Green Hydrogen Timeline 1800–2026: Electrolysis, $130 Billion and the Search for a Buyer
Green hydrogen from the 1800 electrolysis experiment to 2026: $130bn committed, 570+ projects, a shrinking 2030 pipeline and the missing buyers.
Latest Story
Billions of dollars have been committed. Hundreds of projects have moved past the slide deck. Governments from China to Europe, India and the United States have written hydrogen into industrial policy. And yet, through 2024, 2025 and 2026, some of the world’s most ambitious hydrogen projects have been postponed, shrunk or cancelled outright. The problem is not that we do not know how to make green hydrogen — humans first split water with electricity in 1800. The harder question is whether low-emissions hydrogen can be produced cheaply enough, delivered where it is needed, and sold to enough customers willing to pay a premium over the fossil-based hydrogen the world already uses. By September 2026, the Hydrogen Council counts more than $130 billion of committed investment across 570-plus projects, while the IEA counts barely 1 million tonnes a year of actual low-emissions hydrogen demand against a total hydrogen market above 100 million tonnes. That gap — between what has been funded to build and what anyone has agreed to buy — is the real hydrogen story of 2026.
🧠 AI Overview Summary
Green hydrogen is hydrogen made by splitting water with renewable electricity in an electrolyser, so no fossil fuel is used as the feedstock. The technology is two centuries old, but the industry is at an inflection point. As of September 2026 the Hydrogen Council reports more than $130 billion of committed clean-hydrogen investment across over 570 projects and about 6.9 million tonnes per year of committed capacity, with roughly 90% operational or under construction. The IEA reports total hydrogen demand above 100 million tonnes in 2025 — almost all of it conventional refining, ammonia and chemicals — but low-emissions hydrogen demand of only about 1 million tonnes, and a 2030 announced production pipeline that has shrunk to roughly 27 million tonnes because of delays and cancellations. Committed 2030 production is about 4.3 million tonnes. The bottleneck has shifted from whether hydrogen can be produced to whether it can be produced, delivered and purchased at a price that makes economic sense.
Sources: Hydrogen Council / McKinsey, Global Hydrogen Compass 2026 (10 September 2026); IEA Global Hydrogen Review 2025 and 2026; India Ministry of New and Renewable Energy (MNRE); European Commission; company announcements; and Reuters reporting. Figures are the latest published estimates, use different definitions across datasets, and are revised regularly. See the methodology box and the “why the numbers differ” section for what each figure does and does not measure.
Green Hydrogen: Key Questions
What the 2026 hydrogen data actually says
- The science is old. Water was first split into hydrogen and oxygen with an electric current in 1800; industrial electrolysers have run since the 1920s. Technology is not the binding constraint.
- Committed is not operating. The Hydrogen Council’s $130 billion and 6.9 Mt/year cover projects past a final investment decision (FID); about 90% are operational or under construction. That is a subset of a far larger announced pipeline.
- Committed is not spent. “$130 billion committed” means capital allocated to sanctioned projects over their build-out, not money already deployed. Actual 2025 capital spending on low-emissions hydrogen was around $7 billion.
- Demand is the bottleneck now, not supply. Total hydrogen demand is above 100 Mt, but low-emissions hydrogen demand is only ~1 Mt. Developers can build electrolysers; someone has to buy the output.
- The 2030 pipeline is shrinking. The IEA’s announced 2030 low-emissions production pipeline has fallen to about 27 Mt/year, from roughly 37 Mt a year earlier and around 49 Mt in 2023, on delays and cancellations.
- Offtake is mostly non-binding. New offtake agreements held at about 1.7 Mt/year in 2025, but only around one-fifth were firm contracts. The rest are memoranda and letters of intent.
- China is scaling fastest. China accounts for more than half of committed renewable-hydrogen capacity and about three-quarters of new electrolyser installations, helped by equipment that can cost a fraction of Western systems.
- Colour labels are simplifications. Green is not automatically zero-carbon and blue is not automatically clean — actual emissions depend on the electricity source, methane leakage, capture rate and system boundary.
- India has committed early but built little. The National Green Hydrogen Mission targets at least 5 million tonnes a year by 2030 with a ₹19,744 crore outlay; commissioned capacity was around 8,000 tonnes a year in early 2026.
