Global battery energy storage installations will reach 92 gigawatts-247 gigawatt-hours-in 2025, excluding pumped hydro, according to BloombergNEF's latest deployment data. That represents a 23% jump from the 70GW installed in 2024, which itself roughly doubled 2023's figures. The sector has now strung together its fourth consecutive record-breaking year. Compound annual growth projections sit around 14.7% through 2035, at which point annual additions should hit 220GW/972GWh. Cumulative installed capacity worldwide will reach 2 terawatts by the mid-2030s. So yes-the trends grow. The more interesting question is where and how, and who ends up on the wrong side of the growth.
The Year Everything Got Huge
2025 is the year of the gigawatt-hour project.
Not just in the usual places. GWh-scale installations have been commissioned or broken ground in Saudi Arabia, South Africa, Australia, the Netherlands, Chile, Canada, the UK. The Saudis just connected a 7.8GWh project last week across three sites in the southwestern provinces-Sungrow delivered over 1,500 PowerTitan systems in 58 days. Once energized, it becomes the world's largest operational battery system. The previous record lasted about eight months.
BYD signed contracts for 12.5GWh across five Saudi sites earlier this year. Hithium picked up another 4GWh. By mid-2025, roughly 13GWh of grid-scale storage was either operating or finishing construction in the Kingdom, with projections suggesting 33.5GWh by 2026. That would make Saudi Arabia the third-largest storage market globally behind China and the US. A country that essentially didn't exist in energy storage rankings three years ago.
Italy has quietly become Europe's most interesting battery market. Installed systems jumped 23% in the past year by count, but 52% by energy capacity and 40% by power output-the larger format systems pulling up the averages. Two 800MWh utility-scale plants came online in 2024 alone. Aurora Energy Research calls it the most attractive greenfield opportunity in Europe right now, beating even the UK's more mature but increasingly crowded pipeline.
China Still Runs the Table
Over half of 2025's global gigawatt additions will land in China. Not news, but the scale keeps surprising.
Beijing removed storage mandates for new renewable projects earlier this year-a move that should have slowed things down. It didn't. Provincial governments keep announcing their own requirements. In September, a new national storage target appeared, underlining that the mandate removal was administrative housekeeping rather than policy retreat. The shift is toward market-driven growth through spot trading and provincial compensation schemes. Whether that works better than mandates remains an open question. China tends to figure things out.
Chinese manufacturers control roughly 80% of the global battery cell market. The remaining 20% often relies on Chinese components anyway. This dominance extends into every corner of the supply chain-active anode materials, graphite processing, LFP cathodes. CATL, BYD, EVE Energy, CALB, Hithium: these names appear on project announcements from Riyadh to Rome to Queensland.
The overcapacity problem persists. Manufacturing capacity hit 2 TWh against 1.2 TWh of actual demand. Cathode producers have been in loss-making positions for three consecutive years. Some industry observers use the word "involution"-mutually destructive competition that benefits buyers while hollowing out manufacturers. Beijing convened emergency meetings with top producers this year, trying to halt the race to the bottom. LFP technology export restrictions kicked in. The government wants domestic champions, not a crater.
Data Centers Enter the Chat
Here's where things get interesting fast.
Gartner projects data center electricity demand growing 16% in 2025 alone, then doubling by 2030. AI-optimized servers will represent 44% of data center power consumption by 2030, up from 21% today. BloombergNEF revised its US data center power demand projection upward by 36% in December-to 106 gigawatts by 2035.
That's roughly 100 million households' worth of electricity, for context.
Grid connection queues in Virginia have stretched to seven years. Not seven months. Data centers are competing with each other and with everything else for scarce interconnection capacity, transformers, switchgear, and qualified electricians. The bottleneck is no longer capital or construction-it's getting permission to plug in.
Battery storage is emerging as mission-critical infrastructure for this sector. Data centers operate 24/7 with massive swing demands that legacy grids weren't engineered for. Batteries can tame volatility, smooth peaks, provide backup without diesel generators, and-crucially-let facilities energize faster by reducing grid upgrade requirements.
Wood Mackenzie notes that 53GW of "large loads" including data centers and manufacturing facilities will come online over the next decade. Powering them requires, in their words, an "all of the above" approach. Storage is definitively part of that "above."
The Tariff Situation Is a Mess
The US market was supposed to add 15-19GW of utility-scale storage in 2025. That was before April.
Trump raised base tariffs on Chinese imports to as high as 145% in early April. Then a partial pause in May cut some rates to 30% through mid-August. Then additional tariffs on graphite-93.5% on top of existing 25% Biden-era duties-hit in July. Active anode material faces antidumping investigations. The effective tariff rate on Chinese lithium-ion cells for energy storage currently runs somewhere between 54% and 104%, depending on classification and which week you ask.
BNEF's base-case analysis shows 54% tariffs inflating four-hour turnkey system costs by 30% in 2025, to $266/kWh. Their 145% scenario suggests 51-74% lower annual installations between 2025 and 2027 compared to baseline projections.
