Hydrogen Energy Statistics 2026
By Axis Intelligence Research
Co-author: Aidan Jad | Last updated: September 6, 2026 | License: CC BY 4.0
Twenty-five projects entered grant agreement preparation in the European Union’s second hydrogen auction. Six signed. The nineteen that walked away had bid to build 2,731 MW of electrolyser capacity — seven times what the signers actually contracted. That gap between announcement and signature is the single most important hydrogen statistic of 2026, and almost nobody is publishing it.
Quick Answer
Global hydrogen demand passed 100 million tonnes in 2025, but low-emissions hydrogen reached only about 1 Mt — under 1% of supply. Axis Intelligence Research finds the Hydrogen Cost Gap (HCG™) — the volume-weighted subsidy European producers require per kilogram — rose to €0.87/kg in 2026, up 89% from €0.46/kg in 2024. Support is getting more expensive, not less.
Key Findings
- According to the IEA’s Global Hydrogen Review 2026, global hydrogen demand grew almost 3% in 2025 to surpass 100 million tonnes, while low-emissions production reached close to 1 Mt.
- Axis Intelligence Research finds the Hydrogen Cost Gap (HCG™) reached €0.87 per kilogram in the European Union’s third hydrogen auction, up from €0.46/kg in the pilot round — a 89% increase across three auctions.
- Axis Intelligence Research finds that of 25 projects invited to prepare grant agreements in the EU’s second hydrogen auction, only 6 signed — 12.3% of the 3,112 MW of electrolyser capacity that entered the process.
- Axis Intelligence Research calculates that European Hydrogen Bank support across all three auction rounds implies an abatement cost of €93.80 per tonne of CO₂ equivalent avoided, rising from €66.99/t in the pilot round to €126.71/t in 2026.
- According to the IEA’s Global Hydrogen Review 2026, the announced project pipeline for 2030 shrank by 10 Mt to 27 Mt, while production backed by final investment decisions stands at just 4.3 Mt.
How Much Hydrogen Does the World Actually Produce?
Global hydrogen production is still a fossil business with a clean-energy vocabulary bolted on. The IEA’s Global Hydrogen Review 2026, published 18 June 2026, puts total demand above 100 million tonnes in 2025, growing almost 3% year on year, with industry and refining accounting for nearly all of it. Low-emissions hydrogen — electrolytic, or fossil-derived with carbon capture — grew 20% in 2025 to reach close to 1 Mt.
That growth rate sounds impressive until you set it against the denominator. One megatonne against a hundred is under 1%. The IEA expects 2026 to cross the 1% threshold for the first time.
The emissions arithmetic makes the point harder. Hydrogen production generated around 1,000 Mt of CO₂ in 2025 by the IEA’s accounting, and that figure excludes upstream and midstream emissions from fossil fuel supply. Hydrogen is not currently a decarbonisation technology. It is an industrial feedstock that emits roughly as much carbon annually as a mid-sized industrialised economy.
Global hydrogen production and demand, 2025
| Metric | Value | Unit | As-of | Source |
|---|---|---|---|---|
| Total hydrogen demand | >100 | Mt | 2025 | IEA, Global Hydrogen Review 2026 |
| Demand growth | ~3 | % YoY | 2025 | IEA, Global Hydrogen Review 2026 |
| Low-emissions hydrogen production | ~1 | Mt | 2025 | IEA, Global Hydrogen Review 2026 |
| Low-emissions growth | 20 | % YoY | 2025 | IEA, Global Hydrogen Review 2026 |
| Low-emissions share of production | <1 | % | 2025 | IEA, Global Hydrogen Review 2026 |
| CO₂ from hydrogen production | ~1,000 | Mt | 2025 | IEA, Global Hydrogen Review 2026 |
| Installed electrolysis capacity | >4 | GW | end-2025 | IEA, Global Hydrogen Review 2026 |
| Electrolysis capacity under construction | >2.5 | GW | 2026 target | IEA, Global Hydrogen Review 2026 |
| New offtake agreements signed | 1.7 | Mtpa | 2025 | IEA, Global Hydrogen Review 2026 |
Aidan Jad: Installed electrolysis capacity doubled in 2025 to clear 4 GW, and China put in close to three-quarters of the new installations. That is the right number to watch, and it is also the number that gets misread. Four gigawatts of nameplate electrolyser capacity does not run at a gigawatt-scale capacity factor. Electrolysers chase cheap power, which means they run when renewable output is high and idle when it is not. Nameplate capacity is a purchasing statistic. Delivered kilograms are the operating statistic, and the two diverge by more in hydrogen than in almost any other energy technology.
