BRENNEN

Industries / Mining and materials / Mining and quarrying / Uranium mining

Leaching in place

Leaching in place pumps a chemical solution down wells into a buried, water-filled layer of uranium-bearing sand, then pumps the uranium-rich water back up to a plant, without digging the ore out.

7

Milestones achieved, from 1962 to 2024.

2

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1962

    Wellfields tried in Wyoming

    The Utah Construction and Mining Company tested leaching in place at Shirley Basin, Wyoming, from 1961 to 1963. It tried five generations of wellfield layout and more than 100 well patterns with sulfuric acid; a wellfield is the grid of injection and pumping wells laid over the ore. The project ran on a small scale from 1962 to 1970, produced 577 tonnes of uranium, and was then turned into an open pit. By the IAEA's 2016 account, no commercial mine using acid for this method has been developed in the United States since.

    www-pub.iaea.org

    Credit

    Utah Construction and Mining Company

  2. 1980s

    Ammonia leaching abandoned

    By May 1980 the United States had 18 commercial and 9 pilot projects leaching uranium in place, nearly all using alkaline solutions of ammonia or sodium carbonate and bicarbonate. Groundwater at the ammonia sites proved hard to restore, so by the early 1980s operators shifted to sodium carbonate and bicarbonate, or to carbon dioxide and oxygen, by the IAEA's account.

    LessonThe chemical chosen to dissolve the uranium decides how hard the water will be to clean afterwards.

    www-pub.iaea.org

    Credit

    US operators in Texas, Wyoming, Colorado and New Mexico

  3. 1996

    Acid leaching ends at Stráž

    At Stráž pod Ralskem in northern Czechoslovakia, leaching in place with sulfuric acid ran at industrial scale from 1971 and had yielded 16,002 tonnes of uranium by the end of 2010, by the IAEA's count. The operation injected more solution than it pumped out, which spread acid beyond the well patterns. DIAMO, the state company now in charge, counts more than 4 million tonnes of sulfuric acid and other chemicals injected over about 32 years and more than 300 million cubic meters of groundwater contaminated. The Czech government ended chemical leaching on 1 April 1996 and ordered active cleanup, which DIAMO expected in 2008 to finish in 2035.

    LessonA wellfield must pump out slightly more than it injects, or the solution spreads.

    www-pub.iaea.orgimwa.info

    Credit

    Czechoslovak state uranium industry; DIAMO (cleanup)

  4. 2001

    Saline aquifer left to recover

    Beverley in South Australia, discovered in 1969, ran a field trial with sulfuric acid and ion exchange (resin beads that catch the dissolved uranium) in 1998 and was officially opened in 2001, by the IAEA's account. The water in its ore zone holds more than 12 grams of dissolved solids per liter and is unusable before mining begins. Instead of pumping it clean, the operator monitors natural attenuation, the slow return of the water's chemistry as the spent solution reacts with the surrounding rock, following a 2010 Geoscience Australia best practice guide, according to the World Nuclear Association.

    www-pub.iaea.orgworld-nuclear.org

    Credit

    Heathgate Resources

  5. 2009

    Kazakhstan becomes top producer

    Soviet geologists found sandstone uranium deposits in Kazakhstan's Chu-Sarysu basin that suited leaching in place. Tests at Uvanas in 1970 and 1971 succeeded, and production began at Uvanas and Mynkuduk in 1978, all with sulfuric acid. Output fell after independence, then foreign investment in mines such as Inkai and Akdala brought it back, and in 2009 Kazakhstan became the world's largest uranium producer, by the IAEA's account. The World Nuclear Association counts leaching in place as 56% of the world's mined uranium in 2022, up from 16% in 2000.

    www-pub.iaea.orgworld-nuclear.org

    Credit

    Kazatomprom; the Soviet-era Stepnoye, No. 6 and Central mining companies

  6. 2009

    Restored water above baseline

    Susan Hall of the US Geological Survey went through Texas Commission on Environmental Quality records for 27 mines and 77 production areas, the separately permitted blocks of a wellfield. Every production area had been granted relaxed "amended" cleanup goals for at least one element, and only 22 had final measurements. In 68% of those 22, uranium after cleanup was above the pre-mining baseline, although 95% of Texas production areas already held uranium above the US drinking water limit of 0.03 milligrams per liter before mining. A 2015 study of one Wyoming mine by Elizabeth Ruedig and Thomas Johnson found uranium 5.6 times higher after restoration than at baseline in matched wells.

    LessonCleanup goals said little about results until someone compared final measurements with the baseline, wellfield by wellfield.

    pubs.usgs.govosti.gov

    Credit

    Susan Hall, US Geological Survey

  7. 2024

    Acid shortage cuts output

    On 23 August 2024 Kazatomprom cut its 2025 production guidance by 5,000 tonnes of uranium, to 25,000 to 26,500 tonnes on a 100% basis, citing uncertain sulfuric acid supplies and construction delays at new deposits. The World Nuclear Association gives a general figure of about 40 kg of acid per kg of uranium for Kazakh mines, with other figures up to twice that, and notes that acid supply had already constrained production in 2007 to 2010.

