Battery Storage Statistics 2026
By Axis Intelligence Research
Co-author: Aidan Jad | Last updated: August 2, 2026 | License: CC BY 4.0
The world added 108 gigawatts of battery storage in 2025, roughly 40% more than in 2024 and slightly above the all-time peak year for gas-fired capacity additions. According to Axis Intelligence Research, the 161 GW of US storage already holding interconnection agreements represents 6.7 years of construction at the record pace planned for 2026.
Quick Answer
Global battery storage additions reached 108 GW in 2025, up about 40% year over year, with utility-scale systems accounting for roughly 87 GW of that total, according to the International Energy Agency. In the United States, developers plan 24 GW of new utility-scale battery capacity in 2026 against a record 15 GW added in 2025, per the US Energy Information Administration. Axis Intelligence Research estimates that 2026 US build at 48 to 96 GWh of energy capacity depending on duration assumptions.
Key Findings
- Global battery storage capacity additions reached 108 GW in 2025, about 40% above 2024, according to the IEA’s Global Energy Review 2026.
- US developers plan 24 GW of utility-scale battery storage in 2026, up from a record 15 GW in 2025, according to the EIA’s Preliminary Monthly Electric Generator Inventory.
- According to Axis Intelligence Research, China accounted for 58.3% of global battery storage additions in 2025 and the United States 17.6%, based on IEA country data.
- The mid-case overnight capital cost of a 4-hour US lithium-ion battery system is $308/kWh in 2026, within a low-to-high range of $255 to $366, according to the National Renewable Energy Laboratory.
- According to Axis Intelligence Research, US storage capacity holding a draft or executed interconnection agreement stands at 4.0 times everything the country built in the previous five years combined.
How Much Battery Storage Was Added Worldwide in 2025?
Battery storage was the fastest-growing power technology on the planet last year. The IEA’s Global Energy Review 2026 puts 2025 additions at 108 GW, about 40% above 2024, and notes that installed capacity now sits at eleven times its 2021 level.
That 108 GW figure clears a bar worth naming precisely. The IEA’s May 2026 commentary Battery storage is scaling up and taking on a larger system role points out that annual growth at this scale exceeds the historical peak for gas-fired capacity additions, which was around 107 GW in 2002. Batteries beat the best year gas ever had by about 0.9%.
Global battery storage additions by region, 2025
| Market | 2025 additions (GW) | Share of global | Notes | Source |
|---|---|---|---|---|
| China | ~63 | 58.3% | ~55 GW utility-scale, ~8 GW behind-the-meter | IEA (2026) |
| United States | 19 | 17.6% | >16 GW utility-scale, ~3 GW behind-the-meter | IEA (2026) |
| Australia | ~8 | 7.4% | ~4.2 GW utility-scale, ~3.4 GW behind-the-meter | IEA (2026) |
| Europe | ~6.2 | 5.7% | Utility-scale more than doubled to ~4.6 GW | IEA (2026) |
| Middle East | >3 | ~2.8% | Driven almost entirely by Saudi Arabia | IEA (2026) |
| Chile | ~1 | ~0.9% | Absorbing surplus solar generation | IEA (2026) |
| World total | 108 | 100% | ~87 GW utility-scale (~80%) | IEA (2026) |
Regional shares are calculated by Axis Intelligence Research from IEA country figures. Rows do not sum to 108 GW because the IEA reports several markets as approximations and does not publish a complete country breakdown.
Australia is the number that should stop a reader. Additions there surged to nearly 8 GW, almost nine times the previous year, split between roughly 4.2 GW utility-scale and about 3.4 GW behind-the-meter. Batteries now account for around 18% of installed dispatchable capacity in Australia, against 7% in China, 5% in the United States and 4% in Europe.
Chemistry consolidated hard. Lithium-iron-phosphate now accounts for around 90% of deployments, up from well below 50% five years ago. LFP is less energy-dense than the nickel chemistries that dominate electric vehicles, which matters when you are packing a battery into a car and does not matter when you are bolting it to a concrete pad. It is cheaper and it tolerates frequent cycling. For a stationary asset that charges and discharges daily for fifteen years, those are the only two properties that pay.
