Installed electrolyser capacity is not annual hydrogen output

Installed electrolyser capacity is easy to state. Annual usable hydrogen output is harder to determine.

A rated capacity describes the maximum input or output that equipment is designed to sustain under stated conditions. It does not show how much specification-grade hydrogen the complete plant will deliver over a year.

In this article, usable hydrogen means on-specification hydrogen measured at the contractual delivery point, net of agreed start-up, purging, treatment and downstream losses. Annual output depends on operating hours, load profile, system efficiency, technical availability, maintenance, degradation, and reliability across every component between the electrical connection and the hydrogen delivery point.

For industrial buyers, this distinction affects much more than production forecasts. It influences hydrogen cost, storage capacity, process continuity and the expected return on the project.

A megawatt is a rate, not an annual quantity

Electrolyser capacity is commonly stated in megawatts. This usually describes electrical input power, although supplier boundaries can differ. A megawatt is a unit of power: a rate of energy transfer, not a guarantee of yearly hydrogen production.

The International Energy Agency glossary says that nameplate capacity typically refers to the full-load operation intended by the designer and guaranteed by the manufacturer. It defines capacity factor as the average output divided by nameplate maximum capacity over a stated period.

Those definitions expose the missing information in a capacity announcement. A 10 MW rating alone does not justify assuming operation at 10 MW for all 8,760 hours of a non-leap year. Even when it runs, the complete system may not remain at full load or at its best efficiency point.

A meaningful specification, therefore, requires more than MW. It should state:

Hydrogen production rate in kg/h (preferred). If Nm³/h is used, state the exact reference temperature and absolute pressure.
Specific electrical energy consumption in kWh/kg
Whether the figure applies to the stack or the complete system
Whether the rated MW figure represents electrical input or hydrogen energy output; if output, whether it is on an HHV or LHV basis
Hydrogen purity, delivery pressure and temperature at the agreed measurement boundary
Permitted operating range and performance across that range

The European Commission Joint Research Centre’s harmonised testing procedure recommends comparing water electrolysis systems through specific energy consumption per unit of hydrogen output under stated operating and reference conditions. It also notes that system size and the application-specific balance of plant can affect comparisons.

Define availability, utilisation, and capacity factor explicitly

These terms are often used inconsistently. In this article, they are treated as distinct, but every specification or contract should define the numerator, denominator, system boundary, exclusions, and measurement period for each metric.

Technical availability shows whether the system was capable of operating when required. Depending on the agreed definition, planned maintenance, faults, inspections, repairs, and component replacement may reduce it or be excluded from the calculation.

Utilisation shows how much of the available operating opportunity was actually used. It can fall due to electricity prices or supply, hydrogen demand, grid constraints, operating strategy or regulatory conditions.

Capacity factor must use the same basis in its numerator and denominator. An electrical input capacity factor is the annual electrical energy consumed divided by (rated electrical input × period hours). A hydrogen-output capacity factor is annual hydrogen mass divided by (rated hydrogen production rate × period hours). Both can reflect availability and part-load operation; the hydrogen-output basis can also reflect efficiency changes.

Equivalent full-load hours must likewise identify their basis. Input-based equivalent full-load hours equal annual electrical energy consumed divided by rated electrical input. Output-based equivalent full-load hours equal annual hydrogen production divided by rated hydrogen production rate. The two can differ when system efficiency changes with load or degradation.

The exact contractual definitions matter. A technical availability guarantee should state the system boundary, measurement period, excluded events and treatment of planned maintenance. It should also clarify whether unavailable electricity, constrained hydrogen demand or external utility failures are excluded.

This distinction is also visible in the European Commission Joint Research Centre’s 2025 report on electrolyser testing. In an EU-scale illustration, the JRC combines an assumed energy consumption of 51 kWh/kg, full use of installed capacity and 60% time availability (about 5,250 hours) to estimate annual production. This is not a two-scenario plant model, but it shows why installed capacity alone cannot provide annual production.

Calculate output from full-load hours and system energy use

A practical first estimate is:

Annual usable hydrogen (kg) ≈ (P × H ÷ E) × F

In this formula, P is the rated electrical input (kW). H is the input-based equivalent full-load hours in hours. E is the weighted gross-output system energy consumption in kWh/kg. F is the usable-delivery fraction.

Equivalent full-load hours already include shutdowns and partial-load operation. Do not multiply availability into the equation again if the full-load-hour estimate or capacity factor already accounts for downtime. Doing so would count the same lost hours twice.

The energy-consumption figure should represent the agreed system boundary. The usable-product yield should account for hydrogen that does not reach the delivery point because of purging, start-up, off-specification production or downstream losses.

Consider an illustrative 1 MW system with a weighted system energy consumption of 55 kWh/kg. If it could run at full load whenever technically available, 95% availability would represent 8,322 potential operating hours and about 151 tonnes of gross hydrogen per year. At 85% availability, it would represent 7,446 hours and about 135 tonnes.

The ten-percentage-point difference equals 876 hours, or 36.5 days, and about 15.9 tonnes of potential hydrogen per MW in this simplified example. The real production difference may be smaller if electricity or hydrogen demand is unavailable during some of those hours.

This is why the phrase “potential operating hours” matters. Availability creates the ability to produce. It does not guarantee that the project will be dispatched.

The balance of plant is essential for delivering usable hydrogen to the user

The electrolyser stack is the core electrochemical conversion equipment, but it is not the complete hydrogen production system.

The JRC harmonised testing procedure describes balance-of-plant equipment that can include power conversion, water conditioning, pumps, heat exchangers, gas-liquid separation, cooling, drying, purification and compression. Instrumentation, control software and safety sensors also form part of the system.

