There is a specific moment in a hydrogen project's life where a good LCOH stops helping. It is the moment a credit committee asks how much debt the project can carry, and the answer turns out to depend almost entirely on a document nobody in the modelling team owns: the offtake contract.
This piece follows that chain, using the sources that state each link. It does not end with a number, because the number does not exist — and the reason it does not exist is worth more than the number would be.
The paper that will not give you a DSCR
The most authoritative recent treatment of hydrogen bankability is co-written by Henry Rushton, a director in ING's energy sector, and Aliaksei Patonia of the Oxford Institute for Energy Studies. It is unusually direct about what lenders look at:[1]
In practice, senior lenders focus less on standalone LCOH or project IRR and more on cash-flow-based indicators such as minimum DSCR, the size of the debt service reserve account (DSRA) and the proportion of revenue that is fully contracted rather than merchant.
Two pages later it declines to say what those levels are, and explains why:[1]
Whilst exact DSCR and LTV levels will vary depending on the strength of project sponsors, offtakers and regulatory support regimes (amongst other factors), it is likely that levels will be benchmarked on a more conservative basis compared to more established energy sector financings and for this reason, providing numerical thresholds would be misleading.
A banker writing about bankability, in an academic paper, declining to publish a benchmark on the grounds that it would mislead. Take that seriously rather than as a hedge. It means that any figure circulating as "the hydrogen DSCR" did not come from here, and it means the useful work is upstream of the ratio — in what goes into the cash flow the ratio is calculated on.
Link one: the contract caps the debt tenor
The rule is old, it is not specific to hydrogen, and three independent sources state it. Financiers require that the offtake term be at least as long as the debt tenor, plus a tail.[5] Put the other way round, the loan term is generally capped at the term of the contract, less a tail of one to two years.[3] Norton Rose Fulbright puts the consequence in a single sentence: financiers amortise the debt over the length of the agreement, so the longer the agreement, the more debt is available.[6]
So the length of the contract is not a commercial detail. It is the input that sets how many years of cash flow the debt can be spread over.
Figure 1
The contract is shorter than the asset, and shorter than the market it copies
Offtake tenor, in years.
ET52, p. 20. This is the observed European practice, not the design intent.
ET50, p. 6: most agreements are likely to fall within 10 to 15 years.
ET50, p. 5: GSAs and LNG SPAs often exceeded 20 years. The source states a floor; the bar is drawn to 25 for scale.
GaugeH₂ modelling convention. Not a published figure — shown so the gap has something to be a gap against.
The comparison with asset life is OURS, not any single source's. ET50 compares hydrogen tenor to historical gas and LNG tenor; it does not compare it to asset life. The bottom bar is a GaugeH₂ modelling convention, included so the gap is visible, and it is not a published figure.
Source: Oxford Institute for Energy Studies — ET50 (Ason, 2025) and ET52 (Rushton & Patonia, 2026) · Data as of 2026-01
Show the data ▸Hide the data ▾
| Basis | Tenor (years) | Source |
|---|---|---|
| European projects at FID, in practice | 5–7 | OIES ET52, p. 20 |
| Hydrogen offtake, expected range | 10–15 | OIES ET50, p. 6 |
| Gas and LNG sale agreements, historically | over 20 | OIES ET50, p. 5 |
| Asset life in an LCOH model | 20–25 | GaugeH₂ convention, not a source |
Hydrogen offtake agreements are expected to land at 10 to 15 years, against gas and LNG sale agreements that historically often exceeded 20.[2] European projects actually reaching FID have gone shorter still, at 5 to 7 years.[1]
Run that through the rule above and a 5-year contract supports roughly 3 to 4 years of amortising debt on an asset expected to run for twenty. The gap is not a modelling nuance. It is most of the asset's life.
Link two: the contracted volume inside the contract is softer than it looks
A contract term is not the same as contracted cash flow. What the lender can count is the take-or-pay floor — the volume the buyer pays for whether or not it lifts it.
Figure 2
And the contracted volume inside it is softer
Take-or-pay threshold: the share of contracted volume the buyer pays for whether or not it lifts.
ET50 gives three reasons for the lower threshold, and the first inverts the usual causality: hydrogen projects already carry higher equity contributions, so less take-or-pay protection is demanded. The other two are the revenue certainty that public support already provides, and that hydrogen offtake is usually motivated by decarbonisation rather than security of supply, so buyers have alternatives.