- The question has changed. 2020 asked what hydrogen could decarbonise. 2026 is asking which hydrogen uses customers will actually pay for.
Why do hydrogen reports show different numbers?
The Hydrogen Council and the IEA are not contradicting each other. They are counting different things.
Two headline datasets dominate hydrogen coverage, and they are easy to misread as conflicting. They are not — they use different definitions, thresholds and status classes. Never add a number from one to a number from the other.
| Attribute | Hydrogen Council — Global Hydrogen Compass 2026 | IEA — Global Hydrogen Review 2025 / 2026 |
|---|---|---|
| Publisher | Industry body (~140 member companies), with McKinsey & Company | Intergovernmental energy agency |
| Latest release | 10 September 2026 (Tokyo, Hydrogen Energy Ministerial) | GHR 2025 (Oct 2025); GHR 2026 findings released Sept 2026 |
| Headline framing | “Committed investment” — capital in projects past FID | Demand, production, pipeline, FID capacity and offtake, tracked separately |
| Core status class | Committed = FID taken; ~90% operational or under construction | Operating vs under construction vs FID vs announced vs “strong potential” |
| Technology scope | Clean hydrogen: renewable (electrolytic) and low-carbon (fossil + CCS) | Low-emissions hydrogen: electrolysis, fossil + CCUS, and some by-product/biomass routes |
| Capacity basis | ~6.9 Mt/year of committed production capacity | ~27 Mt/year announced for 2030; ~4.3 Mt/year committed for 2030 |
| What “green” vs “clean” means | Distinguishes renewable from low-carbon within “clean” | Distinguishes renewable/electrolytic from fossil-with-CCUS within “low-emissions” |
| Best use of the number | Momentum of the sanctioned build-out | Gap between announcements, sanctioned projects and real demand |
The practical rule: the Hydrogen Council number tells you how much has been decided to build; the IEA numbers tell you how much has been sanctioned, how much is announced but unfunded, and how little has been contracted for sale. A “committed project” is not an “operating project,” and an “announced project” is not a “real project” until it has permits, financing, a buyer and a board decision.
Make hydrogen: pick a production route
The same gas, three very different supply chains and carbon footprints.
Choose a way to make hydrogen
The question this answers: which route avoids fossil fuel as the production feedstock? Only the green / renewable pathway does. Blue hydrogen still starts from natural gas and captures part of the resulting CO₂. Grey hydrogen starts from natural gas or coal and releases the CO₂. “Other low-carbon” routes — such as electrolysis on a low-carbon grid, or by-product hydrogen — can also qualify as low-emissions depending on the standard applied.
How does an electrolyser make hydrogen?
One reaction, run backwards from a fuel cell.
An electrolyser passes a direct current through water between two electrodes. Hydrogen gas collects at one electrode, oxygen at the other, usually with a membrane or diaphragm keeping the two gases apart. That is the whole principle. A fuel cell does the opposite job: it combines hydrogen and oxygen to produce electricity and water. An electrolyser turns electricity into hydrogen; a fuel cell turns hydrogen back into electricity. Every time you convert between them you lose energy.
Four electrolyser families are in use or development, and none is a universal winner:
| Technology | Maturity | Operating notes | Typical strengths | Key challenges |
|---|---|---|---|---|
| Alkaline (AWE) | Mature, decades of industrial use | Runs best at steady load; liquid alkaline electrolyte | Lowest capital cost; no scarce metals; China-dominated supply | Slower to ramp; less suited to highly variable renewables |
| PEM | Commercial, scaling | Handles variable power and fast ramping | Compact; pairs well with wind and solar; high-purity output | Uses iridium/platinum; higher cost per kW |
| SOEC (solid oxide) | Early commercial / demonstration | Runs hot (700–850°C); can use waste heat and steam | Highest electrical efficiency; can co-electrolyse CO₂ | Thermal cycling degrades stacks; needs a heat source |
| AEM (anion exchange) | Pilot / pre-commercial | Aims to combine PEM flexibility with alkaline materials | Potential low cost without precious metals | Membrane durability and lifetime not yet proven at scale |
Grey vs blue vs green hydrogen
Colour is shorthand for the production route, not a complete carbon score.
