Project Dev

Why Green Hydrogen Projects Get Cancelled: Lessons from Seven Exits, and a Bankability Checklist

Seven cancelled green hydrogen projects, five patterns behind them, and a bankability checklist to run before FID. Lessons for developers and EPCs.

Hyzen Engineering Team · October 12, 2026 · 9 min read

Why Green Hydrogen Projects Get Cancelled: Lessons from Seven Exits, and a Bankability Checklist

On 5 October 2026, INOX Air Products told investors in its IPO filing that it would not build its $3 billion green ammonia project in India. Two months earlier, Masdar cancelled a 100 MW electrolyzer for green steel in Abu Dhabi, and a German consortium dropped a 500 MW project for the second time.

These are not isolated stories. The IEA's Global Hydrogen Review 2026 found that the pipeline of low-emissions hydrogen projects announced for 2030 shrank by about a quarter in a single year, to 27 million tonnes, because of delays and cancellations.

Each exit comes with a short press statement. Read together, they say much more. This article looks at seven recent cancellations across four continents, the reasons each one gave, and the patterns behind them. It ends with a checklist that developers, EPCs and consultants can run on their own projects before a final investment decision (FID).

Seven exits, in their own words

The table lists the projects newest first, with the reason each one gave or was reported to give.

Exit	Project	Scale	Reason given
Oct 2026	INOX Air Products green ammonia, Maharashtra, India	500,000 t/y green ammonia, about $3bn	"Techno-commercial feasibility and viability", disclosed in its IPO prospectus
Aug 2026	Masdar / Emsteel green steel, Abu Dhabi	100 MW alkaline electrolyzer	No public reason. FEED was complete and five EPC contractors had bid
Aug 2026	HyScale 100, Schleswig-Holstein, Germany	500 MW by 2030, up to 2.1 GW	Insufficient planning certainty, as reported. Second cancellation after Westküste 100 in 2023
Jan 2026	Nujio'qonik, Newfoundland, Canada	1.2 GW hydrogen and ammonia	Unable to secure offtake agreements
Dec 2025	HyDuqm, Oman	About 5 GW of solar and wind, up to 1.2 Mt/y ammonia, $6.7bn	Assessment of global offtake dynamics and investment frameworks
Dec 2025	BP Duqm, Oman	1.5 GW	Not stated
Jun 2025	CQ-H2, Queensland, Australia	720 MW first phase, $12.5bn as reported	Key offtaker left over higher-than-expected production costs; the state government withdrew support

Three things stand out before any analysis.

These were not paper projects. HyScale 100 had €194 million in state funding allocated in 2022. HyDuqm held a 47-year land concession and was targeting FID in 2027. Masdar had a completed FEED and bids from Envision, Larsen & Toubro, PowerChina, Samsung E&A and Sinopec.
They span every model. Export ammonia (Oman, Canada, Australia), domestic industry (Germany, Abu Dhabi), and a mix (India).
Large sponsors walked away too. Engie, Posco, BP, Ørsted and Masdar all have the balance sheets to carry a project. Balance sheet was not the missing piece.

Five patterns behind the exits

1. No signed buyer, no plant

Offtake is the reason that appears most often, in plain words or between the lines. Nujio'qonik stopped after failing to sign buyers. HyDuqm's partners cited global offtake dynamics. CQ-H2 lost its anchor offtaker before it lost its lead sponsor.

The data agrees. The IEA reports that the volume of low-emissions hydrogen covered by new offtake agreements in 2025 stayed low, broadly unchanged from the year before. A May 2026 study by the Green Hydrogen Organisation (GH2), based on 23 interviews with banks, development finance institutions and developers, found offtake and revenue certainty to be the dominant barrier to FID. Without it, lenders cannot assess revenue, so a project finance structure cannot be assembled at all.

2. The cost gap did not close on schedule

Several of these projects were conceived between 2021 and 2024, on the cost expectations of that time. CQ-H2's offtaker left because production costs came in higher than the 2021 feasibility study expected. INOX cited techno-commercial viability after its own evaluation.

