Editorial Article

Stationary Energy Storage: The $248 Billion Market Powering the Grid

Updated September 12, 2026

Every lithium-ion cell has two possible destinations. One goes into a vehicle and is ultimately sold to a driver. The other stays where it is put — in a container beside a substation, or a rack behind a data centre’s fence — and is sold to a market. The first destination gets the publicity and most of the commentary. The second is becoming a comparable customer, and it behaves almost nothing like the first.

Pheonix Research sizes the global stationary energy storage systems market — grid-scale, commercial and industrial, and residential installations built on lithium-ion cells and, increasingly, on sodium-ion — at USD 64.54 billion in 2025, rising to USD 247.68 billion by 2033 at an 18.31% CAGR. That is a 3.84x expansion in eight years.

The comparison worth holding onto is with the vehicle business. Pheonix expects the global EV battery market to grow from USD 91.70 billion in 2025 to USD 369.50 billion by 2033, a 19.03% CAGR. Stationary storage is therefore worth roughly 70% of the EV battery market in 2025 and about 67% of it by 2033 — the same order of magnitude, growing at a similar rate, drawing on the same cell factories and, in several cases, the same cells. Any view of cell demand for the next decade that leaves storage out is missing a substantial share of it.

The year the power system noticed

2025 settled the question of whether grid storage was a niche. The IEA’s Global Energy Review 2026 puts global battery storage capacity additions for the year at 108 GW, up around 40% on 2024, with cumulative installed capacity now about eleven times its 2021 level. Annual additions at that scale exceed the historical peak for gas-fired capacity additions — roughly 107 GW, set in 2002 — which is worth stating without decoration: in one year the world installed more battery power than it ever installed gas power.

The composition matters as much as the total. About 80% of new capacity, some 87 GW, was utility-scale, with the remainder behind the meter in commercial and residential installations. Roughly 24 GW of the utility-scale additions were co-located with renewables, a share that fell just below 30% as market reforms in China early in 2025 removed broad co-location mandates — a reminder that in this market the configuration of an asset is often a policy artefact rather than an engineering verdict.

Chemistry followed the application. LFP cells now account for around 90% of deployments, against less than half five years ago. Storage is a different duty cycle from driving: it cycles hard, often daily, and cares about cost per cycle more than energy per kilogram. The chemistry that won in cars for range is not the chemistry that wins here — and the two markets are therefore pulling on different parts of the battery supply chain even when they buy from the same supplier.

Power is not energy, and the difference is the market

The unit of account in this industry is the source of most misreadings of it. Additions are reported in gigawatts of power; the cells are bought in megawatt-hours of energy; the two are linked only by duration. Most projects still cluster around two hours, but an increasing number are built for four hours or more, and durations are gradually lengthening.

That matters more than any single deployment figure. A four-hour system uses roughly twice the cells of a two-hour system of the same power rating. If duration drifts from two hours to four across a meaningful share of the fleet, the energy-storage market can roughly double in cell terms while reported GW additions look flat. This is the mirror image of the EV thermal-management market, where the story is content per vehicle; here the story is energy content per megawatt. Forecasts that track megawatts and ignore hours are tracking the smaller half of the market.

Where the megawatts actually went

China remained dominant, adding just over 63 GW, around a third more than in 2024 and roughly 60% of global additions. Utility-scale accounted for some 55 GW of that and behind-the-meter about 8 GW. The United States added 19 GW, up around 60%, with over 16 GW utility-scale. The EIA’s generator inventory shows the US power system holding 43.6 GW of operational battery storage at the end of 2025, after a record year for utility-scale additions, with a further 8.3 GW added in the first half of 2026 to reach nearly 52 GW nameplate, and developers planning about 24 GW of utility-scale battery capacity for 2026 alone.

Europe added around 6.2 GW, slightly below 2024 — but with a clear structural shift, as utility-scale additions more than doubled to about 4.6 GW while smaller systems lagged. Australia was the year’s surprise: additions surged to nearly 8 GW, roughly nine times the 2024 level, with utility-scale rising from under 1 GW to about 4.2 GW and behind-the-meter from roughly 0.2 GW to about 3.4 GW, on the back of state and federal incentives. The Middle East topped 3 GW, more than tripling, almost entirely in Saudi Arabia, and Chile approached 1 GW as utility-scale batteries were deployed to absorb surplus solar and meet evening peaks.

Pheonix identifies the United States, China, Europe and India as the leading regions, supported by renewable deployment, incentives, manufacturing investment and grid-modernisation programmes. India’s case is instructive for a different reason: the report’s connected entities include the transmission operator Power Grid Corporation of India, a reminder that in this market the buyer is frequently a state-owned grid rather than a private developer.

A fragmented market with no winning model

Pheonix characterises stationary storage as fragmented, with high competitive intensity and ten tier-one players, high capital intensity, rising M&A activity, high supply-chain complexity and a vertically integrated operational model — with technology at an emerging maturity stage despite rapid deployment.

The cast of participants the report names explains the fragmentation. It includes integrators such as Fluence, project developers such as Peak Energy, cell and pack ventures including Ultium Cells, a flow-battery maker in Invinity Energy Systems, a second-life specialist in B2U Storage Solutions, and battery-management and controls suppliers such as Dukosi, Nuvation Energy and A123 Systems, alongside iNVERGY and Zenergy Battery. That is not one industry with one business model; it is several candidate industries competing to become one, and no technology or commercial design has yet won decisively.

