With most planned green hydrogen projects never getting built, scientists and startups are turning to a different idea: hydrogen that forms on its own in the Earth’s crust, which could someday cut emissions from the industries that are hardest to electrify.

So-called white hydrogen is generated naturally in rocks and other deposits underground. Its backers say it could sidestep the main problems that have stalled manufactured hydrogen without adding much to the sector’s carbon footprint, though no commercially proven find exists yet.

Green hydrogen, made by splitting water with renewable electricity, was promoted for years as the clean answer for steelmaking and heavy transportation. The numbers have not cooperated. As of 2023, less than one tenth of planned green hydrogen projects were actually built, according to Oilprice.com, and “only 7% of global capacity announcements finished on schedule,” according to a paper published last year in the journal Nature Energy under the title “The green hydrogen ambition and implementation gap.”

The fuel has stayed expensive and outcompeted, and in many cases the renewable electricity it would need serves decarbonization better fed straight into the grid. Hydrogen itself is alluring for industry: it burns hot like coking coal or heavy fuel oil but leaves only water vapor. The trouble is that most of it today is made from fossil fuels, which cancels out much of the point.

The underground reserves, by contrast, are large. Researchers at the University of Oxford, Durham University and the University of Toronto estimate that hydrogen generated in the continental crust over geological time holds energy equal to roughly 170,000 years of current global oil consumption. How much of that sits in deposits anyone can actually recover remains unknown. “The potential is massive,” Pierre Levin, chief executive of the Canadian geologic hydrogen startup Vema, told The New York Times earlier this year. “You can find rocks like this all over the world, enough to produce billions of tons of hydrogen.”

Getting it out is the hard part. A viable operation needs a whole geological system at once: a production source, migration pathways, a reservoir and, ideally, traps that hold large quantities of hydrogen in place for a long time, according to a report from the science outlet Futura. Miss any piece and the accumulation leaks away.

Chris Ballentine of Oxford, the lead author of the recent study, compared the hunt to baking. “Combining the ingredients to find accumulated hydrogen in any of these settings can be likened to cooking a soufflé — get any one of the ingredients, amounts, timing, or temperature wrong and you will be disappointed,” he said.

Startups from Europe to the Philippines are racing to be first to crack that recipe. Mr. Ballentine’s wager, like theirs, is that one repeatable success is all it would take. “One successful exploration recipe that is repeatable will unlock a commercially competitive, low-carbon hydrogen source that would significantly contribute to the energy transition,” he said. “We have the right experience to combine these ingredients and find that recipe.”