Aidan Jad: The gas comparison gets quoted as a milestone and it is one, but the engineering underneath it is more interesting than the headline. A gigawatt of gas turbine and a gigawatt of battery are not the same asset. The turbine runs until the fuel stops arriving. The battery runs for its rated duration and then it is a very expensive steel box until something recharges it. What changed in 2025 is not that batteries replaced gas one-for-one. It is that the duration got long enough, and the cost got low enough, that grid operators stopped treating them as a frequency-regulation curiosity and started dispatching them as bulk energy.
How Much Battery Storage Will the US Add in 2026?
US power plant developers plan to add 86 GW of new utility-scale generating capacity in 2026, a record if it is realized, according to the EIA’s February 2026 analysis of its Preliminary Monthly Electric Generator Inventory. Battery storage makes up 28% of that, behind solar at 51% and ahead of wind at 14%.
In absolute terms: 24 GW of utility-scale battery storage planned for 2026, against a record 15 GW added in 2025. That is a 60% increase on a number that was itself a record. US battery storage capacity has grown steeply over five years, with more than 40 GW added to the grid across that period.
US planned utility-scale capacity additions, 2026
| Technology | Planned 2026 (GW) | Share of additions | 2025 actual (GW) | Source |
|---|---|---|---|---|
| Solar | 43.4 | 51% | 27.2 | EIA (Feb 2026) |
| Battery storage | 24 | 28% | 15 | EIA (Feb 2026) |
| Wind | 11.8 | 14% | NA | EIA (Feb 2026) |
| Natural gas | 6.3 | ~7% | NA | EIA (Feb 2026) |
| Total | 86 | 100% | 53 | EIA (Feb 2026) |
The 53 GW added in 2025 was the largest single-year capacity installation since 2002.
Axis original metric: duration-weighted energy estimate for 2026 US additions
Power capacity in gigawatts tells you how hard a battery can push. Energy capacity in gigawatt-hours tells you how long it can keep pushing. The EIA publishes the first and not the second for planned projects, which leaves a gap that matters for anyone modelling evening peak coverage.
Axis Intelligence Research estimates the energy capacity of the 2026 US utility-scale build at 48 to 96 GWh.
Formula: planned power capacity (GW) x duration (hours) = energy capacity (GWh) Inputs: 24 GW planned 2026 additions (EIA, Preliminary Monthly Electric Generator Inventory, December 2025 data); duration bounds of 2 and 4 hours, with a 3-hour midpoint drawn from the IEA’s finding that the average duration of projects commissioned globally in 2025 rose to three hours from around two hours in 2023. Result: 48 GWh at 2 hours, 72 GWh at 3 hours, 96 GWh at 4 hours.
Limitations, stated plainly. The duration input is a global average and the US build is not globally average. California procurement has historically been anchored to four-hour resource adequacy requirements; ERCOT projects have skewed shorter because they earn on energy and ancillaries rather than a capacity mandate. A build that is 53% Texas by power capacity will sit below the global mean, which is why this is published as a range rather than a point estimate and labelled an estimate rather than a finding.
Which States Are Building the Most Battery Storage?
Three states account for about 80% of planned 2026 US battery storage capacity.
Planned 2026 battery storage additions by state
| State | Planned 2026 (GW) | Share of US total | Source |
|---|---|---|---|
| Texas | 12.9 | 53% | EIA (Feb 2026) |
| California | 3.4 | 14% | EIA (Feb 2026) |
| Arizona | 3.2 | 13% | EIA (Feb 2026) |
| All other states | ~4.5 | ~20% | EIA (Feb 2026) |
Largest US battery storage projects due online in 2026
| Project | Location | Capacity (MW) | Source |
|---|---|---|---|
| Lunis Creek BESS | Jackson, Texas | 621 | EIA (Feb 2026) |
| Clear Fork Creek Solar and BESS SLF | Wilson, Texas | 600 | EIA (Feb 2026) |
| Bellefield 2 Solar & Energy Storage Farm | Kern County, California | 500 | EIA (Feb 2026) |
| Tehuacana Creek 1 Solar and BESS | Navarro County, Texas | 418 | EIA (Feb 2026) |
Three of the four largest are in Texas. Tehuacana Creek 1 is also the largest solar photovoltaic project expected online in 2026 at 837 MW, with the 418 MW of storage attached to it.
