DOE's Third Battery Round Is $500 Million and Seven Projects — Down From $3 Billion and 25. Read the Winners, Not the Press Release.

The Energy Department announced $500 million on August 20 for seven projects spanning critical mineral processing, battery manufacturing, and recycling. Secretary Chris Wright framed it the way every critical-minerals announcement has been framed since 2022: "For too long, America has depended on foreign actors for critical materials essential to modern life."

The framing is familiar. The arithmetic is not.

Round 1, announced October 2022, put $1.82 billion into 14 projects. Round 2, announced September 2024, put over $3 billion into 25 projects across 14 states. Round 3, announced last week, puts $500 million into seven.

That is an 83 percent drop in dollars and a 72 percent drop in project count from the prior round. If you are building a critical-minerals or battery company and you have been waiting for your turn at the Bipartisan Infrastructure Law's Section 40207 money, the shape of that turn has changed fundamentally — and the seven selections are a more honest guide to DOE's current appetite than anything in the press release.

Who won, and what the list is actually saying

Across the two programs, the seven selections break down as follows.

Battery Materials Processing (upstream refining) — $200 million, two projects:

• Waterleaf P1 HoldCo, LLC (Lilac Solutions) — West Promontory, Utah — $100 million. A commercial lithium extraction and refining facility producing battery-grade lithium carbonate, which DOE says would "double current U.S. lithium production." The project returns processed brine to the Great Salt Lake with no net water loss — a design detail that is doing real permitting and political work in a state where lake levels are a live public fight.
• Formation Holdings US, Inc. (Jervois) — site to be determined — $100 million. A commercial cobalt refinery producing battery-grade cobalt sulfate "using proven refining technology."

Battery Manufacturing and Recycling (midstream and downstream) — $300 million, five projects:

• Nth Cycle, Inc. — Southeastern U.S. — $100 million. Electroextraction refining of black mass from end-of-life lithium-ion batteries and manufacturing scrap into high-purity metals.
• Princeton NuEnergy Inc. — Commerce, Georgia — $50 million. Low-temperature plasma processing to recover and rejuvenate nickel-containing cathode material from battery manufacturing scrap.
• Arcanum Ventures, LLC — U.S. Gulf Coast — $50 million. Battery-grade ethylene carbonate, an electrolyte input.
• Elevated Materials, LLC — location to be determined — $50 million. Ultra-thin lithium-metal films and prelithiated materials for advanced batteries.
• Coreshell Technologies Incorporated — San Leandro, California — $50 million. U.S. manufacturing capacity for silicon-anode lithium-ion electrodes and cells.

Read that list twice and three patterns fall out.

First: not a single mine. Every upstream selection is a processing or refining facility. Lilac is brine extraction and refining; Jervois is a refinery. DOE is buying midstream capacity — the step where China's dominance is most complete and most quickly reversible — rather than betting on new hardrock extraction, which carries a permitting timeline no grant program can compress.

Second: three of seven are circular. Nth Cycle, Princeton NuEnergy, and — depending on how you count prelithiation feedstock — Elevated Materials are recovering value from scrap and end-of-life cells rather than from ore. That is $150 million to $200 million of a $500 million round placed on the proposition that the fastest domestic supply of nickel, cobalt, and lithium is the material already inside the country.

Third: the award ladder is flat and short. Two awards at $100 million, five at $50 million. No $150 million or $200 million awards of the kind Round 2 produced (Cirba Solutions took $200 million; American Battery Technology, Clarios Circular Solutions, and Ascend Elements each took $125 million to $150 million). DOE compressed the top of the range and standardized the middle. If you were modeling a $175 million ask, the evidence says the ceiling is now $100 million and the modal award is $50 million.

The context nobody in the press release mentions

Round 3 is not landing on a clean slate. It is landing in the middle of an unresolved fight over Round 1 and Round 2 money.

In late October 2025, DOE confirmed the cancellation of five projects totaling $718 million, on the stated grounds that they were not "economically viable." Four of the five came from these same Battery Materials Processing and Battery Manufacturing grant programs. Broader cancellation lists — some formal, some leaked — named far more.

What followed is instructive for anyone holding a federal award. DOE has formally de-obligated funds for only about 30 projects, under 4 percent of the federal dollars named on those lists, according to reporting by Latitude Media. Most named projects sit in limbo: not cancelled on paper, not proceeding in practice. Some recipients fought and won. American Battery Technology Company's $57 million lithium refinery grant in Nevada was reinstated in June 2026 after the company prevailed on appeal. In January 2026, a federal judge ordered DOE to reinstate nearly $28 million in cancelled grants, finding that the administration had violated the equal protection clause by targeting recipients based on their location in Democrat-led states.

So the honest read of Round 3 is this: DOE is making new selections in a program whose prior selections are still being litigated, appealed, and quietly stalled. That does not make the money unreal — Lilac and Jervois and Nth Cycle will negotiate real agreements. It does mean a selection is not an award, and the gap between the two is wider in this program right now than in almost any other federal grant portfolio. Budget for it.

The policy framing has also shifted. The programs were authorized under the Bipartisan Infrastructure Law — Sections 40207(b) and 40207(c), with the Materials Processing program alone carrying up to $3 billion total, $600 million annually across FY2022 to FY2026. Round 3 was announced under the banner of the current administration's "Unleashing American Energy" executive order. Same statutory money, different justification. For applicants, the practical consequence is that the evaluation narrative has moved from decarbonization and jobs toward supply-chain security and reduced foreign dependence. Proposals still written in 2022 language are being read by reviewers working from a 2026 rubric.

What this means for your next application

If you are preparing for a future round of these programs — or for any of DOE's adjacent critical-minerals vehicles, which we covered when the $500 million Round 3 solicitation first opened in March — the seven selections give you a usable specification.

Size the ask at $50 million and justify anything above it. Five of seven landed exactly at $50 million. That is not a coincidence; it is a program office managing portfolio risk by spreading a smaller pot. A $50 million request with a credible, fully costed scope now reads as the default. A $100 million request needs to clear the bar Lilac cleared — a claim as concrete as "doubles current U.S. lithium production."

Lead with the import-substitution number, not the tonnage. Every winning project has a one-line answer to "what fraction of a Chinese-controlled step does this replace?" Cobalt sulfate, black mass refining, electrolyte carbonate, silicon anode: each is a chokepoint where domestic capacity is near zero. If your project's differentiator is cost or performance rather than substitution, you are competing on the wrong axis for this program.

Bring scrap, not ore. The recycling tilt is the single clearest signal in the round. Feedstock that already exists inside U.S. borders — manufacturing scrap, end-of-life cells, black mass — sidesteps the permitting and offtake risk that makes reviewers nervous about extraction projects. If you have a processing technology, the version of your project that runs on domestic scrap will score better than the version that runs on imported concentrate.

Name the site or explain why you cannot. Three of the seven selections list a location as "to be determined" or a broad region. That is tolerated, but it is tolerated for companies with existing operating history. A first-time applicant with an undetermined site is asking reviewers to underwrite two risks at once.

Assume a long negotiation and a live cancellation risk. Model the period between selection and first drawdown as 9 to 18 months, plan bridge capital accordingly, and read your award terms — particularly the termination provisions — before you sign. The 2025–2026 cancellation wave established that recipients who documented performance and pursued formal appeals recovered funds. Recipients who assumed the announcement was the end of the process did not.

The $3 billion era of this program is over. The $500 million era rewards a narrower, more specific project: midstream, circular, chokepoint-focused, and sized to fit a portfolio that has to stretch.