Electrolysers: Engineering methods
Alkaline, PEM, solid oxide, and the older chlorine and aluminium cells
Four families split water; chlor-alkali, aluminium smelting and plating are older and, for now, far larger users of electrolytic power.
Comparison
| Method | Efficiency | Scale | Economics | Limitations | Reliability |
|---|---|---|---|---|---|
| Alkaline water electrolysis (AWE)37918 | About 65-70% LHV for large commercial units | 20 MW modules of about 300 cells; GW-scale plants | Chinese systems installed in China 500-1,100 USD/kW (2025) | Bulky; slower to follow fast swings in wind and solar | Longest-lived of the low-temperature types |
| Proton-exchange membrane (PEM)18 | 55 kWh/kg system (61% LHV) in 2022 | MW to hundreds of MW | 1,000 USD/kW uninstalled (2022) | Iridium and titanium supply | 40,000-hour stack life (2022), target 80,000 |
| Solid-oxide electrolysis (SOEC)181925 | About 90% LHV on electricity | Pilot to multi-MW (largest units at refineries) | No settled market price yet | Needs a heat source and steady running; ceramics degrade faster | Little long-term fleet data |
| Anion-exchange membrane (AEM)6 | No benchmark in the cited sources | Small units | Promises to avoid iridium and platinum | Membrane durability | Early commercial |
| Chlor-alkali (membrane cell)1014 | 1,979 kWh per tonne NaOH at 6 kA/m2 | Plants above 800,000 t NaOH a year | Electricity about 70% of chlorine cost in Europe | Output tied to chlorine demand | Mature |
| Aluminium smelting (Hall-Heroult)121617 | 12.1-13.1 DC kWh/kg in modern cells; world average 14.1 MWh/t AC | Pot lines of hundreds of cells at 400-600 kA | Electricity is the largest cash cost | Burns 0.43 t of carbon per tonne, giving direct CO2 | Mature |
| Electroplating and electrowinning31 | Set by the share of current that deposits metal rather than making hydrogen | Workshop baths to large copper and zinc refineries | Small energy cost per part | Toxic baths (cyanide, chromium) | Mature |
For hydrogen, alkaline is cheapest per kW and dominant in China; PEM costs more and needs iridium but suits variable power; solid oxide uses least electricity where heat is available but is still early; AEM is a bet on dropping precious metals. In all of them three numbers rule: kWh per kg (voltage), USD per kW (current density and factory scale) and running hours. Chlor-alkali and aluminium show a mature electrolytic industry: the best aluminium cells still use about twice the theoretical minimum, and plants go where power is cheap.
How each method works
Alkaline water electrolysis (AWE)Nickel-coated steel electrodes in 30-40% potassium hydroxide, separated by a porous diaphragm (Zirfon) that passes ions but not gas. Cheapest and most mature.
Proton-exchange membrane (PEM)A solid acid polymer membrane with an iridium anode and a platinum cathode. Compact, high current density, fast response.
Solid-oxide electrolysis (SOEC)Ceramic cells split steam at 700-900 C, so industrial or nuclear heat can replace part of the electricity.
Anion-exchange membrane (AEM)A solid membrane that carries hydroxide ions in a dilute alkaline liquid, aiming for PEM-style compactness with nickel catalysts.
Chlor-alkali (membrane cell)Brine gives chlorine at a coated titanium anode; sodium ions cross a membrane to a nickel cathode, where water becomes caustic soda and hydrogen.
Aluminium smelting (Hall-Heroult)Alumina dissolved in molten cryolite is split by current from carbon anodes into a pool of liquid aluminium.
Electroplating and electrowinningMetal ions are reduced onto a part or cathode; Faraday's law gives about 1.19 g of copper or 1.10 g of nickel per amp-hour (calculated).
Sources
- Technical Targets for Proton Exchange Membrane Electrolysis US Department of Energy, Hydrogen and Fuel Cell Technologies Office, 2023
- Green H2 electrolyser sector enters consolidation: IEA (reporting IEA Global Hydrogen Review 2026; cost figures not on the IEA's free web pages) Argus Media, 2026
- Water Electrolyzers and Fuel Cells Supply Chain: Supply Chain Deep Dive Assessment US Department of Energy, 2022
- How Can India Indigenise Hydrogen Electrolyser Manufacturing? (Annexure) Council on Energy, Environment and Water (CEEW), 2024
- Critical Iridium Demands arising from future Expansion of Proton Exchange Membrane Electrolysis (Wortmann, Stolten, Heinrichs), preprint arXiv 2509.05357, 2025
- thyssenkrupp nucera General Investor Presentation, May 2026 (company claims) thyssenkrupp nucera, 2026
- The right chlor-alkali electrolysis technology for your needs thyssenkrupp nucera, 2024
- Factsheet: Aluminium smelting with use of low-carbon electricity GIZ, 2025
- Why Europe needs local production of chlorine Euro Chlor, 2026
- AP60 and APXe smelting technology factsheet (company data) Rio Tinto AP Technology, 2015
- History and Recent Developments in Aluminum Smelting in China (Gao, Wang, Shi, Hu), 35th ICSOBA Conference ICSOBA, 2017
- The Role of E-fuels in Decarbonising Transport (revised January 2024) International Energy Agency, 2024
- Bloom Energy Corp Form 10-K for fiscal year 2025 Bloom Energy / US SEC, 2026
- Sunfire 2025: scaling, partnerships and European strength (company news) Sunfire, 2025
- CODATA value: Faraday constant US National Institute of Standards and Technology, 2022