Grid-Scale Battery Storage Surges as Sodium-Ion Challenges LFP

A month of large battery projects

October 2026 brought a long run of grid-scale battery announcements from several continents. Taken together, they show that battery storage has stopped being a supporting player in the energy transition. Utilities, banks and governments now plan around it as core grid infrastructure. The bigger shift is less visible: lithium iron phosphate (LFP) has come to dominate the market almost completely, and a well-funded effort is now underway to show that it does not have to stay that way.

The projects themselves are large and widely spread. In Germany, Dutch developer Giga Storage closed €260 million in debt from a group of seven banks for "Albatross." The 350 MW / 1.4 GWh four-hour system near Neuenkirchen in Lower Saxony will connect directly to the 380 kV transmission grid.2 Transmission operator Amprion called it the company's first grid-connection contract for a large battery. CATL will handle engineering, construction and maintenance.2 In New Zealand, Genesis Energy opened a 100 MW / 200 MWh battery at the coal- and gas-fired Huntly power station. It took 14 months to build and cost about NZ$130 million, which Genesis said was below budget.3 A second unit of the same size is approved and due to be fully operating in 2028.8

Japan added two more. Sumitomo Corporation started building a 62 MW / 291 MWh facility in Namie, Fukushima, with financing led by MUFG Bank and commercial operation planned for 2029.4 Trina Storage won the contract to supply a 50 MW / 196 MWh project in Oita Prefecture using its 314 Ah LFP cells.5 In South Australia, the 240 MW / 960 MWh Summerfield battery began commissioning less than two years after construction started. Its title as the state's largest battery may not last, because larger projects such as Alinta's Reeves Plains are already moving forward.10 India's Bondada Engineering received a letter of intent for a 20 MW / 120 MWh vanadium redox flow battery in Gujarat, which the company says will be the country's largest grid-scale flow battery.7

Records, though the trackers disagree on the numbers

The market data shows the same trend, but the totals differ depending on who is counting. The Solar Energy Industries Association and Benchmark Mineral Intelligence put US installations at a record 20.2 GWh in the second quarter of 2026. That brought the first half of the year to about 30 GWh, up 23% from a year earlier.30 Wood Mackenzie and the American Clean Power Association also called it a record quarter but counted 18.9 GWh.25 The gap comes from different methods, and both groups agree that utility-scale projects made up roughly nine-tenths of the total.24

The two camps tell different stories about what is behind the record. Wood Mackenzie points out that installed power actually dropped 7% year over year to 5.4 GW. Energy capacity rose only because developers are building longer-duration systems, as markets like California and Texas start to run short of value for short-duration batteries.21 The firm describes utility-scale deployment in megawatt terms as stagnant through 2027 while supply chains adjust, and expects growth to pick up again in 2028 as domestic cell manufacturing scales.28 Benchmark's analysts say much of the quarter came down to a few very large projects, with seven systems above 1 GWh coming online.26

Both readings hold up. The industry is setting records, but the headline figures depend on a few large projects and on a move toward four-hour and longer systems, not on steadily rising power capacity. That matters because the record GWh figures can make the market look healthier than the flat MW numbers suggest.

Looking globally, BloombergNEF expects additions of 188 GW / 535 GWh in 2026, a 45% rise in power capacity, followed by 209 GW / 638 GWh in 2027.27 China, the US and Australia account for 62% of the 2026 total. India is the big jump, with tender-driven additions expected to pass 6 GW after only a few hundred megawatts the year before.29 BNEF also lists a group of European countries reaching gigawatt-scale additions for the first time, including Bulgaria, Romania, Poland and Greece.29

Policy is shaping when projects get built

The US picture is closely tied to tax policy. SEIA reports that developers safe-harbored more than 200 GW of projects before a July 2026 deadline under the HR1 tax changes.30 Wood Mackenzie dropped an earlier forecast of near-term contraction for the same reason: many projects started construction to secure investment tax credit eligibility. It now expects cumulative US capacity to reach 207 GW and 715 GWh by 2031.25 Home batteries are the weak spot. Residential storage fell 27% year over year in the second quarter after the Section 25D homeowner credit ended.24 Even so, the share of home solar systems sold with a battery rose to 46%.24

Other countries are using long-term revenue contracts rather than tax credits. Both Japanese projects were selected through the country's Long-Term Decarbonization Power Source Auction. The Nozu project's contract gives it inflation-linked capacity payments for 20 years.5 In the US, New York's first bulk storage solicitation awarded contracts to two Flatiron Energy projects.6 Seven States Power is seeking suppliers for a 220 MW distributed battery program across Tennessee, backed by a $439 million federal rural-energy award.9 Across these cases, the common thread is that projects get financed when their revenue is predictable.

Timelines still slip. Giga Storage's Albatross was originally expected to start operating in 2028 and is now scheduled for early 2029.2 Grid connections and financing remain bottlenecks even at this scale.

LFP's dominance, and the sodium-ion challenge

LFP's position is the single biggest fact about this market. SEIA says LFP accounts for 77% of deployed US storage and 90% of what has been installed so far in 2026.30 Almost every project above uses LFP or another lithium chemistry. The Indian flow battery is the main exception.7

That is why General Motors' sodium-ion move deserves attention. On October 23, GM-backed startup Peak Energy said it had raised $108 million in an oversubscribed Series B round led by GM Ventures, with LG Technology Ventures also investing.15 The money builds on a partnership announced in June. GM develops the cells at its Michigan labs and keeps exclusive manufacturing rights, and Peak builds them into its storage systems.13 Peak says it will use the funds to expand US manufacturing, fill more than $1 billion in existing orders, and sell to data centers.15

Peak's case rests on its system design as much as on the cell chemistry. Its NFPP-based sodium-ion systems use passive cooling instead of the active cooling LFP installations need. The company says this cuts energy costs by up to 20% and uses no water.17 It also claims more than 99% uptime and estimates that replacing LFP across the US could save up to 2 TWh of wasted energy a year.11 These are the company's own figures and have not been independently verified.15

Coverage of the deal largely agrees on the framing. GM's Kurt Kelty argues that the use case should decide the battery chemistry, and that sodium-ion's lower energy density rules it out for GM's EVs but suits stationary storage.13 The bigger disagreement is over timing and supply chains. GM's US-made cells are still years away, and commercial production is reportedly not expected until around 2029.17 For now, Peak buys commercial cells from Chinese suppliers.17 The broader cost comparison is also not yet in sodium's favor. Late last year, LFP batteries sold for about $81 per kWh, compared with roughly $100 per kWh for one sodium-ion startup's product.14 The field has also seen high-profile failures, including Natron Energy shutting down in 2025.20

What it means for EV battery makers

The sodium-ion deal is one part of a wider move by carmakers into stationary storage. GM and LG Energy Solution's joint venture, Ultium Cells, started making LFP storage cells at its Spring Hill, Tennessee plant this year.17 GM is also working with Redwood Materials to put roughly 10,000 of its batteries into energy infrastructure, including at an AI data center in Nevada.17 In the same quarter, GM booked $238 million in charges tied to resizing its EV battery supply chain.17

The most likely explanation is that grid storage has become a place to use battery factory capacity that slower EV sales left idle, and the data-center power boom gives that strategy a reason to grow. In the near term, LFP will keep winning nearly every large contract, as this month's projects show. Over a longer horizon, the cell chemistry is starting to be contested. If Peak's cooling and cost claims hold up at scale, sodium-ion could take a real share of the stationary market by the end of the decade. Until then, the gigawatt-hours being installed today are almost entirely lithium.