What is an electrolyzer stack, and what decides how well it works

A good electrolysis cell is the starting point for a good stack. It is not the whole story.

Once multiple cells are assembled into a stack, they begin to share mechanical compression, flow paths, electrical interfaces, and a common thermal environment. At that point, performance is no longer determined by the electrochemical components alone.

The challenge becomes reproducing good cell performance across the entire stack, and keeping it there over time.

What is an electrolyzer stack?

A single electrolysis cell typically operates at only a few volts. To reach a practical power level, multiple cells are electrically connected in series and assembled into one structure. That assembly is the electrolyzer stack.

From the outside, a stack can look like a repeated set of identical cells held between two end plates. But simply repeating a good cell does not guarantee a good stack.

Each cell must operate under sufficiently similar conditions. That means maintaining consistent compression, fluid distribution, electrical contact, temperature, and sealing throughout the assembly. Small differences in any of these can become visible as differences in cell performance.

This is where stack engineering begins.

SourceU.S. DOE Technical Targets for Liquid Alkaline Electrolysis and PEM Electrolysis: cell voltage and current density status

What is inside a stack?

An electrolyzer stack is made up of electrochemical components together with the structures that connect, supply, seal, and compress them.

• Membrane electrode assembly. The membrane and catalyst layers where the electrochemical reactions take place. Their properties strongly influence the efficiency and performance available at the cell level.
• Porous transport layer. Provides electrical contact while allowing reactants and products to move between the electrode and the surrounding flow structure. Electrical conduction and mass transport have to be balanced at the same time.
• Bipolar plate. Separates adjacent cells, conducts current between them, and provides the flow field used to distribute fluids over the active area.
• Seals. Keep gases and liquids within their intended paths and maintain separation between different regions of the cell.
• Compression structure. Holds the repeated cell assembly together and maintains the mechanical load required for electrical contact, sealing, and dimensional stability.

None of these components works independently. Changing the compression can affect electrical contact and porous structures. Changing the flow field can affect temperature and gas removal. Changing one interface can alter the mechanical condition of another.

A stack therefore has to be treated as an integrated structure rather than a collection of separate components.

Stack components, part by part

Every part named above can be bought on its own. The table lists the materials we use in our own stacks, as published on the parts and product pages.

Changing a material changes more than the part. Published performance data applies to the stack as built, which is why a material change comes with its own validation scope.

How cell count scales output

Cells in a stack are connected in series, so the same current passes through every one of them. Each cell makes hydrogen in proportion to that current, which means that at a given current the stack's hydrogen output and its voltage both rise with the number of cells.

That gives a stack designer two levers. Cell area sets how much current each cell can carry at a sensible current density. Cell count multiplies whatever one cell produces.

Adding cells is a problem of assembly and uniformity. Growing the cell area changes what happens inside the cell.

More cells is also where the difficulty lies. When we built the HXS-30 we added cells in steps and measured every cell at each step, because the spread between the best and worst cell is the number that grows with cell count.

What changes when a cell becomes a stack?

Active area becomes a distribution problem

Increasing active area does more than increase the current that a cell can carry. It also increases the distance over which fluids, current, heat, and mechanical load must be distributed uniformly.

A larger cell is therefore not simply a smaller cell scaled geometrically. Maintaining similar conditions over the entire active area becomes a design problem of its own.

Compression has to be uniform

A stack needs enough compression to maintain electrical contact and sealing, but more compression is not automatically better.

Too little can increase contact resistance or reduce sealing reliability. Too much can deform components and change transport through porous structures.

The average compression value alone is not enough. What matters is how evenly that load is distributed over the active area and maintained throughout the stack.

Flow has to reach every region consistently

The average flow rate entering a stack does not tell you how that flow is distributed internally. Each cell, and each part of each cell, should receive sufficiently similar operating conditions.

Poor distribution can create local differences in concentration, temperature, gas removal, and pressure. Those local differences can eventually appear as differences in voltage or degradation.

Repeated interfaces amplify small variations

A single test cell contains only a limited number of interfaces. A stack repeats those interfaces many times.

Small differences in component thickness, surface contact, sealing, or assembly that may be difficult to notice in one cell can accumulate or become more important when repeated throughout a stack.

Average performance is not the whole picture

A stack can have a reasonable average voltage while individual cells behave differently. That is why cell-to-cell consistency matters.

A weak or abnormal region can influence operating limits, efficiency, gas quality, or durability even when the overall stack number still looks acceptable.

