Every gas turbine sold this decade seems to arrive with the words “hydrogen ready” printed somewhere on the brochure. Usually that means the machine can swallow a splash of hydrogen mixed into natural gas, five percent, maybe thirty, before the flame starts doing things nobody wants inside an expensive engine.<\/p>\n
GE Vernova built one that runs on nothing but hydrogen. Then it sold four of them to a power station that was never built.<\/p>\n
The machines are LM6000VELOX packages, 50 megawatts apiece. ATCO Australia ordered them in November 2024, on the sidelines of the COP29 climate conference in Baku, for the Whyalla hydrogen power plant on South Australia’s Upper Spencer Gulf. Two hundred megawatts of firming capacity, fed by a 250-megawatt electrolyzer and a hydrogen storage facility rated at 100 tons, with commissioning slated for early 2026.<\/p>\n
GE Vernova said it would be the first of its power plant projects at commercial scale to run on an aeroderivative turbine capable of 100% hydrogen<\/a>. Early 2026 came and went. There is no plant, there is no electrolyzer, and the state that ordered the turbines is now trying to sell them.<\/p>\n Flame speed is why “hydrogen ready” so often means “hydrogen curious.” A hydrogen flame front travels far faster than a methane one, and in a premixed combustor a fast flame will happily climb back upstream into the fuel nozzle and cook the hardware that was feeding it.<\/p>\n Combustion engineers call that flashback. Everyone else can call it the reason a fuel nozzle turns into a consumable.<\/p>\n When the company validated a separate hydrogen combustor for its industrial B- and E-class frames, carbon solutions leader Jeremee Wetherby put a figure on it in GE Vernova’s January 2025 announcement<\/a>: hydrogen’s “flame speed is roughly eight times higher” than natural gas, and flashback is the consequence.<\/p>\n The LM6000 solves it the old-fashioned way, with water. Its single annular combustor was rebuilt around a redesigned fuel nozzle, a reworked water-injection schedule and new control logic, with the fuel piping and valves upgraded to live with the molecule. Nitrogen purge lines and hydrogen fire detection went on around it, because hydrogen leaks through things and burns with a flame you can barely see.<\/p>\n The work started at New York Power Authority’s 45-megawatt Brentwood plant on Long Island, where GE ran an LM6000 on blends from 5% to 44% hydrogen in 2022. Component testing at the company’s research center in Niskayuna, New York, then pushed the combustor to 100% hydrogen with water injection. Midhat Mirabi, who runs new aeroderivative units for GE Vernova’s Gas Power business, told POWER magazine<\/a> that “a full-scale engine test is planned in Greenville” to finish the job.<\/p>\n The hydrogen milestones GE Vernova has announced since then belong to a different fleet. That dry combustor for the B- and E-class frames was tested at full pressure, flow and temperature in Greenville, holding NOx under 25 parts per million on pure hydrogen with no water at all. Skipping the water is worth 4% to 7% on combined-cycle heat rate, the company says. It is also not the machine that was going to Whyalla.<\/p>\n Strip an LM6000 down and you are looking at a CF6-80C2, the high-bypass turbofan that hauled 747s and 767s across oceans. GE turned it into a power-plant engine in 1988, and more than 1,200 have since gone into 60 countries. Rated output runs from 44.7 to 56 megawatts depending on configuration.<\/p>\n That aircraft ancestry is the whole point. An aeroderivative is light, and light means it starts fast, with GE Vernova rating the LM6000 at about five minutes from cold to full power. A grid pulling three quarters of its electricity from wind and solar does not need a machine that takes an hour to wake up. It needs one that can be at full output before the wind finishes dropping.<\/p>\n Aero engineers keep making the same point elsewhere. Rolls-Royce ran a Pearl 15 business-jet engine to full take-off power on nothing but hydrogen<\/a> at NASA’s Stennis Space Center, and an industrial turbine from Baker Hughes just collected a marine type approval to burn 100% hydrogen on a ship<\/a>. Getting a turbine to eat the stuff is no longer the bottleneck.