The market data tells a compatible story from a different survey. The US installed 9.7 GWh of energy storage in the first quarter of 2026, up 32% year over year and the strongest first quarter on record, according to the US Energy Storage Market Outlook Q2 2026 published by the Solar Energy Industries Association and Benchmark Mineral Intelligence. Utility-scale accounted for 7.8 GWh of that, with 648 MWh of commercial and industrial and 515 MWh of residential.
According to Axis Intelligence Research, utility-scale systems represented 80.4% of US energy storage installed in Q1 2026 by energy capacity, a segment share almost identical to the 80% utility-scale share the IEA reports globally by power capacity.
Texas, Arizona and California led utility-scale installations in the quarter. Thirteen states now have explicit energy storage targets. SEIA also reports that 71% of all utility-scale energy storage capacity installed in Q1 was built in states won by President Donald Trump, which is the kind of figure that quietly reframes storage from a partisan project into an interconnection-and-land question.
Aidan Jad: Texas at 53% of the national build is not a policy story, it is a queue story and a land story. ERCOT is the one large US market where a developer can get an interconnection decision on something close to a project financing timeline, and West Texas has flat, cheap, permittable acreage next to existing transmission. California’s 14% is the opposite case: high value per megawatt-hour, four-hour duration requirements that make each megawatt more expensive, and a much harder path to a signed agreement. When people ask why Arizona is suddenly third, the answer is that it looks like Texas on the inputs that actually gate construction.
How Much Does Battery Storage Cost per kWh in 2026?
The National Renewable Energy Laboratory’s Cost Projections for Utility-Scale Battery Storage: 2025 Update puts the mid-case 2026 overnight capital cost of a 4-hour lithium-ion system at $308/kWh in 2024 dollars, inside a low-to-high band of $255 to $366.
Utility-scale 4-hour battery system capital cost projections
| Year | Low ($/kWh) | Mid ($/kWh) | High ($/kWh) | Source |
|---|---|---|---|---|
| 2024 (base) | 334 | 334 | 334 | NREL (2025) |
| 2025 | 295 | 321 | 350 | NREL (2025) |
| 2026 | 255 | 308 | 366 | NREL (2025) |
| 2030 | 207 | 279 | 354 | NREL (2025) |
| 2035 | 147 | 243 | 339 | NREL (2025) |
| 2050 | 108 | 178 | 307 | NREL (2025) |
All values in 2024 dollars, overnight capital cost for a complete 4-hour system.
Axis original metric: the 2026 cost uncertainty band
According to Axis Intelligence Research, the 2026 cost uncertainty band for US utility-scale battery storage is 36.0%.
Formula: (high case – low case) / mid case Inputs: $366/kWh, $255/kWh and $308/kWh, all from NREL’s 2025 projection table. Result: $111/kWh spread, equal to 36.0% of the mid case.
Read that against the direction of travel and it is the most useful cost number on this page. NREL’s own high case peaks in 2026 and does not return to 2024 levels until the late 2030s, driven by supply chain constraints and tariff impacts. The low case is 23% below 2024 by the same year. A developer signing an EPC contract this year is choosing between two futures that differ by more than a third of the system price, and nothing in the published literature resolves which one they are in.
A limitation NREL states and we will repeat: that work was completed in January and February 2025 and does not include impacts from tariff changes that have occurred since. Any 2026 cost number carrying a 2025 vintage should be read as a projection, not a transaction price.
The bottom-up model shows where the money actually goes. Of the $334/kWh 2024 base cost, $210 is lithium-ion battery cabinets. The bi-directional inverter is $26, electrical balance of system $44, structural balance of system $5. Everything else, including installation labor, permitting, sales tax, contingency, overhead and profit, is soft cost. NREL separates the figure into an energy component of $241/kWh and a power component of $372/kW, which is what lets you price durations other than four hours.
The performance assumptions behind those costs deserve as much attention as the costs. NREL adopts a 15-year lifetime, deliberately at the low end of published values, on the grounds that no 4-hour system has yet reached fifteen years of operation. Round-trip efficiency is set at 85%. Fixed operations and maintenance is set at 4% of the $/kW capacity cost, chosen from the high end of the published range on the assumption that augmentation will counteract degradation so the system performs at rated capacity throughout its life, at roughly one cycle per day.