What Is the Axis Hydrogen Cost Gap (HCG™)?
The Hydrogen Cost Gap (HCG™) is a proprietary metric published by Axis Intelligence Research. It measures, in euros per kilogram, the volume-weighted average fixed premium that hydrogen producers require in order to build — as revealed by their own binding bids into the European Union’s hydrogen auctions.
The logic is simple. The European Hydrogen Bank pays a fixed premium per kilogram of certified hydrogen, for up to ten years, awarded pay-as-bid from the lowest price upward. A producer’s bid is therefore its own estimate of the shortfall between what the hydrogen costs to make and what an offtaker will pay for it. Aggregate those bids, weight them by contracted volume, and you get a market-revealed cost gap rather than a modelled one.
HCG™ formula
HCG = Σ (bid_price_i × bid_volume_i) ÷ Σ (bid_volume_i)
Where bid_price_i is the awarded fixed premium in EUR/kg for project i and bid_volume_i is that project’s contracted hydrogen volume in kilograms over its ten-year support period. Both inputs are published project-by-project by the European Commission. No weights are assigned by Axis Intelligence Research; volume does the weighting, and every input is traceable to a Commission table.
HCG™ readings, 2024–2026
| Auction round | Cohort | Projects | Contracted volume (t / 10 yrs) | Capacity (MWe) | Support (EUR m) | HCG™ (EUR/kg) |
|---|---|---|---|---|---|---|
| IF23 (pilot) | Signed grant agreements, Oct 2024 | 6 | 1,515,000 | 1,442.00 | 694.34 | 0.46 |
| IF24 (second) | Signed grant agreements, Jan 2026 | 6 | 492,000 | 381.25 | 270.89 | 0.55 |
| IF25 (third) | Selected / in grant preparation, Jun 2026 | 13 | 1,434,029 | 1,160.50 | 1,242.68 | 0.87 |
Source: Axis Intelligence Research calculation from European Commission Innovation Fund auction result tables (IF23, IF24, IF25), retrieved 5 September 2026. Full inputs in the accompanying CSV, CC BY 4.0.
The three readings validate cleanly against the Commission’s own published aggregates. The IF24 signed cohort computes to 381.25 MWe and €270.89 million against the Commission’s stated 381.25 MW and €270.6 million. The IF25 selected cohort computes to 1,160.5 MWe and €1.243 billion against the Commission’s €1.09 billion for the nine May selections plus €147.73 million for the four June reserve-list additions. The arithmetic reconciles, which is why we publish it.
The reading to take away: €0.87/kg in 2026, against €0.46/kg in 2024. The gap did not close. It widened by 89% in two years.
Aidan Jad: The received wisdom on hydrogen is a learning curve — build more electrolysers, costs fall, subsidies taper. HCG™ says the opposite is happening in Europe, and the reason is visible in the auction design itself. The pilot round cleared at 37 to 48 cents because a handful of very large Iberian and Nordic projects with cheap renewable power and generous assumptions could bid that low. When the Commission added a completion guarantee and started enforcing timelines, the cheap bids stopped converting. What is left bidding is the honest end of the market, and the honest end costs more. This is not a cost curve failing. It is a cost curve finally being measured.
Why Do EU Hydrogen Auction Winners Keep Walking Away?
Here is the finding that no competitor publishes, because it requires reading three Commission result tables against each other rather than a single press release.
In the second auction (IF24), the Commission initially selected 15 projects requesting €992 million. Several reassessed their bids once the completion guarantee requirement bit, and withdrew. CINEA then pulled additional projects from the reserve list in strict price order. By the time the dust settled on 20 January 2026, 25 projects had entered grant agreement preparation and 6 had signed.
IF24 grant agreement conversion, by measure
| Measure | Entered preparation | Signed | Conversion |
|---|---|---|---|
| Projects | 25 | 6 | 24.0% |
| Electrolyser capacity (MWe) | 3,111.96 | 381.25 | 12.3% |
| Contracted volume (t H₂ / 10 yrs) | 3,131,000 | 492,000 | 15.7% |
| Bid value (EUR m) | 1,900.38 | 270.89 | 14.3% |
Source: Axis Intelligence Research calculation from European Commission IF24 Auction results table, retrieved 5 September 2026.
