Hydrogen & Fuel Cells
Hydrogen Strategy
Not investment advice. This page is for informational and educational purposes only. Sterling is not a registered investment advisor. Nothing here is a recommendation to buy, sell, or hold any security. Markets carry risk and you may lose money. Do your own research and consult a licensed professional before making any investment, securities purchase, or trading decision.
Thesis
Why this sector matters to investors right now. Structural, not market timing.
Hydrogen and fuel cells in 2026 are two businesses. The one that is printing is stationary solid-oxide fuel cells as on-site prime power. Bloom Energy did $2.02 billion of revenue in 2025 (+37%) and raised 2026 guidance to $3.9 to $4.2 billion after a $1.065 billion Q2. AEP converted a 2024 option into 1 GW of SOFC capacity in January 2026. Bloom says the major US hyperscalers have qualified the product for AI factories. That is a generation story about delayed interconnects. Natural gas is the feedstock. Hydrogen capability is optionality.
The molecule is still grey. IEA Global Hydrogen Review 2025 puts 2024 demand at almost 100 million tonnes, up 2%, almost all of it refining (more than 43 Mt) and industry (55 Mt). Low-emissions production is on track for 1 Mt in 2025, less than 1% of supply. Projects at FID reach 4.2 Mtpa by 2030. The announcement pile is 37 Mtpa, already cut from 49. Electrolyzer capacity is just over 3 GW, 65% of it in China. Plug Power, the hydrogen-economy ticker, did about $710 million in 2025 and is still trying to earn. Mixing Bloom's SOFC multiple with Plug's PEM-and-electrolyzer multiple is a category error.
Structural drivers
Forces that shape long-run demand and economics. Each driver is sourced.
- •Data-center interconnect queues turned on-site generation from a C&I nicety into a bottleneck product. Bloom FY2025 $2.02 billion, Q2 2026 $1.065 billion, FY2026 guide $3.9 to $4.2 billion. Source: Bloom FY2025 and Q2 2026 results.
- •AEP converted 100 MW plus a 900 MW option into 1 GW of Bloom SOFC in January 2026, reported as a $2.65 billion 20-year offtake. A utility buying fuel cells at that scale is new. Source: January 2026 AEP/Bloom conversion reporting.
- •Gas turbines and large transformers are sold out well into the decade (GE Vernova backlog on the Data Centers page). SOFC is one of the few dispatchable kW that can be delivered behind the meter on a data-center clock. Source: Data Centers power-suppliers table; Bloom Q2 2026 hyperscaler-qualification claim.
- •Traditional hydrogen demand is large and growing slowly: almost 100 Mt in 2024, +2%, driven by refining and chemicals, not by policy. That is the industrial-gas franchise (Air Products, Linde, Air Liquide). Source: IEA GHR 2025 Demand chapter.
- •Low-emissions hydrogen from FID projects still scales: 4.2 Mtpa by 2030, about 4% of today's total. That is real growth from a tiny base. Source: IEA GHR 2025 executive summary.
- •US 45V and 45Q, EU RFNBO, and the remaining DOE hubs can still move individual FIDs in refining, ammonia, and blue hydrogen. Source: IRS 45V/45Q; EU RFNBO; DOE OCED hubs.
- •China already has 65% of installed electrolyzer capacity and about 60% of manufacturing. Cost-down, if it happens, happens there first. Source: IEA GHR 2025.
Structural risks
Forces that could compress demand, change economics, or break the thesis.
- •Bloom's 2026 guide assumes the data-center offtake converts. A hyperscaler capex air pocket, or interconnects arriving faster than expected, would re-rate the SOFC book. Source: Bloom Q2 2026 guide; Data Centers hyperscaler-capex risk.
- •SOFC runs on natural gas today. A carbon, methane, or interconnection rule that treats behind-the-meter gas generation as equivalent to a new gas plant would hit the product. Source: Bloom 10-K risk factors; fuel is disclosed as natural gas.
- •AEP is one customer at 1 GW. Concentration in a single utility framework is a contract-risk, not a TAM. Source: January 2026 AEP conversion reporting.