    LessonFor acid leaching, the supply of acid limits output as much as the orebody does.

    world-nuclear-news.orgworld-nuclear.org

    Credit

    Kazatomprom (report)

Watch for

  1. Leaching inside a frozen wall

    Denison Mines is building Phoenix, at Wheeler River in Saskatchewan's Athabasca Basin, which the Canadian Nuclear Safety Commission says would be the first uranium mine in Canada to use leaching in place. Denison's plan keeps the solution in by pumping out more than it injects, and adds a freeze wall, a ring of ground frozen by brine pumped through more than 300 drill holes down to the basement rock, as a third barrier. The commission licensed site preparation and construction on 18 February 2026, but not operation. Denison began full construction in July 2026 and targets first production in mid-2028.

    Achieved whenPhoenix produces uranium under an operating license and the regulator's monitoring shows the solution stayed inside the frozen ring.

    api.cnsc-ccsn.gc.caapi.cnsc-ccsn.gc.camagazine.cim.org

    Claimed by

    Denison Mines, July 2026

  2. Carbon dioxide leaching in China

    China National Nuclear Corporation (CNNC) says its National Uranium No. 1 demonstration project in Ordos, Inner Mongolia, produced its first barrel of uranium on 12 July 2025, one year after construction began. It leaches with carbon dioxide and oxygen in a closed loop, which CNNC says avoids water, gas and solid wastes, and CNNC calls it the largest natural uranium production project in its industry in 70 years.

    Achieved whensomeone outside CNNC publishes the project's output and its groundwater monitoring results.

    world-nuclear-news.org

    Claimed by

    CNNC, July 2025

Open

  1. 01

    Blind test for restoration forecasts

    A public forecasting test for wellfield cleanup. Give modelers each Texas production area's pre-mining water chemistry, the leach solution used, the cleanup method and the volume pumped, hide the measured water chemistry after cleanup, and score their predictions of uranium, selenium and other elements against it.

    The brief
    Credit · unclaimed
  2. 02

    Tabletop test of leak alarms

    A glass sand tank holding a small wellfield, run with salt water as a harmless stand-in for leach solution. A ring of monitor wells carries water-level and conductivity sensors, and a deliberately unsealed "old drill hole" lets solution leak into an upper sand layer. Compare how fast and how often different alarm rules catch the leak: the fixed upper limits used in practice against trend and anomaly detection software.

    The brief
    Credit · unclaimed

Open 01 · the brief

Blind test for restoration forecasts

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public forecasting test for wellfield cleanup. Give modelers each Texas production area's pre-mining water chemistry, the leach solution used, the cleanup method and the volume pumped, hide the measured water chemistry after cleanup, and score their predictions of uranium, selenium and other elements against it.

Why it is still open

The USGS compared final results with baseline in 2009 but did not test forecasts. Published studies measure one mine at a time, such as Ruedig and Johnson's Wyoming study, and Los Alamos National Laboratory reported a laboratory leaching and restoration demonstration on deposit cores in September 2026. A search for this brief found reactive transport models run for single pilot or proposed sites, but no test that scores forecasts across many wellfields.

pubs.usgs.govosti.govosti.gov

Built with what exists

The USGS data release of historic Texas wellfield water quality by Hall and Hennesy, which lists baseline values, leach solution, restoration method, pore volumes pumped and post-restoration chemistry; by this brief's count it has post-restoration uranium values for 37 production areas at 16 mines; PHREEQC and PHAST, the USGS's free geochemical and reactive transport codes; Python with pandas and scikit-learn.

sciencebase.gov

  1. First step

    Download the USGS spreadsheet and reproduce Hall's 2009 finding that uranium ended above baseline in 68% of the production areas with final values.

  2. Done looks like

    Forecasts from at least three methods, for example a reactive transport model, a simple statistical rule and a machine learning model, scored on all 37 production areas with whole mines held out, and the scores, code and data split published.

Open 02 · the brief

Tabletop test of leak alarms

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A glass sand tank holding a small wellfield, run with salt water as a harmless stand-in for leach solution. A ring of monitor wells carries water-level and conductivity sensors, and a deliberately unsealed "old drill hole" lets solution leak into an upper sand layer. Compare how fast and how often different alarm rules catch the leak: the fixed upper limits used in practice against trend and anomaly detection software.

Why it is still open

US practice watches rings of monitor wells, keying sideways leaks to water levels and upward leaks to water quality. Oak Ridge National Laboratory's 1986 review for the NRC found that leaks into overlying aquifers through poorly plugged exploration holes "continue to plague" the industry. The closest software, the ISL Guard prototype posted on GitHub in October 2026, runs anomaly detection on synthetic data only. In 1983 Pacific Northwest Laboratory, working for the NRC, found in laboratory tests that chloride, sulfate and total dissolved solids pass through aquifer sediment largely unchanged, and recommended them as signs of a leak, which is why salt water read by a conductivity sensor is a fair stand-in; that work tested which readings to watch, not alarm rules. A search for this brief found field tracer tests and laboratory chemistry tests, but no physical bench test of alarm rules.

osti.govosti.govgithub.comosti.gov

Built with what exists

A glass tank about 1 m long, graded sand with a clay layer, table salt, peristaltic pumps, conductivity and pressure sensors on an Arduino or Raspberry Pi, Python with scikit-learn's Isolation Forest, and MODFLOW 6 with FloPy to model the tank.

  1. First step

    Build the tank with five wells and one monitor ring, and record how long a salt pulse takes to reach the ring.

  2. Done looks like

    20 filmed runs with a leak and 20 without, and a published tally for each alarm rule of hours to detection and false alarms.