Green Hydrogen Project Roadmap: Feasibility to FID

last updated

October 14, 2026

Why do green hydrogen projects stall before investment?

Announced projects far outnumber funded ones. As of October 2026, the International Energy Agency (IEA) reports that announced low-emissions hydrogen projects could supply 27 million tonnes (Mt) a year by 2030, down from 37 Mt a year in its 2025 review. Projects that are committed or have strong potential fell from 10 Mt to just above 6 Mt (IEA, Global Hydrogen Review 2026 and 2025).

Few projects reach FID. About 300,000 tonnes a year of extra production reached FID after the 2025 review, and new FIDs in production projects fell for the first time in 2025 (IEA, Global Hydrogen Review 2026). The IEA warns that 22 Mt of potential production could lose any chance of starting by 2030 if investment decisions are not made by early 2027.

The IEA names high costs, uncertain demand, unclear regulation and slow infrastructure as the causes (Global Hydrogen Review 2025). Those are market forces. Inside a project, four gaps often cause the stall. This is the authors' practitioner view, not an IEA finding:

Feasibility studies that underestimate the work of connecting the electrolyser to power, water, compression and storage. Everything outside the electrolyser package is the balance of plant (BoP).
Cost models that use annual average electricity prices instead of hourly profiles.
Engineering, commercial and policy teams that keep their assumptions in separate files.
Certification and monitoring plans left until after front-end engineering design (FEED), when changes cost the most.

The fix is to have shared assumptions and make earlier decisions. Adding people does not solve it. Late scope changes arrive as change orders under the engineering, procurement and construction (EPC) contract.

What does a stage-gate roadmap look like for a hydrogen project?

A stage-gate process makes teams agree before the project commits money it cannot recover. Each gate has one question, a minimum set of outputs and a list of red flags. It is cheaper to stop a weak project at Gate 1 than at Gate 5.

Treat power as a design input from Gate 1. The IEA's Levelised Cost of Hydrogen Maps choose the cheapest mix of solar, wind and electrolyser size from hourly capacity factors, and add battery storage or curtailment where needed. The IEA has archived the maps and no longer updates them, so use them to learn the method, not to price your project (IEA, Levelised Cost of Hydrogen Maps).

The table below sets out seven gates:

Gate	Main question	Minimum outputs	Red flags that predict a stall
0. Opportunity framing	Is there a real use case?	Use-case screen; demand hypothesis; map of power, water and land limits	No credible buyer; no check on infrastructure limits
1. Pre-feasibility	Does any configuration work?	Block flow diagram; power concept; cost ranges with the main sensitivities	Technology-first story with no cost drivers quantified; single-point cost estimates
2. Feasibility and concept selection	Which design costs the least?	Integrated system model; grid or dedicated renewables choice; utilities and site concept	Intermittency ignored; BoP scope undefined; water supply unconfirmed
3. FEED readiness	Can it be engineered and permitted?	FEED scope; interface specifications; hazard study plan; permit route confirmed	Unclear boundary between the electrolyser package and BoP; hazard studies postponed
4. Commercial structuring	Can the cash flows be contracted?	Offtake term sheet; certification and MRV plan; incentive eligibility confirmed	Certification treated as later paperwork; incentive eligibility unchecked
5. FID package	Can the risks be allocated and financed?	Bankable model; EPC strategy; performance guarantees and liquidated damages defined	Weak guarantees; unclear change-order rules; lender questions unanswered
6. Execution to operations	Can it be built, commissioned and certified?	Commissioning plan; operations and maintenance readiness; audit-quality data trail	Meters missing for certification; no operating regime before commissioning

Where do projects most often fail between gates?

Three transitions carry the most risk in this framework. They reflect the authors' experience and are not drawn from a published dataset.

Gate 2 to Gate 3: the integration burden becomes visible, and scope grows once the BoP is defined properly.
Gate 4 to Gate 5: offtake terms and certification rules clash. A commercial team that agreed to an off-take before reading the certification rules often has to reopen both.
FEED to EPC: interface gaps that FEED left vague return as change orders, at a much higher cost.

Who decides what? Four lenses and one assumption register

Engineers optimise performance. Finance analysts optimise returns. Commercial leads optimise contract flexibility. Policy specialists check compliance, often only at gate reviews. Each does the job well, yet the models drift apart unless everyone works from one list of assumptions. Four lenses should stay in view together:

<table>

<tbody>

<tr>

<td>What it covers</td>

</tr>

<tr>

<td>Engineering</td>

<td>System integration, performance range, interface specifications between the electrolyser package and the BoP, operation across the renewable generation profile</td>

</tr>

<tr>

<td>Commercial</td>

<td>Offtake structure, volume flexibility, price indexation, force majeure and certificate delivery clauses</td>

</tr>

<tr>

<td>Finance</td>

<td>LCOH sensitivities, weighted average cost of capital (WACC) and debt service coverage ratio (DSCR) targets, contingency, incentive eligibility and payment timing</td>

</tr>

<tr>

<td>Policy and certification</td>

<td>Emissions boundary, renewable electricity rules, MRV metering, audit-quality data governance</td>

</tr>

</tbody>

</table>

‍