- •Low-emissions offtake slowed: 1.7 Mtpa signed in 2024 vs 2.4 in 2023, only 20% firm. Electrolyzer OEMs (Nel, ITM, nucera, Plug) sit on that number. Source: IEA GHR 2025 Demand.
- •Announced manufacturing (38 GW/yr in 2024 toward 186 GW/yr in 2030) versus 26 GW cumulative FID-plus-construction is an OEM glut. IEA says the market outside China is consolidating. Source: IEA GHR 2025 executive summary.
- •DOE terminated ARCHES and the Pacific Northwest hub in October 2025. Policy duration on 45V is a political risk. Source: DOE hub status reporting, April 2026 retain list.
- •Plug, FuelCell Energy, and several European electrolyzer names have a history of dilutive capital raises. The hydrogen ticker tape is not Bloom's P&L. Source: Plug FY2025 results; historical FCEL and ITM filings.
Competitive landscape
How to think about the players. Framing along axes (pure play vs diversified, incumbent vs challenger, etc). Not stock picking.
Four archetypes. (1) Stationary SOFC: Bloom Energy. Compare it to on-site generation and delayed interconnects, not to green hydrogen. Ceres is SOFC IP. FuelCell Energy is MCFC/SOFC and distressed; it is on the roster for completeness. (2) PEM fuel cells: Plug (forklifts plus a hydrogen network) and Ballard (mobility stacks). Forklifts print. Trucks, if they print, print on Semi Trucks. (3) Electrolyzer OEMs: Nel, ITM, thyssenkrupp nucera, Plug's GenEco, Bloom's SOEC, plus private Electric Hydrogen and Ohmium. China is the volume. Western names are in an oversupplied factory cycle. (4) Industrial hydrogen: Air Products, Linde (dual-homed with AI Chips), Air Liquide. Grey and blue merchant H2 for refining and ammonia. Chart Industries sells liquefaction kit into whatever color actually gets built. Stegra is the private offtake (hydrogen-DRI steel).
Key metrics to watch
The operational and financial metrics that matter most in this sector. Each one names its source and update cadence.
| Metric | Source | Frequency | Why it matters |
|---|---|---|---|
| Bloom revenue, product revenue, and FY guide | Bloom quarterly results | Quarterly | The only fuel-cell P&L that is inflecting. Product vs installation vs service tells you whether servers are shipping. |
| Bloom backlog (product vs total) | Bloom earnings (FY2025: product ~$6B, total ~$20B) | Quarterly | Converts the AEP and hyperscaler claims into a book. Cancellation language in the 10-K still applies. |
| AEP deployment MW vs the 1 GW framework | AEP and Bloom PRs / 10-Q commentary | Ad hoc | The sized utility contract. If MW installed stalls, the framework was a press release. |
| IEA low-emissions production and FID Mtpa | IEA Global Hydrogen Review (annual) | Annual | The molecule's reality check. Watch the FID path, not the announcement pile. |
| Electrolyzer installed GW and China share | IEA GHR | Annual | OEM volume. 3 GW global is not a shortage. |
| Plug revenue, gross margin, and cash use | Plug quarterly results | Quarterly | Whether the hydrogen-economy ticker can earn. Q4 2025 was the first recent positive gross profit print. |
| Firm offtake Mtpa (IEA) | IEA GHR Demand chapter | Annual | 1.7 Mtpa signed in 2024, 20% firm. If this does not turn up, electrolyzer FIDs will not either. |
Catalysts and milestones
Known upcoming events that could move the sector. Dated where possible.
- •Bloom remaining 2026 quarters vs the $3.9 to $4.2 billion guide. Source: subsequent Bloom 10-Q.
- •First disclosed MW installed under the AEP 1 GW framework. Source: AEP or Bloom commentary.
- •IEA Global Hydrogen Review 2026: does the FID path stay near 4.2 Mtpa or slip. Source: IEA.
- •Plug Q4 2026 EBITDAS target. Source: Plug Q2 2026 commentary.
- •Next DOE hub Phase 2 award or further termination. Source: DOE OCED.
- •45V political duration into 2027. Source: Treasury / legislative calendar.
- •Any hyperscaler that discloses Bloom MW as a named offtake rather than a qualification. Source: issuer 8-K or 10-Q.
