Electrolyser Capacity Doubled in 2025 While New Investment Decisions Fell
Installed water electrolysis passed 4 GW and another 2.5 GW is under construction, but final investment decisions dropped below 0.8 Mtpa and the 2030 pipeline has shrunk by a quarter.
The International Energy Agency's Global Hydrogen Review 2026, published on 18 June, reports that installed water electrolysis capacity doubled during 2025 to surpass 4 GW worldwide. More than 2.5 GW is under construction and targeted to enter operation in 2026, roughly 2 GW of that in Europe and dependent on a small number of large projects.
The second half of the picture is less encouraging. New final investment decisions for low-emissions hydrogen dropped below 0.8 Mtpa, after two consecutive years at around 1 Mtpa. The announced 2030 project pipeline has shrunk to about 27 Mt from 37 Mt previously, of which 4.3 Mt is committed; more than 6 Mt is possible only if further projects reach FID during 2026 and 2027. Low-emissions hydrogen output reached almost 1 Mt in 2025 and is expected to exceed 1% of global hydrogen production this year, concentrated in China, Europe and North America.
Two numbers moving in opposite directions is the useful thing here. Capacity is doubling because projects decided three or four years ago are now being commissioned; investment decisions are falling because the projects that would commission in 2029 are not being sanctioned at the same rate. The build-out visible today reflects a decision climate that no longer exists.
For anyone doing controls and instrumentation, the near-term addressable work is the plants entering operation now: stack-level monitoring, gas analysis, hydrogen area detection, power quality metering and integration into SCADA and a historian. That work is real and dated.
The pipeline number argues for a specific engineering posture, though. A gap between a 27 Mt announced pipeline and 4.3 Mt committed means a large fraction of what is being designed will not be built. Control architectures that are modular and redeployable — capable of being scaled down, relocated or repurposed — survive that better than architectures specified tightly around one mega-project that may never reach a final investment decision.
Source: International Energy Agency