Energy / Grid storage · Deep dive

Form Energy

The iron-rusting battery betting that 100-hour grid storage at $20/kWh beats lithium on cost — if utilities can live with sub-50% round-trip efficiency.

emerging

The question that decides it: Iron-air's <50% round-trip efficiency means Form only pencils when charged on near-free curtailed power. Does enough $0-10/MWh surplus renewable energy actually exist at the sites and hours utilities need 100-hour discharge — or does the efficiency penalty erase the $20/kWh cost advantage the moment charging power isn't free?

My take ★★ Very interesting ★ Interesting − Maybe ✗ Not interesting

Snapshot

Form Energy builds grid batteries out of rusting iron. Its cells discharge for up to 100 hours — over four straight days — at a targeted installed cost near $20/kWh, roughly a tenth of lithium-ion, using iron and air instead of lithium, nickel, and cobalt. It has raised more than $1.2 billion since 2017 (company, Oct 2024) at a ~$3.4-3.5 billion valuation in its October 2024 Series F. It shipped its first commercial batteries from a converted West Virginia steel mill in May 2025, and in February 2026 signed a 300 MW / 30 GWh system with Xcel to power a Google data center in Minnesota — the largest battery by energy capacity ever announced. The catch: iron-air round-trips power at under 50% efficiency versus 90%+ for lithium, so the whole case turns on whether that penalty is tolerable when charging power is near-free surplus renewables.

Founding story

The company began with a 2016 conversation at the CERAWeek energy conference in Houston, where MIT materials scientist Yet-Ming Chiang met Tesla storage executive Mateo Jaramillo and the two agreed that renewables had an unsolved problem: no cost-effective way to store power for days, not hours. In early 2017 Chiang pitched Ted Wiley over coffee on making rechargeable batteries from cheap, abundant materials rather than lithium; Wiley recruited Marco Ferrara and Billy Woodford — both alumni of Chiang’s earlier startup 24M — and they launched Baseload Renewables. Jaramillo, meanwhile, had left Tesla to start Verse Energy. In December 2017 the two companies merged into Form Energy, combining Baseload’s chemistry team with Jaramillo’s commercial and utility credibility.

The pedigree is the thesis. Chiang has founded A123, 24M, American Superconductor, and Desktop Metal — a serial battery entrepreneur who has watched chemistries reach and miss commercial scale. Wiley came from Aquion Energy, a saltwater-battery firm that went bankrupt in 2017, so Form’s operating leadership has personally watched a good-on-paper grid chemistry die on manufacturing cost. Jaramillo ran Powerwall and Powerpack at Tesla and knows where lithium’s economics break down. The founding bet: stop competing with lithium on lithium’s terms and exploit the one axis where iron wins overwhelmingly — raw material cost.

How it works

The chemistry is deliberately humble: it is controlled, reversible rusting. When the battery discharges, metallic iron reacts with oxygen from the ambient air and water in a water-based electrolyte to form iron hydroxide — rust — releasing electrons. When charging, grid electricity reverses the reaction, stripping the oxygen back out and returning the rust to metallic iron. The active materials are iron, air, and a non-flammable water-based electrolyte similar to what is in an AA battery. There is no lithium, cobalt, nickel, or fire risk.

Physically, cells are packed into weatherproof enclosures about the size of a shipping container, and hundreds of those enclosures form modular, megawatt-scale power blocks; a 1 MW system spans half an acre to 3+ MW/acre depending on density. The system is heavy, low-power-density, and slow — and that is fine, because it is not competing on those axes.

The number that governs everything is round-trip efficiency. Form’s disclosures put it below 40-50%, versus 90%+ for lithium — roughly 2.5 MWh in to get 1 MWh back out. That is a brutal penalty unless the input power is nearly free curtailed wind or solar that would otherwise be thrown away. Iron-air is therefore not a general-purpose battery; it is a device for time-shifting cheap surplus renewables across many days, and its economics live or die on how cheap the charging power is.

Product and business overview

Form sells one product: a 100-hour iron-air battery system in utility-scale, modular power blocks. The commercial unit is a project — a utility or corporate offtaker contracts for a system rated in MW of power and GWh of energy (e.g., 10 MW / 1 GWh; the flagship Google/Xcel system is 300 MW / 30 GWh). Around the hardware sits Ferrara’s systems-modeling, which tells a utility how many hours of storage its grid actually needs — a consultative layer that helps justify a technology with no long track record.