| Grey | Blue | Green / renewable | |
|---|---|---|---|
| Feedstock | Natural gas or coal | Natural gas (sometimes coal) | Water |
| Energy input | The fossil fuel itself | Fossil fuel + energy for capture | Renewable electricity |
| CO₂ handling | Vented to atmosphere | Part captured and stored (typ. 60–95% of process CO₂) | No process CO₂; emissions depend on the electricity |
| Main emissions risk | Full lifecycle CO₂ | Upstream methane leakage; uncaptured CO₂; capture-energy emissions | Grid electricity if not genuinely additional renewable power |
| Share of today’s supply | Dominant (~most of >100 Mt) | Small but established | ~1% and rising |
| Indicative 2026 cost | ~$1–2.5/kg | ~$2–3.5/kg | ~$4.5–9/kg (region-dependent; below $2/kg only in the cheapest markets) |
Blue hydrogen should not be called “clean” by default. Its footprint depends on how much CO₂ the plant actually captures, how much methane leaks upstream, how the capture energy is supplied, and where the system boundary is drawn. “Low-emissions” or “low-carbon” is the accurate label, and only when the lifecycle number is verified.
Green hydrogen timeline: 1800–2026
From the first electrolysis experiment to the 2026 reality check. Newest first.
$130 billion committed — and a shrinking 2030 pipeline
What happened: On 10 September 2026 the Hydrogen Council’s Global Hydrogen Compass 2026 reported committed clean-hydrogen investment above $130 billion, 570-plus projects and about 6.9 Mt/year of committed capacity, with roughly 90% operational or under construction and operational capacity having nearly doubled in a year. In parallel, IEA analysis showed the announced 2030 low-emissions production pipeline down to about 27 Mt/year and committed 2030 production at about 4.3 Mt/year.
Why it matters: The two numbers together define the moment — real momentum in sanctioned projects, sitting alongside a large announced pipeline that keeps slipping for lack of firm demand.
The cancellation wave
What happened: ArcelorMittal shelved plans to convert two German steel plants to hydrogen-based production in June 2025 despite about €1.3 billion of offered subsidy. Fortescue cut back its 15-Mt-by-2030 green hydrogen ambition and cancelled projects in Arizona and Gladstone. Repsol, Iberdrola, Woodside and others scaled back, and Queensland withdrew funding for a large liquefied-hydrogen plant. Trade trackers logged dozens of cancelled or postponed projects.
Why it matters: The common thread was not engineering failure. It was cost, missing buyers, power prices, financing and policy uncertainty.
US narrows the 45V hydrogen tax credit
What happened: The One Big Beautiful Bill Act, signed 4 July 2025, moved the deadline for the Section 45V clean-hydrogen production tax credit: projects must now begin construction before 1 January 2028 to qualify, about five years earlier than the original Inflation Reduction Act sunset. The credit itself — up to $3/kg, sliding with emissions intensity, for 10 years after a plant enters service — was retained.
Why it matters: 45V is the main economic lever for US green hydrogen. A tighter window forces developers to sanction quickly or lose the subsidy.
India approves the National Green Hydrogen Mission; world’s largest green-H₂ plant starts in China
What happened: In January 2023 India’s Union Cabinet approved the National Green Hydrogen Mission with a ₹19,744 crore outlay and a target of at least 5 million tonnes a year of green-hydrogen production capacity by 2030. Later in 2023, Sinopec’s 260 MW Kuqa project in Xinjiang — about 20,000 tonnes a year, feeding a refinery — started up as the largest green-hydrogen facility yet built.
Why it matters: Two of the defining 2020s hydrogen stories — India’s policy bet and China’s build-out — take concrete form in the same year.
The US IRA and REPowerEU turn hydrogen into industrial policy
What happened: The US Inflation Reduction Act (August 2022) created the 45V production tax credit of up to $3/kg. The EU’s REPowerEU plan set a goal of 10 Mt of domestic renewable-hydrogen production and 10 Mt of imports by 2030, and the European Hydrogen Bank was launched to auction production subsidies.
