Industries / Mining and materials / Mining and quarrying / Salt and mineral extraction
Solution mining and brines
Solution mining dissolves salt or potash underground with water and pumps the salty water, called brine, to the surface, and brine wells also pump up natural brines that hold lithium, bromine and other elements.
7
Milestones achieved, from 1835 to 2019.
3
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1835
A kilometer-deep brine well
Salt makers in Zigong, Sichuan, finished the Shenhai well at a depth of 1,001.42 m. They drilled by percussion: a few men stood on a plank lever to lift the drill pipe about 1 m, then dropped it so the bit broke the rock, centimeter by centimeter. China's Science and Technology Daily calls it the deepest drilled well in the world at the time and says the method later spread to the West.
Credit
The well drillers of Zigong; the source names no individual
1889
Bromine blown out of brine
Bromine, then used in medicines and photographic chemicals, was made by boiling brine over wood fires. In 1889 in Canton, Ohio, Herbert Dow freed the bromine with bleaching powder instead, dripped the brine over burlap sacks and blew the bromine out with a current of air, which saved the fuel. After he moved to Midland, Michigan, in 1890 he freed the bromine with electricity, and in his lifetime his company drew bromine, chlorine, sodium, calcium and magnesium from the brine under Midland.
Credit
Herbert H. Dow
1894
Melting sulfur underground
Sulfur deposits along the Gulf of Mexico could not be reached by ordinary mining because of their depth and the ground above them. Herman Frasch, working for Union Sulphur in Calcasieu Parish, Louisiana, pumped superheated water (water kept liquid under pressure above its boiling point, at about 165 °C) down a well to melt the sulfur, then lifted it with air. He made his first attempts in late 1894, and the method reached commercial success in 1903, by the US Geological Survey's account. US output by this method peaked at more than 8 Mt in 1974. Sulfur recovered from oil and gas processing overtook it in 1982, and the last US Frasch mine closed on 31 August 2000.
LessonA cheaper by-product can end a mining method that still works.
Credit
Herman Frasch; Union Sulphur Company
1964
Potash by solution mining
Kalium Chemicals began producing potash at Belle Plaine, west of Regina, Saskatchewan, by injecting water into potash beds up to a mile (1.6 km) below the surface and pumping the dissolved potash to a surface refinery and cooling pond. Mosaic, which runs the site today, lists 1964 as the year Kalium built "the first solution mine" at Belle Plaine.
Credit
Kalium Chemicals
1998
A sorbent that picks lithium
Direct lithium extraction (DLE) pulls lithium out of brine with a material that holds lithium and lets other salts pass, instead of waiting for the sun to evaporate the water. Building on work begun at Dow Chemical in the 1970s, John Burba and William Bauman filed sorbent patents in 1992. FMC licensed them in 1994 for its project at the Salar del Hombre Muerto in Argentina, and the plant started up in 1998, by Burba's own account. A January 2025 report by Centro de Energía UC in Chile gives 1997 and calls it the first commercial use of adsorption to extract lithium.
Credit
FMC (later Livent and Arcadium Lithium, now part of Rio Tinto); John Burba and William Bauman
2012
The Bayou Corne sinkhole
On 3 August 2012 a sinkhole opened at Bayou Corne, Louisiana, after the side wall of a cavern more than 3,000 ft (about 900 m) underground collapsed. The cavern, connected to a well run by Texas Brine Company, had been mined closer to the edge of the Napoleonville salt dome than thought. By March 2014 the hole covered about 10 ha and was at least 750 ft (about 230 m) deep. Researchers at NASA's Jet Propulsion Laboratory later found that airborne radar had recorded the ground moving up to 260 mm, mostly sideways, toward the site at least a month before the collapse.
LessonThe ground over a cavern can move weeks before it fails, and a cavern's distance from the edge of its salt has to be measured, not assumed.
Credit
Cathleen Jones and Ron Blom, NASA Jet Propulsion Laboratory (radar study, published in Geology, February 2014)
2019
Maceió's sinking neighborhoods
In 2019 the Geological Survey of Brazil (CPRM) studied unstable ground in the Pinheiro, Mutange and Bebedouro neighborhoods of Maceió, which sit above a field of caverns dissolved in salt by Braskem's solution mining. A 2025 study reports subsidence of the order of meters there, at up to 25 cm a year. It finds that the salt, 900 to 1,200 m down, alternates with shale less than half as strong as rock salt, that the pillars between caverns in the middle of the field were too small, and that roof fractures let several caverns grow to the top of the salt and in places hundreds of meters into the rock above.
LessonThe pillars and roofs of a cavern field have to be sized for the weakest layers in the deposit, not for pure salt.
