The US Department of Energy has awarded $67 million to ElementUSA and Colorado School of Mines to build a rare earth processing facility in Gramercy, Louisiana, extracting critical minerals from decades of alumina refinery waste. The award is the clearest signal yet that Washington has shifted its strategic focus from mine permitting to the processing layer where China holds near-total control. Whether it is enough to close a gap measured in decades and billions of dollars is a different question.
Introduction
For years, the dominant story in US critical minerals policy was about what was in the ground: permits, deposits, royalty regimes, and the slow bureaucratic machinery of opening new mines. That framing was always incomplete. On June 3, 2026, the Department of Energy made its most direct statement yet that Washington knows it has been solving the wrong problem.
The DOE announced $67 million for ElementUSA and Colorado School of Mines to design, construct, commission, and operate a rare earth element processing facility in St. John the Baptist Parish, Louisiana. The facility will recover rare earth elements from alumina tailings, separate them into rare earth oxides, and refine those oxides into rare earth metals. In other words, it will do precisely the midstream work that the United States currently cannot do at meaningful scale, and that China has dominated so thoroughly for so long that most Western policymakers had quietly stopped trying.
The award matters not just for its dollar figure, but for what it targets. As I argued in my analysis of Europe's refining gap in July 2026, the defining vulnerability in Western supply chains is not what is in the ground. It is the industrial capacity to turn raw ore into usable material. The Louisiana project is the US equivalent of that same uncomfortable realisation, expressed in concrete and federal appropriations rather than policy papers.
This piece explains what the project actually involves, why the midstream matters more than the mine, what the realistic objections are, and where the broader US rare earth strategy is heading.
What the Project Actually Does
The Gramercy site holds exclusive access to approximately 34 million tonnes of bauxite residue stored at the former Atalco alumina refinery, one of the largest single-site concentrations of this material in North America. The refinery began operating in 1958, and the residue it left behind, a reddish slurry known in the industry as red mud, has been sitting in tailings impoundments on the banks of the Mississippi ever since. ElementUSA, a company founded in 2021 and specialising in what it describes as waste-to-market solutions, holds exclusive rights to this material.
Red mud is not conventionally thought of as a resource. It is an environmental liability: alkaline, voluminous, and expensive to manage. But it contains meaningful concentrations of rare earth elements and other critical minerals, and ElementUSA's integrated hydrometallurgical and pyrometallurgical process is designed to recover them simultaneously with pig iron production. The target elements include dysprosium, terbium, yttrium, gadolinium, neodymium, praseodymium, samarium, and lanthanum, covering both the light rare earths used in high-volume magnet applications and the heavy rare earths that matter most for defence and high-performance EV motors.
At the demonstration scale funded by this DOE award, the plant is projected to produce between 150 and 1,000 metric tonnes per year of rare earth elements for domestic use. At full commercial scale, with an estimated capital expenditure of approximately $1.1 billion, the resource has potential to produce between 45 and 385 percent of current US annual demand for gallium, scandium, yttrium, germanium, ytterbium, dysprosium, and gadolinium. Initial production from the demonstration facility is targeted for the third quarter of 2028.
The project is part of a broader DOE commitment: a total of $134 million across two facilities, with Phoenix Tailings receiving $66 million for a separate facility in Ardmore, Oklahoma, focused on rare earth separation from domestic industrial waste streams. Both projects are partnered with leading research universities, Phoenix Tailings with MIT and the University of Minnesota, and both represent the DOE's Rare Earth Demonstration Facility Program at work. Combined with a prior $29.9 million Department of Defense award to ElementUSA for gallium and scandium recovery, total federal investment in the Gramercy project now stands at $96.9 million.
Why the Processing Layer Is the Strategic Chokepoint
To understand why this award matters, it helps to think of the rare earth supply chain as a relay race with four legs: mining, separation and refining, alloying, and magnet manufacturing. Western policy has spent most of the past decade focused on the first leg while conceding the remaining three almost entirely to China. That concession now looks extraordinarily costly.
IEA data are stark. China controls between 47 and 87 percent of global refining capacity across critical mineral supply chains, depending on the mineral. For rare earths specifically, China accounts for approximately 91 percent of global separation and refining output. For sintered permanent magnets, which are the finished product that actually goes into an EV motor or a precision-guided munition, China's share reached 94 percent in 2024, up from roughly 50 percent in 2005. The IEA's Global Critical Minerals Outlook found that China leads refining for 19 of the 20 strategic minerals it tracks, with an average market share of 70 percent. None of those numbers are improving. The average market share of the top three refining nations rose to 86 percent in 2024 from 82 percent in 2020, with almost all supply growth coming from China.
