Clean Hydrogen Offtake Trap: Compliance Credit Or New Demand?

The Hydrogen Council published its Global Hydrogen Compass 2026 on 10 September. It reports USD 130 billion of committed investment across more than 570 projects. Committed capacity is 6.9 million tonnes a year, against 1.7 Mtpa operating. The release announcing it says operational capacity has nearly doubled. But where does the new clean hydrogen offtake actually go?

The clean hydrogen offtake table on page 41 carries the harder number. Binding offtake reached 4.2 Mtpa, up about 16% in twelve months, and ammonia and refining take 71% of it. Those two drove nearly all of the year’s growth. So most clean hydrogen offtake is a credit created by policy rather than demand created by a fuel market. Nothing in that 71% is a new use. It is substitution inside plants that have run on hydrogen for decades.

Where clean hydrogen offtake actually goes

Exhibit 19 splits the 4.2 Mtpa by end use. Ammonia for chemicals and fertiliser takes 43%. Refining takes 28%, power 9%, methanol and chemicals 7%, steel 5%, and e-fuels for aviation 5%. But geography divides clean hydrogen offtake as sharply as end use does. North America holds 1.7 Mtpa and Europe 1.0 Mtpa, so three quarters of the total sits outside European rules.

The instrument differs by region. Atome’s 145 MW project in Paraguay will make 260,000 tonnes a year of renewable ammonia-based fertiliser under a binding ten-year Yara agreement. That contract runs on development-finance debt and equity, and on a buyer who already purchases the molecule.

Europe runs on a compliance quota instead, Asia on import support, and other ammonia projects in the cohort on production tax credits. Each is a policy instrument or a financing structure. None of them is a fuel market.

New demand does exist, and the report names it. Project Roadrunner in Texas puts a 100 MW electrolyser inside Infinium’s e-fuels plant, making eSAF, eDiesel and eNaphtha at 23 Ktpa. One project of that size against 4.2 Mtpa is a signal rather than a market.

In Europe, a penalty sets the price

Article 25 of RED III sets a combined target for advanced biofuels and renewable fuels of non-biological origin. Together they must reach 5.5% of transport energy by 2030. At least one percentage point of that must be RFNBO, and the obligation falls on fuel suppliers.

A supplier has three ways to comply. It can put hydrogen directly into vehicles. It can sell e-fuels into aviation and shipping. Or it can substitute renewable hydrogen for unabated hydrogen inside a refinery. The third route needs no new distribution network and no change at the vehicle.

So the price of clean hydrogen offtake in Europe is the sum of three things a refinery avoids. It stops buying grey hydrogen. It stops paying ETS on that hydrogen. And it escapes the fine for missing its quota. Hydrogen Europe’s example puts those first two near 2.70 euros a kilogram against a delivered cost of 9. So the fine has to clear 6.30 before the route works at all.

Those fines now sit in national law. A Hydrogen Europe survey dated 26 May 2025 records three. Finland legislated 6.6 euros a kilogram, Czechia 9.6 and Romania set a ceiling at 6.0, below the break-even. Germany had not transposed at all. That matters, because the Compass puts the German penalty at 14 euros and sources it to its own advocacy paper.

These penalties do not index. Each is a flat figure in national law, so none of them moves when the cost of hydrogen does. ReFuelEU works the other way. Its e-SAF penalty is at least twice the gap between synthetic and conventional jet fuel, so it rises with the fuel.

The cheapest way to comply is the one being capped

Refining is the only industrial process that sits inside a transport mandate. Steel and cement face the ETS alone. In refining the ETS is worth 0.70 euros a kilogram against penalties running from 6 to 9.6, and carbon pricing at that level moves nothing on its own.

RED III tries to pull hydrogen into transport. But the refinery route takes the tonnes and changes nothing in transport. No refuelling stations, no fuel-cell vehicles, no e-fuel plants. Green hydrogen costs the same whichever route it takes. Only this one needs nothing built around it.

The Compass names that drawback itself. Being the cheapest way to comply is what attracts a cap. European refineries use more hydrogen than the whole quota needs, so this route can absorb the mandate by itself and leave the other two unbuilt. A regulator who wants any of those has to shrink it, not close it.

In fact two limits already apply. Finland caps hydrogen used as a refining intermediate at 1% in 2030. The Netherlands discounts it with a correction factor. A third limit sits in the accounting. Only dedicated, non-byproduct hydrogen qualifies as RFNBO, and refineries already make byproduct hydrogen that does not count.

The national numbers look like a range and are not one. Finland has legislated 4%, but only for road transport. Denmark has 0.9% and leaves aviation out of the denominator while counting it in the numerator. Czechia has 1%, which becomes 0.5% once its multipliers are applied. Each state writes its own denominator, so the percentages cannot be laid side by side.

Whether the limits are wise is arguable. Capping the only route that clears its break-even may produce fines rather than e-fuels.

There is no fallback bid

Clean hydrogen offtake financed against the refinery route has revenue created by a fine. That fine exists only where a state has transposed, and only at the level it chose. It holds only up to whatever cap that state sets, and only until the open RED III revision moves the terms.

The risk is not a thinner margin. It is losing the buyer while the buyer still has to buy. ReFuelEU names synthetic aviation fuel, so a weaker mandate costs an e-SAF plant money but never its place. RED III names no route. It sets a quota and allows three ways to meet it, and two member states have already narrowed the refinery one. Narrow it again and the obligation stays put. The buyer simply fills it elsewhere. More than half of Europe’s post-FID projects sit on that route, and nothing downstream can absorb them. Aviation e-fuels hold 5% of global clean hydrogen offtake, maritime under a percentage point.

The lag decides who finds out

Of the 6.9 Mtpa committed, 5.2 Mtpa still sits at FID or under construction. The report puts a three-to-four-year lag between commitment and operation. So clean hydrogen offtake signed this year is delivered into the rules of 2029, not the rules that justified it.

Momentum is already slowing. Committed investment growth has fallen to about 20% a year, against roughly 45% on average since 2021. The series runs 10, 20, 30, 45, 75, 110 and 130 billion dollars from 2020 to 2026. Roughly 30 billion advanced to FID or beyond over the year, about five times the 6 billion delayed or discontinued. For example, Plug Power suspended Limestone, removing 850 million dollars.

What to underwrite in a clean hydrogen offtake

A project that earns its money replacing grey hydrogen in a refinery is not selling fuel. It is selling compliance. That is the trap: it prices like a fuel offtake until the route closes. 

Four questions test where it breaks. Establish which member state the offtaker sits in, and whether transposition is law or draft. A directive binds nobody until a parliament acts. Check what penalty level the business case assumes, and whether it comes from a statute or a model. Also test how much quota the non-byproduct rule leaves once refinery byproduct hydrogen comes out. Then find out whether a cap or a correction factor already bites in that market.

None of this makes refinery substitution a poor use of the molecule. It displaces unabated hydrogen tonne for tonne, and it is among the lowest-cost real abatement in heavy industry. The policy is doing what it was written to do. The open question, rather, is who carries the risk and on what terms. A fuel price can be hedged. Someone will quote you a price for 2029 diesel. Nobody will quote you a price on a Finnish penalty. 

So protection has to sit in the contracts instead: change-in-law clauses, volume floors that survive a cap, and termination rights if the route closes. Clean hydrogen offtake at this scale is a compliance credit, and the 5.2 Mtpa now being built will find that out by 2030.

MVAventures reviews green fuel offtake and project structures for developers, offtakers and lenders. If a contract is on your desk and you want to know which instrument is actually paying for it, contact us.