The strategic anchor is manufacturing. Form is not licensing chemistry; it is building the factory. “Form Factory 1” occupies a repurposed steel mill in Weirton, West Virginia — skilled industrial labor and existing infrastructure in a shuttered mill town. That vertical, made-in-America posture is both a competitive moat and a capital sink.

Business model and pricing

Form’s model is capital-goods manufacturing: design, build, and sell battery systems to utilities and, increasingly, corporate offtakers, with revenue booked on project delivery rather than subscription. The company does not publish per-unit price lists, but the target economics are explicit and repeated: an installed system cost around $20/kWh, roughly one-tenth of lithium-ion’s per-kWh cost. Reported deal sizes give a rough check — The Information reported (Feb 2026) that Google paid Form about $1 billion for the 30 GWh system, implying very roughly ~$33/kWh at the system level, above the long-run $20 target as first-of-kind projects usually are.

Two features define the economics. First, the cost edge only exists at long duration: for a 4-hour battery lithium is cheaper and far more efficient, but as duration stretches toward days lithium’s cost scales linearly with every added hour while Form’s marginal cost of adding iron is trivial. Second, the model is heavily policy-exposed: Form’s factory and early projects lean on DOE grants, IRA-era tax credits, and state procurements that materially improve returns — the Maine project alone won a ~$147M federal grant (Aug 2024). This is not a business that pencils today on merchant economics alone.

Traction over time

By early 2026 the commercial pipeline passed 750 MW / 75 GWh under agreement (company/trade press, 2026), spanning Great River Energy (MN), Xcel (CO, MN), Georgia Power (15 MW / 1,500 MWh), Dominion (VA), California (PG&E), New York, and an 85 MW / 8,500 MWh system in Lincoln, Maine. Weirton headcount reached ~450 (2025-26), with plans for 750+ factory jobs and 500 MW/year capacity by 2028; the Eighty-Four, PA pilot was closed and consolidated in. Product revenue remains minimal — shipments only began May 2025 — so this is a backlog story, not an income one.

Market analysis

The long-duration energy storage (LDES) market is real but early. Third-party sizing clusters around $4-5B in 2024-25 growing to roughly $10-17B by the mid-2030s at ~13% CAGR (multiple firms incl. MarketsandMarkets, GMInsights, 2025-26) — respectable, not explosive, on those headline numbers. The more relevant frame is physical: the LDES Council estimates worldwide LDES capacity must grow ~50x by 2040, potentially 85-140 TWh of energy capacity, to decarbonize power grids dominated by intermittent wind and solar. WoodMac reported LDES deployments rose 49% in 2025.

The structural tailwind: as grids add renewables, both oversupply (cheap/curtailed power) and multi-day undersupply (windless, cloudy stretches) grow — precisely the arbitrage iron-air is built for. The newer, possibly larger driver is AI data-center load; the Google deal shows corporates willing to anchor utility-scale LDES for 24/7 carbon-free power. The counter-force is lithium getting cheaper, which pushes the “duration crossover” where iron-air wins further out. Form’s TAM is not all storage — only the slice where duration is long enough that lithium’s linear energy cost loses to iron, and cheap surplus charging power actually exists.

Competitive intel

The honest competitive picture is that Form’s main rival is not another LDES startup — it is lithium-ion getting cheaper. Lithium (Tesla Megapack, CATL, LFP) held ~96% of US/Canada storage installs from 2020-2023 (WoodMac, cited 2026) and wins decisively on efficiency and flexibility. Form only exists in the duration band where lithium becomes uneconomic; that band is the whole business, and lithium’s cost curve is trying to erase it.

Among LDES peers, only three firms raised over $1B in 2021-25: Form, Eos Energy (zinc, public, showing financial stress in 2025-26 per pv magazine), and Hydrostor (compressed air). ESS Inc, the closest chemistry analog with its iron-flow approach, is the cautionary tale — public since 2021, struggling badly with deployment and burn. Antora attacks from thermal storage; pumped hydro remains the dominant real-world long-duration incumbent, cheaper where geography cooperates but geographically stuck. Form’s differentiation is the combination its rivals lack: abundant-material chemistry, a built and shipping US factory, a hyperscaler anchor, and strategic backers (ArcelorMittal for iron, GE Vernova for grid integration) that de-risk supply chain and channel.

