Sodium-Ion and Iron-Air Batteries Reshape Grid Storage in 2026

The year battery hype met the production line

For most of the past decade, "next-generation battery" usually meant a press release and a promised date. In 2026 that changed in a specific and uneven way. Some chemistries reached factories, customers and the grid. Others are still waiting on manufacturing. Across the coverage, the main finding is that the most important battery developments this year are not the ones that got the most attention. Sodium-ion has moved ahead of solid-state on the road to commercial use. Lithium iron phosphate (LFP) has become the standard for grid storage. And a group of long-duration storage startups, led by iron-air developers, is turning data-center electricity demand into large order books.

The reporting largely agrees on a short list of chemistries that matter for vehicles: solid-state, sodium-ion and silicon anodes.1 It disagrees much more on how far each one has actually progressed.

Sodium-ion arrives, with a price premium

The main milestone for EV batteries this year came in February. CATL and Changan unveiled the Changan Nevo A06, presented as the first mass-production passenger car with a sodium-ion pack.12 CATL rates its Naxtra cells at up to 175 Wh/kg. That is near LFP, and the company wants full LFP-level density within three years.1 According to the International Energy Agency, the best sodium-ion cells sit around 175 Wh/kg, against roughly 205 Wh/kg for LFP and 255 Wh/kg for NMC.8

The case for sodium rests less on energy density than on what the cells avoid and where they work well. Sodium is far more abundant than lithium. The pack in the Nevo A06 is specified to operate from −40°C to +70°C. CATL says the cells keep about 90% of their capacity at −40°C.1 On the less favorable side, sodium atoms take up more than twice the volume of lithium atoms and weigh three times as much. As a result, sodium-ion packs currently hold about 30% less energy than lithium equivalents.2

This is where the coverage splits. Some enthusiastic reporting says sodium cells already match LFP density at potentially 30% lower cost. It also cites CATL's projection that sodium-ion could eventually take 30% to 40% of the battery market.6 More cautious analysis cites the only public price test so far. In a September storage tender in Renqiu, Hebei, sodium-ion systems drew average bids of about $153 per kWh, against $79 per kWh for LFP in the same project.3 The same analysis notes that no completed CATL sodium-ion storage delivery had been independently reported by early October. It also describes the 60 GWh CATL–HyperStrong supply agreement as a contract, not proof of shipments.3

The cautious reading is the better-supported one. Sodium-ion is a real product in 2026, but it costs more today, and its supply depends mostly on one company. Policy could close the gap faster than chemistry. Hard-carbon anode costs are reported to have fallen sharply since 2024. China's new battery consumption tax applies to lithium-ion from September 2026 but exempts sodium-ion until the end of 2028.3 CATL has also launched a stationary product, TENER Sodium, with a stated 15,000-cycle life.5 That suggests grid storage, rather than cars, may be where sodium's economics work first.

Solid-state: real shipments, but semi-solid

Solid-state batteries are the other half of the EV story, and the gap between claims and evidence is wider here. One account says Ganfeng has begun small-scale production of a 500 Wh/kg solid-state cell. It adds that the first production vehicle on a solid-state pack is a motorcycle.1 That matches reporting that QuantumScape and PowerCo showed solid-state cells in a Ducati motorcycle at IAA Mobility.3

The industry data is less exciting. Global solid-state shipments grew 51.9% to 8.0 GWh in 2025. All of it was semi-solid, meaning the cells still contain some liquid electrolyte. All-solid-state mass production is dated to 2027 at the earliest, with Samsung SDI targeting the second half of that year.3 The IEA's position, as summarized in one guide, is that all-solid-state cells remain at prototype or small testing scale.5 Meanwhile, LFP made up 81.2% of China's domestic traction-battery installations in May 2026.5 Solid-state is a technology for later in the decade. The batteries going into cars right now are still overwhelmingly lithium-ion, and increasingly LFP.

LFP takes over the grid

Grid storage is where batteries are changing the industry fastest, and the chemistry doing it is not new. The IEA reports that LFP now accounts for about 90% of storage deployments, up from under 50% five years ago.1118 A separate quarterly figure puts LFP at roughly 72% of US installations by energy capacity in Q2 2026.13 The two numbers measure different markets and periods, but both show LFP pulling away from other chemistries.