Why it matters: Government money moved from research grants to per-kilogram production support — the first serious attempt to close the cost gap with grey hydrogen.
The announcement boom
What happened: The EU published its Hydrogen Strategy in July 2020, targeting 40 GW of renewable-hydrogen electrolysers by 2030. Dozens of countries followed with national hydrogen strategies. India announced a National Hydrogen Mission in its August 2021 Independence Day address. Export projects were unveiled across Australia, the Middle East, Chile and North Africa.
Why it matters: This is the moment announced pipelines ballooned — and where the gap between announced and sanctioned capacity opened up.
The Hydrogen Council forms at Davos
What happened: Thirteen energy, transport and industrial companies launched the Hydrogen Council at the World Economic Forum to promote hydrogen in the energy transition. Membership later grew to well over 100 companies.
Why it matters: It marks the point where large incumbents — carmakers, oil majors, gas firms, industrial-gas companies — publicly committed to a hydrogen strategy.
The Paris Agreement changes the question
What happened: Nearly 200 countries agreed to hold warming well below 2°C. Attention shifted to “hard-to-abate” sectors — steel, cement, chemicals, shipping, aviation — where electrification is difficult.
Why it matters: The hydrogen question moved from “can it power a car?” to “can low-emissions hydrogen decarbonise industries that electricity cannot easily reach?”
The fuel-cell car era
What happened: Carmakers and governments invested heavily in hydrogen fuel-cell vehicles and refuelling stations. The US launched a hydrogen fuel initiative; Japan, Korea and California built early station networks. Toyota, Hyundai and Honda released fuel-cell models.
Why it matters: Battery-electric vehicles ultimately captured the overwhelming majority of passenger-car deployment. Hydrogen persists in selected heavy-duty, fleet and regional applications rather than as a mass car fuel.
“The hydrogen economy” is named
What happened: Electrochemist John Bockris used the phrase “hydrogen economy” around 1970 to describe an energy system in which hydrogen, made from non-fossil electricity, would carry energy across sectors. The 1970s oil shocks pushed the idea into wider circulation, and the International Association for Hydrogen Energy was founded in 1974.
Why it matters: The vision of hydrogen as a universal fuel dates from here — and so does the recurring gap between that vision and its economics.
Fuel cells go to space
What happened: NASA used hydrogen–oxygen fuel cells for onboard power on the Gemini and Apollo missions, with drinking water as a useful by-product. This proved fuel cells could work reliably outside the laboratory.
Why it matters: Space programmes kept fuel-cell engineering alive for decades when there was no commercial market.
Industrial-scale electrolysis on hydropower
What happened: Norsk Hydro operated large alkaline water electrolysers powered by hydroelectricity in Norway to produce hydrogen for ammonia and fertiliser, at outputs measured in tens of thousands of cubic metres per hour.
Why it matters: Renewable-powered electrolysis at industrial scale is not new — it ran for decades wherever cheap hydropower existed, until cheap natural gas displaced it from the 1960s.
Haber–Bosch makes hydrogen an industrial feedstock
What happened: Fritz Haber and Carl Bosch developed and industrialised the synthesis of ammonia from hydrogen and nitrogen. Ammonia became the basis of synthetic fertiliser, and hydrogen became a bulk industrial chemical.
Why it matters: Modern hydrogen demand grew because industry needed the molecule — for fertiliser, then refining and chemicals — long before climate policy existed. That existing demand is the most obvious place to use low-emissions hydrogen.
Grove builds the first fuel cell
What happened: William Robert Grove demonstrated a “gas voltaic battery” that combined hydrogen and oxygen to generate an electric current — electrolysis run in reverse. Christian Friedrich Schönbein had described the underlying effect the year before.
Why it matters: Grove’s device is the ancestor of every fuel cell since, from Apollo to today’s hydrogen trucks and buses.
Water is split with an electric current
What happened: Within weeks of Alessantro Volta announcing the electric pile, William Nicholson and Anthony Carlisle passed its current through water and observed hydrogen and oxygen forming at the two wires. Johann Wilhelm Ritter reported similar results in Germany the same year.
Why it matters: This is the founding experiment of green hydrogen — the exact reaction a modern gigawatt electrolyser runs, more than two centuries later.
