The study, published in Nature Communications by researchers at the Technical University of Denmark, DynElectro, SolydEra, and Aalborg University, describes a 70-cell solid oxide electrolysis stack operated in steam electrolysis mode for nearly three years of continuous testing. After an initial conditioning period, the stack settled into a stabilized regime in which its average voltage degradation rate fell to just 0.05 percent per thousand hours, equivalent to 0.65 millivolts per thousand hours for each individual cell. That rate was sustained for 22,000 hours following the break-in phase. According to the authors, this represents the longest and lowest-degradation operation ever reported for a complete solid oxide electrolysis stack, a claim that, if confirmed by the community, repositions durability from the central bottleneck of the technology to a solvable engineering problem.

The key innovation lies in what the team calls frequency-dependent electrochemical operation, or AC:DC operation. Instead of feeding the stack a steady direct current, the conventional approach that has defined virtually every long-duration test to date, the researchers periodically inverted the polarity of the stack, alternating between electrolysis mode, in which steam is split into hydrogen and oxygen, and fuel-cell mode, in which the device briefly operates as a power-producing fuel cell. This rhythmic reversal modulates the electrochemical polarization experienced by the electrodes and redistributes the heat generated within the stack. The effect, the researchers report, is consistent with mitigating the thermal and electrochemical gradients that accumulate during sustained one-way DC electrolysis, gradients that are widely believed to drive the slow decay of cell voltage over time.

To appreciate why this matters, it helps to understand what happens inside a solid oxide electrolysis cell. Each cell is a thin sandwich of ceramic layers operating at roughly 700 to 850 degrees Celsius. Steam flows over a nickel-based ceramic cathode, where electrons supplied by the circuit reduce water molecules to hydrogen gas, while oxide ions migrate through a dense ceramic electrolyte, typically yttria-stabilized zirconia, to an anode where they are oxidized to oxygen. At these temperatures, every component is in constant chemical conversation with its neighbors. Nickel migrates and can agglomerate, electrodes sinter and coarsen, oxygen partial pressure gradients build up across the electrolyte, and thermal stresses accumulate wherever heat is generated unevenly. Under steady DC load, these processes proceed in one direction, relentlessly, and the stack’s voltage climbs as ever more energy is needed to push the same current through the degrading assembly.

The AC:DC strategy interrupts that one-way march. By periodically flipping the stack into fuel-cell mode, the researchers reverse the electrochemical forces acting on the electrodes, allowing polarization-induced concentration gradients to relax and redistributing Joule heating across the cell components. The team characterizes this as modulating electrochemical polarization and heat generation in a way that suppresses the degradation signatures commonly observed under DC operation. Crucially, their analysis of stack-level impedance evolution over the 25,000-hour test showed that the electrochemical performance not only stopped deteriorating but stabilized, and post-mortem microstructural examination of the cells confirmed that the degradation pathways typically activated during DC electrolysis were not engaged under the alternating regime. In other words, the failure mechanisms were not merely slowed; they appear not to have been switched on at all.

The magnitude of the achievement becomes clear when set against prior full-stack benchmarks. Long-duration demonstrations of solid oxide electrolysis stacks have historically reported degradation rates an order of magnitude or more higher, often several percent per thousand hours at industrially relevant current densities, with tests typically ending after a few thousand hours when voltage losses crossed acceptable thresholds. Economic models of hydrogen production are acutely sensitive to these numbers, because stack lifetime determines how the capital cost of the electrolyzer is amortized over the hydrogen it produces. A stack that lasts 25,000 hours at near-zero degradation, rather than 10,000 hours at meaningful degradation, changes the arithmetic of green hydrogen economics in a way that few materials-science advances have managed. The authors frame this explicitly: by shifting durability control from materials development to operational strategy, AC:DC operation offers a scalable pathway toward long-lived solid oxide electrolysis for economically competitive green hydrogen.