Credit
Geological Survey of Brazil (2019 study); Wolfgang Minkley, Christoph Lüdeling, Dieter Brückner and Dirk Naumann (2025 analysis, Rock Mechanics and Rock Engineering)
Watch for
DLE at full rate
Eramet says its Centenario-Ratones plant in Salta, Argentina, which uses Eramet's own DLE process, reached its monthly nameplate rate of 2,000 t of lithium carbonate in October 2026, 15 months after ramp-up began. By the company's figures it produced about 6,700 t in 2025 and 8,440 t in the first half of 2026, against a nameplate capacity of 24,000 t a year.
Achieved whenEramet's audited figures show a full year near 24,000 t, or an independent party reports the plant's lithium recovery.
Claimed by
Eramet, November 2026 (reported by Panorama Minero)
Oilfield brine lithium in Arkansas
Smackover Lithium, a joint venture of Standard Lithium and Equinor, says its South West Arkansas Project will use DLE on brine from the Smackover formation, expects a final investment decision in 2026 and first production in 2028, and has signed a 10-year offtake for 8,000 t a year of battery-quality lithium carbonate with Trafigura.
Achieved whenthe plant sells battery-quality lithium carbonate and reports its output in audited filings.
Claimed by
Smackover Lithium (Standard Lithium and Equinor), March 2026
Lithium from hot Rhine brine
Vulcan Energy says its Lionheart project in Germany's Upper Rhine Valley will pump naturally hot, lithium-rich brine, use its heat to run the plant, and make 24,000 t a year of lithium hydroxide, with commercial production targeted for the second half of 2028. It says the project has an AUD 3.9 billion financing package and is in construction.
Achieved whenVulcan reports commercial lithium hydroxide sales and an independent party confirms the output.
Claimed by
Vulcan Energy, November 2026 (project page)
Open
01
Same brine, same scorecard
A public test for DLE materials. Write recipes for synthetic brines that match published analyses of four brine types (an Andean salar, a Smackover oilfield brine, an Upper Rhine geothermal brine and a Salton Sea geothermal brine), set one test procedure, and run at least three published sorbents through it.
The briefCredit · unclaimed02
Radar warnings, scored
A list of every reported sinkhole, collapse or sudden subsidence over solution-mined caverns and brine wells since 2015, each matched to a time series of free Sentinel-1 satellite radar images of the ground before the event, and a published tally of which events showed movement first and how many days ahead. Then a monthly public watch map for active cavern fields.
The briefCredit · unclaimed
Open 01 · the brief
Same brine, same scorecard
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public test for DLE materials. Write recipes for synthetic brines that match published analyses of four brine types (an Andean salar, a Smackover oilfield brine, an Upper Rhine geothermal brine and a Salton Sea geothermal brine), set one test procedure, and run at least three published sorbents through it.
Why it is still open
A 2023 review in Nature Reviews Earth & Environment found that only 30% of DLE test experiments were run on real brines, so the effect of other dissolved salts on results is often not measured, and said the fresh water DLE uses "needs to be urgently quantified".
Built with what exists
PHREEQC, the US Geological Survey's free water-chemistry program, to design the recipes; published sorbent recipes such as lithium manganese oxide and lithium aluminum layered double hydroxide; a university lab's ICP-OES instrument to measure the metals; the open dle-simulation model on GitHub to scale lab results to water and reagent use per tonne; a large language model such as Claude to pull past test results from papers into one table (see the Note "Seven more flagged").
First step
Collect published analyses of the four brine types and write the first PHREEQC recipe.
Done looks like
Recipes and procedure published under an open license, with results for at least three sorbents on all four brines: lithium recovery, lithium-to-magnesium and lithium-to-sodium selectivity, capacity after 200 load-and-strip cycles, and fresh water and acid used per kilogram of lithium. A second lab repeats one result to within 10%.
Open 02 · the brief
Radar warnings, scored
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A list of every reported sinkhole, collapse or sudden subsidence over solution-mined caverns and brine wells since 2015, each matched to a time series of free Sentinel-1 satellite radar images of the ground before the event, and a published tally of which events showed movement first and how many days ahead. Then a monthly public watch map for active cavern fields.
Why it is still open
The Bayou Corne warning was found by chance, in airborne radar flown to study coastal sinking, and its authors wrote only that radar "could offer a monitoring technique for identifying at least some sinkholes" if data were collected routinely. Searches for this brief found single-site studies, such as radar work at brine wells (Rucker and others, 2013) and at Maceió, but no scored test across many sites. A 2021 master's thesis at Delft University of Technology, by M. A. Felius, designed automatic warning methods and tested them on one simulated sinkhole, leaving tests on real collapses to future work.
Built with what exists
Sentinel-1 radar images from the European Space Agency, downloaded through the Copernicus Data Space Ecosystem or the Alaska Satellite Facility; the open radar-interferometry programs ISCE2 and MintPy; state registers of brine and storage wells to locate cavern fields; Python.
First step
Build the event list from regulator reports and news, and download Sentinel-1 images for the first three sites.
Done looks like
At least 10 events scored, each with hit or miss and lead time in days, plus the false-alarm rate over fields with no event, with code and data public.