Anza Renewables reported in June that delivered AC system prices are 56-68% higher than January levels, depending on market segment. That's not a typo. Projects are pausing, contracts renegotiating, supply chains reshuffling.
The chaos has a perverse silver lining: deployment rushed forward to beat deadlines. Imports from South Korea surged 225% quarter-over-quarter as developers sought non-Chinese supply. Stockpiles accumulated. Late-stage projects sprinted to commissioning. The 2025 numbers look okay precisely because everyone expected 2026 to look terrible.
Korean manufacturers-LG Energy Solution, Samsung SDI, SK On-are positioning to fill gaps the tariffs create. LG operates the only LFP cell facility currently producing in the US, at its Michigan plant. Samsung and SK are converting lines or signing supply deals. Domestic content requirements may eventually force localization anyway. The question is timing, scale, and whether these companies can match Chinese cost structures. Current evidence: not really.
Europe Wakes Up
Italy added over 6GWh of new storage in 2024-a 58% year-on-year surge-with utility-scale accounting for 3.4GWh. That's a 15-fold increase in large-scale additions compared to the previous year. By September 2025, the country had 848,814 connected electrochemical storage systems representing nearly 17.5GWh of energy capacity.
The UK reached roughly 5.9GW/8.2GWh of operational BESS capacity earlier this year, with another 6.3GW/13.2GWh approved and working through development. Aurora Energy Research projects 10.3GW installed across Europe today growing to 55GW by 2030 and 126GW by 2050.
Germany crossed two million connected battery systems in July, with 14.5GW of installed power and 22GWh of available capacity. Finland introduced 20% tax reductions for PV-plus-storage installations. Romania exempted storage systems from various grid fees, cutting operational costs 20-30%. The Netherlands allocated €100 million for battery deployment subsidies this year.
Grid-scale overtook distributed storage in Europe for the first time in 2024. The growth engine is shifting.
InfoLink projects the European market adding 27GWh in 2025, up 41% year-on-year. Italy keeps its crown, Germany ranks second, and utility-scale installations in Spain, Belgium, and other countries are diversifying the landscape away from the traditional UK-Germany-Italy triangle.
What About Alternatives?
Flow batteries got some buzz. Deserve some.
Eos Energy and Invinity Energy Systems have benefited from the UK's LDES Cap and Floor scheme. ESS Tech commissioned a 10MW/100MWh iron-flow system in California in September, supporting 12-hour discharge. Form Energy broke ground on its first commercial iron-air manufacturing facility in West Virginia, targeting 100-hour discharge systems.
The problem: none of these companies has seen operating profits yet. Eos jumped revenue 240% year-on-year in H1 2025 but also deepened its net loss to $207.8 million. The non-lithium chemistry market exists but remains limited. Most developers stick with what works at scale.
Lithium-ion already accounts for 70% of the 64.7GWh of inter-day (8-12 hour) long-duration projects targeting operations by 2030. Compressed air is second at 20%. Everything else-flow batteries, liquid air, thermal, gravity-splits the remaining 10%.
BNEF notes that thermal and compressed air storage have closed the capex gap with lithium-ion for longer durations: $232/kWh and $293/kWh respectively, versus $304/kWh for four-hour lithium systems in their 2023 survey. But closing a gap and winning a race are different things. Lithium-ion keeps improving. The target moves.
California's 2027 centralized procurement for up to 1GW of 12-hour storage explicitly excludes long-duration lithium-ion batteries-a deliberate policy choice to give emerging technologies a proving ground. Whether they prove anything remains to be seen.
So, Yes-Obviously
The market grows. That question was barely a question.
Annual additions hit another record. Cumulative capacity accelerates. Emerging markets contribute increasing shares-Middle East and Africa jumped 381% in projected 2025 capacity, becoming the fastest-growing region globally. Pack prices fell 45% for stationary storage in 2025, hitting $70/kWh and making batteries the cheapest lithium-ion segment for the first time ever.
The growth sits atop fragile foundations, though. Chinese manufacturing dominance means geopolitical tension translates directly into price shocks. US tariffs have already demonstrated how quickly costs can spike when trade policy shifts. Grid interconnection queues stretch years in both the US and UK. Data center demand could overwhelm infrastructure before infrastructure catches up.
But growth? Yes. The numbers are unambiguous. Mordor Intelligence projects the market reaching $465 billion by 2030. Fortune Business Insights sees battery storage hitting $114 billion by 2032. GM Insights forecasts the broader energy storage systems market exceeding $5 trillion by 2034. The forecasting firms can't agree on exact sizing-their methodologies diverge substantially-but the direction is unanimous.
What happens inside that growth matters more. Whether Saudi Arabia's ambitions materialize on schedule. Whether European grid-scale catches up to residential. Whether American manufacturing ever approaches Chinese costs. Whether data center operators can secure power before customers walk away. Whether alternative chemistries carve out niches or fade into footnotes.
Those are the interesting questions. Growth just provides the backdrop.