Electrolyser stack · Glossary

An electrolyser stack is the cell stack of an electrolyser. In it, water is split electrochemically into hydrogen and oxygen. The stack consists of many individual cells. In the common bipolar design they are electrically connected in series; other designs connect them differently. At the same time, the cells are supplied hydraulically with water, electrolyte or steam. The stack converts most of the electrical energy into chemically bound energy in the hydrogen. The remainder is released as waste heat, which has to be removed. In high-temperature electrolysis, by contrast, heat is deliberately supplied from outside and replaces part of the electricity demand.

The term therefore does not refer to the entire electrolyser. The complete system also includes pumps, valves, cooling, water treatment, gas separation, drying, transformer and rectifier, sensors, controls, safety systems and, where applicable, a compressor. Together these parts are called the balance of plant. The stack is the electrochemical core of the plant. How much it delivers and how long it lasts, however, depends strongly on how these surrounding components are designed and operated.

Design: from the cell to the stack

The single cell is the smallest electrochemical functional unit. It contains two electrodes with catalysts, a membrane or a separator (diaphragm) to separate the gas compartments, as well as porous transport layers and flow channels. A single cell delivers only just under two volts. Thermodynamically, at least about 1.23 V is required; in operation, cell voltages typically lie roughly between 1.7 and 2.2 V, depending on the technology and current density. Only when many cells are connected in series does a technically usable stack voltage result. The current, by contrast, is determined by the active cell area and the chosen current density, given in amperes per square centimetre. The electrical power drawn is the product of stack voltage and current. One can imagine the stack as a battery made of many thin slices, except that here current flows in and gas comes out.

The stack combines the cells into a mechanically stable, tight and electrically conductive assembly. Bipolar plates, seals, current collectors, tie rods and end plates are therefore no minor matter. They determine the pressure resistance, the contact resistances, the heat removal, the gas purity and how easy the stack is to maintain. Even small leaks or unevenly distributed clamping forces can impair the performance of individual cells. Because the same current flows through all cells, one weak cell can limit the entire stack.

Distinction from the electrolyser and the balance of plant

Stack and electrolyser are often treated as the same thing. This obscures important differences. A manufacturer may offer a very efficient stack while the overall system needs considerably more electricity because of auxiliary units, rectifier losses, cooling, gas purification or compression. An example from a JRC analysis: for a large plant, about 49 kWh per kilogram of hydrogen was stated at stack level and about 55 kWh/kg at system level. Conversely, a stack that performs well in the laboratory may perform worse in an industrial plant if water quality, temperature control, pressure control or load changes are not reliably managed.

The balance of plant is also more than an accessory. It keeps the stack within its permissible operating conditions. This includes water treatment, heat removal, separation of hydrogen and oxygen, monitoring of gas crossover, pressure and temperature, and the connection to the power grid via transformer and rectifier. European funding programmes now explicitly require optimising the components and architecture of the balance of plant so that they promote stack degradation as little as possible.

Key figures and their reference bases

A stack is mainly assessed using rated power in kilowatts or megawatts, current density, cell voltage, efficiency, operating temperature, pressure level, hydrogen purity and degradation rate. In the hydrogen economy, efficiency is usually stated as electricity demand per kilogram of hydrogen. The energy content of hydrogen serves as the benchmark: the lower heating value is 33.3 kWh/kg, the higher heating value 39.4 kWh/kg. An efficiency in percent is therefore meaningful only together with this reference.

For every figure it must be clear whether it applies to the stack or the system. The European Clean Hydrogen Partnership states its reference values, for example, at system level: with alternating-current and tap-water input, including cooling, and with hydrogen according to ISO 14687-2 as output. For 2020 it gives as the state of the art around 50 kWh/kg for alkaline electrolysis and around 55 kWh/kg for PEM and AEM electrolysis. By 2030 all three are to reach 48 kWh/kg. For high-temperature electrolysis it gives about 40 kWh/kg of electricity plus just under 10 kWh/kg of heat. These are programme targets and reference values, not guaranteed values of individual products.

