Why FOAK Climate Projects Are Hard to Finance

Sections

H2 Green Steel, since renamed Stegra, is building a green steel plant at Boden in Sweden. In January 2024 it closed more than €4 billion of debt financing for the first phase of that plant, alongside equity and public grants. The credit committees behind that number were lending against a first-of-a-kind asset. They could benchmark parts of the construction, equipment and feedstock plan, and nobody could show them a commercial plant with the same process, configuration and operating history. What they could see was demand in contractual form, because the company had already reported pre-selling more than 1.5 million tonnes of its planned output under multi-year offtake agreements, contracts in which buyers commit to take future output.

In 2026 Stegra had to raise further capital to complete the plant after a funding gap opened during construction. A financial close of that size did not end the financing story. First plants keep generating new decisions as construction and ramp-up reveal what earlier models could only estimate, and each new decision is taken before the asset has produced the evidence that would make it easier.

The capital becomes project-sized before the evidence does

At pilot stage, venture equity, grants and strategic support can finance learning. A first commercial plant changes the shape of the problem. The Canadian innovation hub MaRS, launching its first-of-a-kind lab in 2026, noted that complex first commercial projects can fall outside traditional investor and programme mandates. J.P. Morgan’s 2026 climate-tech report describes a missing middle in which a first facility may need growth equity, catalytic capital, equipment finance and project finance together. The US Department of Energy’s review of FOAK financing approaches describes project debt for first commercial projects as possible while emphasising strong protections for lenders and guarantors and long diligence before conditional commitments.

Investors in our interview archive describe the same gap from inside their funds. Jonny Everett of Marble, a Paris venture studio, told us in July 2024 that “While the early stage in climate tech is well-supported, the middle stage is more challenging at the moment.” A month earlier, Myke Näf of Übermorgen Ventures, a pre-seed and seed investor, had put the hardware problem in one line. “The path may be clear but financing the capex is challenging.” Carlota Ochoa Neven Du Mont of the climate fund Extantia Capital described its role later that year as “the capital that helps those companies move from a demo or out of the lab to a first-of-a-kind solution.” Each was speaking from their own fund's position at the time of the interview, and Extantia's cheques then ran from €1M to €4M, a fraction of a plant. Venture can carry a company to the edge of its first asset, and the asset itself needs capital of another kind. That is why a company can be too capital intensive for venture and still too novel for ordinary project debt at exactly the same moment.

A reference class gives uncertainty a shape

Finance can price uncertainty once there is enough structure around it to set terms. A lender looking at a familiar asset class can draw on histories of construction cost, schedule, equipment failure, output and default. The future remains uncertain, and the lender still has a range against which to test assumptions.

Large nuclear shows a reference class being built in public. Georgia Power announced in 2024 that Unit 4 at its Vogtle plant had entered commercial operation, using Westinghouse's AP1000 reactor design. Idaho National Laboratory then published an analysis of potential cost reduction in new nuclear deployments based on recent AP1000 experience. The report treats the builds that have happened as the basis for estimating the cost of the builds that follow. The technology is far from a fermentation plant or a carbon removal site, and the discipline transfers.

FOAK removes part of that comfort. The civil works may be conventional while the integrated process is new. Individual pieces of equipment may have operating histories while the complete configuration does not. A product can meet specification at pilot scale while nobody has produced it at the intended commercial volume.

Daniel Ellsberg’s 1961 paper on risk and ambiguity separates uncertainty with known or estimable odds from situations where the odds themselves are unclear. The paper concerned decision-making in general and says nothing about first plants, so applying it here is an analogy. The distinction still names what financiers confront at a first plant. Some uncertainties can be benchmarked. Others have too little comparable evidence for a confident probability distribution. Waiting for somebody else to build first can therefore be rational for an individual financier. The collective problem appears when every financier reaches the same conclusion.

How Stegra surrounded the novel part with precedent

The Stegra commitments changed the evidence available to the capital providers while the plant itself remained a first-of-a-kind asset. Demand had a contractual form, and Mercedes-Benz was among the buyers that announced supply agreements with the company. The car maker decided on supply and specification. The lenders decided on repayment and security. Public support was visible. The financing stack spread the exposure across different kinds of capital, so no single lender had to carry the whole of the ambiguity.

Spreading the exposure has a cost in time, because each counterparty decides on its own part and on its own clock. Carbios, which recycles PET plastic with enzymes, shows the company and the asset being financed separately. In March 2026 it described financing for its Longlaville PET biorecycling plant, with an estimated construction cost of €230 million, through bank debt, equity from French partners and a Carbios contribution, with the project held in a separate operating entity. In a later 2026 update it reported that several banks had approved the plant at credit committee while its earlier financial-close target would not be met. Our reading is that the banks and the other parties were each deciding on a different exposure. Our essay on why pilots do not convert follows that sequence in full.

The weakest FOAK case presents the entire plant as one innovation and forces the reader to treat too much of the project as unknown. A stronger case separates what has precedent, what has been demonstrated, what depends on a contract and what is still unresolved, which our guide to presenting FOAK risk sets out as a register.

The technology can be credible and the project can still be unfinanceable

A first plant can be technically sound and still fail a financing test because the offtake is weak, the feedstock is uncertain, the site is unresolved, the buyer lacks credit quality, the construction contract leaves too much exposure with the project or the economics do not work at the capital cost on offer. Different capital providers will also read the same plant differently. Company equity takes corporate and development risk. Grants may support demonstration or policy aims. Strategic capital may value supply or market access. Project equity takes asset-level exposure. Debt needs a path to repayment and protections around completion. One generic fundraising deck cannot answer all of those decisions equally well.