What is hydrogen actually used for today?
Above 100 Mt of demand — almost none of it low-emissions.
This distinction is the one most often lost in coverage. The world does not consume 100 million tonnes of green hydrogen. It consumes more than 100 million tonnes of hydrogen in total, and the overwhelming majority is made from unabated natural gas and coal, for three long-established uses:
- Oil refining — removing sulphur from fuels and upgrading heavy crude.
- Ammonia — overwhelmingly for nitrogen fertiliser, plus some industrial and emerging fuel uses.
- Methanol and other chemicals — plastics precursors, solvents and industrial feedstocks.
Low-emissions hydrogen demand — hydrogen that is actually renewable or low-carbon and used as such — was only about 1 million tonnes in 2025, though growing around 20% a year. The single most bankable near-term opportunity is not inventing new demand; it is replacing the fossil hydrogen already consumed in refineries and ammonia plants with a low-emissions version.
The hydrogen project funnel
Why “570 projects” and “27 Mt announced” describe very different things.
- Announced — a press release, an MoU, a feasibility study. The largest bucket. ~27 Mt/year of low-emissions production is aimed at 2030 on paper.
- Feasibility & engineering — site, grid connection, water, front-end design.
- Offtake — a buyer commits to purchase the output. Most projects stall here.
- Permits — environmental, planning, grid and safety approvals.
- Financing — debt and equity raised, usually contingent on firm offtake.
- Final Investment Decision (FID) — the board sanctions the spend. ~4.3 Mt/year of 2030 production has reached this stage.
- Construction — roughly 90% of the Hydrogen Council’s “committed” 6.9 Mt/year is here or operating.
- Operating — producing hydrogen. Low-emissions output is around 1 Mt/year today.
Announcement does not equal FID. FID does not equal construction. Construction does not equal production. Each step filters out projects, and in 2024–26 the filter tightened sharply between “announced” and “FID.”
The $130 billion question: who will buy it?
Supply has momentum. Demand does not.
🏗 Committed supply
- $130B+ committed investment
- 570+ projects
- ~6.9 Mt/year committed capacity
- ~90% operating or under construction
🏭 Demand that has actually been contracted
- ~1 Mt/year low-emissions demand today
- ~1.7 Mt/year new offtake signed in 2025
- ~one-fifth of that is firm
- ~6 Mt/year could be supported by existing policies by 2030
The Hydrogen Council estimates that around 6 Mt/year of low-emissions hydrogen demand could be supported by policies already in force by 2030, potentially roughly doubling if governments fully implement the measures they have announced. Even the optimistic figure is a fraction of what has been announced on the supply side. This is why the industry describes 2026 as its “spreadsheet phase”: the constraint is no longer whether a molecule can be produced, but whether a customer will sign a contract to buy it at a price that covers its cost.
Offtake and FID: the two words that decide projects
What is a hydrogen offtake agreement?
An offtake agreement is a commitment by a buyer to purchase future hydrogen, or a hydrogen-derived product such as ammonia or methanol, usually over many years. It is what gives a project revenue visibility. With a firm, long-term offtake contract, a developer can raise debt and take a final investment decision. Without one, banks hesitate, and the project does not get built — no matter how good the engineering.
With firm offtake
- Buyer signs binding contract
- Revenue is predictable
- Banks lend
- Board takes FID
- Construction starts
With only an MoU
- Buyer signs letter of intent
- Revenue is uncertain
- Banks wait
- FID slips
- Project stalls or is cancelled
Why the offtake numbers look weak
New offtake agreements for low-emissions hydrogen totalled about 1.7 Mt/year in 2025, unchanged from 2024. But only around one-fifth of newly signed volumes were firm contractual commitments; the rest were memoranda of understanding and preliminary agreements. A memorandum is a statement of interest. A firm offtake contract is a bankable obligation. Coverage that treats the two as equivalent overstates how much demand has actually been secured.
What does FID mean?
A final investment decision is the point at which a company’s board formally commits the capital to build a project. Before FID, spending is on studies and options. After FID, the project is “sanctioned” and shows up in the committed-investment totals. During 2025, only about 0.3 Mt/year of additional low-emissions production capacity reached FID — a sign of how cautious investors had become.