What would change the view
Conditions or evidence that would invalidate the thesis or materially shift the risk picture.
- •Bloom misses the $3.9 billion low end of 2026 guidance without a stated supply or accounting reason. The SOFC inflection would be a one-year spike.
- •AEP discloses a pause, cancellation, or material delay on the 1 GW framework.
- •A named hyperscaler signs a multi-hundred-MW SOFC offtake in an 8-K. Concentration risk would fall.
- •IEA FID path for 2030 low-emissions hydrogen falls below 3 Mtpa. The molecule thesis would be slipping, not lagging.
- •Plug reports two consecutive quarters of positive GAAP gross margin and declining cash use. The hydrogen-economy ticker would be a business.
- •A Western electrolyzer OEM announces a factory closure or take-private. The OEM glut would be resolving through capacity destruction.
- •Behind-the-meter gas generation is barred or tariffed as a new emitting plant in Texas or PJM. SOFC's interconnect advantage would shrink.
What we are not covering
Sub-areas, technologies, or companies we are deliberately excluding from the analysis, and why.
- •FCEV Class 8 trucks, HD hydrogen stations, and DOE hub truck relevance (Semi Trucks).
- •Toyota Mirai and Hyundai Nexo (EV).
- •GE Vernova, Caterpillar, Cummins, Generac gensets (Data Centers). Bloom is dual-noted there as context.
- •AEP as a regulated utility (Utilities). AEP is the customer, not the fuel-cell company.
- •Natural gas production and LNG (Oil & Gas). Gas is the SOFC feedstock.
- •Batteries and Megapack (Batteries). Storage is not generation.
- •Nuclear PPAs for data centers (Nuclear).
- •Consultant hydrogen TAM and paywalled electrolyzer share tables.
- •Doosan Fuel Cell, PowerCell Sweden, and Chinese alkaline OEMs (later roster, not this ship).
- •Nikola (bankrupt) and HYLA stations.
Sources
Primary sources cited in this analysis. Links open in a new tab.
- IEA Global Hydrogen Review 2025 executive summaryAccessed 2026-08-24
- IEA GHR 2025 DemandAccessed 2026-08-24
- IEA GHR 2025 Production highlightsAccessed 2026-08-24
- Bloom Energy FY2025 resultsAccessed 2026-08-24
- Bloom Energy Q2 2026 resultsAccessed 2026-08-24
- Bloom Energy FY2024 resultsAccessed 2026-08-24
- AEP 1 GW Bloom conversion reporting (Jan 2026)Accessed 2026-08-24
- Plug Power FY2025 resultsAccessed 2026-08-24
- Plug Power Q2 2026 resultsAccessed 2026-08-24
- DOE OCED Regional Clean Hydrogen HubsAccessed 2026-08-24
- Environment+Energy Leader: hub terminationsAccessed 2026-08-24
- DOE alumni network: five hubs retained April 2026Accessed 2026-08-24
- IRS hydrogen production credit (45V)Accessed 2026-08-24
- IRS carbon oxide sequestration credit (45Q)Accessed 2026-08-24
- EU RFNBOAccessed 2026-08-24
Audit trail
Record of the last review and what changed. Required on every refresh.
Last reviewed: 2026-08-24
Change log
- 2026-08-24Initial publication. Thesis split between Bloom SOFC (FY2025 $2.02B, Q2 2026 $1.065B, AEP 1 GW) and the still-grey hydrogen molecule (IEA GHR 2025: ~100 Mt demand, <1% low-emissions, 4.2 Mtpa FID path, >3 GW electrolyzers).
Unresolved questions
- •Bloom MW deployed under the AEP 1 GW framework, as opposed to contracted.
- •Hyperscaler MW totals. Qualification is not a disclosed MW book.
- •Plug 2026 full-year revenue vs the 15-16% growth guide.
- •Nel, ITM, and nucera 2025 revenue in native currency, for a later OEM chart.
- •Air Products, Linde, and Air Liquide hydrogen-segment revenue as a disclosed line.
- •IEA 2026 review: FID path and China electrolyzer share.
- •Whether Bloom SOEC revenue is ever broken out of the Energy Server P&L.
