The Definition Is The Purse

Two governments signed the same declaration on 19 June 2026. Australia committed EUR 200 million through the Department of Climate Change, Energy, the Environment and Water. Germany committed EUR 200 million through the Federal Ministry for Economic Affairs and Climate Action, out of the Climate and Transformation Fund. Four hundred million euro for one fuel, delivered as a joint tender and run from Hamburg by HINT.CO. First deliveries to Germany are pencilled for 2031. Annual auctions are targeted from 2027 or 2028. The announcement was reported as an opening. Read the annexe and it is a definition with a budget attached. The budget is the part everyone quotes. The definition is the part that decides.

The mechanism is a double-sided auction, and it only makes sense with both sides on one page. On the supply side a reverse auction awards long-term Hydrogen Purchase Agreements to the most competitive renewable hydrogen producers in Australia. On the demand side, sales auctions in Germany award short-term Hydrogen Sales Agreements to offtakers at the highest price. The gap between a producer price in Australia and an offtaker price in Germany is the spread, and the state pays it. The named products are renewable hydrogen, ammonia and methanol. That is a subsidy for a corridor, not a grant for a plant. The probability that a corridor instrument fits a single site is close to zero, because the instrument is priced over a route between two continents rather than over one gate.

Eligibility runs on RFNBO, renewable fuels of non-biological origin. In plain terms, hydrogen made from renewable electricity by electrolysis. The EU window behind the mechanism carried a minimum electrolyser input capacity of 5 MW. Three products are named, and two routes were struck out during consultation: coal gasification, and steam methane reforming with carbon capture. Say the second exclusion out loud and the shape of the tender appears. Biomass gasification makes syngas from biological material, so it cannot be an RFNBO. Not because the carbon arithmetic fails. Because the definition says non-biological, and a definition is a gate rather than a measurement. The purse opens on a sentence, and that sentence was drafted before the auction was announced.

The timeline was published, which is the only reason it can be read. 13 September 2024, the first joint declaration of intent, signed in Brisbane. September 2025, consultation opened and closed on 30 September. 5 January 2026, the German ministry published its market consultation. 10 April 2026, the evaluation of that consultation went out. 19 June 2026, the second declaration, signed by Australian Assistant Minister Josh Wilson and German Parliamentary State Secretary Stefan Rouenhoff. The first supply-side auction sits in launch phase, with no award date on the public pages as of 10 August 2026. Two years from the first signature to the first auction, and the design decisions were visible the whole way. Most applicants read the announcement. Almost none read the eight months of consultation behind it.

An earlier instrument ran the same corridor, and it is closed. The German-Australian Hydrogen Innovation and Technology Incubator, HyGATE, held one round. AUD 50 million from the Australian Renewable Energy Agency plus EUR 40 million from Projekttraeger Juelich, roughly AUD 110 million together. Expressions of interest ran 8 March to 9 June 2022, full applications were due 12 October 2022, and the awards landed in January 2023. Four projects. Hysata with Fraunhofer IPT, AUD 8.98 million for a capillary-fed electrolyser in the Illawarra. Edify with Siemens Energy, AUD 20.74 million for a 17.5 MW electrolyser and 21 MW of solar. Vast Solar with Fichtner, AUD 19.48 million for a 10 MW electrolyser and solar methanol at Port Augusta. ATCO with Fraunhofer IST, AUD 0.8 million for a 1 GW pathway and a feasibility study on 800,000 tonnes a year of ammonia. Four awards, one round, no successor announced.

AUD 25,000 does not sound like leverage, so read the mechanism rather than the number. The Universities Australia and DAAD joint research cooperation scheme, agreed in 2014, funds return airfares and living costs only, capped at AUD 12,500 a year across two years, and it requires an early career researcher on the proposal. All 39 universities in Universities Australia can apply, and the German partner files the same project title with DAAD. The 2026 round opened on 1 April and closed on 12 June, outcomes land in November 2026 and funding starts in 2027. The next cycle is expected to reopen around April 2027. The published contact is Annika Hipke at DAAD in Bonn. Small money, and it is the only instrument on this corridor that puts a German research institute and an Australian university on one contract.

The rest of the bilateral map is a list of no, and the list is worth more than the yes. ZIM International has no agreement with Australia, and its published partner list runs Argentina, Brazil, Canada, Catalonia, Finland, France, Japan, Singapore, South Korea, Sweden, Taiwan and Vietnam. A foreign partner can join a ZIM cooperation project only self-funded, and ZIM has been under a temporary application stop since 7 July 2026, with exceptions only for bilateral agreement partners. Eurostars is harder again: Australia is not one of the 37 participating countries, and Call 11 closed on 10 September 2026. The German Research Foundation has no Australia-specific programme. Horizon Europe association does not include Australia. Four instruments, four closed doors, every one of them documented in advance.

The four hundred million euro is real, and it will fund electrolysis at megawatt scale on a corridor into Germany. It will not bend its own definition. That is the transferable lesson, and it repeats across every programme in this series. The money is published last and understood first. The definition is published early and read last, if it is read at all. I have built companies across twelve countries, restructured a EUR 75 million industrial group and deployed energy systems on three continents, and the pattern has not moved once: the operators who read the rules before the announcement collect, and the ones who read the announcement collect nothing. The probability that a programme is written for the thing you already own is low. The probability that it says so, in writing, months in advance, is close to one.