What makes the result especially compelling is its industrial pedigree. The stack was supplied by SolydEra, a commercial manufacturer of solid oxide cell technology, and the testing was carried out with the involvement of DynElectro, a Danish company specializing in electrolysis operation. This was not a button-cell experiment in a laboratory furnace, nor even a short-stack demonstration; it was a full 70-cell stack of the kind that would be arrayed into megawatt-scale electrolyzer modules. The operational protocol was developed and validated at that scale, meaning the results translate directly into the engineering context where durability actually determines commercial viability. The work was supported by the Danish Energy Technology Development and Demonstration Programme, the Pioneer Center for Accelerating P2X Materials Discovery, and the European Clean Hydrogen Partnership, reflecting the strategic weight that European energy agencies have placed on solving exactly this problem.

The scientific implications extend beyond the immediate headline numbers. The finding suggests that a substantial fraction of what the field has treated as intrinsic materials degradation may in fact be operationally induced, a consequence of the steady polarization and thermal gradients imposed by conventional DC testing. If periodic polarity inversion can keep nickel-based electrodes, zirconia electrolytes, and ferritic steel interconnects in a stabilized state for tens of thousands of hours, then research priorities across the field may shift toward understanding the electrochemical relaxation mechanisms that AC:DC operation exploits. Impedance spectroscopy, the diagnostic technique the team used to track the stack’s internal resistances over time, emerges as a powerful tool for distinguishing genuine materials decay from reversible operational stress, and the paper’s impedance evolution data provide a template for how future long-duration tests should be evaluated.

There are, of course, questions that the broader community will want to explore. Operating a stack in a regime that alternates between electrolysis and fuel-cell modes means the system spends part of its time not producing hydrogen, and the net hydrogen output and system-level efficiency under AC:DC duty cycles will need careful accounting in techno-economic analyses. The frequency and amplitude of the polarity inversions, described by the team as frequency-dependent operation, likely represent a tunable design space whose optimum may vary with stack architecture, current density, and steam utilization. Integrating AC:DC-capable power electronics into plant designs also poses engineering challenges, though the authors argue the approach is scalable. None of these caveats diminish the central result: a full industrial stack has now demonstrated that near-zero degradation is achievable with today’s materials, provided the operating strategy is right.

For the hydrogen economy, the timing could scarcely be better. Green hydrogen is projected to be essential for decarbonizing steel, ammonia, shipping, and long-duration energy storage, and solid oxide electrolysis offers the highest electrical efficiency of any electrolysis technology, particularly when coupled with industrial waste heat or high-temperature nuclear sources. The barrier has always been lifetime, and lifetime has always been assumed to be a materials problem requiring decades of ceramic chemistry research. This study demonstrates that a substantial part of the answer may already be sitting in the control room rather than the laboratory. A 70-cell stack that ran for 25,000 hours while losing almost nothing is not just a record; it is an argument that the age of durable, economical solid oxide hydrogen production may arrive far sooner than the field dared expect.

Subject of Research: Long-duration durability of solid oxide steam electrolysis stacks under alternating current operation for green hydrogen production

Article Title: Record stack durability in industrial solid oxide steam electrolysis through operational control

Article References: Bilalis, V., Sloth, O. F., Smitshuysen, T. E. L., Beyrami, J., Traulsen, M. L., Nielsen, M. N., Montinaro, D., Ouweltjes, J. P., Frandsen, H. L., Chen, M., Mogensen, M. B., Jensen, S. H., & Esposito, V. (2026). Record stack durability in industrial solid oxide steam electrolysis through operational control. Nature Communications. https://doi.org/10.1038/s41467-026-78001-1

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78001-1

Keywords: solid oxide electrolysis, green hydrogen, stack durability, AC:DC operation, polarity inversion, electrolyzer degradation, steam electrolysis, impedance spectroscopy, fuel cell mode, hydrogen production, ceramic cells, operational strategy