Three things fall out of that table.
Big projects did not convert. Mean capacity among signers was 63.5 MWe. Mean capacity among the nineteen withdrawals was 143.7 MWe — 2.3 times larger. The withdrawn cohort includes a 560 MWe Dutch electrolyser, a 367.5 MWe German project and two 252 MWe and 198 MWe Spanish developments. Every project above 210 MWe that entered preparation withdrew.
Cheap bids did not convert either. The lowest price in the round, €0.20/kg for 252 MWe in Spain, withdrew. So did the second lowest at €0.25/kg. A bid price is not a cost estimate if there is no penalty for being wrong, and until the completion guarantee, there wasn’t much of one.
The pilot round looked healthier than it was. In IF23, six of seven selected projects signed and €694,521,237 was committed against €720 million awarded — a 96.5% money conversion. That number is now routinely quoted as evidence the auction model works. Across the first two rounds together, €1.712 billion was awarded and €965.1 million signed: 56.4%. One round carried the average.
Aidan Jad: Withdrawal is not failure of the auction. It is the auction working — late. A completion guarantee that forces a developer to price its own delivery risk is exactly the discipline this sector has lacked since 2021. The cost is four years of announced gigawatts that were never going to exist, sitting in every market forecast in the industry, inflating every 2030 projection built on top of them. The IF25 round now has 13 projects in preparation and signatures expected between September and November 2026. That result is the one that matters, and it is not in yet.
What Does Hydrogen Policy Support Actually Cost Per Tonne of CO₂?
Every awarded project publishes an expected greenhouse gas avoidance figure, calculated by the Commission against the 2021–2025 ETS benchmark of 6.84 tonnes CO₂e per tonne of hydrogen. Dividing committed support by avoided emissions produces an implied abatement cost — a number the Commission does not publish and that Axis Intelligence Research computes here for the first time.
Implied abatement cost of European Hydrogen Bank support
| Round | Support (EUR m) | GHG avoided (t CO₂e / 10 yrs) | EUR per t CO₂e | Source |
|---|---|---|---|---|
| IF23 signed | 694.34 | 10,365,000 | 66.99 | Axis calculation, EC IF23 table |
| IF24 signed | 270.89 | 3,367,000 | 80.45 | Axis calculation, EC IF24 table |
| IF25 selected | 1,242.68 | 9,806,950 | 126.71 | Axis calculation, EC IF25 table |
| All three rounds | 2,207.90 | 23,538,950 | 93.80 | Axis calculation |
Source: Axis Intelligence Research, from European Commission Innovation Fund auction tables, retrieved 5 September 2026. GHG avoidance as published by the Commission using the 2021–2025 ETS benchmark of 6.84 t CO₂e/t H₂, which excludes substitution effects in end use and is therefore conservative.
We are deliberately not netting this against a carbon price or an ETS allowance value. The Commission’s avoidance figure is a benchmark calculation, not a measured outcome, and the projects have not yet produced a kilogram. What the table supports is a comparison across rounds under identical methodology, and on that basis the cost of buying a tonne of avoided CO₂ through hydrogen auctions has risen 89% since the pilot.
Where Are the Green Hydrogen Projects Actually Being Built?
Concentration is the defining feature of the low-emissions hydrogen map, and it runs deeper than the headline geography.
The IEA reports that installed electrolysis capacity doubled in 2025 to surpass 4 GW, with China behind nearly three-quarters of new installations. China accounts for more than 60% of committed electrolysis capacity by 2026 but only around 25% of estimated investment. Europe is the mirror image: under 20% of capacity, 45% of investment. The per-megawatt capital intensity difference between the two is the whole European competitiveness argument in a single ratio.
Capital spending on low-emissions hydrogen projects reached nearly USD 7 billion in 2025, roughly double 2024, and equal to 0.7% of global investment in energy supply. The IEA expects close to USD 10 billion in 2026, with electrolysis taking around 70% of it — the first year electrolysis investment overtakes CCUS-based hydrogen.
