Power to Hydrogen and Repsol tested a five cell AEM electrolyzer for 1,250 hours at 40% to 100% loads and pressures of 2 to 30 bar, recording approximately 2 mV of irreversible degradation per 1,000 hours and hydrogen purity of 99.9% at 30 bar.
The results are significant for AEM electrolysis because durability and scale up remain central to its commercial case. P2H2 estimates that the measured degradation rate could support a stack lifetime of more than 50,000 hours, although that figure is a projection derived from the test results rather than a lifetime demonstrated through 50,000 hours of operation. The companies are now assessing a larger demonstration in the 500 kW to megawatt range.
The testing was conducted through Repsol’s All4Zero industrial innovation hub using commercial scale P2H2 cells configured in a five cell short stack. The test program included parametric and durability testing as well as simulated renewable energy load profiles, with repeated operation between 40% and 100% load. The system also operated across a pressure range of 2 to 30 bar.
That operating profile addresses a practical challenge for renewable hydrogen projects. Electrolyzers connected to wind and solar generation can face substantial changes in available electricity, making the ability to operate dynamically an important part of system economics. A technology that performs well only under relatively stable operating conditions can lose utilization or incur additional degradation when exposed to frequent load changes.
P2H2 says the system maintained dynamic performance during the test and also resumed operation after an unplanned facility power outage. These results provide evidence of operational resilience, but they do not by themselves establish how the technology will perform after years of repeated cycling at full commercial scale.
The distinction is important because a five cell short stack is not equivalent to a complete industrial electrolyzer plant. Scaling from individual commercial scale cells to hundreds of kilowatts or megawatts introduces additional engineering considerations involving thermal management, power electronics, gas handling, water treatment, controls and balance of plant.
The reported 2 mV per 1,000 hours of irreversible degradation is the most consequential result from the test. P2H2 says average cell voltage remained virtually unchanged between the beginning and end of the testing period and uses the measured degradation rate to project a lifetime exceeding 50,000 hours.
For AEM technology, durability has historically been closely linked to the question of whether its potential material cost advantages can translate into commercially viable systems. AEM electrolyzers are being developed as an alternative to conventional alkaline and proton exchange membrane systems, with manufacturers seeking to combine the dynamic characteristics associated with PEM with a lower cost materials platform.
P2H2’s architecture is designed around that proposition. The company says its hybrid AEM approach avoids iridium and PFAS, reducing exposure to materials that can increase cost or create supply and regulatory concerns. The company also says its design combines an alkaline separator with the membrane and oxygen producing electrode.
The test therefore addresses more than electrical efficiency. For project developers, the relevant question is whether an electrolyzer can sustain performance while operating flexibly enough to make effective use of variable renewable electricity.
The stack produced hydrogen with purity of up to 99.9% at 30 bar, according to P2H2 and Repsol. Performance was reported across the full 2 to 30 bar pressure range.
Operating at elevated pressure can reduce the amount of downstream compression required before hydrogen is delivered to storage, pipelines or industrial users. That can affect both capital expenditure and electricity consumption at the plant level, although the economic benefit depends on the pressure required by the eventual application.
The pressure result is particularly relevant as electrolyzer projects move closer to industrial deployment. Hydrogen production cost is determined not only by electrolyzer efficiency, but also by electricity prices, utilization, compression, water treatment, financing and plant integration. Improvements in one component therefore do not automatically translate into a proportional reduction in the delivered cost of hydrogen.
P2H2 has modeled that broader economics using operating inputs from Repsol and the U.S. Department of Energy’s H2A Lite framework. The company estimates a levelized cost of hydrogen of €3.86 per kilogram, potentially declining to €2.82 per kilogram under an optimized renewable energy scenario using hybrid power purchase agreements. These figures are modeled projections, not costs demonstrated by the 1,250 hour test.
The move toward a 500 kW to MW scale demonstration will provide a more meaningful test of whether the short stack results can translate into an industrial system. P2H2 has already begun commercial operation of a 500 kW system at the Port of Antwerp Bruges, comprising two 250 kW AEM stacks. The company says the system is expected initially to produce up to 70 tonnes of hydrogen annually for regional customers.
That deployment provides an early bridge between component validation and larger industrial projects, but the Repsol test and the Antwerp system serve different purposes. The All4Zero program generated controlled durability and dynamic performance data, while the Antwerp installation provides operational experience from a complete commercial system.
For Repsol, the technology also fits into a wider renewable hydrogen development strategy. P2H2 says the companies are considering a larger demonstration as a next step toward Repsol’s target of 600 MW to 800 MW of renewable hydrogen capacity by 2030.
The commercial question will therefore shift from whether an AEM cell can demonstrate low degradation to whether complete systems can maintain that performance at hundreds of kilowatts and eventually megawatt scale, while achieving competitive utilization and operating costs.
That transition is where the technology will face its more demanding test. A projected 50,000 hour lifetime based on 1,250 hours of data is encouraging evidence, but it remains a projection. Demonstrating sustained degradation performance across larger stacks, repeated renewable cycling and long duration commercial operation will determine how far AEM electrolysis can move from an emerging alternative toward a significant share of industrial hydrogen capacity.