Every required component creates an operational dependency. A healthy stack cannot deliver usable hydrogen if the rectifier has failed, cooling is insufficient, water quality is outside specification, the dryer is unavailable or the compressor cannot meet delivery pressure.

System-level performance is therefore the relevant measure. The system boundary must be clearly agreed upon, reflecting whether performance is guaranteed at the stack outlet, the system boundary, or the final point of use. Stack-only efficiency or availability can still support technical analysis, but it should not be mistaken for plant output.

This boundary also affects comparisons. One supplier may quote stack consumption on the DC side. Another may include rectification, cooling and gas treatment on the AC side. The lower number is not automatically the better system if the excluded auxiliaries must still be purchased and powered.

Operating conditions change annual performance

Rated performance is typically specified under defined reference conditions, often at steady state. A real project follows an hourly and seasonal operating profile.

Several factors change annual output:

Electricity supply and dispatch

A grid-connected plant may reduce load when electricity prices rise. A system coupled to wind or solar follows resource availability unless the project adds grid supply, storage or renewable oversizing. The JRC’s 2025 report notes that intermittent renewable energy supply must be carefully aligned with energy demand. It also calls for long-term testing under real-world renewable power profiles to assess performance, degradation, and reliability.

Part-load efficiency

System energy consumption is not constant across the operating range. The JRC policy brief on electrolyser energy efficiency explains that stack and balance-of-plant efficiencies together determine the power range in which the whole system performs best. Variable power can move the plant away from that range.

An annual model should therefore use an efficiency curve or operating map, not only the rated-load value.

Start-ups, shutdowns and standby

Frequent cycling can create standby time, start-up losses, and periods when hydrogen has not yet reached specification. Technology selection should reflect the expected operating profile, including minimum load, ramping, restart time and tolerance of repeated on-off operation.

Delivery pressure and purity

Hydrogen is useful only when it meets the receiving process requirement. Drying, purification and compression consume energy and may constrain flow. Buyers should compare output at the same pressure, purity, temperature, and delivery boundary.

Degradation

Performance changes over the operating life. Degradation may increase the electricity required to maintain a given hydrogen output, reduce maximum output, or trigger stack intervention.

A JRC review of FCH 2 JU-funded electrolyser projects warns that short-term degradation tests are difficult to extrapolate because several degradation processes operate over different timescales. It also reports different durability outcomes under flexible and steady-state operating conditions. More recent JRC accelerated-stress-testing guidance likewise emphasises defining operating profiles and stressors when evaluating durability under fluctuating renewable electricity.

The annual production model should therefore cover more than year one. It should include performance degradation, maintenance intervals and planned stack work across the project life.

Annual averages can hide operational risk

An availability percentage does not describe the pattern of downtime.

Under the same calendar-hours definition, 95% annual availability means 438 hours of downtime. That could consist of many short interruptions or, in an extreme case, a single continuous outage lasting more than 18 days. Both cases produce the same annual availability figure, but they do not create the same consequences for an industrial user.

A process with continuous hydrogen demand may tolerate brief interruptions with a modest buffer. A long outage may require much more storage, a backup supply contract or a change to the production schedule. The buyer therefore needs information about failure frequency, expected repair time, maintenance duration, spare-part strategy and support response, not only the annual percentage.

Storage should be designed against credible outage events and the hourly balance between production and demand. It should not be sized solely on annual output.

Questions buyers should ask before selecting an electrolyser

Procurement should translate the required annual output into measurable system obligations.

What exactly does the rated MW figure represent (e.g., stack DC power or total plant AC grid input), and what auxiliary loads are included?
What hydrogen flow is guaranteed at the required delivery pressure and purity, and at which reference temperature and absolute pressure is volumetric flow stated?
Is specific electrical energy consumption stated for the stack or the complete system, and at which load points? If efficiency is quoted, is it stated on an HHV or LHV basis?
What is the minimum stable load, ramp rate, start-up time and standby consumption?
How are technical availability and planned maintenance defined?
Which events are excluded from the availability calculation?
What degradation metric is guaranteed, over which period and under which operating profile?
Which balance-of-plant components can stop production, and where is redundancy justified?
What spares, remote diagnostics and service response are available, and what is the expected repair time?
How much hydrogen is expected to be off specification or lost during starts, stops, purging and treatment?
How was the annual output forecast built from hourly electricity supply and hydrogen demand?
Where will performance be measured during acceptance testing and commercial operation?

The answers should appear in technical schedules, acceptance criteria and the operating model. They should not remain assumptions in a spreadsheet.

Design the annual output into the project

Reliable hydrogen supply starts with the required delivery profile, not with an installed MW target.

The project team should first define hourly and seasonal hydrogen demand at the delivery point. It should then establish the system boundary and product specification. An hourly model can combine electricity availability, dispatch rules, part-load efficiency, start-up behaviour, maintenance, and degradation. The resulting production profile can guide electrolyser sizing, modular configuration, redundancy, storage, and backup supply.

Hydrogenera applies this systems perspective when integrating electrolysers into industrial processes. The objective is not simply to install conversion capacity. It is to connect power, water treatment, hydrogen production, gas conditioning, controls and the receiving process around a measurable delivery requirement.

Installed capacity remains an important project parameter. Annual usable hydrogen is one of the key outcomes that determines whether the investment works.

Plan hydrogen output around the real process

Hydrogenera designs and integrates hydrogen systems tailored to the required production profile, delivery conditions and industrial process. Contact Hydrogenera to discuss an electrolyser project built around defined, measurable annual hydrogen output.