We Connect Asia

Hydrogen

Hydrogen Project Risk Allocation: Who Carries Which Risk Before FID

10 min read  ·  Jan 2, 2026  ·  By Alex

hydrogen project risk

Sorting out who carries which risk in a hydrogen file is the first thing I do before a lender is approached. A risk left with the sponsor, the developer that commits the equity, and with no instrument behind it is the usual reason a project never reaches the final investment decision, the point at which that equity is committed. Hydrogen project risk allocation means deciding, before that point, who carries the offtake risk, meaning the buyer failing to take the output, the political and regulatory risk, the technology risk and the power supply risk, and with which instrument, because the cash flow behind the debt has to survive all four. I work inside the hydrogen, power-to-liquid and e-fuels branch, and I read every file from the lender’s chair: what is contracted, what is only announced, and who takes the loss when one of those four things happens.

This article orders the four risk categories a lender separates, the revenue side behind them, the public instruments that carry part of the gap, the Southeast Asian position, and the questions I ask before I call a risk allocated. Throughout, I keep announced figures and contracted figures apart.

In short

  • Offtake: only around 13 percent of contracted volumes are binding, and 94 percent of capacity carries no identified offtaker (2024).
  • Politics: offtake and market risk, then political and regulatory uncertainty, are the two constraints on debt in emerging markets (2024).
  • Technology: technology and environmental risk sit in the equity funded development phase, before any debt is raised (2024).
  • Instruments: contracts for difference and capital expenditure support are matched to asset type rather than offered to every file (2024).
  • Southeast Asia: 6 percent of announced regional capacity had reached a final investment decision against 4 Mt of demand in 2024.
Facebook
Twitter
LinkedIn

In this article

What risk allocation means in a hydrogen file

A 2026 paper separates four large categories: technology and engineering, execution and development, financial and market, political and regulatory. The first two are normally carried by equity before debt is raised, so risk allocation is the written decision about who carries the rest, through which contract, for how long.

  • Sponsor: the developer committing equity and absorbing whatever loss the contracts leave uncovered.
  • Non-recourse debt: debt repaid only from project cash flow, with no claim on other assets.
  • Offtake: the contract by which a buyer commits to purchase the output.
  • Take-or-pay: payment for a set share of volume whether or not the buyer lifts it.
  • Contract for difference: a payment closing the gap between a reference price and the realized cost.
  • Credit enhancement: a letter of credit, guarantee or reserve standing behind a weaker offtaker.
  • Investment grade: a rating from BBB- upwards, read as a solvent buyer.
  • Final investment decision: the moment the equity is committed.

What this means for you: I want these words written into your term sheet as defined here, because a term left vague is priced as unallocated.

What a lender ranks first

The Oxford Institute for Energy Studies paper Bankability of Hydrogen Projects of January 2026 defines bankability as the degree to which a project is judged able to service its debt through project cash flow, a workable contract structure and a risk allocation solid enough to attract outside capital. Projects that reached a final investment decision combined stable policy with long-term visibility, a solid sponsor and an offtake that anchors the financing.

On World Bank data from 2024 adapted in that same paper, offtake and market risk and then political and regulatory uncertainty are the most critical obstacles to debt in emerging markets, while technology and environmental risk sit in the equity funded development phases. Only around 13 percent of contracted volumes sit on fully binding purchase agreements, and 94 percent of capacity carries no identified offtaker.

The metrics I track include the minimum debt service coverage ratio, the reserve account size, the contracted volume share, merchant exposure, contract tenor and the offtaker’s rating. Blended finance, subordinated concessional tranches and guarantees absorb first losses without removing the market or offtake risk beneath them. Long-term usually means more than 10 years, while some projects reached a final investment decision on 5 to 7 year contracts only by contributing around 50 percent equity. I turn those numbers into contract language through contract drafting and risk allocation work.

What this means for you: if two of those metrics are missing from your file, I do not call the risk allocated, I call it pending.

How revenue risk is written into the contract

The paper Hydrogen Offtake Agreements of August 2025 records that hydrogen sale and purchase agreements typically carry long-term commitments of 10 to 15 years, about half the length of the first international gas and liquefied natural gas contracts. Price models run cost-plus, netback, fixed or partly fixed, against the absence of a transparent market price for the fuel.

Electricity and gas act as price proxies and import their volatility, while public support enters the price through formulas that split the value of the subsidy between the parties. Expected take-or-pay volume provisions run 60 to 80 percent, against more than 90 percent in gas and liquefied natural gas. Change of law clauses come first in the negotiation, and broad contract reopenings are unlikely because they undermine bankability. In March 2025 a delivery of 30 000 tonnes a year for 15 years from 2030 was agreed, which is the scale a signed file actually carries. I set these clauses against the financing route I map for a hydrogen file.

What this means for you: I read the take-or-pay percentage before the price, because that percentage pays the debt in a year the buyer would rather skip.

Who actually carries the revenue risk

The Green Hydrogen Organisation work on offtake agreements of 2024 reports that only 10 percent of announced projects have identified a buyer, and that of contracted volume 13 percent is contractually binding, 7 percent is documented pre-contractual and the remaining 80 percent is memoranda of understanding or unspecified.

An 8.4 billion USD project reached financial close in May 2023, was oversubscribed more than two times and was financed on a non-recourse basis by 23 banks and investment institutions. The first public subsidy auction of 800 million EUR was oversubscribed several times in February 2024. Both show appetite on paper; neither shows volume sold.