The risk is not where the rest of the chain’s risk is

Stationary storage carries the highest overall risk rating in the battery cluster Pheonix covers, with high geopolitical exposure and — unusually — high substitution risk, where EV thermal management sits at low on both counts.

The geopolitical exposure is straightforward enough. LFP cell manufacturing and the processing steps behind it are heavily concentrated, and export controls and trade policy reach them. The substitution risk is the more interesting one, because it comes from outside the battery industry entirely: a market that sells flexibility competes with gas peakers, pumped hydro, transmission reinforcement, demand response and grid-operating practices that reward not building anything at all. Within chemistry, sodium-ion and long-duration non-lithium designs pressure the same spend from below. Storage has no protected demand.

The revenue question is a policy question

A battery that stores electricity has no intrinsic revenue. It earns from wholesale arbitrage, ancillary services, capacity payments and behind-the-meter bill savings — every one of which is defined by market design, not by physics. That is why the same asset can be a strong business in one jurisdiction and a stranded one in the next, and why the year’s most consequential storage event may not have been a gigafactory opening but China’s early-2025 reform of co-location mandates, which immediately changed both the configuration and the economics of new projects.

It is also why utilisation deserves the same attention here that it does in EV charging infrastructure. A storage asset with three cycles a week is a different investment from one with two cycles a day, and the difference is set by tariff structures and dispatch rules rather than by the hardware. Pheonix lists the growth drivers as renewable integration, grid modernisation, energy resilience, supportive policy, falling battery costs, manufacturing capacity expansion, and rising electricity demand from data centres and electrification — of which the last is the newest and, in several markets, the fastest-moving. Data-centre load does not merely increase demand for power; it increases demand for power that arrives on a schedule.

What to watch

Three things will determine whether this market delivers its 18.31% trajectory or beats it. First, duration: whether four-hour systems become the default as solar shares rise, because that decides cell volumes far more than gigawatt additions do. Second, market design: whether merchant revenue stacks become deep and bankable enough to fund the pipeline, or whether storage stays a policy-supported asset whose returns are rewritten every time a capacity mechanism is reviewed. Third, substitution: whether LFP’s dominance is entrenched by sodium-ion at the low end and long-duration non-lithium technologies at the top end, or whether lithium-ion simply absorbs both applications as it has absorbed most others.

The market also deserves to be read against its sibling in this series. Charging infrastructure and stationary storage have the same underlying problem — a capital-intensive asset whose return depends on how often it is used and at what price — and the same underlying advantage: they sell into a system that must change, rather than to a consumer who can wait. The difference is who signs the contract. A charger’s customer is a driver. A storage asset’s customer is a grid.

That is the quieter half of the battery business, and on Pheonix’s numbers it will soon be worth close to two-thirds of the EV battery market it shares its cells with. The batteries that never move are not the ones the industry talks about. They are increasingly the ones it is building for.

Sources and further reading

Pheonix Research market intelligence:

Primary and reference sources:

IEA, Global Energy Review 2026 — battery storage chapter and accompanying commentary: 2025 capacity additions, deployment shares by segment and market, LFP share, and the comparison with historical gas-fired additions.
US Energy Information Administration, Today in Energy and the Preliminary Monthly Electric Generator Inventory — US operational battery storage capacity, annual additions and planned 2026 capacity.

Frequently Asked Questions

How big is the stationary energy storage market?

Pheonix Research sizes it at USD 64.54 billion in 2025, rising to USD 247.68 billion by 2033 at an 18.31% CAGR — a 3.84x expansion over eight years. The figure covers grid-scale (utility-scale), commercial and industrial, and residential installations, built on lithium-ion cells and increasingly on sodium-ion.

How does stationary storage compare with the EV battery market?

Closely. Pheonix expects the EV battery market to move from USD 91.70 billion in 2025 to USD 369.50 billion by 2033, a 19.03% CAGR. Stationary storage is therefore worth roughly 70% of the EV battery market in 2025 and about 67% by 2033 — the same order of magnitude, growing at a similar rate, and drawing on the same cell factories. Any cell-demand forecast that leaves storage out is missing a substantial share of demand.

Which regions lead the market?

China remained dominant in 2025, adding just over 63 GW — roughly 60% of global additions, of which about 55 GW was utility-scale. The United States added 19 GW, up around 60%, and held 43.6 GW of operational battery storage at the end of 2025 on EIA figures, reaching nearly 52 GW nameplate by mid-2026. Europe added around 6.2 GW, slightly below 2024 but with utility-scale more than doubling to about 4.6 GW. Australia surged to nearly 8 GW, roughly nine times 2024, on state and federal incentives. The Middle East topped 3 GW, more than tripling, almost entirely in Saudi Arabia, and Chile approached 1 GW. Pheonix names the United States, China, Europe and India as the leading regions.

How does a storage project actually make money?

A battery that stores electricity has no intrinsic revenue. It earns from wholesale arbitrage, ancillary services, capacity payments and behind-the-meter bill savings — every one of which is defined by market design rather than physics. That is why the same asset can be a strong business in one jurisdiction and a stranded one in the next, and why China's early-2025 reform of co-location mandates was among the year's most consequential events for the sector.