The default picture of grid storage these days is a Tesla Megapack: a steel box about the size of a shipping container, parked in a field, soaking up cheap solar in the afternoon and pushing it back onto the grid a few hours after sunset. That model is winning. American utilities bolted down a record amount of battery capacity last year and they are not slowing down.<\/p>\n
But out in the high desert of central Utah, in a town called Delta, a joint venture between Chevron and Mitsubishi Power has built something that works nothing like a battery, holds its energy for months instead of hours, and, according to the company that supplied its electrolyzers, banks two to three times more energy than every grid-connected battery in the United States combined.<\/p>\n
The trick is that the energy isn’t stored as electricity at all. It’s stored as hydrogen gas, pumped into two enormous caverns hollowed out of an underground salt dome. And as of early 2026, the machine that makes the hydrogen is finally running flat out.<\/p>\n
The hub is called Advanced Clean Energy Storage, or ACES Delta, and its backers bill it as the world’s largest green hydrogen production and storage facility. The storage half of that claim is two solution-mined salt caverns, each engineered to hold roughly 5,500 metric tons of working hydrogen, or about 11,000 tons across the pair. Translate that into the unit utilities actually care about and each cavern is rated at around 150 gigawatt-hours<\/a> of dispatchable energy, which puts the combined site north of 300 GWh. The whole thing is backed by a $504 million federal loan guarantee from the Department of Energy.<\/p>\n Numbers that big stop meaning anything without a yardstick, so here’s one. The entire US battery industry added a record 57.6 gigawatt-hours of new storage capacity across all of 2025, according to the Solar Energy Industries Association<\/a>. A single one of these Utah caverns is rated at 150. HydrogenPro, the Norwegian firm that built all 40 of the project’s electrolyzers, goes further and says the combined storage holds two to three times the energy<\/a> of every grid-connected battery in the country put together. That’s the company’s own framing, and the national battery fleet is expanding fast enough that the multiple shrinks a little every year. Even on the conservative end of it, though, you are looking at one desert site that out-stores a serious slice of the national grid.<\/p>\n The reason for storing energy as hydrogen instead of in batteries comes down to time. A battery is brilliant at moving power across a few hours, charging at noon and discharging at dinner. It is hopeless at moving power across a few months, because it slowly bleeds its charge and you can’t build one big enough to matter at that timescale without spending a fortune. The Western grid has the opposite problem. It makes far more solar and wind power in spring than anyone needs, then sweats through summer demand peaks when the panels can’t keep up.<\/p>\n That seasonal gap is what a salt cavern fills, in roughly the same way a pumped-hydro reservoir banks power with water and gravity<\/a>. When there’s surplus renewable power on the grid, the project runs it through 220 megawatts of alkaline electrolyzers, the 40 HydrogenPro units, which split water into hydrogen and oxygen and crank out up to 100 metric tons of hydrogen a day.<\/p>\n The gas gets compressed and injected into the caverns, where it sits, more or less losslessly, for as long as it needs to. Months later, when demand spikes, the hydrogen comes back out and gets burned to make electricity. John Ward, a spokesperson for the Intermountain Power Agency, told Utah News Dispatch<\/a> the system “works like a battery, except batteries are good for hours.”<\/p>\n The caverns themselves are a quirk of geology. Central Utah happens to sit on one of the only Gulf-Coast-style salt domes in the western US, and you carve a cavern out of it by pumping in water to dissolve the salt and pulling the brine back out, a process called solution mining. Each finished cavern is roughly 4.5 million barrels of empty space, sealed inside rock that holds pressurized gas about as well as anything in nature. The site has room for dozens more of them if the demand ever shows up.<\/p>\n None of this would mean much if the hydrogen had nowhere to burn. It does, right next door. For nearly four decades the Intermountain Power Project ran two big coal units that sent most of their output to the Los Angeles area down a long high-voltage DC line. California spent years legislating its way off coal, so the plant’s owners replaced those units with IPP Renewed, a pair of Mitsubishi M501JAC gas turbines rated at 840 megawatts combined<\/a>, both of which came online during 2025.<\/p>\n Here’s the part that makes the loop close. Those turbines are built to burn hydrogen. They started up on a blend of about 30% hydrogen and 70% natural gas, which Mitsubishi says cuts carbon emissions more than 75% compared with the coal they replaced, and the plan is to climb to 100% hydrogen no later than 2045.<\/p>\n The same combustion-turbine path is showing up in places nobody expected: a Baker Hughes unit was just certified to drive a ship across the ocean on pure hydrogen<\/a>. Burning hydrogen in a gas turbine, in other words, is quietly turning into a real piece of hardware rather than a slide in a pitch deck.<\/p>\n The coal units, for the record, did stop. They went quiet the day before Thanksgiving in 2025. And the new setup keeps Los Angeles supplied for a long time. The power-purchase agreement that was scheduled to expire in 2027 has been extended all the way out to 2077.<\/p>\n A few things keep this from being a clean fairy tale. The turbines are not running on pure hydrogen yet. They’re on a blend, and the road to 100% runs to 2045, which is a long time to hold a target in the energy business. The headline comparison, the two-to-three-times-all-US-batteries line, belongs to HydrogenPro, not to an independent referee, and the national battery fleet is growing fast enough that it gets less flattering every year.<\/p>\n The coal plant isn’t gone so much as mothballed, either. Worried about losing dispatchable power and the local jobs that came with it, Utah lawmakers passed laws keeping the old coal units operable and wired to the switchyard while the state hunts for a buyer. The units aren’t burning anything right now, but the door has been left deliberately cracked.<\/p>\n What’s genuinely new here isn’t a gadget, it’s the scale and the timeframe. Storing surplus renewable energy as hydrogen and holding it for a whole season has mostly lived in research papers and small pilots, right down to a Chinese lab cell that packs hydrogen and electricity into the same solid-state device<\/a>.<\/p>\n The Utah project is that idea built at industrial size, with 11,000 tons of gas underground and a turbine wired into a real city’s grid. Whether seasonal hydrogen storage ever pencils out across the rest of the country depends on electrolyzer costs, hydrogen prices, and how many more salt domes turn out to be sitting in exactly the right place. Central Utah got lucky on that last one. The rest of the map is the hard part.<\/p>\n","protected":false},"excerpt":{"rendered":" The default picture of grid storage these days is a Tesla Megapack: a steel box about the size of a … <\/p>\nSpring surplus goes in, summer power comes out<\/h2>\n
The hydrogen has somewhere to go, and it used to be a coal plant<\/h2>\n
It’s a blend today, and the coal isn’t fully buried<\/h2>\n