Regional and structural indicators, low-emissions hydrogen
| Indicator | Value | Unit | As-of | Source |
|---|---|---|---|---|
| Capital spending on low-emissions hydrogen | ~7 | USD bn | 2025 | IEA, Global Hydrogen Review 2026 |
| Share of global energy supply investment | 0.7 | % | 2025 | IEA, Global Hydrogen Review 2026 |
| Expected capital spending | ~10 | USD bn | 2026 | IEA, Global Hydrogen Review 2026 |
| China share of committed electrolysis capacity | >60 | % | 2026 | IEA, Global Hydrogen Review 2026 |
| China share of estimated investment | ~25 | % | 2026 | IEA, Global Hydrogen Review 2026 |
| Europe share of committed electrolysis capacity | <20 | % | 2026 | IEA, Global Hydrogen Review 2026 |
| Europe share of estimated investment | 45 | % | 2026 | IEA, Global Hydrogen Review 2026 |
| Hydrogen venture capital, share of energy VC | 4 | % | 2025 | IEA, Global Hydrogen Review 2026 |
| Hydrogen VC peak share | 13 | % | 2023 | IEA, Global Hydrogen Review 2026 |
| International public finance commitments | ~3.3 | USD bn cumulative | Q1 2026 | IEA, Global Hydrogen Review 2026 |
| Africa low-emissions hydrogen production | ~6 | kt | 2026 | IEA, Global Hydrogen Review 2026 |
| Africa announced 2030 projects reaching FID | 1 of 31 | count | 2026 | IEA, Global Hydrogen Review 2026 |
Venture capital tells the sentiment story cleanly. Hydrogen’s share of energy VC fell from a 13% peak in 2023 to 4% in 2025. Aviation absorbed almost a quarter of what remained.
Africa is the clearest illustration of announcement risk. The continent used 3.1 Mt of hydrogen in 2024, roughly 3% of global demand, overwhelmingly for ammonia made from natural gas. Low-emissions production is around 6 kt. Of 31 projects announced for 2030, one has reached final investment decision.
What Happened to the US Hydrogen Hubs?
The United States committed $7 billion to seven Regional Clean Hydrogen Hubs under the Bipartisan Infrastructure Law, against $8 billion appropriated by Congress. Then the program became a political object.
On 2 October 2025 the Department of Energy announced the termination of 321 financial awards supporting 223 projects, stating savings of approximately $7.56 billion. DOE said the projects did not adequately advance national energy needs and would not deliver a positive return on taxpayer dollars. Of the terminated awards, 26% had been made between Election Day and Inauguration Day, valued at over $3.1 billion. That action followed a May 2025 termination of 24 Office of Clean Energy Demonstrations awards worth over $3.7 billion.
H2Hubs federal cost share as originally awarded
| Hub | Lead region | Production route | Federal cost share (USD m) |
|---|---|---|---|
| California Hydrogen Hub (ARCHES) | California | Electrolysis | 1,200 |
| Gulf Coast Hydrogen Hub (HyVelocity) | Texas | Electrolysis + natural gas with CCS | 1,200 |
| Midwest Alliance for Clean Hydrogen (MachH2) | IL, IN, MI | Mixed | 1,000 |
| Pacific Northwest Hydrogen Hub (PNWH2) | WA, OR, MT | Electrolysis | 1,000 |
| Appalachian Hydrogen Hub (ARCH2) | WV, OH, PA | Natural gas with CCS | 925 |
| Heartland Hydrogen Hub (HH2H) | MN, ND, SD | Mixed | 925 |
| Mid-Atlantic Hydrogen Hub (MACH2) | PA, DE, NJ | Green and pink | 750 |
| Total | 7,000 |
Source: DOE Office of Clean Energy Demonstrations awardee fact sheets and program announcements. ARCHES phase 1 award of $30 million against a total federal cost share of up to $1.2 billion confirmed in the OCED ARCHES Awardee Fact Sheet, July 2024. Retrieved 5 September 2026.
Two clarifications matter for anyone citing this table. First, these are ceilings, not disbursements — the hubs were structured in phases, with initial tranches between $18.8 million and $30 million. Second, the live status of each hub as of September 2026 is contested and moving through litigation and appropriations, and DOE has not published a consolidated post-termination status page. Axis Intelligence Research publishes the award ledger, which is documented, and declines to publish a status column that would require sourcing from press leaks rather than agency records.
The tax side is more settled and more consequential. Section 45V of the Internal Revenue Code offers up to $0.60 per kilogram, multiplied by 5 where prevailing wage and apprenticeship requirements are met, for hydrogen with lifecycle emissions below 0.45 kg CO₂e/kg — a headline $3.00/kg before inflation adjustment. Public Law 119–21, enacted 4 July 2025, substituted “January 1, 2028” for “January 1, 2033” as the construction-start deadline. Five years of runway removed in a single amendment.
