Green Hydrogen

Green hydrogen is hydrogen made by splitting water with renewable electricity. Production emits no CO2, and the product can replace fossil hydrogen in fertiliser and refining, reduce iron ore without coal, and feed synthetic fuel plants. The world already uses close to 100 million tonnes of hydrogen a year (about 97 Mt in 2023, according to the IEA), but less than 1% of it is low-emission. Nearly all of it comes from natural gas and coal and releases on the order of 900 Mt of CO2 a year, more than aviation. The first job for green hydrogen is to clean up that existing demand; after that come sectors electricity can’t reach directly.

How it works

An electrolyser runs direct current through water between two electrodes separated by an electrolyte or membrane. The water splits into hydrogen at the cathode and oxygen at the anode:

2 H2O + electrical energy → 2 H2 + O2

Thermodynamics sets the theoretical minimum: 39.4 kWh per kilogram on a higher-heating-value basis, or 33.3 kWh/kg on the lower heating value (LHV) basis normally used for fuels (see calorific value). Real systems need more. There are overpotentials at the electrodes and ohmic resistance, plus balance-of-plant loads: rectifiers, pumps, cooling, gas drying and water purification. Commercial plants today use roughly 50–55 kWh per kg at the system level.

At 52 kWh/kg, efficiency is about 64% LHV. The last term of the cost formula shows why electricity price dominates: at $40/MWh, the power alone costs about $2.1 per kg before any capital is repaid. Our green hydrogen calculator runs this calculation for a specific electrolyser, power source and capacity factor.

Main electrolyser types

Alkaline (AEL)

Alkaline electrolysis is the oldest and cheapest technology and has been used in industry for a century. Nickel-based electrodes sit in a potassium hydroxide solution, separated by a porous diaphragm. The stacks are durable and need no precious metals, but they have traditionally responded more slowly to load changes and run at lower current density. Chinese manufacturers dominate global alkaline output and sell at the lowest prices.

Proton exchange membrane (PEM)

PEM electrolysers use a solid polymer membrane under acidic conditions. That allows compact stacks, high current densities, fast ramping and output at elevated pressure, all useful when pairing with variable wind and solar. They depend on iridium (anode) and platinum (cathode). Only a few tonnes of iridium are produced each year, so loadings have to fall sharply before PEM can scale to many gigawatts.

Solid oxide (SOEC)

Solid oxide electrolysers run at 700–850 °C and split steam using a ceramic electrolyte. Because part of the energy comes in as heat, electricity use can drop to about 40 kWh/kg when waste heat from industry or a nuclear plant is available. The stacks degrade faster and handle cycling less well. The technology is moving from demonstration to early commercial scale.

Anion exchange membrane (AEM)

AEM tries to pair the cheap materials of alkaline systems with the membrane design of PEM. Units are still small (kilowatt to low-megawatt), and durability has yet to be proven.

Ranges based on IEA (2019) The Future of Hydrogen, IRENA (2020) and IEA Global Hydrogen Review 2024.

The colours of hydrogen

Industry shorthand sorts hydrogen by how it is made: grey (steam methane reforming without capture, most of today’s supply), brown/black (coal gasification, common in China), blue (gas reforming with carbon capture), green (renewable electrolysis), pink (nuclear electrolysis), turquoise (methane pyrolysis, which leaves solid carbon) and white (natural geological hydrogen). Regulators are dropping colours in favour of lifecycle thresholds. The EU’s rules for renewable fuels of non-biological origin require emissions below 3.38 kg CO2e per kg H2 (a 70% saving against the fossil comparator), and the US 45V credit is tiered by lifecycle intensity below 4 kg CO2e/kg.

Efficiency and capacity factor

Two efficiencies matter. Conversion efficiency, 60–67% LHV for today’s systems, sets how much electricity each kilogram takes. Utilisation, meaning the electrolyser’s capacity factor, sets how many kilograms each dollar of capital produces. An electrolyser running on solar alone at a 20–25% capacity factor spreads its capital over few hours. Hybrid wind-solar sites in Chile, Australia or the Middle East can reach 50–70%, and grid-connected plants more, but grid power brings cost and carbon problems of its own.

Converting electricity to hydrogen and back through a turbine or fuel cell returns only about 30–40% of the original electricity. So hydrogen is treated as a last-resort seasonal storage medium, not a rival to battery storage for daily cycling.

Costs

The levelized cost of hydrogen (LCOH) depends on electricity price, electrolyser capital cost, utilisation and financing. According to the IEA Global Hydrogen Review 2024:

• Grey hydrogen usually costs about $1–2.5/kg at normal gas prices (more during the 2022 gas crisis, under $1/kg where gas is cheap, such as the Middle East).
• Green hydrogen from new projects generally costs $3–8/kg. The low end is at excellent renewable sites with cheap financing; the high end is in Europe and Japan.
• Installed electrolyser system costs run from roughly $750/kW in China to $2,000–2,500/kW or more in Europe and North America. Balance-of-plant, engineering and inflation have kept them from falling as fast as earlier projections expected.