Aidan Jad: That last assumption is the one that quietly decides project economics, and it is worth saying out loud what it means. The 4% FOM buys cell replacement. If your revenue model requires two cycles a day instead of one, which is increasingly what arbitrage-plus-ancillaries stacking looks like in ERCOT, the degradation curve steepens and that 4% no longer holds the rated capacity flat. NREL says this in the report. It gets dropped from every summary I read. Cycle count is a cost input, not a free lunch, and the fifteen-year lifetime is an assumption about a fleet that has not yet lived long enough to prove it.
How Long Do Batteries Store Power?
Duration is where the technology quietly changed shape. In 2025 the average duration of projects commissioned globally rose to three hours, from around two hours in 2023, according to IEA analysis. Most projects still cluster around two hours, but a rising share are being deployed at four hours or more.
The reason is a shift in what batteries are paid to do. Energy shifting, meaning storing large volumes of power for later deployment, has gone from around 40% of new projects in 2015 to more than 90% in 2025. Over the same period, projects primarily targeting ancillary services fell from around 45% to about 7%, even as the absolute volume of ancillary projects kept growing.
| Primary application | Share of new projects 2015 | Share of new projects 2025 | Source |
|---|---|---|---|
| Energy shifting | ~40% | >90% | IEA (2026) |
| Ancillary services | ~45% | ~7% | IEA (2026) |
The early business case was frequency regulation, a lucrative but shallow market that saturates fast. The current business case is moving bulk energy across the day, which has no ceiling worth worrying about as long as there is surplus solar at noon and demand at eight in the evening.
Build speed is the other structural advantage. Median construction time for utility-scale batteries is around 275 days, close to solar PV at about 220 days and far below gas at over two years or nuclear at more than six. Total time to market runs longer, at roughly two to two-and-a-half years in Europe, the United States and Japan, because permitting, financing and grid connection dominate the schedule rather than construction. Non-construction phases often account for more than half of the total timeline.
How Much Battery Storage Is Stuck in the Interconnection Queue?
This is the section most storage statistics pages skip, and it is where the constraint actually lives.
As of the end of 2025, around 8,200 projects were actively seeking grid interconnection in the United States, representing 1,312 GW of generation and approximately 749 GW of storage, according to Lawrence Berkeley National Laboratory’s Queued Up: 2026 Edition. Active storage capacity fell 16% year over year. Total active queue volume fell 10%.
The more decision-relevant figure sits one layer deeper. Of the 549 GW that already holds a draft or executed interconnection agreement without having reached commercial operations, 161 GW is storage.
US interconnection queue, end of 2025
| Category | Active capacity (GW) | Year-over-year change | Holding draft/executed IA (GW) | Source |
|---|---|---|---|---|
| Solar | 773 | -19% | 256 | LBNL (2026) |
| Storage | 749 | -16% | 161 | LBNL (2026) |
| Natural gas | 253 | +86% | 45 | LBNL (2026) |
| Wind | 220 | -19% | 76 | LBNL (2026) |
Axis original metric: Storage Queue Conversion Ratio (SQCR) v1.0
Baseline reading: 6.7 years, as of the 2026 data vintage.
According to Axis Intelligence Research, US battery storage capacity holding a draft or executed interconnection agreement represents 6.7 years of construction at the record pace planned for 2026.
Formula: SQCR = storage capacity with draft or executed interconnection agreement not yet operational (GW) / planned utility-scale battery storage additions for the current year (GW) Inputs: 161 GW (LBNL, Queued Up: 2026 Edition, data through end of 2025) / 24 GW (EIA, Preliminary Monthly Electric Generator Inventory, December 2025 data, published February 2026) Result: 6.7 years.
A companion ratio: 161 GW at interconnection-agreement stage is 4.0 times the more than 40 GW of battery storage the United States added across the entire preceding five years.
Methodology version: v1.0. This is a baseline reading. No prior readings exist and the metric makes no claim about its own history.
Stated limitations. Capacity at interconnection-agreement stage still withdraws, and LBNL notes late-stage withdrawals appear to be rising. The two inputs carry different vintages and different survey boundaries: LBNL covers transmission-connected projects across seven ISOs/RTOs and 50 non-ISO utilities representing about 98% of installed US generating capacity, while EIA’s inventory covers planned utility-scale units reported through its generator surveys. The ratio is a backlog-to-throughput indicator, not a forecast. It does not assume the 161 GW will be built, and it should not be read as one.