History and evolution

• 2016 — Chiang and Jaramillo meet at CERAWeek and frame the multi-day storage problem.
• 2017 — Chiang, Wiley, Ferrara, and Woodford launch Baseload Renewables; Jaramillo starts Verse Energy.
• Dec 2017 — The two merge to form Form Energy.
• 2018-2019 — Seed ($2M, The Engine), Series A ($9M), and Series B (~$40M, Eni Next) fund early R&D; Breakthrough Energy Ventures backs the company.
• Nov 2020 — $76M Series C (Coatue); Form discloses its “aqueous air” long-duration chemistry.
• Aug 2021 — $240M Series D led by steelmaker ArcelorMittal, locking in an iron supplier as investor.
• Jan 2023 — First utility contracts, with Xcel Energy in Colorado and Minnesota.
• Oct 2022 — $450M Series E (TPG Rise Climate) with GIC, Temasek, CPP.
• 2023 — Groundbreaking on Form Factory 1 in Weirton, WV.
• Aug 2024 — Groundbreaking on first commercial install with Great River Energy, Cambridge, MN; Maine project wins ~$147M federal grant.
• Oct 2024 — $405M Series F (T. Rowe Price), GE Vernova joins with a strategic MOU; total raised passes $1.2B.
• May 2025 — First commercial batteries ship from Weirton.
• Jul 2025 — One Big Beautiful Bill Act reshapes IRA credits and rescinds DOE Loan Programs Office funding — a direct hit to the policy scaffolding under Form and its peers.
• Feb 2026 — Google/Xcel 30 GWh deal announced; ~$1B; world’s largest battery by energy capacity.

What people say

The case for. Believers argue Form has cleared the hardest gates for climate hardware: a working chemistry, a factory actually shipping (May 2025), a pipeline past 75 GWh, and — most persuasively — a ~$1B commitment from Google, the most sophisticated possible buyer, validating that a sub-50%-efficient battery can still be the cheapest way to firm a data center 24/7 (energy-storage.news, Feb 2026). The material argument is strong: iron is the cheapest metal on earth, the electrolyte is non-flammable, and there is no lithium supply-chain or fire risk. Strategic investors chosen for function — ArcelorMittal for iron, GE Vernova for grid integration — suggest the supply chain and channel are being de-risked deliberately. On Glassdoor, employees give CEO Jaramillo a high approval rating and cite mission and technical caliber (~149 reviews, accessed 2026).

The complaints. The skeptic case is blunt and technical. Round-trip efficiency below 50% (the company concedes under 40% in some framings) means the battery wastes more than half its input power — critics (RealClearEnergy, Feb 2025) argue this only works if charging power is genuinely near-free, and that enough $0-10/MWh surplus renewable energy may not reliably exist where and when 100-hour discharge is needed. Winter performance is a recurring worry: cold slows the electrochemistry. The ESS Inc precedent looms — an iron-chemistry peer that reached public markets then foundered on deployment — as does Eos’s 2025-26 stress, evidence the category’s unit economics are unproven at scale. Then the subsidy dependence the deck omits: Form’s factory and early projects lean on DOE grants and IRA-era credits, and the July 2025 One Big Beautiful Bill Act rescinded DOE Loan Programs Office funding and began phasing down clean-electricity credits. On the workforce side, Glassdoor reviews and the closure of the Eighty-Four, PA pilot point to pre-commercial-hardware volatility: job-security anxiety and a business years from self-funding.

Outlook: the open question

Form has done the hard, capital-intensive thing most climate startups never reach: it has a chemistry that works, a factory that ships, and a marquee customer. That earns it a real shot. But the verdict is not yet earnable, because the technology’s defining weakness has not been stress-tested at scale. Form works if two things prove true: that abundant near-free surplus renewable power actually exists at the specific sites and hours where utilities need multi-day discharge — making sub-50% round-trip efficiency economically irrelevant — and that installed system cost genuinely converges toward the ~$20/kWh target as volume ramps, rather than sticking near the first-of-kind ~$33/kWh implied by the Google deal. If both hold, Form owns a duration band lithium physically cannot reach economically, and the 75 GWh pipeline is the front edge of a very large market.

