Green hydrogen is made by splitting water into hydrogen and oxygen using renewable electricity. For industrial users, the choice of electrolysis technology is not a minor detail. It shapes how much power you need, how quickly the plant can follow variable wind and solar output, and the cost per kilogramme of hydrogen. Three routes dominate: alkaline, PEM (proton exchange membrane) and solid oxide. Each has a distinct character.

Alkaline electrolysis: proven and robust

Alkaline electrolysis is the most mature technology. It uses a liquid electrolyte, usually potassium hydroxide, and a diaphragm. Cells run at 60 to 80°C. Efficiency is typically 60 to 70% on a lower heating value (LHV) basis, so you need around 50 to 55 kWh per kilogramme of hydrogen.

Strengths? Durability, with stack lifetimes often beyond 60,000 hours. Scalability, from tens of megawatts to hundreds when combined. And the lowest capital cost of the three. That makes alkaline attractive for steady industrial loads such as ammonia production or refining.

Trade-offs exist. Lower current density means a larger footprint. Response to power changes is slower, so pairing directly with an off-grid solar farm may need battery storage. The corrosive electrolyte also demands careful maintenance.

PEM electrolysis: responsive and compact

PEM uses a solid polymer membrane. Higher current densities allow compact stacks. Efficiency is similar to alkaline, around 60 to 70% LHV, but PEM can ramp up and down in seconds. That flexibility suits variable renewable supply or grid balancing services.

The membrane produces very pure hydrogen, ideal for fuel cells or electronics. Operating pressures of 30 bar or more can cut downstream compression costs.

Cost is the catch. PEM stacks rely on platinum, iridium and titanium. Capital costs are typically 1.5 to 2 times alkaline. Stack lifetimes are improving but often shorter. Choose PEM when you need fast response, a small footprint, or high purity. It is less attractive for continuous, lowest-cost baseload hydrogen.

Solid oxide electrolysis: high-temperature efficiency

Solid oxide electrolysis cells (SOECs) run at 700 to 850°C. Part of the energy comes from heat rather than electricity, so electrical efficiency can reach 75 to 85% LHV. You might need only 40 to 45 kWh per kilogramme.

This appeals to sites with high-grade waste heat, such as steel, glass or cement plants. SOECs can also co-electrolyse water and carbon dioxide to make syngas, a useful chemical feedstock.

But SOECs are the least mature. Thermal cycling can crack ceramic cells. Start-up takes hours, not seconds. They suit steady, continuous operation. Materials and sealing challenges limit stack lifetimes, and commercial deployments remain small. Costs are high, though falling.

Comparing efficiency, cost and scalability

Side by side: alkaline offers 60-70% LHV, 50-55 kWh/kg, low capital cost, high scalability, slow response. PEM offers 60-70% LHV, 50-55 kWh/kg, medium-high capital cost, medium-high scalability, fast response. Solid oxide offers 75-85% LHV, 40-45 kWh/kg, high capital cost, low-medium scalability, slow but steady response.

Actual performance depends on temperature, pressure, current density and heat recovery. For a 10 MW plant running 8,000 hours yearly, the gap between 50 and 45 kWh/kg is 400,000 kWh. At industrial prices that matters. But capital cost and maintenance also count. A lower-efficiency alkaline system with a 20-year life may beat a higher-efficiency solid oxide system that needs frequent stack replacements.

Practical selection for industrial sites

Start with your electricity supply. Firm, low-cost renewable power and steady hydrogen? Alkaline is usually safest. Variable power or grid balancing? PEM's fast response justifies the premium. Waste heat above 200°C? Solid oxide deserves a serious feasibility study, especially if you also need syngas.

Consider scale. Under 5 MW, PEM's compactness often wins. Above 20 MW, alkaline's lower capital cost per megawatt is hard to ignore. Solid oxide remains best for pilots or industrial clusters with abundant heat.

Think about purity. PEM delivers 99.99% purity without extra processing. Alkaline needs a deoxo unit. Solid oxide produces hydrogen with steam, easy to condense, but high temperature adds complexity.

No technology wins everywhere. Balance electricity price, load profile, heat availability and risk tolerance. Model the levelised cost of hydrogen. A phased approach — starting with a PEM pilot, then scaling with alkaline for baseload — is how many industrial decarbonisation journeys begin.