What gives the ratio its weight is the historical conversion rate sitting alongside it. Only 13% of capacity that submitted interconnection requests between 2000 and 2020 had reached commercial operations by the end of 2025. Seventy-five percent was withdrawn and 10% remains active. For regions with available data, the median duration from interconnection request to commercial operation exceeded five years for projects built in 2025.
Aidan Jad: The queue numbers get read as a bullish pipeline and they are closer to the opposite. A 749 GW active storage queue against a 24 GW annual build is not evidence of 749 GW of storage arriving. It is evidence that the studies, the network upgrade cost allocations and the deposits are doing the filtering that the market used to do. The 16% year-over-year drop in active storage capacity is the healthier signal in the dataset, because it means speculative requests are exiting before they consume study cycles that a financeable project needs. Show me executed agreements converting to commercial operation inside three years and I will move the ratio. FERC Order 2023 is being implemented and it is too early to measure its full effect.
How Much Grid Balancing Do Batteries Actually Do?
Capacity statistics describe what is installed. Dispatch statistics describe whether it matters. On the second measure, the change over five years is larger than the capacity growth implies.
California is the clearest case. State solar capacity has grown to over 55 GW, which is greater than peak load, meaning net load on sunny days runs close to zero and at times below it. Battery capacity in the state has grown from less than 1 GW in 2019 to over 17 GW. Batteries covered more than 40% of the state’s power load on the evening of March 29, 2026.
Ramping is the metric that shows the structural shift. Battery storage has gone from contributing less than 1% of California’s hour-to-hour ramping needs five years ago to above 60% in the first quarter of 2026, according to IEA analysis of CAISO data.
Battery contribution to grid ramping, selected markets
| Market | Battery share of ramping | Period | Source |
|---|---|---|---|
| California (CAISO) | >60% | Q1 2026 | IEA analysis of CAISO data |
| Texas (ERCOT) | >40% | April 2026 | IEA (2026) |
| South Australia | >30% | February-March 2026 | IEA analysis of AEMO data |
The IEA notes this analysis uses transmission-level data and excludes ramping from behind-the-meter resources, so the true contribution is understated.
The single-evening records add texture. California’s battery fleet discharged a record 12.99 GW on the evening of July 9, 2026, covering 36.3% of demand across the CAISO region at 8:20 PM PST, according to SEIA. In Texas, batteries supplied a record 20% of evening electricity demand on March 13, 2026, and all renewable resources combined met a record 79% of demand on March 14.
A discrepancy we are flagging rather than resolving. SEIA’s July 2026 summary carries a headline stating that batteries covered 44% of demand on the evening of July 9, while the accompanying chart and body text state 36.3% at the 12.99 GW peak. Axis Intelligence Research cites 36.3% because it is the figure attached to a stated timestamp and a stated fuel mix. The 44% figure may refer to a different hour or a different denominator; SEIA does not specify. We have logged both and cite the specified one.
Why Are Data Centers Buying Battery Storage?
The battery numbers and the AI infrastructure numbers have started converging, and the convergence runs through two distinct product categories that are routinely conflated.
The first is uninterruptible power supply. Battery-based UPS capacity additions, primarily in data centers, rose 30% to 45 GW in 2025 according to the IEA’s Global Energy Review 2026. That is a very large number sitting next to the 108 GW of grid storage, and it measures something different. UPS systems generally provide only short-duration backup to bridge outages until other backup sources start. They are not dispatchable grid assets and they do not shift energy across the day.
The second is grid-connected and behind-the-meter storage procured to firm up renewable supply for data center load. SEIA and Benchmark Mineral Intelligence identify data centers as a driver of demand growth, citing procurement deals from Google, Meta and other technology companies covering tens of thousands of megawatt-hours announced during 2026.
Put the two categories side by side and the total battery capacity flowing toward digital infrastructure in 2025 is larger than the grid-storage figure alone suggests. Axis Intelligence Research is not merging them into a single number, because a 45 GW UPS fleet sized for minutes of bridging power and an 87 GW utility-scale fleet sized for hours of energy shifting do not measure the same thing, and averaging them would produce a figure that means nothing. Anyone who publishes a combined total without separating duration is selling a headline.
