AEM, PEM or alkaline electrolysis: how to choose an industrial system

An anion-exchange-membrane (AEM) electrolyser is a credible choice when its complete system delivers the required hydrogen reliably at a competitive lifetime cost under the site’s electricity profile. Compare it with proton-exchange-membrane (PEM) and conventional alkaline systems using power consumption, water conditioning, gas purification, degradation, replacement and supplier support. Catalyst cost and cell current density alone are insufficient.

Hydrogen requirements and complete electrolyser systems

A procurement comparison begins with the hydrogen that the process needs: delivery rate, pressure, purity and permitted interruptions. An electrolyser advertised at a particular electrical capacity can produce a different useful output once auxiliaries and purification are included. Define the point where compliant hydrogen enters the customer process, then compare all proposals at that boundary. This avoids selecting a cheap stack that requires expensive compression, storage or standby supply elsewhere on site.

The US Department of Energy explains the underlying distinction: alkaline electrolysis transports hydroxide ions, while a proton-exchange membrane (PEM) transports protons. Anion-exchange-membrane (AEM) systems use a solid membrane in an alkaline environment. Their attraction includes the possibility of avoiding expensive precious-metal catalysts. That material advantage does not establish an installed cost advantage: membranes, electrodes, gas handling and manufacturing quality remain part of the purchased system.

A catalyst result must be kept separate from integrated performance. A 2019 Los Alamos project report documented how interactions between an AEM catalyst and its ionomer could reduce activity compared with a simpler electrode experiment. This historical example explains a continuing evaluation problem without setting a current performance benchmark. The buyer needs measurements from the offered cell and stack, using the specified membrane, electrode area, water quality and operating temperature.

Power profile, efficiency and stack life

The power profile should be represented as a time series rather than an annual electricity price. Continuous operation and frequent shutdowns place different demands on thermal management, gas separation and controls. Request the useful hydrogen output and electricity consumption across the proposed operating range, including standby and restart. An efficiency measured only at the most favourable load cannot describe a plant following intermittent power. Equally, a highly flexible unit may add little value at a site with steady baseload demand.

The economic influence of efficiency can be made explicit without assuming a market forecast. In an illustrative comparison, a difference of 5 kWh per kilogram at an electricity price of €60 per MWh changes electricity cost by €0.30 per kilogram. This calculation excludes water, capital, maintenance and compression. It shows the sensitivity that a purchasing team should run with its own contracted power prices and measured system consumption, rather than treating the lowest equipment quotation as the lowest hydrogen cost.

Durability evidence needs the same operating context. A long constant-current test demonstrates survival under that experiment; it does not automatically predict lifetime through repeated pressure, temperature and load changes. Ask how degradation affects output, energy consumption and the replacement date. The replacement assumption should include the labour, crane or access requirements, commissioning time and lost production associated with changing a stack. A warranty is useful only if its permitted operating conditions match the intended plant.

Water treatment and supplier support

Water treatment can change both cost and operating responsibility. Establish whether the offered equipment requires pure water or supporting electrolyte, the limits for relevant contaminants, and what monitoring or conditioning is supplied. A proposal that shifts water-quality compliance to the buyer should price that obligation visibly. Large-area electrode consistency also matters: request acceptance records and operating references for the offered stack architecture rather than extrapolating from small research cells or a different product generation.

Supplier support belongs in the technical decision because an unavailable spare part can interrupt hydrogen supply regardless of cell efficiency. Define service response, consumable access, replacement availability and the handling of design changes. Where uninterrupted hydrogen is essential, compare storage and backup supply under credible outages. The strongest AEM proposal is therefore a supported, qualified system for a defined duty cycle. PEM or conventional alkaline equipment remains preferable where its demonstrated operating envelope and contractual support better match the application.