The scale of what this means for any country trying to build an independent supply chain is illustrated by the IEA's pipeline analysis. By 2035, existing and announced non-Chinese capacity is expected to cover only around half of mining requirements, a quarter of refining needs, and less than a fifth of magnet demand. The further down the value chain you go, the worse the gap gets. A country can announce ten new mines tomorrow and still find itself shipping ore to Chinese refineries for separation, recreating exactly the dependency it was trying to escape. As one analysis from CSIS put it: countries that succeed in opening new mines but lack processing infrastructure will find themselves shipping ore to China for refining.
The economic logic reinforces the strategic logic. Industry analysts at Rare Earth Exchanges estimate that nearly 80 percent of total rare earth supply-chain costs sit in the refining and separation stages, which is precisely where China holds near-monopoly control. That control was not accidental. It reflects five decades of accumulated expertise, scale economies, and a tolerance for the chemical and environmental complexity of rare earth separation that Western countries largely chose to avoid. Rebuilding that capacity is genuinely difficult, genuinely expensive, and genuinely slow, which is why demonstration-scale facilities like the one in Gramercy represent an important step rather than a solution.
Market Conditions: Why Prices and Policy Are Aligned Right Now
The DOE award lands at a moment when rare earth markets are doing the policy argument's work for it. NdPr oxide, the key input for high-performance permanent magnets, opened 2026 at approximately $53 per kilogram and surged to roughly $137 to $140 per kilogram by late April, a gain of approximately 160 percent in four months. A correction brought it back toward $88 to $91 per kilogram through May and June, but BMI, a Fitch Solutions company, has revised its 2026 average price forecast upward to $90,000 per tonne and forecasts a second consecutive year of market deficit. The Critical Minerals Platform's physical supply risk rating for NdPr remains at the highest category.
The price story is not simply a function of demand. It reflects supply restraint that is deliberate and policy-driven on the Chinese side. China's Ministry of Industry and Information Technology has not announced the staged 2026 quota increases that some market participants expected. The effect of China's October 2025 export controls, even after a one-year suspension in November 2025, continues to shape buyer behaviour. Exports of yttrium, dysprosium, and terbium from China remain approximately 50 percent below pre-April 2025 levels. Dysprosium and terbium prices outside China have risen four to five times since April 2025, while yttrium prices increased by roughly 140 times.
On June 22, 2026, China's Ministry of Commerce added ten US companies to its export control list, including MP Materials and USA Rare Earth, two of the companies receiving the largest federal investment in domestic rare earth capacity. The timing was pointed. Washington is funding the companies that Beijing views as the most direct competitive threat to its processing dominance, and Beijing is responding with the instruments of commercial pressure available to it.
This context matters for the Louisiana project because the investment case and the policy case have converged. The DOE's price support mechanisms in the broader US rare earth strategy, including the DoD-backed price floor of $110 per kilogram for NdPr under the MP Materials contract, create the long-term revenue visibility that processing facilities need to attract private capital. ElementUSA's $1.1 billion full-scale development pathway will require commercial financing well beyond its federal grants, and that financing will be easier to secure if buyers can point to structural deficit conditions and a policy architecture designed to support domestic processing output.
Objections and Limitations
The case for celebrating this award deserves to be tempered by a clear-eyed view of what it does not achieve.
The first objection is one of scale. One hundred and thirty-four million dollars across two DOE demonstration facilities, however strategically well-targeted, is a rounding error against the investment required to meaningfully shift the global rare earth processing balance. China's rare earth processing infrastructure represents decades of cumulative capital expenditure, supported by state financing, subsidised energy, and a regulatory tolerance for chemical processing complexity that Western jurisdictions do not replicate. The IEA projects that by 2030, China will still control 51 percent of rare earth element production and 76 percent of refining. The Louisiana facility is a necessary proof point, not a structural solution.
The second objection is about timelines. Initial production is targeted for Q3 2028, following groundbreaking in mid-2026 and construction commencing in 2027. The January 2027 deadline for the US DFARS Rule 252.225-7052, which prohibits defence contractors from using magnets containing materials mined, refined, separated, or manufactured in China, arrives before the Gramercy facility produces its first tonne of separated oxide. The industrial policy is running behind the regulatory timeline it is supposed to support.