Technologies and operating behaviour

How a stack is built depends on the electrolysis technology. Alkaline electrolysis works with a liquid alkaline electrolyte and a diaphragm and has long been proven industrially. PEM electrolysis uses a proton exchange membrane, achieves considerably higher current densities and can usually be operated more dynamically. However, it requires noble-metal catalysts such as iridium and platinum. AEM electrolysis combines elements of both approaches and is considered less mature, but promises fewer critical raw materials. SOEC stacks work with solid oxide cells at high temperatures and with steam. With integrated waste heat they need particularly little electricity, but place high demands on material durability, heat-up times and thermal operation management.

This operating behaviour matters for the power system. Electrolysers are often regarded as flexible consumers because they can take up electricity when a lot of renewable power is generated and prices are low. Whether this flexibility is actually available technically and economically, however, does not depend only on a stack being present. The Clean Hydrogen Partnership gives, as of 2020, very different times to reach rated power. From warm standby it is a few seconds for PEM, about one minute for alkaline systems and around ten minutes for SOEL. A SOEL cold start, by contrast, takes many hours. Minimum load, ramp rate, start-stop capability, permissible temperature gradients, pressure control and gas quality additionally limit the usable operating range.

Degradation and lifetime

Degradation refers to the ageing of the stack during operation. It usually shows up as rising cell voltage at the same current density, that is, as higher electricity demand per kilogram of hydrogen. Possible causes are aged catalysts, membrane damage, corrosion, impurities in the water, mechanical stresses and deteriorating contacts. A stack that is frequently ramped up and down may age differently from one in steady continuous operation. Load changes, standstills, pressure differences and local hot spots stress membranes, electrodes and seals.

Degradation rates are usually stated in percent per 1,000 operating hours. The definition is decisive here. For low-temperature electrolysis, the Clean Hydrogen Partnership understands this as the efficiency loss at rated load. Its worked example: 0.125 percent per 1,000 hours results in around ten percent more energy demand over ten years with 8,000 operating hours each. For SOEL, by contrast, it measures the loss of production rate at constant efficiency. According to a JRC analysis from 2025, average degradation in funded PEM projects fell from about 0.25 to around 0.13 percent per 1,000 hours. The stack degradation targets for 2024, however, were still not met overall by newly added projects.

The lifetime of a stack is usually stated in operating hours. In 2020, IRENA used 80,000 hours in cost analyses and gave ranges, depending on the technology, from under 20,000 hours for SOEC to about 60,000 to 80,000 hours for PEM and alkaline stacks. Such figures are meaningful only if the operating profile, load changes, temperature, water quality, pressure, maintenance and the end-of-life criterion are also stated, for example ten percent additional consumption. 60,000 or 80,000 hours therefore do not mean that a stack remains economically equivalent in use for that long in every application.

Efficiency, design and costs

The stack determines a large part of the efficiency, but not the entire energy demand of the plant. As current density rises, so does cell voltage. Higher current densities allow smaller active areas and thus lower investment costs per kilowatt. However, they usually increase losses and heat generation. A plant can thereby be cheaper to buy but more expensive to run if electricity costs weigh heavily. At low electricity prices a different design point may make sense than at high prices and many operating hours.

Stack replacement is a cost block of its own. The levelised cost of hydrogen (LCOH) includes, among other things, investment costs, operating hours, degradation, electricity price, grid charges, levies and maintenance. In the Clean Hydrogen Partnership's cost indicators, stack replacement is explicitly not included for low-temperature electrolysis. It must therefore be considered separately. According to IRENA, at the then low production volumes the stack accounted for about 45 percent of electrolyser costs. Large-scale series production could almost halve its manufacturing costs. A cheap stack can become expensive if it has to be replaced often. A more expensive stack can pay off if it lasts longer or needs less electricity.

