The playbook — potash · October 19, 2026
Sage Potash's Paradox Basin resource jumps 259% to 771 Mt
A Utah potash body now carries 226 Mt Measured and Indicated at 41.36% KCl, plus 545 Mt Inferred — big and high-grade, but lightly drilled.
Sage Potash released an updated resource estimate for its Sage Plain project in Utah’s Paradox Basin, and the headline number is big: 226 million tonnes in the Measured and Indicated categories grading 41.36% KCl — that’s potassium chloride, the potash salt that becomes fertilizer, and 41% is a rich, consistent grade. Below that sits another 545.4 million tonnes Inferred at 41.53% KCl. Together that’s 771 million tonnes of in-place sylvinite, a 259% jump in tonnage from the 2025 preliminary economic assessment. Trading was halted the same day, standard practice for material news. The company plans to file the supporting NI 43-101 technical report within 45 days.
Grade is the real story here. Across different beds and confidence categories, the potash runs between 40.7% and 42.27% KCl — remarkably even. That consistency matters for the mining method Sage proposes: solution mining, where you pump water underground to dissolve the salt and bring the brine back up. Steady grade makes that approach more predictable. One number to sit with carefully: the reported Upper 18 Measured bed shows 7.2 metres at 40.73% KCl. The company is led by CEO Pat Varas, a geologist who co-founded Western Potash and raised more than $240 million for its Milestone project in Saskatchewan — a project that never reached commercial production and entered court-supervised insolvency in 2024 after defaulting on over US$108 million in debt. He owns roughly 0.7–0.8% of Sage. The stock fell 15.4% over three months while a base-metals peer index rose 3.1%, an 18-point divergence that tracks the company, not the sector.
Now the honest part. These tonnes are gross in-place sylvinite at 100% extraction and 0% plant loss — the mining and economic deductions have explicitly not been applied yet, so the recoverable figure will be materially lower. The estimate rests on very few boreholes, with Peterson 1 and Johnson 1 named, and the Inferred radius of influence stretches to 4,000 metres — a wide reach from sparse data. No 3D seismic has been done, and a 25% area exclusion was applied for unknown seismic anomalies, acknowledging structural uncertainty. The Inferred 545 Mt dwarfs the M&I 226 Mt, so most of the resource sits in the lowest-confidence bucket. Some grades were read from gamma-ray logs where core was lost. The release claims “consistent lateral continuity” but gives no step-out distance, depth, or defined strike length. Next up: the technical report within 45 days, then solution-mining engineering and a Pre-Feasibility Study to show whether the economics hold.