gate-fee income into something close to a contractual annuity, which is why lenders often treat it, rather than the power purchase agreement, as the real security. Indonesia achieves the equivalent by obliging local governments to commit a fixed daily waste volume for the life of operations.
Can a waste to energy plant sell carbon credits? Only the biogenic fraction of the waste, the food, paper and wood, carries a defensible renewable and carbon claim. Landfill diversion is typically credited at 0.5 to 1.0 tonnes of CO2 equivalent per tonne diverted. Booking it requires credible biogenic measurement and a registry willing to certify the method, so treat it as an upside line that needs its own due diligence rather than a revenue assumption at financial close.