Highlights

• ReElement Technologies announced a major expansion of its Marion, Indiana refining campus, targeting over 16,000 metric tons per year of separated oxides with initial production expected in Q3 2026, focusing on Nd, Pr, Dy, and Tb to address a critical U.S. supply chain gap currently dominated by China (85–90% of global refining).
• The company’s multi-feedstock approach (combining recycling and primary inputs) with modular scaling shows strategic promise, though its proprietary separation process still requires independent validation at industrial scale for throughput, cost, and purity versus proven solvent extraction methods.
• Key execution risks remain around NdPr-equivalent output values, secured feedstock and offtake agreements (including known partners Pensana, POSCO, and Vulcan Elements), and real-world yields at commercial scale.

ReElement Technologies has announced a major expansion of its Marion, Indiana refining campus, targeting over 16,000 metric tons per year of separated oxides, with initial production expected in Q3 2026. The focus—Nd, Pr, Dy, and Tb—addresses a critical U.S. supply chain gap dominated by China (~85–90% of global refining).

The news comes via American Resources  an affiliated minority holder in ReElement Technologies.

Strategically, the model makes sense. A multi-feedstock approach (recycling + primary inputs), modular scaling, and proximity to end markets align with Western efforts to rebuild domestic capacity. If executed, this could represent meaningful midstream progress.

The company claims improved efficiency over traditional solvent extraction (SX). Yet SX remains the only commercially proven large-scale separation method. ReElement’s process still needs independent validation at industrial throughput, cost, and purity—raising execution risk.

The 16,000 tpa figure also blends multiple streams. Key unknowns include:

• NdPr-equivalent output (true economic value)
• Secured feedstock and offtake agreements (some are known—Pensana upstream, POSCO, Vulcan Elements downstream)
• Real-world yields and scalability

REEx Bottom Line: This targets the right problem—and remains in the early stages of execution.