The airstrip comes first. On the Patterson Lake peninsula in northern Saskatchewan, NexGen Energy has spent 2026 finishing a 5,840-foot airstrip, crushing more than 575,000 tonnes of aggregate, and building camp accommodation for the workforce that will sink two shafts — one to about 630 metres, one to about 530 — into one of the richest uranium deposits ever found. The Canadian Nuclear Safety Commission issued the licence to prepare the site and construct the Rook I project on March 5, 2026. Official construction began that July. If the schedule holds, first production comes around 2030, and the mine will then produce up to 30 million pounds of uranium a year — roughly a fifth of everything the world's mines produce today.
The deposit underneath justifies the scale. The Arrow deposit, discovered in February 2014, holds probable reserves of about 239.6 million pounds of uranium at an average grade of 2.37% U3O8, with measured and indicated resources of about 256.7 million pounds plus another 80.7 million pounds inferred, according to NexGen's 2021 feasibility study. The study sketched an 11-year mine life producing about 233.6 million pounds of yellowcake, an after-tax net present value of C$3.5 billion at an 8% discount rate, a 52.4% internal rate of return, and a payback period of less than a year. The planned annual output of about 30 million pounds would exceed the licensed capacity of Cameco's McArthur River–Key Lake operations, which stands at 25 million pounds. Rook I is not an incremental mine. It is the largest uranium mine under construction anywhere, and the industry's supply gap now depends on it working.
That dependence is not theoretical. The World Nuclear Association's World Nuclear Fuel Report 2025 projects reactor requirements rising from 175 million pounds of U3O8 in 2024 to 391 million pounds by 2040 — more than double. Against that, the two largest producers are struggling to hold their ground. Cameco's share of uranium production fell 20% year over year to 3.5 million pounds in the third quarter of 2026, after an extended maintenance shutdown at the Key Lake mill ran into the fourth quarter, and the company cut its 2026 production guidance to 19.5–20.5 million pounds from 19.5–21.5 million. Kazatomprom, the world's largest producer, mined 67.18 million pounds in 2025 and guides 71.49–75.39 million pounds for 2026 — but its 2027 production depends on sulphuric acid it has not yet secured, after Russia's Resolution No. 1158 of September 12, 2026 put acid exports under prime-ministerial approval. Each kilogram of uranium mined in Kazakhstan takes 70 to 80 kilograms of acid. The gap between what reactors will need and what mines can supply is measured in tens of millions of pounds a year, and Rook I is the single largest answer on the drawing board.
The market has noticed. In December 2024, NexGen signed uranium sales contracts with major US utilities committing to supply 1 million pounds a year from 2029 to 2033, and in 2025 it signed offtake agreements covering 2 million pounds a year for the first five years of production — 10 million pounds in total, all with US counterparties. A mine that will not produce until about 2030 already has buyers for its first decade of output. The C$2.2 billion construction cost, spread over a four-year build, is large for a single company, but the offtake gives lenders something concrete to underwrite: contracted pounds, at prices set against a term market that has just broken its 2007 ceiling.
None of that makes Rook I a sure thing, and the honest case against it is strong. Four years is the plan; mining schedules slip, and this one involves sinking shafts more than half a kilometre deep into basement rock in a remote corner of Saskatchewan, where the construction season is short and the logistics run through a single airstrip and a road that needs upgrading. The CNSC licence authorises site preparation and construction — it does not authorise production, and a separate operating licence will be needed before a single pound is milled. The feasibility study's 52.4% return was calculated at uranium prices and costs from 2021; the C$2.2 billion budget has already been tested by the same cost inflation that has hit every large mining project of the last five years. And the demand forecast it rests on assumes a reactor build-out — 81 reactors under construction worldwide as of September 30, 2026, 78 GW under construction per the International Energy Agency — that has slipped before, in country after country, whenever budgets or politics intervened.
There is also the American counterweight. US uranium concentrate production reached 2.1 million pounds in 2025, more than triple the 657,000 pounds of 2024 and the most since 2017, according to the Energy Information Administration, and the first half of 2026 added another 2.13 million pounds. That is a real recovery — and it is still about 5% of what US reactors buy in a year. Domestic production covers 2% to 4% of national consumption. The United States cannot mine its way out of the gap, and neither can any single project. Rook I at full output would cover roughly a sixth of annual reactor requirements. The gap needs Rook I, and it needs more than Rook I.
