Aclara’s DOE Selection Turns Its Digital Twin Into a Separation Scale-Up Test

Aclara Resources has won a place in a U.S. Department of Energy programme aimed at using artificial intelligence to improve rare earth separation, adding federal validation to a strategy that depends less on discovering another deposit than on making a complicated solvent-extraction circuit work reliably at commercial scale.

The Toronto-listed developer said on July 23 that its U.S. subsidiary was selected for Phase I funding under the DOE’s Genesis Mission initiative. The project, called “AI-Enabled Process Optimization for Multi-Feed Rare Earth Separation,” will use an AI-assisted digital twin to improve process efficiency, product quality and operating stability. The amount and timing of the award remain subject to negotiation, and successful Phase I projects may compete for additional Phase II funding.

That distinction matters. This is not yet financing for Aclara’s planned Louisiana separation plant, nor does it remove the construction and commissioning risk attached to a new Western heavy rare earth circuit. It is an attempt to reduce one of the less visible risks in rare earth processing: keeping dozens or hundreds of solvent-extraction stages stable when feed chemistry changes.

Aclara is working with Argonne National Laboratory, which is applying its SolventX modelling platform and advanced computing capability to operational data generated at Aclara’s pilot plant at Virginia Tech. The aim is to create a high-fidelity digital representation of the separation circuit that can simulate changes, predict performance and support process control before the company commits fully to industrial-scale operation.

The practical test is already under way. Aclara’s Virginia pilot plant was commissioned in March and designed to produce separated neodymium-praseodymium, dysprosium and terbium oxides. The plant was prepared for continuous operations, with light rare earth output targeted first and heavy rare earth production expected later in 2026. Those campaigns are intended to generate the data needed to train and validate the digital twin.

The commercial case rests on flexibility. Aclara plans to feed its U.S. facility with mixed rare earth carbonate from ionic-clay projects in Brazil and Chile, while also leaving room to process third-party material. That multi-feed ambition is valuable because Western heavy rare earth supply is fragmented, but it makes process control harder. Changes in impurity levels, rare earth distribution and feed concentration can alter reagent consumption, recovery and product purity across a solvent-extraction train.

Aclara’s Carina project in Brazil gives the strategy scale. Its 2026 feasibility study outlined an 18-year mine life and average annual output of 4,378 tonnes of rare earth oxides contained in mixed concentrate. The product mix was reported to contain about 27.2% neodymium-praseodymium and 4.2% dysprosium-terbium, giving the project exposure to the heavy rare earths that are most difficult to source outside China.

The company is also advancing the Penco Module in Chile, where environmental approval was formalised in June. Penco is designed around ionic-clay extraction and production of a high-purity rare earth carbonate, with a feasibility study expected later in 2026. Together, Carina and Penco are intended to provide feed for separation in the U.S., followed by metals and alloys for permanent magnets.

Louisiana is therefore the point where the strategy must become an operating business. The planned separation facility received final approval for an industrial tax exemption expected to save about $20.8 million over its first five years. Aclara has estimated a broader capital requirement in the hundreds of millions of dollars for the U.S. separation build, meaning tax incentives and federal research support help, but do not replace project finance.

The market opportunity is clear. Dysprosium and terbium allow high-performance permanent magnets to retain magnetic strength at elevated temperatures, making them important in electric-vehicle motors, defence systems and other demanding applications. China remains dominant in heavy rare earth processing, and Western magnet projects still face a shortage of qualified non-Chinese feedstock and separation capacity.

Aclara’s advantage is that it is trying to connect heavy rare earth-rich ionic-clay feed with separation, metals and alloys rather than stopping at a mixed concentrate. Its risk is that vertical integration multiplies the number of technologies and projects that must work on schedule.

The DOE selection improves Aclara’s credibility with U.S. industrial-policy institutions and may lower scale-up risk by turning pilot data into a predictive operating model. It does not yet prove recovery, purity, reagent consumption or uptime at commercial throughput. The next meaningful evidence will come from heavy rare earth output at the Virginia pilot plant, completion of engineering for Louisiana, and demonstration that the digital twin can control a changing feed rather than merely reproduce a laboratory flowsheet.

For investors, the story has shifted from whether Aclara possesses heavy rare earth resources to whether it can make a flexible separation platform bankable. The DOE has endorsed the technical question. Aclara still has to answer it at industrial scale.