Form Energy’s Iron-Air Batteries: Reversible Rusting for 100-Hour Grid Storage (2026)

Most grid batteries today are lithium-ion packs sized for a few hours of evening peak. That works when the sun sets on a clear day. It does not cover a week of cloudy, still weather, a polar vortex that freezes wind turbines in place, or a heat wave that stretches demand across several nights. For that, utilities need something cheaper per kilowatt-hour than lithium, even if it is heavier and slower: multi-day storage.

Form Energy, a Somerville, Massachusetts company founded in 2017, is betting that the answer is iron and air. Its batteries store electricity for about 100 hours by "reversible rusting" iron metal into iron hydroxide and back again. In October 2026, MIT Technology Review named Form one of its Climate Tech Companies to Watch, citing commercial traction at a West Virginia factory, a first Minnesota project due online in 2027, and a deal that could put roughly 30 gigawatt-hours of iron-air storage behind a new Google data center.

Here is how the chemistry works, why multi-day storage matters next to lithium and sodium-ion packs, where Form’s factory and contracts stand in 2026, and what still has to go right.

Reversible rusting, explained

Iron-air batteries are not a new idea. NASA experimented with rust-based cells in the 1960s, and alkaline iron electrodes have been studied for decades. Form’s commercial pitch is to make the chemistry cheap enough for multi-day grid duty, not for cars or phones.

In discharge, the battery "breathes in" oxygen. At the air electrode, oxygen is reduced in a water-based alkaline electrolyte to hydroxide ions. Those ions move to an iron electrode and oxidize metallic iron to iron(II) hydroxide, often written Fe(OH)₂ — the first stage of rust. Electrons leave through the external circuit as usable current. In charge, the process runs in reverse: applied current reduces the hydroxide back to iron metal, and oxygen is released at the air electrode. Form describes the whole cycle as reversible rusting.

The electrodes and electrolyte are built around abundant materials. Iron is one of the cheapest and most common metals on Earth. The electrolyte is water-based rather than the flammable organic solvents used in many lithium cells. Energy density by weight is low compared with lithium-ion, which is why Form is not chasing electric cars. For stationary storage, cost per kilowatt-hour of capacity and the ability to discharge for days matter more than packing energy into a light pack.

Researchers writing in One Earth have argued that only a short list of chemistries can hit the roughly $20 per kilowatt-hour installed-cost target often cited for multi-day storage, because chemical costs alone for lithium-ion and vanadium flow batteries already sit above that floor. Iron-air, with cheap iron and "free" air, is on that short list. Form’s public cost goal matches that figure; the company has declined to publish a current system price.

Why multi-day storage is different from lithium

Lithium-ion grid batteries have scaled fast for a reason. They are energy-dense, efficient, and well understood. Most large installations are designed for about four to six hours of discharge: enough to shift afternoon solar into the evening peak or to firm a few hours of wind. They are expensive to oversize for a hundred-hour event. You would be paying lithium prices for capacity that sits idle most of the year.

Multi-day storage is meant for a different duty cycle. Think of a winter week when a high-pressure system parks over a region, wind drops, and demand stays high for heating. Or a summer stretch when solar is strong at noon but overnight air-conditioning keeps the grid stressed for several nights in a row. Gas peakers traditionally cover those gaps. Long-duration batteries aim to cover them with energy stored from earlier wind and solar surpluses.

That is a complement to, not a replacement for, shorter-duration lithium — and to other long-duration bets such as pumped hydro, compressed air, flow batteries and gravity systems. Sodium-ion packs are also moving into grid niches where cost and cold-weather behavior matter; iron-air is aimed even further out on the duration axis. Solid-state cells, meanwhile, are still mostly an electric-vehicle story. The grid needs a portfolio: fast, dense chemistry for daily cycling, and cheap bulk capacity for rare multi-day shortfalls.

Storage type	Typical duration	What it is good at
Lithium-ion (grid)	About 4–6 hours	Daily peak shifting, fast response
Sodium-ion (emerging)	Hours-scale, cost-focused	Grid niches, cold climates, cheaper materials
Iron-air (Form)	Up to ~100 hours	Multi-day renewable gaps, firming over weather systems
Pumped hydro	Hours to days	Bulk storage where geography allows

Lithium-ion packs win the daily peak. Iron-air is built for the week when the wind does not blow.

Form Factory 1 in Weirton, West Virginia

Form builds its batteries at Form Factory 1 in Weirton, West Virginia, on the historic grounds of the former Weirton Steel mill — a Rust Belt site that once defined American tinplate production. The company describes a roughly one-million-square-foot manufacturing campus that was built in about a year, now employing more than 500 people, with plans to expand floor space and headcount toward more than 750 employees and at least 500 megawatts of annual battery production capacity by 2028.

The symbolism is deliberate: a steel town remade for iron-based energy storage. West Virginia and U.S. Department of Energy manufacturing grants have backed the plant; Form’s total capital raised, including those grants, has been reported above $2 billion after an August financing round highlighted in the MIT Technology Review profile. CEO Mateo Jaramillo has said the factory has already started commercial production and is on a path to that 500 MW-per-year rate.