The implication for anyone financing uranium is to watch the build, not the price. The spot price will move on weekly trading volumes and fund flows. Rook I will move on shaft sinking in 2027, on the operating licence, and on whether the C$2.2 billion holds. If it delivers around 2030 at something close to 30 million pounds a year, the supply gap of the 2030s narrows by a fifth in a single stroke, and the term price that just broke its record will have been the market paying in advance for exactly this mine. If it slips, the gap widens, and the buyers who signed offtake for 2029 will be back in a market with fewer pounds than they planned for. Either way, the airstrip in northern Saskatchewan is now the most important construction site in the uranium business.
Sources: Carbon Credits, “NexGen Energy Begins Construction of Rook I as Uranium Demand Rises” (2026; C$2.2 billion project; about 30 million pounds a year; probable reserves about 239.6 million pounds at 2.37% U3O8; measured and indicated about 256.7 million pounds plus 80.7 million inferred; 11-year mine life, about 233.6 million pounds total, after-tax net present value C$3.5 billion at 8%, 52.4% internal rate of return, payback under one year; full 5,840-foot airstrip by end of 2026; CNSC approval March 2026), https://carboncredits.com/nexgen-energy-begins-construction-of-rook-i-as-uranium-demand-rises/. Canadian Mining Report, “NexGen Rook I Construction Begins” (2026; CNSC licence March 5, 2026; construction from July 2026; about 29 million pounds a year in the first five years; offtake 2 million pounds a year for the first five years with US counterparties), https://www.canadianminingreport.com/blog/nexgen-energy-crosses-the-rubicon-construction-begins-on-the-world-s-largest-uranium-mine. Skillings Mining Review (2026; shaft depths about 630 metres and 530 metres; first production around 2030; licence valid to March 31, 2036; about 459 full-time operating jobs over a 24-year life), https://www.skillings.net/nexgen-rook-i-2-2b-saskatchewan-uranium-build-begins/. Zacks (2026; NexGen contracts with major US utilities for 1 million pounds a year 2029–2033, signed December 2024; Rook I targeting 30 million pounds a year at C$13.86 per pound), https://www.zacks.com/stock/news/2905706/denison-mines-vs-nexgen-energy-which-uranium-stock-to-buy-now. Aranca Research (August 2026; World Nuclear Association World Nuclear Fuel Report 2025 — reactor requirements 175 million pounds in 2024 rising to 391 million pounds by 2040), https://www.aranca.com/assets/docs/Uranium-Structural-Supply-Gap-Creating-the-Next-Investment-Opportunity.pdf. Zacks, “Should You Buy, Hold or Sell Cameco Stock Post Q3 Earnings?” (November 2026; Cameco Q3 2026 production 3.5 million pounds, down 20% year over year; 2026 guidance cut to 19.5–20.5 million pounds; McArthur River/Key Lake 10–10.5 million pounds), https://www.zacks.com/stock/news/3023132/should-you-buy-hold-or-sell-cameco-stock-post-q3-earnings. World Nuclear News (February 17, 2026; Kazatomprom 2025 production 67.18 million pounds, up 10–11%; 2026 guidance 71.49–75.39 million pounds), https://www.world-nuclear-news.org/articles/Cameco-Kazatomprom-release-2025-Figures. agency.trading, “Uranium's 2027 acid problem” (October 8, 2026; Russia Resolution No. 1158, September 12, 2026; 70–80 kilograms of acid per kilogram of uranium; 2027 terms unsigned), https://agency.trading/takes/uranium-s-2027-acid-problem. International Energy Agency, Global Energy Review 2026 — Nuclear (78 GW under construction in 15 countries), https://www.iea.org/reports/global-energy-review-2026/technology-nuclear. US Energy Information Administration, 2025 Domestic Uranium Production Report (June 2026; concentrate 2.1 million pounds in 2025, most since 2017, more than triple 2024; mine production 1.388 million pounds), https://www.eia.gov/uranium/production/annual/.