This connects directly to the demand side documented in our AI data center energy consumption statistics and the infrastructure spending patterns in our AI data center statistics. For the competing firm-capacity option, our nuclear energy statistics report covers the reactor pipeline that hyperscalers are contracting against.
Aidan Jad: A data center operator has two problems that look similar and are not. Problem one is a grid outage lasting seconds to minutes, and the answer has been UPS batteries plus diesel for decades. Problem two is that the grid cannot deliver the interconnection at all, or cannot deliver it on the schedule the buildout requires. Storage only solves problem two when it is paired with generation the operator controls. That is why the interesting procurements pair solar with multi-hour storage rather than buying batteries alone. A battery with nothing behind it is a timing device, not a power source, and the reason storage keeps showing up in these deals is that a solar-plus-storage package can reach commercial operation in roughly two years while the queue position for a grid injection runs past five.
Methodology
Collection. Every figure in this report was retrieved from a primary source document during production on August 2, 2026. No statistic in this article originates from model memory or from a secondary aggregator. Six source documents were fetched and read in full:
- US Energy Information Administration, New U.S. electric generating capacity expected to reach a record high in 2026, February 20, 2026, based on the Preliminary Monthly Electric Generator Inventory (December 2025 data)
- International Energy Agency, Global Energy Review 2026, Technology: Battery storage chapter
- International Energy Agency, Battery storage is scaling up and taking on a larger system role, commentary, May 29, 2026
- National Renewable Energy Laboratory, Cost Projections for Utility-Scale Battery Storage: 2025 Update, NREL/TP-6A40-93281, June 2025
- Lawrence Berkeley National Laboratory, Queued Up: 2026 Edition, data through end of 2025
- Solar Energy Industries Association and Benchmark Mineral Intelligence, US Energy Storage Market Outlook Q2 2026, released May 21, 2026
Axis-calculated figures. Three metrics in this report are calculated by Axis Intelligence Research, each with formula and inputs disclosed inline at the point of use: the Storage Queue Conversion Ratio (6.7 years), the duration-weighted energy estimate for 2026 US additions (48-96 GWh), and the 2026 cost uncertainty band (36.0%). Regional shares of global 2025 additions are simple arithmetic on IEA country figures. Every calculated row is flagged axis_calculated = yes in the accompanying CSV with a method note.
Source reconciliation. The IEA and the EIA report different figures for US utility-scale battery additions in 2025. The IEA, drawing on Benchmark Mineral Intelligence data, states the United States added 19 GW total with over 16 GW utility-scale. The EIA states a record 15 GW of utility-scale capacity was added. We publish both rather than averaging them. The gap is consistent with differing survey boundaries and commissioning-date conventions: EIA’s generator inventory records units by reported commercial operation through its survey forms, while Benchmark’s tracking captures project energization on a different schedule. Readers modelling US capacity should choose one source and stay with it rather than mixing the two.
A second reconciliation: SEIA’s Q1 2026 segment figures (7.8 GWh utility-scale, 648 MWh commercial and industrial, 515 MWh residential) sum to 8.963 GWh against a stated quarterly total of 9.7 GWh, leaving 7.6% unallocated in the published summary. We report the segment figures as published and do not attempt to distribute the remainder.
Units. Battery capacity is reported in gigawatts (power) or gigawatt-hours (energy), and the two are not interchangeable. Where a source reports one and not the other, we say so. Cost figures are overnight capital costs in 2024 dollars unless stated.
Limitations. Planned capacity additions are developer-reported intentions, not commitments; historical slippage between planned and realized additions is material. NREL’s cost projections were completed in early 2025 and exclude subsequent tariff changes. Interconnection queue data covers transmission-connected projects only and excludes distribution-connected and behind-the-meter systems. IEA regional figures for 2025 are based on Benchmark Mineral Intelligence data and are described by the IEA as approximate for several markets.
About This Dataset
File: battery-storage-statistics.csv
Contents: 158 observations covering global and US battery storage capacity additions, regional shares, state-level 2026 plans, project-level capacity, cost projections by year and scenario, duration and application shares, interconnection queue volumes, and grid ramping contributions.