Form breaks if either fails: if cheap curtailed power is too scarce or mistimed to charge the batteries — so the efficiency penalty becomes real fuel cost — or if lithium’s falling price pushes the duration crossover far enough out that few utilities ever need 100 hours. The wildcard is policy: current returns depend materially on federal subsidy, and the July 2025 rollback of IRA credits and DOE loan funding removed part of it. The clearest tell is not the next raise but operating data from the first live installs (Minnesota, Colorado, California) and the Google system — real round-trip efficiency, charging-cost economics, and installed cost per kWh. Those numbers, not the pipeline headline, decide whether iron-air is the cheapest firm power on the grid or an elegantly engineered subsidy sink.

How a challenger would attack it

Attack the gap between $33 and $20. Form’s first-of-kind pricing — roughly ~$33/kWh implied by the ~$1B Google deal against a $20/kWh long-run target — is the umbrella a challenger camps under. Form chose the hardest possible route: a vertically owned US factory in a converted steel mill, ~450 workers, capacity ramping to 500 MW/year by 2028. A capital-light rival that licenses an alternative long-duration chemistry to existing battery or industrial manufacturers — or a sodium-ion or thermal player like Antora that piggybacks on established production lines — can reach bankable cost faster without financing a Weirton. The second vector is efficiency: iron-air’s sub-50% round-trip is Form’s structural concession, so any chemistry that delivers 60-70% efficiency at even $35-40/kWh flips the economics wherever charging power isn’t free — which is most sites, most hours, per Form’s own critics. Third, the policy flank is open: the July 2025 OBBBA rescinded DOE loan funding and is phasing down the IRA credits Form’s early project returns lean on, so a challenger whose model pencils on merchant economics — or one arbitraging cheaper non-US manufacturing for non-US grids — attacks the subsidy scaffolding Form cannot quickly replace. Form’s moat is a shipped factory and the Google validation; its exposure is that every dollar of it assumes the 100-hour band stays wide while lithium’s cost curve works to close it.

Same playbook, new buyer

Form sells 100-hour storage to regulated US utilities and one hyperscaler — a sales motion of multi-year procurements, pilot regulatory approvals, and federal grant stacking. The same iron-air proposition has buyers with faster clocks and worse alternatives. Island and remote grids running diesel — Caribbean, Pacific, Alaska — pay $200-400/MWh for generation, so sub-50% efficiency against free solar beats imported fuel outright, and no LDES player owns that niche. Mining operations with the identical stranded-diesel problem are a second variant, and ArcelorMittal’s presence on Form’s own cap table gestures at industrial off-grid demand Form isn’t productizing. Geographically, Form’s made-in-America, Weirton-anchored posture — its core political asset domestically — is precisely what makes it slow abroad; a licensor bringing iron-air-style chemistry to India or Southeast Asia, where solar curtailment is exploding and US content rules mean nothing, faces no Form response for years. The incumbent won’t follow because its capital, its factory, its federal grants, and its 75 GWh utility pipeline all chain it to large-format US grid projects; small, fast, unsubsidized deployments are the deals its cost structure can’t chase.

Sources and further reading

• Form Energy — About / founding story (Form Energy, accessed July 2026)
• Form Energy Secures $405M in Series F Financing (Form Energy, October 2024)
• Form Energy raises $405M for its 100-hour iron-air batteries (Canary Media, October 2024)
• Google: Minnesota data centre energy deal includes 30GWh iron-air batteries from Form Energy (Energy-Storage.news, February 2026)
• Google bets big on 30GWh of Form Energy’s iron-air storage despite efficiency trade-offs (Energy-Storage.news, 2026)
• 2024 Climate Tech Companies to Watch: Form Energy and its iron batteries (MIT Technology Review, October 2024)
• Iron-Air Batteries: Green Energy Advocates’ Newest Shiny Object (RealClearEnergy, February 2025)
• Form Energy Powers Up To Ship First Batteries From Weirton Factory (The Intelligencer, May 2025)
• Long-duration energy storage deployments rose 49% in 2025: WoodMac (Utility Dive, 2026)
• Medium duration storage endures woes, while Form’s long duration grows (pv magazine USA, April 2026)
• 2025 Reconciliation Debate: One Big Beautiful Bill Act Energy Provisions (Bipartisan Policy Center, 2025)
• Form Energy — Wikipedia (Wikipedia, accessed July 2026)