Scaling remains the hard part. MIT Technology Review noted that Form’s existing production capacity, on the order of 2 gigawatt-hours of storage per year in that account, is still a fraction of the roughly 80 gigawatt-hours the company has promised through commercial agreements. Success hinges on electrode yields, cost trajectory and the ability to outfit more lines at Weirton.

Great River Energy: the first 150 MWh commercial system

Form’s first commercial deployment is the Cambridge Energy Storage Project with Great River Energy, a Minnesota generation and transmission cooperative. The pilot is rated at about 1.5 megawatts of power and 150 megawatt-hours of energy — a 100-hour system at that power rating. Great River Energy and Form broke ground in Cambridge, Minnesota, in August 2024, with Minnesota contractor Mortenson as the engineering, procurement and construction partner.

MIT Technology Review reported in October 2026 that Form had begun producing batteries at the West Virginia factory for that Great River Energy system, which is set to come online in 2027. Latitude Media separately reported in late 2025 that Form had started deploying its first commercial batteries for the project. The cooperative plans to study how multi-day discharge works in real grid operations: storing surplus when production is high, then releasing energy over multiple days of high demand or low renewable output, including emergency conditions when other resources are scarce.

Xcel, Google and a ~30 GWh Minnesota bet

The largest announced Form project sits inside a broader clean-energy package for a new Google data center in Minnesota. Utility Xcel Energy plans to add roughly 1,400 megawatts of wind and 200 megawatts of solar, plus a 300 megawatt / 30 gigawatt-hour Form iron-air system — about 100 hours at that power rating. Xcel has called it the largest battery project by energy capacity announced to date. Phased operation is expected between 2028 and 2031.

Google is covering costs for its new electric service under the agreement and has said it worked with Xcel so existing customers are not left holding the bill for the large-load expansion. The package also includes a $50 million investment in Xcel’s Capacity*Connect program for strategically placed batteries that ease distribution congestion. Form’s modules for the project are to be made at Weirton. Xcel has not finalized whether the 300 MW will be one site or several, and does not expect to put that capacity behind the meter at the data center itself; the storage sits on the utility’s Upper Midwest grid.

Data-center load is one reason multi-day storage is suddenly more than a niche research topic. AI training and inference are driving new electricity demand just as grids try to retire coal and lean harder on wind and solar. Cheap long-duration storage is one way renewables can cover more of that growth without defaulting entirely to new gas plants. Form’s Google-linked project is a high-profile test of whether iron-air can show up at that scale on a utility timeline.

Caveats: footprint, competition and idle capital

Iron-air is not magic. Form’s containerized modules need a large physical footprint — MIT Technology Review cited roughly an acre for every two to three megawatts, which would put a Google-scale installation on the order of dozens of football fields. The long-duration market is crowded with pumped hydro, compressed air, flow batteries, fuel cells and gravity devices. Economics are awkward: multi-day batteries require large capital outlays yet may sit at low utilization for much of the year, waiting for rare weather events.

Form also has to deliver against a backlog that far exceeds current factory output. First projects arriving late or over budget would not be unusual for a first-of-a-kind manufacturing scale-up. Investors and utilities will watch Great River Energy’s operating data as closely as the glossy capacity announcements.

Where iron-air sits next to other cleantech bets

On aglena.com, the storage and generation story already includes sodium-ion batteries leaving the lab for grid storage, perovskite tandems racing toward commercial roofs and fields, research on why perovskite solar cells wear out outdoors, and solid-state EV batteries still waiting for cars you can buy. First power from hot dry rock at Fervo’s Cape Station, Iceland’s giant air filter finally hitting design pace, and the reality that permitted is not built for TVA’s first U.S. SMR are separate lanes of the same decarbonization push. Iron-air belongs with the grid-storage cluster: not a phone battery, not a car battery, but a weather-resilience tool for a renewable-heavy power system.

If Form hits cost and manufacturing targets, 100-hour iron batteries become a practical answer to a problem lithium was never designed to solve. If scaling stalls, utilities will keep leaning on gas and on whatever other long-duration options clear their own cost curves. The 2026 scorecard is clearer than it was a few years ago: a working factory in Weirton, modules moving toward Minnesota, and a data-center-sized contract on the books. The next scorecard is whether those systems run as advertised through real multi-day weather.

Form Energy iron-air FAQ

How long can Form Energy’s iron-air batteries discharge?

Form designs its systems for about 100 hours of discharge at rated power — roughly four days — far longer than typical four-to-six-hour lithium-ion grid batteries.

What is "reversible rusting"?

During discharge, oxygen from air helps convert iron metal to iron hydroxide (rust’s early stage) while releasing electrons. Charging uses electricity to turn that hydroxide back into iron metal and release oxygen.

Where are the batteries made?

At Form Factory 1 in Weirton, West Virginia, on the site of the former Weirton Steel mill. The plant is Form’s first high-volume manufacturing facility.

When does the Great River Energy project come online?

The Cambridge, Minnesota, 1.5 MW / 150 MWh pilot broke ground in 2024. Reporting in 2025–2026 points to commercial operation around 2027 as modules from Weirton are installed and commissioned.

What is the Xcel–Google project?

Xcel Energy plans about 300 MW / 30 GWh of Form iron-air storage as part of a renewables package serving a new Google data center in Minnesota, with phased online dates between 2028 and 2031, pending regulatory approvals and construction.

Further reading