Hydrogen reality check: projects delayed or cancelled
Illustrative cases from 2024–26. Status and reasons are as reported by the developers.
| Project / plan | Country | Developer | Type | Status as reported | Stated reason |
|---|---|---|---|---|---|
| Bremen & Eisenhüttenstadt hydrogen DRI conversion | Germany | ArcelorMittal | Green steel | Shelved (June 2025) | Cost and competitiveness; hydrogen supply and power prices |
| Arizona Hydrogen (80 MW) & PEM50 Gladstone (50 MW) | US / Australia | Fortescue | Green hydrogen | Cancelled (2025) | Cost; renewable-power requirements; strategic reprioritisation |
| 15 Mt/year green hydrogen ambition | Australia / global | Fortescue | Green hydrogen | Target dropped | Economics and scale of renewables needed |
| Large liquefied-hydrogen plant | Australia (Queensland) | State-backed project | Green hydrogen / export | Funding withdrawn | Government funding pulled on cost grounds |
| Spanish electrolyser projects (~350 MW frozen) | Spain | Repsol | Green hydrogen | Frozen / delayed | Fiscal and policy uncertainty |
| 2030 green-hydrogen production target cut ~two-thirds | Spain / global | Iberdrola | Green hydrogen | Scaled back | Funding delays; weak demand |
| Green hydrogen projects in Australia and New Zealand | Australia / NZ | Woodside | Green hydrogen | Shelved | Commercial viability |
Trade trackers recorded on the order of 60 major green-hydrogen projects cancelled during 2025 alone, representing several million tonnes a year of would-be capacity. Cancellations do not mean hydrogen “failed” — they mean the market is sorting projects with real demand and cheap power from projects that had neither.
Where should we use hydrogen?
Pick an application and see how the evidence stacks up.
Select a use case
The underlying principle: hydrogen is likely to matter most where a molecule is genuinely needed as feedstock, or where direct electrification is difficult, expensive or impractical. It is not likely to “replace fossil fuels everywhere,” and in most cases where you can simply use electricity directly, that will win on cost and efficiency.
Hydrogen vs direct electrification
Why “just use the electricity” is often the right answer.
Direct electrification
- Renewable electricity
- Battery or grid
- Electric motor / heat pump
- End use
Hydrogen pathway
- Renewable electricity
- Electrolyser
- Compression / liquefaction
- Transport & storage
- Fuel cell or combustion
- End use
Every conversion step loses energy. Turning electricity into hydrogen, moving it, storing it and turning it back into useful work can deliver only a minority of the original electricity to the wheel or the furnace, where a battery-electric path delivers most of it. That is why, for a passenger car, the battery route is far more efficient than making hydrogen, compressing it, trucking it and running it through a fuel cell. Hydrogen earns its place when the alternative is not “use electricity directly” but “keep burning fossil fuel,” or when you need the molecule itself.
Why green hydrogen is expensive: the cost stack
Move the sliders. This is a directional teaching model, not a project quote.
Green hydrogen cost estimator
Assumes a system electricity use of about 50 kWh per kg of hydrogen and a 20-year plant life. Output is indicative only.
The model captures electricity, annualised electrolyser capital, fixed operating cost and a flat allowance for compression, water and stack replacement. It ignores storage, long-distance transport, conversion to ammonia, and any subsidy. Real projects vary widely; treat the number as a direction, not a decimal.
Two things dominate the stack: the price of electricity and how many hours a year the electrolyser runs (its capacity factor). Cheap power used at a low capacity factor still produces expensive hydrogen, because the electrolyser capital is spread over fewer kilograms. This is why the cheapest green hydrogen tends to come from places with abundant, cheap renewables and a way to run the electrolyser most of the year — and why a single “global green hydrogen price” is misleading.