The US deployment numbers are striking. The Energy Information Administration expected developers to add 24 GW of utility-scale storage in 2026, up from a record 15 GW in 2025. Storage makes up 28% of all planned US capacity additions, second only to solar.1120 SEIA and Benchmark Mineral Intelligence reported a record 20.2 GWh installed in Q2 2026, bringing the first half of the year to 30.8 GWh.18 Utility-scale projects accounted for 17.9 GWh of that quarter. Commercial and industrial storage, driven largely by data centers, added 1.8 GWh.18 Batteries supplied more electricity to the US grid in the first eight months of 2026 than in all of 2025.12 Texas leads the planned 2026 pipeline with 12.9 GW.11

The growth is global. China connected a 1 GW / 4 GWh grid-forming LFP plant in Inner Mongolia in August, described as the world's largest single-site facility of its kind.14 BloombergNEF expects global additions of 158 GW / 459 GWh in 2026, a 41% increase over 2025.20 One weakness stands out. Most US systems still use cells made in China or South Korea, which leaves the buildout exposed to supply-chain risk.13

Long-duration storage finds a buyer: data centers

LFP works well for four-hour daily balancing, but its costs rise steadily with duration. That leaves room for chemistries built to last for days. BloombergNEF expects long-duration storage additions to quadruple to 2 GW in 2026, mostly from non-lithium technologies.20

Form Energy is the clearest example. Its iron-air battery stores energy for 100 hours through "reversible rusting." Its largest deal is with Xcel Energy, which will back a Google data center in Minnesota with 30 GWh of Form batteries. If built, that could be the world's largest battery project by energy capacity.29 Form's first commercial system, a 150 MWh installation for Great River Energy, is due online in 2027.29 Its total funding, counting grants, now exceeds $2 billion.29

The main problem is manufacturing capacity, not demand. Form's factory produces about 2 GWh of storage per year, against roughly 80 GWh promised through commercial agreements.2926 Its long-stated $20/kWh cost goal also needs a caveat. Some analyses present it almost as an achieved figure.15 Form itself declined to disclose current system costs.29 Iron-air also has lower round-trip efficiency than lithium-ion, around 40–50% by one estimate, and needs a lot of land.15

Europe is following. Dutch startup Ore Energy signed a 1 GWh iron-air deal with supplier Budget Thuis, starting with a 400 MWh phase in 2028. That follows a grid-connected pilot with EDF in France.17 Form also has a 1,000 MWh project with Futur Energy Ireland due by 2029.19 Other long-duration startups are raising money too. UK-based Certain Energy raised $14 million for manganese flow batteries.27 Thermal storage company Antora raised $550 million.26

Where carbon capture fits

Carbon capture is a smaller and more volatile part of the climate-tech funding picture. One estimate gives carbon removal and capture about 10% of 2026 climate venture dollars and labels the category volatile. Energy and grid companies take about 35%.21 Carbon-management firms such as Twelve, Svante and Climeworks are valued around $1 billion, well below leading battery and storage companies.23

The capture deals that do close tend to sit next to energy systems. Mantel raised $18 million for molten-salt carbon capture.27 Reverion's $175 million Series B funds plants that combine dispatchable power, storage and biogenic CO2 capture.24 Captura and ElectraLith are pairing ocean carbon capture with lithium extraction.27 Remora sells CO2 captured from exhaust to breweries and greenhouses, with the aim of not depending on tax credits or carbon credits.22 The pattern is that capture companies get funded more easily when they have a second source of revenue, often tied to energy or battery materials.

Money is moving to scale

The funding data points the same way as the deployment data. Investment is concentrating in fewer, larger companies. One analysis finds that 2025 climate deal counts fell 18% while total capital rose.26 Base Power, a distributed-battery company founded in 2023, was valued at $13 billion after its latest raise.26 Redwood Materials is turning used EV packs into grid storage before recycling them.26 Northvolt is the warning case: it raised more than $10 billion before going bankrupt.23

The overall picture for 2026 is fairly clear. The changes in the battery industry are coming from scale, policy and new buyers more than from breakthrough chemistry. LFP is winning the grid, sodium-ion is entering through the cold-climate and stationary markets, and data centers are paying for multi-day storage. Solid-state may still be important later in the decade. This year, though, the cheaper and more widely available chemistries are the ones actually being deployed.