The third objection concerns workforce and technical knowledge. Rare earth separation is not a process that can be stood up quickly. It requires specialists in hydrometallurgy, hazardous chemical handling, and separation chemistry who are in genuinely short supply in the United States. Rebuilding that capability is partly a matter of capital, but it is also a matter of time, curriculum development, and the slow accumulation of operational experience. Colorado School of Mines' Waste to Value Center, with its stated goal of a 10 percent reduction in mine waste and a 10 percent reduction in critical mineral imports within a decade, reflects a realistic rather than optimistic timeline for the knowledge rebuild.
Finally, there is the integration question. A separation and refining facility that produces rare earth metals is a necessary but not sufficient link in the domestic supply chain. The White and Case analysis of mine-to-magnet integration from June 2026 makes this point clearly: security of supply is not achieved at the level of raw material alone. It requires the alignment of mining, separation, refining, and finished product manufacturing within a single coherent commercial structure. Processing facilities that are not connected upstream to domestic feedstock and downstream to domestic magnet manufacturing will find their strategic value limited.
Where This Fits in the Broader US Strategy
The Gramercy award is one piece of a larger mosaic that is taking shape across multiple agencies and instruments. The DOE has committed close to $1 billion to critical mineral supply chain development, described by some analysts as the largest government intervention in rare earth markets since the 1980s. The DOMINANCE Act, passed by the House in June 2026 and awaiting Senate action, would formalise an outward-facing strategy including allied supply chain development. The Export-Import Bank has issued letters of intent across the rare earths supply chain totalling nearly $4 billion. The DoD-MP Materials price floor contract establishes a policy-supported domestic reference price. USA Rare Earth has secured $277 million from the Department of Commerce and up to $1.3 billion in CHIPS Act loans for its mine-to-magnet strategy.
The allied dimension is also developing. The US-Australia Critical Minerals Framework signed in October 2025 is backed by a $1.25 billion government loan to Iluka Resources for a refinery tied to allied offtake. In Europe, Carester's Caremag recycling and refining project in France, backed by approximately EUR 216 million in financing, remains on track for a late-2026 startup. Neo Performance Materials is advancing its Narva, Estonia magnet plant toward mass production. These are not yet at the scale required to displace Chinese dominance, but the infrastructure of an allied processing network is becoming real in a way it was not two years ago.
The feedstock approach represented by the Gramercy project deserves particular attention as a policy model. Recovering rare earths from industrial waste streams avoids the permitting timelines and community opposition that have slowed new mine development in the United States for decades. The 34 million tonnes of bauxite residue at the Atalco site are already there. The environmental liability already exists. The question is whether it can be converted into a resource, and this DOE award is a structured bet that it can. If ElementUSA demonstrates commercial viability at the scale funded by this award, the replication potential across other industrial waste streams in the United States and allied countries is substantial.
What to Watch Next
The Gramercy project will break ground in mid-2026 and begin construction in 2027, with demonstration-scale production targeted for mid-2028. The most important near-term milestone is not the construction start but the pilot process data that will determine whether the economics of red mud rare earth extraction are genuinely commercial at scale. ElementUSA's hydrometallurgical and pyrometallurgical process is innovative; the combination of pig iron and rare earth co-production is a meaningful differentiator. But co-production economics are sensitive to commodity price assumptions for both iron and rare earth products, and those assumptions will need to hold across a capital raise an order of magnitude larger than the current federal awards.
On the policy side, the Senate's handling of the DOMINANCE Act will signal how much appetite exists for the kind of sustained, multi-year federal commitment that processing infrastructure actually requires. Demonstration facilities are fundable in single appropriation cycles. Commercial-scale rare earth processing plants are not. The gap between the $96.9 million committed to Gramercy and the $1.1 billion needed for full commercialisation is not a financing gap that federal demonstration funding was designed to close; it is one that will require patient private capital, long-term offtake commitments, or a deeper federal financing instrument.
The January 2027 DFARS deadline for defence procurement remains the most concentrated point of near-term pressure. If US defence contractors find themselves unable to source non-Chinese magnets in sufficient volume for compliant systems, the political environment for both accelerated investment and potential interim waivers will shift quickly. The Louisiana facility will not help them in January 2027. Whether it helps them in 2029 depends on execution decisions being made right now.
The broader trajectory is fairly clear. Washington has identified the processing layer as the strategic priority. The instruments being deployed, grants, price floors, allied frameworks, demonstration facilities, are better calibrated to the actual problem than anything deployed in the previous decade. The remaining question is whether the scale and pace of deployment will be sufficient against a competitor that has a fifty-year head start and a domestic demand base growing fast enough to absorb an increasing share of its own production. The Louisiana red mud is a start. It is not a finish line.