Source: Oxford Institute for Energy Studies — ET50, Hydrogen Offtake Agreements (Ason) · Data as of 2025-08
Show the data ▸Hide the data ▾
| Contract type | Take-or-pay threshold |
|---|---|
| Clean hydrogen, in negotiation | 60–80% |
| Early LNG and gas, historically | over 90% |
Clean hydrogen contracts are being negotiated at thresholds of 60 to 80%, where early LNG and gas projects routinely exceeded 90%.[2] Twenty points of take-or-pay is twenty points of base-case revenue that has to be found somewhere else, or done without.
The reasoning behind the softer threshold is worth reading in full, because the first of the three reasons inverts the causality most people assume. It is not that a weak take-or-pay forces more equity; it is that hydrogen projects already carry higher equity contributions, so buyers are asked for less protection. The second is that public support schemes already supply revenue certainty. The third is that hydrogen offtake is usually motivated by decarbonisation targets rather than security of supply, so buyers have alternatives and negotiate accordingly.[2]
There is an exception the paper is careful to state: in a greenfield project with no alternative supply, higher thresholds will still be required.[2]
Link three: what falls outside the contract does not count
Whatever the project produces above the contracted, take-or-pay-protected volume is merchant. Lenders treat it conservatively — haircut, or excluded outright from base-case revenue projections — and the working assumption is that it stays either out of the debt-sizing case or in as upside, not as core cash flow.[1]
The historical anchor for that is uncomfortable and precise. In LNG, it took more than three decades before lenders were willing to underwrite any meaningful share of merchant revenue.[1]
This is where a well-built LCOH model and a bank credit paper stop describing the same project. The model computes a cost over the full nameplate output across the full asset life. The credit paper sizes debt on a fraction of that output, for a fraction of that life, at a price fixed in a contract. The two are not in disagreement — they are answering different questions, and only one of them is the question at the committee.
Link four: the shortfall is paid in equity
This is the part that is actually quantified, and it is the single most useful figure in the literature.[1]
Some European projects have reached FID with contracts of 5-7 years, though this typically requires a higher equity contribution of approximately 50% to compensate for the shorter revenue certainty period.
Figure 3
A short contract is paid for in equity
Debt as a share of total investment. The three grey bars are other sectors, offered as precedent — not hydrogen outcomes.
ET21, p. 20. Analogue, not hydrogen.
ET21, p. 20. Analogue, not hydrogen.
ET21, p. 20. Analogue, not hydrogen.
ET21, p. 2: USD 1,250–1,450m of debt against USD 2,100m of capex. A model, not a closed deal.
ET52, p. 20: a higher equity contribution of approximately 50% to compensate for the shorter revenue certainty period.
The grey bars matter because they are what hydrogen is being benchmarked against, and they are the reason a 50% gearing reads as expensive. No lender has published a gearing outcome for a hydrogen project: the NEOM ratio that circulates is arithmetic on a disclosed debt figure and an undisclosed equity figure, and we do not use it.
Source: Oxford Institute for Energy Studies — ET21 (Craen, Hove & Meidan, 2023) for the sector precedents and the hydrogen archetype; ET52 (Rushton & Patonia, 2026) for European practice · Data as of 2026-01
Show the data ▸Hide the data ▾
| Case | Debt share | Source |
|---|---|---|
| Offshore wind, mature | up to 80% | OIES ET21, p. 20 (analogue) |
| LNG, typical | ~70% | OIES ET21, p. 20 (analogue) |
| Offshore wind, first projects | 60–65% | OIES ET21, p. 20 (analogue) |
| Hydrogen export archetype, modelled | 60–70% | OIES ET21, p. 2 |
| European hydrogen, 5–7 year contracts | ~50% | OIES ET52, p. 20 |
Set that against what hydrogen is implicitly benchmarked against and the cost becomes legible. Offshore wind started at 60–65% debt and reached up to 80% over fifteen years of precedent; LNG typically raises around 70%.[3] A modelled hydrogen export archetype assumes 60–70%.[3] A European project on a 5–7 year contract gets about 50%.
Moving from 70% to 50% gearing on the same capital cost does not shave the equity cheque. It raises it by two thirds. That is the price of the short contract, and it is charged before anyone argues about a coverage ratio.
What it looks like when it actually closes
Stegra is the largest hydrogen-based project financing Europe has completed, and it is worth looking at because its disclosure is unusually full.