Temporal coverage: 2015-2050 (historical observations 2015-2026; cost projections to 2050)
Spatial coverage: Global, with country and US state breakdowns
Provenance columns: every row carries source organization, source document, source URL, retrieval date, primary-source flag, Axis-calculated flag, and a method note for calculated rows.
License: CC BY 4.0
Citation: Axis Intelligence Research, Battery Storage Statistics 2026, 2026, axis-intelligence.com/battery-storage-statistics/
How to Cite This Report
APA Axis Intelligence Research. (2026). Battery storage statistics 2026: Global capacity, costs and grid data. https://axis-intelligence.com/battery-storage-statistics/
MLA Axis Intelligence Research. “Battery Storage Statistics 2026: Global Capacity, Costs and Grid Data.” Axis Intelligence, 2 Aug. 2026, axis-intelligence.com/battery-storage-statistics/.
Chicago Axis Intelligence Research. “Battery Storage Statistics 2026: Global Capacity, Costs and Grid Data.” Axis Intelligence, August 2, 2026. https://axis-intelligence.com/battery-storage-statistics/.
Frequently Asked Questions
How much battery storage was installed globally in 2025?
Global battery storage capacity additions reached 108 GW in 2025, about 40% more than in 2024, according to the IEA’s Global Energy Review 2026. Utility-scale systems accounted for roughly 87 GW, or four-fifths of the total, with the remainder installed behind the meter by commercial and residential consumers.
How much battery storage does the US have?
The United States has added more than 40 GW of utility-scale battery storage over the past five years, according to the EIA. Developers plan a further 24 GW in 2026, following a record 15 GW added in 2025.
What does grid-scale battery storage cost per kWh in 2026?
NREL’s mid-case projection puts the 2026 overnight capital cost of a 4-hour utility-scale lithium-ion system at $308/kWh in 2024 dollars, within a low-to-high range of $255 to $366/kWh. That equals roughly $1,232/kW for a 4-hour system at the mid case. The projection was completed in early 2025 and excludes subsequent tariff changes.
Which country installs the most battery storage?
China led global deployment in 2025 with just over 63 GW of additions, which Axis Intelligence Research calculates as 58.3% of the world total, followed by the United States at 19 GW (17.6%) and Europe at around 6.2 GW.
How long do grid batteries last per charge?
The average duration of battery projects commissioned globally rose to three hours in 2025, from around two hours in 2023, according to the IEA. Most projects still cluster around two hours, with a rising share deployed at four hours or more as energy shifting displaces ancillary services as the primary application.
Which US state has the most battery storage?
Texas accounts for 53% of planned 2026 US battery storage additions at 12.9 GW, followed by California at 14% (3.4 GW) and Arizona at 13% (3.2 GW), according to the EIA. Those three states represent about 80% of the national build.
How long does it take to build a battery storage project?
Median construction time for utility-scale batteries is around 275 days, compared with about 220 days for solar PV, over two years for gas and more than six years for nuclear, according to IEA analysis. Total time to market runs two to two-and-a-half years in Europe, the United States and Japan, because permitting, financing and grid connection dominate the schedule.
How much battery storage is waiting in the US interconnection queue?
Approximately 749 GW of storage was actively seeking grid interconnection at the end of 2025, of which 161 GW already holds a draft or executed interconnection agreement, according to Lawrence Berkeley National Laboratory. Axis Intelligence Research calculates that 161 GW as 6.7 years of construction at the pace planned for 2026. Historically, only 13% of capacity entering queues between 2000 and 2020 had reached commercial operation by the end of 2025.
What battery chemistry dominates grid storage?
Lithium-iron-phosphate accounts for around 90% of deployments, up from well below 50% five years ago, according to the IEA. LFP is less energy-dense than the chemistries used in electric vehicles but is cheaper and better suited to frequent cycling, which matters more for a stationary asset than energy density does.
Do data centers use battery storage?
Yes, in two distinct forms. Battery-based uninterruptible power supply capacity, primarily in data centers, rose 30% to 45 GW in 2025 according to the IEA, but provides only short-duration backup to bridge outages. Separately, data center operators are procuring grid-connected and behind-the-meter storage paired with generation, which SEIA and Benchmark Mineral Intelligence identify as a driver of 2026 demand growth.