Significance for the power system and the hydrogen economy

In the power system, the stack helps determine how well surplus electricity, low exchange prices or grid-serving operation can actually be converted into hydrogen. The connection capacity of an electrolyser only says how much power it can take up. It says little about how fast it can regulate, how often it may start, which partial load is still efficient and what costs fluctuating operation causes.

This is especially important when electrolysers are planned as off-takers for wind and solar power. A plant that runs mainly in very cheap hours achieves fewer full-load hours than one in continuous operation. Its investment costs are then spread over fewer kilograms of hydrogen. IRENA describes that investment costs per unit decrease significantly only at about 3,000 to 4,000 operating hours per year. The stack must therefore not only cope technically with changing load. Its cost structure must also fit the operating model. This is a core question for power-to-X.

Grid issues also depend on it. In regions with a lot of renewable generation, large electrolysers can take up load and thus reduce curtailment. However, they can also aggravate grid congestion if location, connection capacity and mode of operation do not suit the grid. The stack itself does not resolve this alignment. Whether the conversion is system-serving, market-driven or grid-straining depends on site selection, electricity procurement, grid connection rules, levies, funding conditions and operating strategy.

Another misunderstanding concerns scaling. More megawatts of electrolysis capacity do not automatically mean more usable green hydrogen. This also requires renewable electricity, water, grid connection, permits, storage, off-takers, transport infrastructure and suitable operation management. The stack is one bottleneck among several. Its industrial manufacture, the availability of materials such as iridium, quality assurance and interchangeability, however, help decide whether electrolysis projects become plannable and financeable.

The term electrolyser stack shows that hydrogen production is not an abstract conversion box. In the stack, electrical operation, electrochemical reaction, material ageing and economic design meet directly. Anyone talking about the costs, efficiency or flexibility of electrolysers should therefore say whether they mean the stack, the complete electrolysis system or the entire hydrogen plant, and which energy content the values refer to. Otherwise technical figures easily lead to wrong conclusions about system costs, availability and the contribution to the power system.

Sources

Clean Hydrogen JU – SRIA Key Performance Indicators (KPIs) — Clean Hydrogen Partnership (Clean Hydrogen Joint Undertaking, EU); unbekannt (beruht auf der SRIA 2021–2027 vom Februar 2022). Abschnitt Renewable Hydrogen production, KPI-Tabellen AEL, PEMEL, SOEL und AEMEL samt Anmerkungen
Hydrogen brief: Energy efficiency of water electrolysers for hydrogen production — Gemeinsame Forschungsstelle der Europäischen Kommission (JRC); unbekannt. Abschnitt zu Heizwert und Brennwert (Figure 5), Beispiel zum Energieverbrauch auf Stack- und Systemebene, Abschnitt Energy efficiency and the levelised cost of hydrogen
Green Hydrogen Cost Reduction: Scaling up Electrolysers to Meet the 1.5°C Climate Goal — International Renewable Energy Agency (IRENA); 2020-12. Zusammenfassung (Figure ES1, Anmerkungen zu Annahmen) sowie Kapitel zu Stackkosten und Lebensdauer; Technologietabellen auch in der IRENA-Publikation Making the Breakthrough
Programme Review Report 2025 — Gemeinsame Forschungsstelle der Europäischen Kommission (JRC) für die Clean Hydrogen Partnership; 2025. Kapitel zur Wasserstofferzeugung (Auswertung der TRUST-Daten zu AEL und PEM)
Clean Hydrogen JU Annual Work Programme 2025 — Clean Hydrogen Partnership (Clean Hydrogen Joint Undertaking, EU); unbekannt (Arbeitsprogramm 2025). Ausschreibungsthemen zur Elektrolyse, Abschnitte zu KPIs und zur Balance of Plant

Related glossary entries

Edited by eCMS on 11.11.2026 19:44