A report by Rystad Energy projects US data centre developers to spend up to US$30 ($39) billion by 2030 on fuel cells, driven by demand for AI computing power. This represents a tenfold increase from US$2.8 billion in 2025.
Rystad believes that major operators are gaining confidence in fuels cells as a viable long-term power source, pointing to a contracted order book of approximately nine gigawatts (GW), including framework agreements with Oracle, AEP, Equinix and Brookfield.
Fuel cells offer operational flexibility — they can be deployed quickly and run on natural gas presently before transitioning to run on biogas, renewable natural gas or hydrogen as supply matures, while producing lower on-site emissions than combustion alternatives — suggests Rystad for their demand.
Meanwhile, the tripling of US grid interconnection timelines since 2015 to three to six years also seems to be a factor pushing data centres to pursue dedicated on-site power generation.
North America is expected to account for 91% of installed global on-site power generation capacity, thanks to a combination of grid delays, federal tax incentives and an established domestic supply chain, according to Rystad.
Rystad vice president of clean tech supply chain research Lein Mann says that power availability has become one of the defining constraints on data centre growth, with operators increasingly looking beyond the grid for solutions. “Fuel cells have moved from a niche application to a measurable part of the firm power mix,” he adds. “The question now is whether the supply chain can scale at the same pace as demand.”
See also: STT GDC expands Jakarta data centre campus, training efforts for Indonesia’s AI buildout
In response to rising demand, fuel cell manufacturers are expanding capacity. Rystad understands that aggregate operational and planned manufacturing output is on track to reach 4GW per year by 2030, up from 1.8GW today.
Trends include the dominance of solid oxide fuel cells (SOFC) — which has supplied around 53% of cumulative stationary deliveries to date — for always-on data centre power.
Rystad points out concentration risk in the SOFC supply chain, with Bloom Energy holding virtually every primary-load SOFC contract in the visible order book. The concentration risk also extends to materials, with Bloom Energy’s SOFC technology depending on scandium, a rare earth element used in its electrolyte chemistry.
See also: Keppel secures land for data centre development in South Korea
At full utilisation of its planned 2GW manufacturing expansion, Bloom’s theoretical scandium requirement would approach the size of the entire global market, currently estimated to be around 60 tonnes annually. Compounding the supply chain issue is strong control of scandium by China. Thus, competitors using alternative electrolyte chemistries do not share this exposure, and a sustained supply constraint could influence how market share develops as the sector scales.
Rystad Energy projects SOFC system costs will fall 20 to 25% by 2030, though the pace will depend on manufacturers’ ability to reduce costs across the full delivered system, not the fuel cell stack alone.