To close the gap you need electricity well below $30/MWh, high utilisation and cheaper electrolysers, or else carbon pricing and subsidies. The US production tax credit (45V) pays up to $3/kg for the cleanest hydrogen, though 2025 legislation limited it to projects starting construction before 2028. In its first round, the EU’s Hydrogen Bank auction cleared at premiums of roughly EUR 0.4–0.5/kg. You can check the electricity component for your planned power source with the LCOE calculator.

Uses and the hydrogen ladder

Oil refining (hydrocracking and desulphurisation, roughly 40% of demand) and ammonia for fertilisers (around a third) account for most of today’s demand, followed by methanol and direct reduced iron (DRI) for steel. Swapping out grey hydrogen in these uses is the obvious early market, since the infrastructure and the buyers are already there.

Michael Liebreich’s widely cited Hydrogen Ladder ranks new uses by how well clean hydrogen competes with the alternatives:

• Most compelling: ammonia and fertiliser, refining, methanol, green steel via DRI (projects such as HYBRIT and Stegra in Sweden), long-duration grid storage and e-fuels for shipping and aviation, including ammonia as a marine fuel.
• Contested: long-haul trucks, high-temperature industrial heat, seasonal power storage.
• Least compelling: passenger cars, home heating and light-duty vehicles. Here heat pumps and battery vehicles use energy two to four times more efficiently than the hydrogen route.

Storage and transport

Hydrogen is the lightest element. It carries a lot of energy per kilogram but little per volume: about 3 kWh per normal cubic metre, roughly a third of natural gas. The options are compression to 350–700 bar for vehicles; liquefaction at −253 °C, which uses roughly 10–13 kWh/kg (about 30–40% of its LHV); salt cavern storage, the cheapest bulk option where the geology allows (used for decades in Texas and Teesside); conversion to ammonia (17.6 wt% hydrogen, easy to liquefy) or liquid organic hydrogen carriers for shipping; and pipelines, new or converted from natural gas. Over land, pipelines are by far the cheapest way to move large volumes. Shipping it as ammonia adds large conversion losses.

Land, water and materials

The electrolyser itself is compact: a 100 MW plant fits on a few hectares. The land goes to the renewable plants feeding it. Producing 1 Mt of hydrogen a year takes around 50–55 TWh of electricity, roughly the output of 20–25 GW of solar in a sunny region. Water use is about 9 kg per kg H2 in theory and 10–15 litres per kg in practice, which is modest. In arid export hubs, seawater desalination adds well under 1% to the cost of hydrogen. The critical materials are iridium and platinum for PEM, nickel for alkaline, and rare earths and zirconia for SOEC.

Lifecycle CO2 emissions

Most of green hydrogen’s lifecycle emissions come from building the renewable plant and the electrolyser. With wind or solar power at 10–40 g CO2e/kWh (IPCC AR5 WGIII Annex III; UNECE 2021) and 52 kWh/kg, lifecycle intensity works out at roughly 0.5–2 kg CO2e per kg H2. For comparison:

• Grey hydrogen from steam methane reforming emits about 9–12 kg CO2e/kg including upstream methane emissions.
• Blue hydrogen ranges from about 1.5 to over 6 kg CO2e/kg, depending on the capture rate and upstream methane leakage.
• Grid electrolysis using average electricity at 400 g CO2e/kWh produces around 20 kg CO2e/kg, which is worse than grey hydrogen. That is why EU rules require additionality and matching with renewable supply in both time and location.

Leaks matter too. Sand et al. (2023) estimate hydrogen’s global warming potential at about 12 over 100 years and about 37 over 20 years, through indirect effects on methane, ozone and stratospheric water vapour. Leakage of a few percent across a future supply chain would wipe out part of the benefit.

Safety and environmental impacts

Hydrogen burns across a wide range of concentrations (4–75% in air), ignites with very little energy and has a nearly invisible flame. In open air it rises and disperses quickly, which lowers some risks, but indoors it collects under roofs. It makes some steels brittle, a concern for converted pipelines. Ammonia carriers bring their own toxicity hazards, and burning ammonia emits NOx and N2O. Industry handles these risks well today; the hard part is doing so in new, more scattered settings.

Outlook

Ambition far outruns investment. According to the IEA Global Hydrogen Review 2024, announced projects could deliver close to 50 Mt a year of low-emission hydrogen by 2030, yet only a few percent of that volume had reached final investment decision. Global installed electrolysis capacity was around 1.4 GW at the end of 2023, and the IEA expected it could reach about 5 GW by the end of 2024. Electrolysis capacity at FID roughly doubled in a year to about 20 GW, with China taking the largest share. Flagship projects include NEOM in Saudi Arabia (around 2 GW of electrolysis for ammonia export) and large alkaline projects in Xinjiang and Inner Mongolia.

Growth will most likely be concentrated: existing ammonia and refinery demand, green steel and shipping fuels, at sites where wind and solar are exceptionally cheap and, ideally, pipelines or ports already exist. The things to watch are electrolyser costs, offtake contracts and certification rules. Regions with strong sun and wind show up on the solar potential map and the wind potential map, and the technology comparison sets hydrogen against other options.