"Battery Storage Outruns Governments: 108 GW Added in 2025, 1,500 GW Pledge Still Far Off"

Something unusual is happening in the energy transition: the hardware is moving faster than the governments that are supposed to plan it. A new report from the think tank Ember, published in October 2026, finds that the world added a record 118 gigawatts of energy storage in 2025, of which 108 GW were batteries. That is a 40% jump on 2024, and it brings total installed storage capacity to 430 GW. The striking part is not just the size of the number but the speed of the climb: annual battery additions rose from 9.5 GW in 2021 to 108 GW in 2025, a seventeenfold increase in five years. Solar power took nine years to grow from 10 GW to 100 GW of cumulative capacity; batteries covered a similar distance in about four.

What makes this more than a good year for one technology is the gap it exposes between what is being built and what is being promised. National storage targets compiled by Ember now total 822 GW by 2030, nearly triple their pre-COP29 level of 284 GW, yet still 45% below the 1,500 GW pledge that 74 countries endorsed at COP29. Ember's arithmetic is blunt: keeping up the 2025 pace every year through 2030 would take cumulative storage to about 1,020 GW, comfortably above current national targets but still well short of the pledge. In other words, governments have raised their ambitions a lot and are still being outrun by installers, while the international pledge itself remains out of reach without a further step-change.

The engine behind the surge is not a mystery, and it is not a subsidy story alone. Battery system costs have fallen by roughly 20% a year over the past decade, according to Ember, which turns storage from a grid luxury into a practical way to make solar and wind usable after sunset. That is why the comparison with solar matters: solar's story was first about falling panel prices, then about everything built around them, and storage is now going through the same second act, where the bottleneck shifts from the device itself to connections, markets and rules.

There are two honest caveats, and both cut against easy optimism. The first is concentration: China accounted for 54% of 2025 storage additions and the United States for 16%, so the global total rests heavily on two markets and their policies. The second is that deployment outrunning targets is not the same as deployment being enough. A world that adds storage at the 2025 pace still misses the 1,500 GW pledge, and pledges, unlike targets, are what the climate arithmetic actually demands. What Ember is really documenting, I think, is a transition whose technology curve has bent faster than its political curve, in both directions at once.

My prediction: global battery additions in 2026 will reach at least 130 GW, 62% confidence. The case is momentum, 40% growth last year, falling costs and deep pipelines in the two biggest markets; the case against is that growth rates this steep rarely hold, and a policy or supply shock in either leading market would show up in the global total almost immediately.