<\/p>\n The VELOX half of the name is packaging, not thermodynamics. GE Vernova assembles the LM6000 into factory-built modules to cut site installation to roughly 90 days, about 40% off the usual schedule and some 4,000 labor hours. You bolt it down, plumb it, run it.<\/p>\n The Hydrogen Jobs Plan was Labor’s first pledge at the 2022 state election, with a A$593 million price tag. (Every dollar figure here is Australian.) Whyalla was to get the electrolyzer, the storage, the 200-megawatt power station, and a supply of green hydrogen for a steelworks trying to get off coal.<\/p>\n In February 2025 the steelworks collapsed instead. The state pushed it into administration, assembled a A$2.4 billion rescue with Canberra, and took the hydrogen money to pay for it. The power plant was “deferred,” which turned out to mean it stopped existing.<\/p>\n By April 2025 the 200-megawatt plant had been pulled from the Australian Energy Market Operator’s project pipeline. A month later the state dissolved the Office of Hydrogen Power South Australia<\/a> and folded its work into the Department for Energy and Mining. The auditor-general later reported the office had spent A$285.2 million by June 30, 2025.<\/p>\n The turbines were already paid for. More than A$250 million of that spending went into the four machines, and Tom Koutsantonis, who serves as both treasurer and energy and mining minister, told parliament they would be on-sold at the original price or better, on the condition that they “be operated in South Australia to provide additional generation capacity.”<\/p>\n Nobody took the announcement down. It is still on the premier’s website, headlined as the world’s first 100% hydrogen-capable gas turbines<\/a>. It lives in the news archive now.<\/p>\n Read that condition again. The turbines have to end up in South Australia. Nothing in it says they have to burn hydrogen.<\/p>\n They almost certainly will not. GE Vernova has said the hydrogen build can be expected to start and stop on natural gas anyway, and with no electrolyzer, no storage and nobody in the state making green hydrogen at scale, a buyer is acquiring four fast-start gas peakers that happen to carry a very unusual combustor.<\/p>\n The government has stopped pretending otherwise. Asked in February whether it would revive the green hydrogen plan, Koutsantonis told InDaily<\/a> that it had no plans to, and that for the steelworks, “gas is going to be king.” The turbine sale, the outlet reported at the time, was still running, with the state hoping to recoup its money.<\/p>\n The steelworks itself is down to two shortlisted bidders, with a sale promised by August 2026. One of them wants to make hydrogen at Whyalla after all, by cracking natural gas into hydrogen and solid graphite rather than splitting water. Which is a different molecule with the same name, arriving four years late, for a different customer.<\/p>\n What those four machines are actually worth got decided in May, by an auction designed to buy exactly what a 200-megawatt peaker sells. The Firm Energy Reliability Mechanism went looking for capacity that could hold up the grid when the wind stops, and it was open to flexible gas generators and to batteries with at least eight hours of storage. Bids came in from both.<\/p>\n On May 29, 2026, the results landed<\/a>. Six projects won contracts. All six were batteries, totaling 517 megawatts of output and 4,136 megawatt-hours of storage, led by two stages of Neoen’s Goyder North project at 75 megawatts and 600 megawatt-hours each. The round came in under its own 700-megawatt target, and not one gas project made the cut.<\/p>\n Gas is not finished in the state. Further tender rounds are expected, and South Australia has contracted AGL to keep two units running at Torrens Island B while the EnergyConnect transmission link to New South Wales runs late. Existing gas is still load-bearing. New gas keeps losing auctions.<\/p>\nHydrogen burns about eight times faster than natural gas, and that is the entire problem<\/h2>\n
The LM6000 spent its first life on the wing of a widebody<\/h2>\n
South Australia bought the turbines, then buried the plant<\/h2>\n
The condition on the sale is geography, not fuel<\/h2>\n
Batteries just swept South Australia’s own firming auction<\/h2>\n