Bloom Energy Positioned as AI Data Center Fuel Cell Boom Nears

What happened

Rystad Energy projects fuel cell market revenues will surge from roughly $2.8 billion in 2025 to approximately $30 billion by 2030, driven almost entirely by AI data center demand for on-site power generation. The contracted order book already totals about 9 GW, with cumulative demand forecast at 10.4 GW from 2026 to 2030. Major operators including Oracle, AEP, Equinix, and Brookfield have signed agreements contributing to this pipeline. US grid interconnection timelines have tripled since 2015, now averaging three to six years for large loads, prompting data center developers to pursue dedicated on-site solutions. Rystad estimates around 40% of projected 2030 US data center capacity may rely on these alternatives rather than grid connections. North America is expected to capture 91% of global on-site fuel cell capacity, supported by federal tax incentives and an established domestic supply chain. Bloom Energy holds virtually every primary-load SOFC contract in the visible order book, making it the dominant player among solid oxide fuel cell providers. SOFCs account for approximately 53% of cumulative stationary fuel cell deliveries to date and are the preferred technology for continuous, uninterrupted baseload power in data centers. Fuel cell manufacturers are expanding aggregate operational and planned output to reach 4 GW per year by 2030, up from 1.8 GW currently.

Why it matters

For Bloom Energy, this market expansion represents a significant growth opportunity, as the company controls the vast majority of visible SOFC contracts for data center primary loads. However, Bloom's SOFC technology depends heavily on scandium, a rare earth element used in its electrolyte chemistry. At full utilization of its planned 2 GW manufacturing expansion, Bloom's theoretical scandium requirement would approach the size of the entire global market, currently estimated at around 60 tonnes per year. China controls the majority of global scandium supply, creating a critical dependency that could constrain scaling if demand accelerates faster than supply chain capacity. Competitors developing alternative electrolyte chemistries that reduce or eliminate scandium usage do not face this exposure, which could influence market share dynamics as the sector scales. Rystad projects SOFC system costs will fall 20 to 25% by 2030, though achieving this depends on manufacturers reducing costs across the full delivered system, not just the fuel cell stack. The concentration of contracts in one manufacturer also presents supply chain risk if demand growth outpaces Bloom's production capacity.

Bigger picture

The shift toward on-site fuel cells reflects broader infrastructure constraints in the US power grid, where interconnection delays have become a defining barrier to data center expansion. The fuel cell market's projected tenfold revenue increase assumes AI compute demand continues its current trajectory, which could be disrupted by slower AI adoption, economic headwinds, or advances in energy-efficient chip design. Grid modernization efforts, while slow, could eventually reduce the urgency for on-site solutions if interconnection timelines improve. The repurposing of cryptocurrency mining infrastructure for AI data centers, exemplified by IREN's $3.65 billion agreement with Microsoft, highlights how energy-hungry facilities built in remote locations are finding new demand from AI operators willing to pay a premium for reliable power. Geopolitical tensions around rare earth supply chains, particularly China's control of scandium, introduce risk not just for Bloom Energy but for the broader SOFC sector if alternative materials are not developed or diversified quickly.

What to watch

Monitor Bloom Energy's ability to secure diversified scandium supply or develop alternative electrolyte chemistries to reduce dependency on a single critical material. Track whether competitors using different SOFC chemistries gain market share as the sector scales. Watch for updates on Bloom's manufacturing capacity expansion and whether production can keep pace with the projected 10.4 GW demand through 2030. Observe federal policy developments around tax incentives for on-site power generation, as these subsidies are a key driver of economic viability. Follow grid interconnection timeline trends; any meaningful acceleration in approvals could reduce the urgency for on-site fuel cell solutions. Keep an eye on AI compute demand growth, as slower adoption or improvements in chip energy efficiency could dampen projections. Look for announcements of new framework agreements or contracted capacity beyond the current 9 GW order book.