For stack validation, understanding the spread within the stack can therefore be as important as understanding the average.

Why single-cell results do not automatically scale

Single-cell testing is essential. It allows membranes, electrodes, catalysts, interfaces, and operating conditions to be studied under controlled conditions. But it answers a different question from stack testing.

• A single-cell result asks: can this electrochemical configuration perform well?
• A stack test asks: can that performance be reproduced across a multi-cell assembly?
• Manufacturing adds a third question: can the same stack performance be reproduced from unit to unit?

A strong single-cell result is important, but it is not sufficient evidence of a mature stack.

These are different stages of development. Moving from the first to the second introduces flow distribution, compression uniformity, repeated interfaces, electrical contact, thermal gradients, and cell-to-cell variation.

Moving from the second to the third introduces manufacturing tolerances, assembly repeatability, and quality control.

Durability also changes meaning at stack level

Durability is often presented as a single degradation rate. But that number only becomes useful when the conditions behind it are known. Current or current density, temperature, fluid conditions, pressure, test duration, and the basis used to calculate degradation all matter.

Stack durability adds another layer: whether all parts of the stack are aging in a similar way.

A stable average voltage does not necessarily mean every cell is behaving identically, just as a change in stack voltage does not by itself explain where the change originated. Long-term stack validation therefore needs both time and context.

What should an electrolyzer stack datasheet tell you?

A useful datasheet should allow the reader to understand not only what performance was measured, but under what conditions.

• Active area
• Rated current and voltage
• Rated hydrogen production
• Operating temperature
• Fluid or electrolyte conditions
• Operating pressure range
• Gas purity or crossover information where relevant
• Efficiency and the basis used to calculate it
• Durability or degradation data with the corresponding test conditions

The operating conditions are not footnotes. An efficiency measured at one temperature or operating point cannot be compared directly with a number measured somewhere else under different conditions.

The same is true for degradation. A durability number without the current, temperature, pressure, test duration, and other relevant operating conditions is difficult to interpret.

It is also useful to know whether a result comes from a single cell, a short stack, or a full stack. The larger the test object becomes, the more of the actual stack engineering is included in the result.

Stack vs system

A stack only makes hydrogen when something supplies it with DC power, water and electrolyte at the right temperature, takes the gases away and watches the whole process. That surrounding equipment is the balance of plant, and a stack plus its balance of plant is an electrolysis system.

The difference shows up in the energy figures. Our HXS-2 stack uses 44 kWh per kilogram of hydrogen. The complete HXB-V1 system built around it uses 56 kWh/kg, and the gap of about 12 kWh/kg is the balance of plant. A stack number and a system number are therefore different claims, and should be compared only with their own kind.

Status and targets are consistent with an end-of-life performance loss of 10%.

SourcesU.S. DOE Technical Targets for Liquid Alkaline Electrolysis: separate stack and system energy rows (2022 status 51 and 55 kWh/kg) · U.S. DOE Technical Targets for PEM Electrolysis

Integrating a stack into your own plant

Buying a bare stack moves the balance of plant to your side of the boundary. For the HXS-2 that means a DC supply sized for the operating current range, an electrolyte circulation loop and pump at a standard 4 L/min, cooling and temperature control, gas-liquid separation, outlet piping with vent and drain, instrumentation, control and site hydrogen safety.

None of these is unusual, but each has limits the stack sets: the ports, the heat load, the pressure the stack is rated for and the sequence for commissioning it. Those are written down in the integration guide, so the design can start from them rather than discover them.

Buying an electrolyzer stack: from test cell to 30 kW

At HydroXpand, we work across several stages of AEM water electrolyzer development.

HXS-0 is our test-cell platform for evaluating membranes, electrodes, catalysts, and other electrochemical components over a range of active areas.

HXS-2 is our 2 kW-class AEM water electrolyzer stack, available to order for integration with your own balance of plant. The product page provides operating conditions, performance and durability data, interface details and current lead time.

HXS-30 is our 30 kW-class stack. The 46-cell full stack, designed for 6,900 NL/h (15 kg/day), is available to order now; it completed first operation in September 2026 and is currently under long-term evaluation.

These platforms address different questions, but they are part of the same development path. A test cell tells us whether an electrochemical concept works. A multi-cell stack tells us whether that performance can be reproduced across a larger assembly. Long-term operation tells us whether it can be maintained. And repeated builds tell us whether it can be manufactured consistently.

Good cell performance is where stack development starts.

Making that performance uniform, durable, and repeatable is what turns it into stack engineering.