Aidan Jad: Put the two policy instruments side by side and resist the urge to convert currencies. The EU’s number is a contracted average across projects that competed for it. The US number is a statutory ceiling available to anyone who qualifies and starts construction in time. They are not the same kind of number and dividing one by the other produces a comparison that means nothing. What is comparable is the direction of the constraint. Europe’s binding constraint is price — the premium keeps rising. America’s binding constraint is the calendar — the credit is generous and the door closes in fifteen months. A developer in Texas is not optimising against cost per kilogram right now. They are optimising against a shovel in the ground before 1 January 2028.
Is the 2030 Hydrogen Target Still Reachable?
No, and the IEA now says so in its own numbers rather than in hedged prose.
Announced production targets reach almost 27 Mt per year by 2030. Projects that have secured investment decisions suggest just over 4 Mt, with a further 2 Mt possible if enabling conditions improve. Committed production grew 3% to 4.3 Mt — 15.9% of the announced pipeline. The pipeline itself shrank by 10 Mt over the year, driven by post-2030 delays, pauses and cancellations.
More than 100 GW of announced electrolysis capacity loses any realistic chance of operating by 2030 if investment decisions are not taken before the end of 2027. Around 22 Mt of announced production sits in that same window, two-thirds of it in Europe, North America and Latin America.
The IEA’s assessment of national progress is one sentence long and worth quoting the substance of: only China and the Netherlands are on track to meet their objectives.
New FIDs fell below 0.8 Mtpa in 2025 after two consecutive years at around 1 Mtpa — the first decline the sector has recorded.
Pipeline reality check, 2030 outlook
| Metric | Value | Unit | As-of | Source |
|---|---|---|---|---|
| Announced production pipeline for 2030 | 27 | Mt | 2026 | IEA, Global Hydrogen Review 2026 |
| Pipeline contraction vs 2025 review | 10 | Mt | 2026 | IEA, Global Hydrogen Review 2026 |
| Committed production (post-FID) by 2030 | 4.3 | Mt | 2026 | IEA, Global Hydrogen Review 2026 |
| Committed as share of announced | 15.9 | % | 2026 | Axis calculation from IEA figures |
| Additional potential if FID taken by 2027 | ~2 | Mt | 2026 | IEA, Global Hydrogen Review 2026 |
| Announced electrolysis capacity at risk | >100 | GW | 2026 | IEA, Global Hydrogen Review 2026 |
| Announced production at risk without FID by early 2027 | 22 | Mt | 2026 | IEA, Global Hydrogen Review 2026 |
| New FIDs | <0.8 | Mtpa | 2025 | IEA, Global Hydrogen Review 2026 |
| Countries on track for 2030 targets | China, Netherlands | — | 2026 | IEA, Global Hydrogen Review 2026 |
Set the EU auction results against the European ambition and the scale problem is stark. Across all three European Hydrogen Bank rounds, contracted and selected volume totals 3,441,029 tonnes over ten years — 344,103 tonnes per year, or 3.4% of the REPowerEU domestic renewable hydrogen ambition of 10 Mt per year by 2030. The 2,984 MWe of electrolyser capacity across those rounds is a real number. It is not a 2030-scale number.
Who Is Actually Buying Low-Emissions Hydrogen?
Demand, not supply, is the constraint every developer names first.
New offtake agreements for low-emissions hydrogen totalled around 1.7 Mtpa in 2025, unchanged from 2024. Only about 20% of newly signed volumes carried firm contractual commitments, concentrated in refining, industry and power generation. For the first time, trade-oriented agreements overtook agreements for domestic use, driven by policy in Japan and the European Union.
Procurement tenders grew marginally to more than 1 Mtpa, of which more than 0.3 Mtpa had been contracted as of Q1 2026, mostly in refining and fertilisers.
Based on projects that have reached FID, 2.5 Mt of low-emissions hydrogen is expected to be consumed in refineries and industrial facilities by 2030 — 60% of committed global production. Refining and chemicals are the lead markets, and they are lead markets because they already use hydrogen. Nobody has to be persuaded that the molecule is useful; they have to be persuaded to pay more for a cleaner version of what they already buy.
The IEA’s cost acceptability analysis puts a number on that persuasion problem. Absent policy support, the maximum acceptable hydrogen cost is below USD 2/kg for most sector-region combinations. For steel production, the acceptable cost is negative — the hydrogen would have to come with an incentive attached to reach parity with the incumbent route. Even with policy incentives, the cost gap does not close in most regions and applications. Only China comes close, and only with sufficiently high carbon pricing across all sectors.