Public and catalytic capital sits between company equity and project debt. The US Department of Energy's Title 17 loan programme exists to help deploy innovative clean energy technologies at commercial scale. Prime Coalition launched Trellis Climate to accelerate first-of-a-kind climate projects through catalytic capital, money that accepts more risk or a lower return so that other investors can follow. Public funders can take exposures private markets decline because a project produces learning or public value. They answer to their own mandates, so they ask why their support is additional, which market gap it fills and what private capital it could unlock later. A project that describes public money as free capital loses that support quickly.

Fundable and financeable are different tests

A FOAK project is fundable when some investor or funder will put money in on today's risk. It is financeable when the asset itself can support the debt or project finance structure the developer needs, on terms a lender can defend to its own credit committee. Neither word is a formal category. The distinction still changes what you say to each capital provider. Many climate companies raise equity, strategic, grant or catalytic capital for years while their first plant still lacks the contracted cash flow, precedent and risk allocation a lender requires.

A venture investor backs a company before any asset produces stable cash flow. A project lender asks a narrower question. It wants to know whether this asset will generate enough dependable cash to service the debt, and what happens when it does not. A founder who brings the venture deck to a lender tends to hear polite interest, followed by requests for signed offtake, a fixed-price construction contract and a completion guarantee.

LanzaJet, a sustainable aviation fuel producer, shows where grant money can sit. The first project funding from Catalyst, a Breakthrough Energy programme, was a $50 million grant to LanzaJet for its first alcohol-to-jet plant. LanzaJet later opened Freedom Pines Fuels, which it described as the world's first ethanol to sustainable aviation fuel production facility. A grant carries no repayment obligation, so it absorbs exposure that debt could not price. It made the project more fundable. The announcements say nothing about debt in the stack. Whether the next plant becomes financeable depends on the operating record this one produces, which our FOAK guide sets out.

"Bankable" is too vague to plan against. Name the lender you are aiming at, the contracts it will lend against, the technology exposure it can accept and the security it expects. Then stage the ask to the evidence. A lender should not be asked to commit construction debt while front-end engineering is incomplete.

Development
Engineering
Commercial commitments
Final investment decision
Construction
Operation
Refinance or repeat deployment

The gates vary by project. Each one should name the evidence that unlocks the next form of capital. List every capital provider you plan to approach and write one line for each on the exposure it will take and the evidence it needs. Where a line demands evidence you lack, you have found the next milestone. Where no provider will take a given exposure, you have found the gap that catalytic or public capital must fill. A line such as “we need £80 million” is a number. Each provider needs the reasoning that shows why its part of that number fits its mandate, and that reasoning should be built with qualified finance advisers.

Where communication stops

This is also where communication has to know its boundary. Brighter Future can make the evidence states, dependencies and capital logic easier to inspect. We will not describe a structural financing gap as an investor-understanding problem when the project lacks the contract, economics or mandate fit required for the capital it is seeking.

When first-plant financing has stalled and the cause is unclear, The Belief Diagnosis finds where the case stops moving in two to three weeks. If the gap is structural, the diagnosis says so.

Who keeps the value of the first plant's evidence

FOAK finance contains a circularity that better language cannot remove. The market wants evidence from a commercial plant before treating the technology as established, and the commercial plant needs financing before that evidence exists. Some projects break the loop by surrounding the novel part with enough precedent, contracts, guarantees, public capital or strategic commitment for a first group of capital providers to act. Others wait because the remaining uncertainty is still too large or too poorly allocated.

The financiers who do act pay for something they do not fully own. Delivered cost, construction schedule, uptime and ramp-up data from the first plant become the reference class that the second plant's lenders will use, and so will the company's competitors and their lenders. The first group carries the ambiguity, and much of the evidence it buys flows to people who carried none of it. Our essay on why nobody wants to become the precedent sets out what that gap can cost in money and years.

A first-plant financing case is stronger when it says where that value goes. A company can show which follow-on projects the first plant's data will support, offer early capital a defined place in the financing of later plants, and keep the operating record in a form that the first group's credit teams can use before anyone else. The financing belongs in that record too. Note which capital providers took which risks, which conditions were hardest to satisfy, which public or catalytic funds made the difference and what should not need repeating. Project two may reach different forms of capital because project one exists, and the terms it wins depend on how clearly the first financing was documented. The first plant is expensive partly because it has to create the evidence that lets the second plant enter a different conversation, and the capital that pays for that evidence has a fair claim to some of the conversation that follows.

Sources and further reading

Tell us what needs to move.

Bring the brief if it is clear. If it is unclear, tell us where the work is stuck.

Bring us the problem

Why Are FOAK Climate Projects Hard to Finance? Risk, Ambiguity and the Missing Reference Class

The capital becomes project-sized before the evidence does

A reference class gives uncertainty a shape

How Stegra surrounded the novel part with precedent

The technology can be credible and the project can still be unfinanceable

Fundable and financeable are different tests

Where communication stops

Who keeps the value of the first plant's evidence

Sources and further reading

New interviews and studio thinking, by email.

Continue in Commercialisation, Pilots, FOAK & Scale

The capital becomes project-sized before the evidence does

A reference class gives uncertainty a shape

How Stegra surrounded the novel part with precedent

The technology can be credible and the project can still be unfinanceable

Fundable and financeable are different tests

Where communication stops

Who keeps the value of the first plant's evidence

Sources and further reading

New interviews and studio thinking, by email.

Continue in Commercialisation, Pilots, FOAK & Scale

Why Nobody Wants to Become the Precedent

How to Present FOAK Risk Without Pretending the Risk Is Gone

What Is an Offtake Agreement and Why Does It Matter for FOAK Climate Tech?