The gap is still wide. MEED puts green hydrogen at roughly $3–8/kg against $0.5–2/kg for fossil-based hydrogen. The IEA's 2026 analysis goes further: without policy support, the maximum hydrogen cost most sectors and regions can absorb is below $2/kg.

The lesson for engineers is uncomfortable: a feasibility study built on optimistic CAPEX, efficiency or capacity factor assumptions does not fail at the study stage. It fails years later, when a buyer compares the price to the one promised.

3. Policy moved while the project was being built on paper

HyScale 100 was cancelled over insufficient planning certainty, and none of its allocated state funding was ever spent. In Queensland, the state government withdrew support, refusing to match federal funding for CQ-H2.

In Europe, the uncertainty is now regulatory too. The rules that define renewable hydrogen are being revised, two years ahead of schedule, and the proposal has slipped from June to the autumn. The GH2 study found EU regulatory uncertainty extending the pre-FID phase even for projects outside Europe. We covered what could change in our analysis of the RFNBO delegated act revision.

4. Capital arrives after the riskiest stage, not during it

The GH2 study makes a point that deserves more attention: capital is not scarce, it is late. Development banks and lenders are willing to finance projects once they approach bankability. But the most expensive risk sits earlier, when developers pay for feasibility studies, engineering and commercial structuring with no certainty of reaching FID.

The study calls it a vicious cycle: projects must be bankable to access the instruments intended to make them bankable. That is why cost of capital is often not even testable: most projects never get far enough for a bank to price them. Our CAPEX vs discount rate analysis shows how much that number moves LCOH once it is set.

5. Integrated demand helps, but does not guarantee anything

The GH2 study found that projects tied to existing industrial facilities progress further. Its example is Egypt Green Hydrogen: a 100 MW electrolyzer beside Fertiglobe's existing ammonia plant, backed by Fertiglobe's win in the first H2Global auction to supply 397,000 tonnes of green ammonia between 2027 and 2033.

But Masdar's project had a captive buyer too, a steelmaker owned by the same Abu Dhabi state ecosystem, and it was still cancelled after FEED. A buyer on paper is not the same as a buyer at a price that closes the gap.

A bankability checklist to run before FID

None of these questions is new. What the exits show is that skipping any one of them can end a project years later. Run them early, and again before every major spend.

Revenue

Is there a named buyer with a signed term sheet, not a memorandum of understanding?
Does the contract fix volume, price, duration and what happens if either side cannot perform? Our guide to offtake agreement structure covers the clauses lenders read first.
Is the price the buyer will pay above your LCOH, including a margin for cost overruns?
If the price depends on a subsidy, mandate or auction, what happens to the project if that support changes?

Cost

Is CAPEX based on current quotes, not a published curve from three years ago?
Is LCOH calculated from an hourly simulation of your actual site, with electrolyzer degradation and stack replacement included?
Have you tested LCOH against a range of discount rates, not one assumed WACC?

Regulation and policy

Does the design meet the certification rules of the market you sell into, under today's rules and the most likely revision?
Is any part of the business case dependent on a single government decision or funding round?

Execution

Are grid, water and port infrastructure confirmed, with an owner and a date?
Is the technology choice one that lenders and insurers have already financed at this scale?

A project that reaches FEED with several boxes still empty is carrying risk that tends to surface later, as a delay, a higher cost of capital or a cancellation.

Conclusion

The exits of 2025 and 2026 do not mean green hydrogen has failed. They mean the industry is sorting projects that can close from projects that cannot. The ones that fall out share the same gaps: no firm buyer, numbers from an earlier cost era, and a dependence on policy that moved.

Most of those gaps were visible long before the press release. They sat in a feasibility study that assumed a price, a capacity factor or a cost curve that the market later refused.

The best protection is unglamorous: honest numbers, early. Model the site hour by hour, test every assumption against a range, and know your real cost per kilogram before you ask anyone to buy it.

Hyzen links renewable input, electrolyzer sizing, balance of plant and LCOH in one model, so you can stress-test a project in hours and find the weak assumption before your buyer or your lender does.