Aidan Jad: Steel is where the hydrogen thesis either survives or doesn’t, and the cost acceptability work explains why it keeps stalling. Steel has high specific emissions but a low share of energy in total product cost, which makes it maximally sensitive to carbon pricing and minimally sensitive to hydrogen price. That is an awkward combination. It means the lever that moves steel is not a cheaper electrolyser — it is a carbon price that most jurisdictions have not legislated. Every green steel announcement should be read as a bet on future carbon policy, not on future hydrogen cost.
What the Middle East Conflict Did to Hydrogen Supply Chains
The region holds around one-sixth of global hydrogen production, most of it dedicated to chemicals, fertilisers and refined products. It accounts for more than 10% of global refining capacity, ammonia and urea production, and close to 17% of methanol production.
Trade exposure is higher than production exposure. The Middle East supplies over one-quarter of global ammonia trade, almost 40% of urea trade and almost 45% of methanol trade, with a third of its refining capacity export-oriented. The closure of the Strait of Hormuz disrupted all of it.
Urea prices doubled between January and May 2026. Ammonia production in Bangladesh, India and Pakistan — a quarter of which runs on Middle East natural gas — was directly affected. Morocco imports all of its ammonia, 40% from the region. Brazil, Australia, South Africa and Thailand import all their urea, 40–85% of it from the Middle East.
This is the energy security argument for hydrogen stated in its strongest available form, and the IEA’s own conclusion is that existing and committed low-emissions production is not sufficient to provide an immediate response.
Methodology
Collection. All figures were retrieved from primary source documents during production on 5 September 2026. No figure in this article originates from model memory, a secondary aggregator, or a market-research summary. Where a source publishes a rounded or approximate value, that value is reproduced as published rather than given false precision.
Primary sources used. International Energy Agency, Global Hydrogen Review 2026 (published 18 June 2026), chapters on Executive Summary, Key Questions, Demand, Production, Cost Acceptability, and Investment and Innovation. European Commission DG Climate Action, Innovation Fund IF23, IF24 and IF25 hydrogen auction result pages, including project-level bid tables. European Commission DG Energy, European Hydrogen Bank programme page. US Department of Energy, termination announcements of 30 May 2025 and 2 October 2025, and Office of Clean Energy Demonstrations awardee fact sheets. US House Office of the Law Revision Counsel, 26 U.S.C. §45V, text in effect 3 September 2026.
HCG™ construction. The Hydrogen Cost Gap is a volume-weighted mean of awarded fixed premiums: HCG = Σ(bid_price × bid_volume) ÷ Σ(bid_volume). Inputs are the Commission’s published bid price (EUR/kg) and bid volume (tonnes H₂ over ten years) for each project in the relevant cohort. No Axis-assigned weights, no normalisation, no adjustment. A competent outsider can recompute every reading from the Commission’s tables and the CSV accompanying this page.
Cohort definitions. IF23 and IF24 readings use signed grant agreements only, because an award that is never signed is not a contracted premium. The IF25 reading uses projects invited to grant agreement preparation as of June 2026, because signatures are expected between September and November 2026. This is stated on the reading and in the CSV. The IF25 figure will be recomputed on the signed cohort when the Commission publishes it.
Verification performed. Three independent reconciliations were run against Commission-published aggregates. The IF24 signed cohort computes to 381.25 MWe against the Commission’s stated 381.25 MW; to €270.89 million against the Commission’s stated €270.6 million; and to 492,000 tonnes against the Commission’s stated “around 500 kilotonnes”. The IF23 signed cohort computes to €694.34 million against the Commission’s stated €694,521,237 and 1,515,000 tonnes against the stated 1.52 million tonnes. The IF25 cohort separates cleanly into the nine projects announced 7 May 2026 (1,083.5 MWe, 1,322,253 t, €1,094 m against Commission-stated “almost 1.1 GW”, “over 1.3 million tonnes”, “€1.09 billion”) and the four reserve-list additions of June 2026 (77 MWe, 111,776 t, €148.6 m against Commission-stated 77 MW, 111,000 t, €147.73 m). Residual differences reflect the Commission’s rounding of bid volumes to whole kilotonnes in the IF23 and IF24 tables. All arithmetic was computed twice, in floating point and in exact rational arithmetic.