Financiers require long-term fixed price offtake with creditworthy offtakers, and where the offtaker lacks an investment grade rating they ask for letters of credit, bank guarantees, parent guarantees or retention of sums, an investment grade rating starting at BBB-. Buyer side take-or-pay obligations secure the revenue that pays the debt. I place those requirements in the project finance route for green hydrogen in Southeast Asia.

What this means for you: a memorandum of understanding is not an offtake, and I treat that 80 percent as unsold until a binding document says otherwise.

What public instruments absorb

The H2Global Foundation report of December 2024 puts the investment needed to line the hydrogen economy up with net zero at 1 000 billion USD worldwide by 2030, of which 335 billion USD is not yet allocated, with infrastructure accounting for the largest share of the shortfall at 190 billion USD against 145 billion USD for end uses.

Four support instruments are evaluated in it: capital expenditure support, a fixed premium, contracts for difference and anchor capacity reservations. A survey of donors ranked market risk, political and regulatory risk and permitting and compliance risk as the most critical obstacles to hydrogen infrastructure. Public private partnerships, regulated asset models and special purpose vehicles have worked elsewhere but do not answer hydrogen’s specific problems, which is why a contract for difference is recommended for high risk, long duration assets such as pipelines and underground storage, and capital expenditure support for simpler ones.

Uncertain future volumes raise the financing cost against conventional energy projects, and this support only pays once the fuel meets its certification rules, including the RFNBO matching rules.

What this means for you: I match the instrument to the asset, because a fixed premium on a pipeline does nothing for the volume risk that makes the pipeline expensive.

Where Southeast Asia sits

The IEA’s Southeast Asia presentation of the Global Hydrogen Review 2025 counts 4 Mt of regional hydrogen demand in 2024 against only 6 percent of announced regional capacity at a final investment decision, where Mtpa means million tonnes a year of production.

It puts regional potential at nearly 11 Mtpa below 6 USD/kg by 2030, with no certification scheme in place and work going on in three countries, and names high cost of capital as the main barrier. Over 2021 to 2025 it counts 7 Mtpa of offtake agreements announced of which only 1.6 Mtpa are firm, and 6.6 billion USD of development finance institution loans.

Its recommendations to governments are to prioritise renewable deployment, cut the cost of capital, build pilot projects, anchor demand with large applications and develop certification. I read that position through the investment conditions I track across Southeast Asia.

What this means for you: in this region the barrier I price is the cost of capital and the missing certification scheme, not the electrolyser.

What I would not sign without

Five things sit in my signature block. An offtaker rated investment grade, or a credit enhancement that behaves like one. A take-or-pay percentage written into the contract rather than expected. A reserve account sized on the same cash flow the model shows, not on a happier case. A change of law clause that survives a change of subsidy regime.

The fifth is certification I can still meet when the contract starts. Under the European rules published in 2024 the production must fall in the same month as the contracted generation until 31 December 2029 and in the same hour from 1 January 2030, against annual matching until 2027 and hourly matching from 2028 in the United States (2024). Each answer above is written, dated and testable; none of them is a promise that policy will hold.

What this means for you: I sign only where the document answers the risk, and I walk away where the answer is a plan to be written later.

The four questions I ask before I call a risk allocated

Four questions, in this order. Who pays if the offtaker stops lifting. Who pays if the policy changes. Who pays if the equipment underperforms. Who pays if the electricity is not there. If the answer to any of them is the sponsor alone, the risk is not allocated, it is parked.

The regional numbers make the point. Announced projects in Southeast Asia equal 12 percent of current regional demand while only 6 percent of announced capacity has reached a final investment decision (2025), and 7 Mtpa of offtake was announced over 2021 to 2025 against 1.6 Mtpa firm (2025).

What this means for you: for a developer I extract one rule. Write the answer to each question into a contract before the equity is committed, and let the instrument carry the loss rather than the expectation. Where no instrument exists yet, the file is not ready for debt.

Sources

Frequently asked questions

The decision, taken before the final investment decision, of who carries the offtake, political, technology and power risks and through which contract or instrument, so that the cash flow behind the debt survives each of them. I take that decision before the equity is committed, because a risk left with the sponsor and no instrument behind it is what usually stops the debt. It is a contract question before it is a finance question.

In emerging markets the offtake and market risk and then political and regulatory uncertainty are the most critical constraints on debt, while technology and environmental risk sit in the equity funded development phase (2024 data). A file can fix its equipment and still fail if nobody is obliged to buy the output, which is why I test the offtake document first.

Only 13 percent of contracted volume is contractually binding and 10 percent of announced projects have an identified buyer (2024), while 7 Mtpa of offtake was announced over 2021 to 2025 against 1.6 Mtpa firm (2025). The remainder is memoranda of understanding or unspecified, which I do not count as sold volume.

A creditworthy offtaker or credit enhancement, a take-or-pay percentage in the text, a reserve account, a change of law clause and certification rules that the fuel will still meet when the contract starts. All five are documents I can read before signing rather than intentions held by a counterparty. If one is missing, the lenders will price the gap as risk sitting on the sponsor.

Not sure your project would survive a credit conversation?

A short call is usually enough to see where your project stands.

If your four answers are written but not yet tested, I run the file through contracting and risk management support, so every risk has a clause, a named party and an instrument before the lenders see it.

I do not provide investment advice or arrange investments. I act as a strategic advisor and facilitator.

Alex, Founder of WeConnect Asia

Alex, Founder of WeConnect Asia

I help low-carbon fuel and carbon removal project developers in Southeast Asia secure the commercial and financial partners needed to reach FID.

More about me·Get in touch

Facebook
Twitter
LinkedIn

Related articles