The global hydrogen race: 2026
Same technology, very different economics and strategies.
| Region | Position in 2026 | Main use / strategy | Main challenge |
|---|---|---|---|
| China | More than half of committed renewable-H₂ capacity; ~three-quarters of new electrolyser installs | Domestic ammonia, methanol, refining; low-cost equipment; export of electrolysers | Some flagship plants under-producing; grid and curtailment issues |
| Europe | Second in investment; binding RFNBO industrial target of 42.5% by 2030 | Steel, chemicals, refining, shipping fuels; European Hydrogen Bank subsidy auctions; imports | High power and capital costs; several steel and export projects cancelled |
| United States | Leads low-carbon (CCS-based) hydrogen deployment | Refining, ammonia, hubs; 45V production tax credit | 45V construction-start deadline pulled forward to end-2027; policy volatility |
| India | Large 2030 target, early awards, minimal operating capacity | Refinery and fertiliser demand; green ammonia exports; domestic electrolyser manufacturing | Under 1% of the 2030 goal commissioned; cost gap vs grey hydrogen |
| Middle East | Large export-oriented projects in Saudi Arabia, Oman, UAE | Green ammonia for export using cheap solar and land, plus ports | Conversion and shipping costs; securing overseas offtake |
| Australia | Huge renewable potential; many announcements, high cancellation rate | Export ammonia and domestic industry | Cost, water, transmission; several flagship projects cancelled or shelved |
Why is China so important?
China accounts for more than half of the world’s committed renewable-hydrogen capacity and roughly three-quarters of new electrolyser installations. Its advantages compound: the largest electrolyser manufacturing base (dominated by alkaline systems), equipment that can cost a fraction of comparable Western units, large existing industrial demand for hydrogen in ammonia and methanol, a vast renewable build-out, and policy support at provincial and national level. The Sinopec Kuqa project in Xinjiang — 260 MW, feeding a refinery — became the largest green-hydrogen plant when it started in 2023, though it reportedly produced only about a third of its expected output in its first year and did not reach full load until 2026. China’s cost structure does not automatically transfer to Europe, the US or India, where power, capital and equipment all cost more.
Europe
Europe is second in committed investment and has the most demand-side regulation: under the Renewable Energy Directive, industry must source 42.5% of its hydrogen as renewable fuels of non-biological origin (RFNBOs) by 2030. The European Hydrogen Bank runs subsidy auctions — its third auction awarded about €1.09 billion to nine projects covering roughly 1.1 GW of electrolysers and 1.3 Mt of hydrogen over ten years — and a fourth auction is planned for the end of 2026. Yet Europe is also where the highest-profile cancellations have landed, including ArcelorMittal’s German green-steel conversion, because power and capital costs are high and few buyers will pay the premium without a mandate forcing them to.
United States
The US leads deployment of low-carbon (CCS-based) hydrogen and built its policy around the Section 45V production tax credit of up to $3/kg. The One Big Beautiful Bill Act of July 2025 kept the credit but moved the construction-start deadline forward to 1 January 2028. US hydrogen policy has changed quickly and more than once, so any project economics that depend on federal support should be checked against the current rules rather than 2023–24 assumptions.
India’s green hydrogen bet
India’s National Green Hydrogen Mission, approved by the Union Cabinet in January 2023, has an outlay of about ₹19,744 crore and a target of at least 5 million tonnes a year of green-hydrogen production capacity by 2030, alongside a large associated renewable-energy build-out and an aim to serve domestic industry and export markets. The core incentive programme, SIGHT (Strategic Interventions for Green Hydrogen Transition), has a total outlay of about ₹17,490 crore, split between domestic electrolyser manufacturing and green-hydrogen production support.
India 2026 status
- Electrolyser manufacturing: about 3,000 MW/year of capacity awarded to around 15 companies.
- Green-hydrogen production: roughly 8,62,000 tonnes/year of capacity awarded across tranches to around 18 companies.
- Green ammonia: SECI has discovered prices for about 7,24,000 tonnes/year of green-ammonia supply to fertiliser units.
- Refineries: about 20,000–30,000 tonnes/year of green hydrogen awarded for supply to public-sector refineries (IOCL, BPCL, HPCL).
- Operating reality: as of early 2026, commissioned green-hydrogen capacity was on the order of 8,000 tonnes/year — under 1% of the 2030 target — and MNRE reported roughly ₹292 crore of mission funds spent so far.