Unit handling note. The Commission’s IF25 result tables label bid volume as “kilotonne” in two of three topic tables while the values are stated in tonnes. Axis Intelligence Research treats all IF25 bid volumes as tonnes. This is confirmed by internal consistency: Cloudberry’s 508,915 at 500 MWe and NJK’s 445,000 at 300 MWe are consistent with tonnes and off by three orders of magnitude as kilotonnes, and the tonnes reading reconciles exactly with the Commission’s own stated aggregate of “over 1.3 million tonnes”.
Scope. HCG™ measures European auction-revealed premiums. It is not a global hydrogen production cost. It reflects one policy instrument in one jurisdiction, which is precisely what makes it a clean signal — the auction design is constant across rounds, so movements in the reading reflect changes in bidder economics rather than changes in methodology. HCG™ does not capture national auction-as-a-service awards, which are funded and reported separately by member states. Volumes are contracted, not produced; no project in any round has yet delivered a kilogram of subsidised hydrogen.
Figures we declined to publish. We do not publish a live status column for the seven US hydrogen hubs. As of September 2026 the post-termination status of individual hubs is subject to litigation and to a “retain or modify” list that has circulated in the press but has not been published as an agency record. A status table sourced from leaked documents would be an unverifiable table with our name on it. We publish the award ledger, which is documented, and will add status when DOE publishes it. We also do not convert the US 45V credit into euros for comparison with HCG™: a statutory ceiling and a contracted volume-weighted average are different objects, and the ratio between them would carry an authority it has not earned.
About This Dataset
The complete dataset behind this page is published as a single CSV under CC BY 4.0. Every numeric claim in the prose above has a corresponding row with its value, unit, as-of date, source organisation, source document, source URL, retrieval date, primary-source flag, Axis-calculation flag and method note.
The dataset covers global hydrogen demand and production indicators from the IEA Global Hydrogen Review 2026; project-level bid data for all 44 project entries across three European Hydrogen Bank auction rounds, including withdrawals; HCG™ readings and their inputs; implied abatement costs; and the US H2Hubs award ledger.
Citation format.
- APA: Axis Intelligence Research, & Jad, A. (2026). Hydrogen energy statistics 2026: Production, costs, and the subsidy gap that decides the decade. Axis Intelligence. https://axis-intelligence.com/hydrogen-energy-statistics/
- MLA: Axis Intelligence Research, and Aidan Jad. “Hydrogen Energy Statistics 2026: Production, Costs, and the Subsidy Gap That Decides the Decade.” Axis Intelligence, 6 Sept. 2026, axis-intelligence.com/hydrogen-energy-statistics/.
- Chicago: Axis Intelligence Research and Aidan Jad. “Hydrogen Energy Statistics 2026: Production, Costs, and the Subsidy Gap That Decides the Decade.” Axis Intelligence. September 6, 2026. https://axis-intelligence.com/hydrogen-energy-statistics/.
- HCG™ metric: Axis Intelligence Research, Hydrogen Cost Gap (HCG™), 2026 reading €0.87/kg, as of September 6, 2026, axis-intelligence.com/hydrogen-energy-statistics/. CC BY 4.0.
Frequently Asked Questions
Why do so many winners of EU hydrogen auctions never sign their grant agreements?
Because winning an auction and financing a project are different tests. Of the 25 projects that entered grant agreement preparation under the EU’s second hydrogen auction, 6 signed. The Commission requires a signed completion guarantee within two months of invitation, and that guarantee forces developers to price their own delivery risk before committing. Several of the 15 originally selected projects reassessed their bids and maturity and withdrew, after which CINEA invited reserve-list projects in strict price order. Axis Intelligence Research calculates the conversion at 12.3% by electrolyser capacity.
What does green hydrogen cost to produce in 2026, and why is the subsidy rising?
The Hydrogen Cost Gap (HCG™) — the volume-weighted premium producers required in the EU’s third auction — reached €0.87/kg in 2026, against €0.46/kg in the 2024 pilot round. The rise is not a cost curve reversing. It reflects the exit of very low, weakly collateralised bids after the completion guarantee was enforced, leaving a bidder pool whose prices better reflect deliverable project economics. The IEA separately notes that fossil-based production will remain cheaper than renewable hydrogen in most of the world in the near term.
Is blue hydrogen currently cheaper than green hydrogen?