Figure 4
What that looks like in a real capital stack
Stegra, January 2024. Europe's largest hydrogen-based project financing to date.
- Senior debt
- Junior debt
- Equity raised to date
The equity figure is the amount accumulated at that date, not a committed sponsor contribution set against this debt. And the stack is not the whole answer: the senior debt carries a Swedish state green credit guarantee and Euler Hermes cover reported at 80% and 95% of the loan amount respectively. Half of the initial 2.5 Mt/year of output was sold on binding five- to seven-year customer agreements.
Source: Stegra (then H2 Green Steel), financing announcement · Data as of 2024-01
Show the data ▸Hide the data ▾
| Tranche | Amount (€m) |
|---|---|
| Senior debt | 3,500 |
| Junior debt (up to) | 600 |
| Equity raised to date | 2,100 |
| Total | 6,200 |
Half of the initial 2.5 Mt/year of output was sold on binding five- to seven-year customer agreements.[8] That is precisely the profile ET52 describes, in a deal signed two years before ET52 described it.
What closed the remaining gap was not a coverage ratio. It was guarantees: a Swedish state green credit guarantee and Euler Hermes cover, reported at 80% and 95% of the loan amount respectively.[8] Read that again as a credit structure. On the guaranteed portion, the lender's exposure is largely to a sovereign and an export credit agency, not to the hydrogen project.
NEOM's structure carries the same signature from the other end. USD 8.4 billion of total investment, USD 6.1 billion of non-recourse debt across 23 institutions, senior and mezzanine tranches, a Euler Hermes tranche, Saudi state institutions — and an exclusive 30-year offtake with Air Products for 100% of the ammonia produced.[10] A thirty-year contract for all of the output is the condition that makes that debt quantum possible, not a footnote to it.
Why this is getting harder, not easier
Figure 5
The input to all of this is not growing
New low-emissions hydrogen offtake agreements signed, million tonnes per year.
2023
2.4
2024
1.7
2025
1.7
The headline volume is not the underwritable volume. The IEA reports that only around 20% of the volumes newly signed in 2025 were backed by firm contractual commitments — roughly 0.34 Mtpa on a 1.7 Mtpa headline.
Source: IEA — Global Hydrogen Review 2025 and 2026 · Data as of 2026-06
Show the data ▸Hide the data ▾
| Year | New offtake signed (Mtpa) |
|---|---|
| 2023 | 2.4 |
| 2024 | 1.7 |
| 2025 | 1.7, of which ~20% firm |
New offtake agreements signed fell from 2.4 Mtpa in 2023 to 1.7 Mtpa in 2024, and stayed at 1.7 in 2025.[11][12] The headline is not the underwritable number: only around 20% of the 2025 volume was backed by firm contractual commitments — roughly 0.34 Mtpa.[12]
Against a 2030 pipeline the IEA measures in tens of millions of tonnes, with projects that have reached FID representing 9% of it.[11] The binding constraint is on the demand side of the contract, which is the same thing as saying it is on the debt.
What to do with this before your next committee
None of the following requires a DSCR benchmark, which is convenient, because there is not one.
- Model the contract, not the plant. Base-case revenue is contracted volume × take-or-pay floor × contracted price, for the contract term. Everything else is a separate line labelled upside.
- Cap the debt tenor at the contract term minus a two-year tail before sizing anything. If that leaves four years of amortisation, you have found the real constraint, and it is not the electrolyser.
- Price the equity consequence explicitly. Going from 70% to 50% gearing raises the equity cheque by two thirds. Put that number in front of the sponsor early, because it is the number that kills projects quietly.
- Name the counterparty's credit. Financiers require a creditworthy buyer, or a buyer with a creditworthy parent.[6] An unrated offtaker is an unfunded project with extra steps.
- Ask what guarantee is doing the work. In both deals examined here, an export credit agency or a sovereign carries a large share of the risk. If your structure has no equivalent, you are asking commercial lenders to take a risk that the two reference transactions did not ask them to take.
- Expect lower gearing and oversized contingency, and shorter tenors than a comparable renewables project.[7] How much contracted revenue is required is not a published constant — it varies project by project.[7]
The LCOH still matters. It decides whether anyone wants the hydrogen at all, and a project that cannot clear a buyer's price threshold never reaches a credit committee. But clearing that threshold is the entry condition, not the decision. The decision is made on a contract, and it is made by people who will read that contract more carefully than they read your model.