Hydrogen had a grueling 2025, particularly in the United States. If we take a step back and review other technologies that are powering the energy transition, however, the data suggest that the industry’s struggles have precedent.
In September, the Hydrogen Council, announced that more than 500 hydrogen projects worldwide had passed final investment decision, amounting to $110 billion in investment. This represents a survival rate of roughly 30%, considering that the trade group tracked more than 1700 hydrogen project announcements in the past five years.
Evidence of a collapsing bubble, or in line with other energy deployment? Context is key — Mt. Fuji towers over Tokyo but would be the runt of the Himalayas — and is provided by Lawrence Berkeley National Laboratory. The 2024 edition of the lab’s Queued Up reports, makes it clear that completion rates vary wildly by sub-sector; what might be considered normal for wind would be worrying for solar-plus-storage.
The stacked horizontal bar charts on the left represent the status at the end of 2023, of U.S. energy projects that made grid transmission interconnection requests between 2000 and 2018. The Hydrogen Council survival rates can be mapped onto the chart, with the caveat that its data is global while the LBNL dataset is for the United States only.
The 70% withdrawal rate for worldwide hydrogen projects is broadly in line with project withdrawal rates LBNL observed not just for utility-scale batteries, interconnection-seeking solar (without storage), and wind — each near 75% — but for natural gas power plants as well, at 67%. This means that even before Trump administration tribulations, most types of transmission-facing electricity projects in the United States roughly followed a version of the Pareto Principle: Rather than 20% of efforts causing 80% of results, in the energy sector it seems that roughly 20% of projects (or slightly more, depending on the sector) represent all of a sector’s success — while 80% fail to materialize.
That said, the Trump 2.0 era has certainly complicated things for hydrogen projects in the U.S., prompting some companies to move their operations abroad. In October, Latitude Media reported that Danish manufacturer Topsoe paused work on its $400 million electrolyzer facility in Virginia in light of poor demand for the green hydrogen; earlier this year, the Norwegian Nel Hydrogen decided to cancel its own planned electrolyzer factory in Michigan. And in November, Exxon announced plans to freeze development of a planned blue hydrogen plant in Texas.
These announcements, linked as they are to policy shifts in the U.S., are worrying. It’s possible, given the global nature of the Hydrogen Council data, that some companies are picking up and leaving the U.S. — but still getting their projects built elsewhere. But the data suggests that hydrogen projects haven’t necessarily been more prone to cancellation than those for other clean energy technologies.
In any industry building new infrastructure, there are startups who shut down even before reaching the announcement stage. A blunt Bluesky post by Tim Latimer, CEO and co-founder of advanced geothermal pioneer Fervo Energy, hints that an 80-20 dynamic might emerge in advanced geothermal also.
That said, hydrogen can not be compared directly to grid-connected projects. A key metric for success, as Topsoe demonstrated, is finding buyers for the electrolyzers or hydrogen that a company produces. According to BloombergNEF data, as of November 2023, only a small minority of near-term hydrogen projects — 13% of announced capacity — had announced offtakers. They estimated a further 7% of announced hydrogen production capacity was likely to enter into binding contracts, for a total of 20% — right in the heart of the Pareto zone.
Solar-plus-storage is the outlier in the LBNL dataset because of its large “active projects” bar: 37% by project count that were neither operational or withdrawn. This is likely because that sub-sector saw a rapid scale-up in the past decade. LBNL only records seven solar-plus-storage projects as requesting interconnection in 2014; in 2018, that had increased to 200; and by 2022, they counted 791 requests.
This juxtaposition of sectors begs the question of whether the Pareto principle could apply for the biggest infrastructure build-out currently underway: data centers. It will certainly take several years to know for sure, but there are a few pieces of evidence that it might.
Earlier this year, Camus Energy CEO Astrid Atkinson said that she was seeing five to 10 times as many data center transmission interconnection requests in the United States as data centers being built: phantom data centers that have complicated building timelines. At best this would amount to a Pareto-abiding 20% project completion rate.
This is not to say that this Pareto rule for successful energy- or energy-adjacent projects would be predictive. The industry’s goal, of course, should be to increase its share of successful projects, both for commercialized technologies and more nascent ones. Advanced geothermal, small modular reactors and other technologies may enjoy more success — and if they do, the reasons will deserve study.
Matthew Klippenstein is an energy-focused engineer whose career has spanned hydrogen, wind, EVs and EV infrastructure. The opinions represented in this contributed article are solely those of the author, and do not reflect the views of Latitude Media or any of its staff.
