Xcel Energy just announced the largest battery project by energy capacity ever contracted on the planet: 300 MW and 30 GWh of iron-air storage, destined to power a Google data center in Pine Island, Minnesota. The active ingredient is iron powder. The chemistry is rust. Stop scrolling and sit with that for a second.
The intermittency problem has haunted renewable energy for as long as we have taken it seriously. Solar panels generate nothing at 2 a.m. Wind turbines go silent during a high-pressure system that can park itself over a region for four, five, six days straight. Lithium-ion technology's primary constraint is its duration: while effective for short power bursts lasting between two and four hours, it becomes economically unfeasible when attempting to bridge multi-day gaps in renewable generation. That two-to-four-hour ceiling is not an engineering failure. It is a chemistry ceiling. You cannot lithium-ion your way out of a five-day wind drought. We have known this for years and mostly looked away, because the alternative was not yet ready.
It is ready now.
When Rust Became a Grid Asset
The physics underneath iron-air storage are almost insultingly simple. During discharge, the battery absorbs oxygen from the air, which converts iron pellets into rust and releases energy. To charge, an electrical current converts the rust back into metallic iron and the battery releases oxygen. That is the whole trick. A process your garden gate has been doing unintentionally for centuries, harnessed and reversed on demand. Think of it as capturing the energy locked inside oxidation, the same slow burn that has been degrading infrastructure since the Iron Age, but now running it backward and forward on a grid operator's schedule.
Form Energy's first commercial product is an iron-air battery system that can cost-effectively store and discharge energy for up to 100 hours. One hundred hours. That is more than four days of continuous discharge. The cost per kilowatt-hour targets a price point below $20 per kWh, a fraction of lithium-ion's cost. And critically, it is made from some of the safest, cheapest, and most abundant materials on the planet: low-cost iron, water, and air. No cobalt. No lithium. No geopolitical chokepoints.
This is not a lab result. Form Energy has begun deploying its first commercial batteries, which will be installed at Great River Energy's multi-day storage project in Minnesota. Since the beginning of 2025, Form has produced approximately 100,000 electrodes, representing about 100 kilometers of material that has flowed through its West Virginia factory. The Ore Energy pilot, connected to the grid in Delft in July 2025, represents a crucial proof of concept for the European market. Two continents. Grid-connected. Operational. This is no longer theoretical.
The safety profile is worth noting separately, because it changes the deployment calculus entirely. Form Energy's iron-air battery system successfully completed UL9540A safety testing. The test evaluates a battery's potential for thermal runaway and its ability to prevent the spread of heat or fire. The cells were subjected to simulations of fault and abuse conditions known to trigger thermal runaway in lithium-ion. The results were consistent across all scenarios: no uncontrolled heating, no thermal runaway, no dendrite formation, and no fire. For context, the International Association of Fire Chiefs strongly advises that no firefighter enter a burning lithium-ion battery facility. Iron-air batteries cannot catch fire the way lithium-ion can. That is not a minor footnote.
The AI Economy Accidentally Built the Case for Iron
Here is the part that the energy press has mostly underreported. The AI data center boom did not just create new electricity demand. It created demand for a specific kind of electricity: firm, always-on, 24/7 power that cannot tolerate a cloudy week. At 30 GWh, the Google-Xcel-Form Energy deal is the largest battery system by energy capacity ever announced globally. It also marks Form Energy's first direct deployment for a data center, demonstrating the unique value of a 100-hour firm capacity resource in meeting the 24/7 energy needs of the AI economy.
The AI economy, somewhat by accident, just handed iron-air batteries its best argument. Data centers need what renewables struggle to guarantee, and iron-air delivers exactly that. While Form was focused on manufacturing execution, the AI boom caused a significant shift in the energy storage market. The company found itself with more near-term demand than it had planned for. Volumes from projects currently in the pipeline are more than what Form was projecting for 2028, 2029, 2030.
Utilities are reading the same signal. PacifiCorp is planning to install 3,073 MW of iron-air battery storage by 2045, including 605 MW of 100-hour iron-air battery storage by 2032. A U.S. Department of Energy report suggests 225 to 460 GW of long-duration energy storage will be needed by 2060. The scale of what needs to be built is staggering. The fact that the material to build it is iron, the most common metal in the Earth's crust, is either the universe's best joke or its most elegant solution.
One real-world validation is not knowledge. A single commercial pilot in Minnesota and a grid hookup in Delft are promising, not proven at scale. Long-term reliability and cycle life still need real-world validation, as the technology remains in early commercial stages. The next five years of operational data will tell us whether the electrochemistry holds up across thousands of charge cycles in genuine grid conditions. That is the honest scientific position: the hypothesis is strong, the early evidence is compelling, and the experiment is now running live on the grid.
The universe does not care about your timeline. But for once, the timeline is actually cooperating. The grid needs multi-day storage now. The factories are running. The batteries are shipping. And the active ingredient costs less per ton than a decent steak. This is bigger than you think.