In most regions, yes, and the IEA expects that to persist in the near term. Fossil-based production remains less costly than renewable hydrogen in most of the world, though fossil price volatility can narrow the gap during some periods. Deployment of CCUS-based hydrogen is running behind schedule: one large project started operation in 2025, two smaller ones are expected in 2026, and several have been postponed because of delays in CO₂ transport and storage infrastructure. Investment in electrolysis overtook investment in CCUS-based hydrogen in 2025 on the strength of a stronger pipeline and faster project progress.
How much of the world’s hydrogen is genuinely low-emissions?
Under 1% in 2025. Low-emissions production reached close to 1 Mt against total demand above 100 Mt, and the IEA expects the 1% threshold to be crossed for the first time in 2026. Most low-emissions hydrogen today is fossil-derived with carbon capture rather than electrolytic. Hydrogen production generated around 1,000 Mt of CO₂ in 2025, excluding upstream and midstream fossil supply emissions.
What is the current status of the seven US regional hydrogen hubs?
The seven hubs were awarded up to $7 billion in federal cost share under the Bipartisan Infrastructure Law, ranging from $750 million to $1.2 billion each. On 2 October 2025 the Department of Energy announced the termination of 321 awards supporting 223 projects, citing approximately $7.56 billion in savings. Individual hub status as of September 2026 is contested and moving through litigation and appropriations, and DOE has not published a consolidated post-termination status record. Axis Intelligence Research publishes the award ledger and declines to publish a status column that would require leaked documents as its source.
Does the 45V clean hydrogen tax credit still exist?
Yes, with a much shorter runway. Section 45V provides up to $0.60/kg multiplied by 5 where prevailing wage and apprenticeship requirements are met — $3.00/kg before inflation adjustment — for hydrogen with lifecycle emissions below 0.45 kg CO₂e/kg, over a ten-year production period. Public Law 119–21, enacted 4 July 2025, moved the construction-start deadline from 1 January 2033 to 1 January 2028. Projects that begin construction before that date retain eligibility.
Why is the hydrogen project pipeline shrinking if demand keeps growing?
Because the demand that is growing is the wrong kind. Total hydrogen demand rose almost 3% in 2025, but that growth is in conventional industrial and refining use served by unabated fossil hydrogen. Demand for low-emissions hydrogen specifically — backed by firm offtake at a price that covers the premium — is the constraint. New offtake agreements were flat at around 1.7 Mtpa and only about 20% carried firm contractual commitments. Without firm offtake, developers cannot reach FID, and the announced pipeline for 2030 contracted by 10 Mt to 27 Mt.
Which countries are on track to meet their 2030 hydrogen targets?
China and the Netherlands. Announced targets total almost 27 Mt per year by 2030 while post-FID production stands at 4.3 Mt — 15.9% of the announced pipeline. More than 100 GW of announced electrolysis capacity, and around 22 Mt of announced production, lose any realistic chance of operating by 2030 if investment decisions are not taken by early 2027. Two-thirds of that at-risk production sits in Europe, North America and Latin America.
How does the EU’s hydrogen subsidy compare with the US 45V credit?
They are not directly comparable, and Axis Intelligence Research declines to convert one into the other. The EU figure of €0.87/kg is a volume-weighted average of what competing projects actually bid and were awarded. The US figure of $3.00/kg is a statutory ceiling available to any qualifying producer. What is comparable is the shape of the constraint: Europe’s binding limit is price, and it is rising; America’s is the calendar, with construction required to start before 1 January 2028.
What would make the Hydrogen Cost Gap fall?
Three things, in order of realism. Cheaper firm renewable power, which drives the largest single input to electrolytic production cost. Higher carbon pricing, which raises the maximum acceptable hydrogen cost by penalising the incumbent route — the IEA finds that even with policy incentives the gap does not close in most regions, and only China comes close, and only with high carbon pricing across all sectors. And electrolyser cost reduction through manufacturing consolidation, which is now under way in China through excess capacity and pricing below manufacturing cost, but which the IEA characterises as unsustainable internal competition rather than durable learning.
Related Axis Intelligence Research
Hydrogen sits inside a wider low-carbon energy build-out that Axis Intelligence Research tracks across several datasets. For the electricity that any electrolyser depends on, see Nuclear Energy Statistics 2026 and Solar Energy Statistics 2026 and Offshore Wind Statistics. On the transport demand side — where hydrogen fuel cells continue to lose ground to batteries — see Electric Vehicle Statistics 2026 and EV Adoption Statistics 2026. The full catalogue of sourced datasets is at Axis Intelligence Statistics.