Grid operators and power planners across the world now have access to roughly 108 gigawatts of new battery storage capacity that came online in 2025, a 40 percent increase over the prior year and the single largest annual jump ever recorded. The International Energy Agency confirmed the figure in its Global Energy Review 2026, calling battery storage the fastest-growing power technology on the planet. The surge is reshaping how electricity systems absorb wind and solar output, but gaps in duration data and regional detail leave open questions about how quickly batteries can shoulder the work now done by fossil-fuel peaker plants.
Why 108 gigawatts of new storage changes the grid calculus
The scale of the 2025 buildout matters because batteries are no longer a niche supplement to renewable generation. They are becoming a structural part of electricity supply. Roughly 80 percent of the new capacity was utility-scale, meaning large installations connected directly to transmission networks rather than small units behind a home or business meter. That translates to about 87 gigawatts of utility-scale additions in a single year, a volume that dwarfs what most countries had installed cumulatively just a few years ago.
China drove the majority of the expansion, accounting for roughly 60 percent of all new battery storage deployed in 2025, according to the IEA’s review. The United States came in as the second-largest market, adding approximately 19 gigawatts. That figure represents a year-over-year increase of about 60 percent for the U.S. alone, with more than 16 gigawatts at utility scale and nearly 3 gigawatts of behind-the-meter systems. U.S. regulators had already documented a sharp increase in storage capacity during 2024, so the 2025 acceleration extends a pattern of compounding growth rather than a one-off spike.
For electricity consumers, the practical effect is that grid operators can store cheap midday solar power and dispatch it during expensive evening peaks. Each gigawatt of four-hour battery storage can, in theory, replace a similarly sized gas peaker that runs only during those high-demand windows. In markets with high solar penetration, such as parts of China and the American West, this arbitrage is already flattening price spikes and reducing the need to start up older, less efficient fossil plants.
If average battery durations keep lengthening at the pace the IEA has described, utility-scale storage could begin to displace a meaningful share of U.S. natural-gas peaker capacity within the next few years. Whether that share reaches 15 percent or more by 2028 depends on project completions that can be tracked through planned-addition filings, making the hypothesis testable in real time. The same logic applies in China’s provincial grids, where planners are pairing large battery blocks with new solar and wind bases to firm output and avoid curtailment.
IEA data and EIA filings anchor the 108-gigawatt count
Two primary institutions provide the evidentiary backbone for the headline number. The IEA’s battery storage chapter in its 2026 review is the source of the 108-gigawatt global total, the 40 percent year-over-year growth rate, the 80 percent utility-scale share, and the 60 percent China share. The agency’s separate commentary adds the U.S.-specific breakout: 19 gigawatts total, over 16 gigawatts utility-scale, and nearly 3 gigawatts behind the meter, with a 60 percent annual increase.
On the U.S. side, the Energy Information Administration’s generator inventory confirmed that battery capacity had already grown 66 percent in 2024. That figure serves as a baseline against which the 2025 numbers can be compared. Together, the two datasets show that the American storage fleet roughly doubled in size across two consecutive years of aggressive installation, moving from a marginal contributor to a resource with system-level significance.
The strength of these sources is that they draw on interconnection records and generator surveys rather than industry press releases. The IEA aggregates data from national energy agencies, while the EIA tracks every utility-scale generator in the United States through mandatory reporting forms. Both organizations publish their underlying datasets, allowing independent analysts to replicate the totals and cross-check them against regional grid operator statistics.
Even so, the methodologies are not identical. The IEA’s global view must reconcile different national definitions of what counts as utility-scale storage, while the EIA’s reporting thresholds exclude the smallest behind-the-meter systems. That means the 108-gigawatt global figure and the 19-gigawatt U.S. subset are best understood as conservative estimates of hardware that is clearly connected to the power system, rather than exhaustive tallies of every battery installed.
Duration gaps and missing regional detail limit the full picture
The 108-gigawatt figure measures power capacity, meaning the maximum instantaneous output the batteries can deliver. It does not, by itself, reveal how many hours each system can sustain that output. A four-hour battery and a one-hour battery look identical in a gigawatt tally but perform very differently on the grid. The IEA has noted that durations are lengthening, yet neither its Global Energy Review nor its commentary breaks out the 2025 additions by average duration at the regional or project level. Without that detail, it is difficult to calculate how much total energy storage, measured in gigawatt-hours, the world actually gained.
Duration matters because it determines which grid services batteries can reliably provide. One- to two-hour systems excel at fast frequency response, short-term balancing, and shifting solar output into the early evening. Four-hour or longer systems can cover the full evening peak and begin to compete directly with gas peakers. Multi-day storage, which is still rare at scale, would be needed to ride through prolonged wind lulls or cloudy periods without leaning on fossil backup.
Regional granularity is similarly thin. China and the United States together account for the vast majority of reported additions, but the IEA’s headline release does not provide country-level figures for Europe, India, Australia, or other markets that have active storage programs. Analysts tracking those regions will need to wait for supplementary data tables or national agency reports to fill in the picture. In the meantime, local grid operators are left to extrapolate from partial information when planning transmission upgrades and reserve margins.
Cost data present another blind spot. The IEA’s high-level narrative describes rapid cost declines over the past decade, driven largely by lithium-ion manufacturing scale, but the 2026 review does not attach project-level price tags to the 2025 cohort of installations. Without standardized reporting on capital costs and operating lifetimes, it is hard to benchmark whether today’s battery projects are truly undercutting new gas peakers on a levelized cost basis in every region, or only in markets with strong policy support and high fuel prices.
What the boom means for policy and planning
Despite those gaps, the 108-gigawatt milestone carries clear implications for policymakers and grid planners. First, storage is now large enough to warrant the same long-term planning treatment historically reserved for conventional generation. Integrated resource plans that omit batteries risk misjudging future capacity needs, especially in systems adding large amounts of variable renewables.
Second, market rules need to catch up with the technology. Many wholesale power markets still compensate batteries primarily for energy arbitrage, even though they can provide capacity, reserves, and fast-response services. Aligning payment structures with the full range of services would make it easier for storage developers to finance projects with longer durations and more flexible operating profiles.
Third, regulators will need better data. The IEA’s global tallies and the EIA’s national filings are essential, but they are not sufficient on their own. Standardized reporting of duration, state-of-charge constraints, degradation rates, and cycling patterns would allow planners to distinguish between a fleet that can reliably backstop renewables and one that merely shaves a few peak hours. More transparent cost and performance data would also sharpen debates over whether to prioritize batteries over alternatives such as demand response, transmission expansion, or flexible gas plants.
The IEA has already signaled that batteries are “taking on a larger system role” in its recent commentary on storage growth, and the 2025 buildout validates that assessment. The challenge now is to translate headline gigawatt numbers into a more nuanced understanding of how, where, and for how long these new assets can deliver power. Until duration and regional data catch up, the 108-gigawatt figure should be read as both a landmark achievement and an invitation to ask harder questions about what kind of storage future grids actually need.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