The pattern mirrors the global one: India has committed on paper and awarded incentives early, but firm offtake, financing and commissioned plants are the bottleneck. Refinery and fertiliser demand — replacing existing grey hydrogen — is the most credible near-term market, which is why the first tenders target exactly those users.
Infrastructure, water and carbon accounting
Hydrogen cannot simply reuse all fossil infrastructure
Hydrogen is the smallest molecule, it can embrittle some steels, and it carries about a third of the energy of natural gas per unit volume. Moving it means new or modified pipelines, higher compression, dedicated storage, and — for long distances — either liquefaction at very low temperature or conversion to ammonia and back. The IEA has noted that announced hydrogen pipeline projects run to tens of thousands of kilometres by the mid-2030s, but only a small share is operational or backed by committed investment; operational and committed pipeline length grew about 70% over the past year from a low base.
Does green hydrogen use too much water?
The chemistry needs roughly 9 kg of water per kg of hydrogen as feedstock; real plants use somewhat more for cooling and purification, often cited around 20–25 kg per kg. That is small next to power-plant cooling or irrigation nationally, but it is a genuine local issue in water-stressed regions, where projects may need desalination or treated wastewater. Location matters more than the global average.
Colour is not a complete carbon footprint
Green hydrogen is only as clean as the electricity behind it — drawing from a fossil-heavy grid without genuinely additional renewables can make “green” hydrogen higher-emitting than grey. Blue hydrogen’s footprint depends on capture rate, upstream methane leakage, the energy used for capture, and the system boundary. This is why regulators increasingly certify hydrogen by measured lifecycle emissions per kilogram rather than by colour.
2030: boom or reality check?
Three scenarios, not a forecast.
- Announced 2030 low-emissions production: ~27 Mt/year (IEA) — and falling
- Committed (post-FID) for 2030: ~4.3 Mt/year, up to ~6 Mt if strong-potential projects take FID in 2026–27
- Operating today: ~1 Mt/year
Rapid scale
Cheaper electricity and electrolysers, firm mandates, real offtake and infrastructure combine. Deployment accelerates toward the higher end of committed capacity and beyond.
Selective success
Hydrogen concentrates where it clearly works: refining, fertiliser, some steel, shipping fuels, industrial clusters in low-cost regions. Growth is real but narrower than the 2020 vision.
Project shakeout
High cost, weak demand and policy uncertainty persist. Most announcements are cancelled; surviving projects cluster in the cheapest regions and the strongest use cases.
Demand decides
In every scenario, the binding variable is not electrolyser capacity. It is how much low-emissions hydrogen customers are contracted to buy, and at what price.
Fun facts & context
- The reaction inside a modern gigawatt electrolyser is the same one Nicholson and Carlisle ran with two wires in 1800.
- Renewable-powered electrolysis is not new — Norway ran hydro-powered electrolysers for fertiliser from the 1920s until fossil gas undercut them.
- Hydrogen is invisible and colourless; the “colours” refer only to how it is made.
- Ammonia made from hydrogen underpins the fertiliser that feeds a large share of the world’s population.
- A fuel cell is an electrolyser run backwards — both were invented in the same decade-scale window in the 1830s.
- The cheapest green hydrogen in 2026 can dip below $2/kg in a few favourable markets, but $4.5–9/kg is closer to the global norm.
Explore More Timelines
How we track hydrogen projects
Methodology
This article separates four things that are often merged. Announced capacity is any publicly stated intention, including MoUs and feasibility studies — the largest and least reliable bucket. Committed / post-FID capacity has a board decision and usually financing behind it. Under construction means physical build has started; operational means it is producing. We distinguish renewable/green hydrogen (electrolysis on renewable power) from the broader low-emissions category (which can include fossil hydrogen with carbon capture if verified lifecycle emissions are low). We distinguish hydrogen from hydrogen-derived products such as ammonia and methanol, and we report annual production capacity (tonnes per year), not one-off volumes. We treat a memorandum of understanding as non-binding and only count firm offtake separately. Different databases produce different totals mainly because they draw these lines in different places and use different project-size thresholds — so figures from the Hydrogen Council and the IEA are reported side by side, not added together. All figures are dated and attributed; where sources disagree, we give a range.