Magnesium recovery from brine: routes, product grades and reagent economics

Home Magnesium recovery from brine: routes, product grades and reagent economics

Magnesium recovery from brine almost always means one reaction: raise the pH with lime, dolomitic lime or caustic soda until magnesium hydroxide, Mg(OH)2, precipitates, then wash it, dewater it and, if the buyer wants magnesia, calcine it to MgO. The chemistry is proven at industrial scale. The US Geological Survey still lists US companies recovering magnesium compounds from seawater in California and Delaware, from well brines in Michigan and from lake brines in Utah (USGS Mineral Commodity Summaries 2026, magnesium compounds).

What decides a project is narrower: which product grade your brine can reach, and whether the reagent costs less than the product is worth. This page covers both, with dated 2025 prices and two worked examples.

How much magnesium is in the brine

Water Magnesium (mg/L) Source
Standard seawater 1,290 Saudi Water Authority brine mining report, 2025
Arabian Gulf seawater at Jubail and Kuwait 1,530 to 1,570 Saudi Water Authority, as above
Reverse osmosis desalination brine, range cited in a review 1,860 to 2,880 Fontana et al., technical review, 2022
Perth Seawater Desalination Plant brine, calculated from design recovery 2,540 Gong, CEED seminar proceedings for Water Corporation, 2018

Seawater reverse osmosis brine usually carries about 1,860 to 2,880 mg/L of magnesium, the range in the review cited below, and Gulf seawater starts richer than standard seawater. Reverse osmosis brine is the most common feed discussed, but the same chemistry applies to industrial brines, evaporation pond bitterns and nanofiltration reject. Calcium travels with the magnesium. In seawater-derived brine it runs at about a third of the magnesium concentration, and it is the main source of impurity in the product. For the wider picture of what else desalination brine contains and which minerals have markets, see our page on mineral recovery from desalination brine.

The precipitation routes compared

Every route rests on the same stoichiometry: one magnesium ion takes two hydroxide ions. Per tonne of Mg(OH)2 that means 1.37 t of sodium hydroxide (100% basis) or 0.96 t of calcium oxide, which is about 1.16 t of a quicklime containing 83% CaO.

Reagent Reagent per tonne of Mg(OH)2 Purity reported Strengths Problems
Caustic soda (NaOH) 1.37 t (100% basis) 98 to 100% in most runs of one study cited in the 2022 review Highest purity; adds no calcium Fine precipitate that settles slowly and holds liquor; reagent cost
Lime, Ca(OH)2 0.96 t as CaO About 97% in a membrane crystalliser study (2022 review); 61.7 to 91% in tests on Perth brine (CEED, 2018) Cheapest alkali Calcium carbonate and gypsum report to the product; calcium returns to the brine
Dolomitic lime (calcined dolomite) Not reported in the sources reviewed Not reported Used industrially; part of the magnesium comes from the dolomite itself Same calcium issues as lime
Ammonia Not reported in the sources reviewed 93.5% in one study (2022 review) Soluble reagent, easy dosing Ammonium left in the brine is a hazard if the brine later feeds chlor-alkali electrolysis, because it can form explosive nitrogen trichloride

The Perth work is a useful warning about lime. With unscreened lime, product purity was 68%, and across lime doses of 60 to 120% of stoichiometric it stayed between 61.7% and 66.3%. Screening the lime slurry through 100 µm raised it to 91%. The impurities were calcium carbonate at about 13 to 20% and gypsum at about 7 to 13%, with gypsum rising as the lime dose rose. The brine had already been acidified to pH 4 and stripped of carbon dioxide before precipitation.

Three design steps do most of the work on purity:

  1. Remove carbonate first by acidifying and stripping carbon dioxide, so calcium carbonate cannot form alongside the hydroxide.

  2. Deal with calcium before the magnesium step, either by precipitating it with soda ash or by splitting the brine with nanofiltration. The Saudi Water Authority’s brine mining flowsheet uses nanofiltration this way: the reject, rich in magnesium and sulfate, becomes the magnesium feed.

  3. Grow filterable crystals with seeding, controlled reagent addition and ageing. Caustic-precipitated Mg(OH)2 tends to settle slowly, and flakes that hold mother liquor carry its chloride into the product.

Our ZLD Guide chapter on product-specific pathways sets these steps in the context of a full valorisation flowsheet.

Product grades and what buyers check

Magnesium from brine is sold as a slurry, a powder, a calcined oxide or, rarely, a metal. Each grade has a different gatekeeper:

Product What the buyer checks A current benchmark
Mg(OH)2 slurry for wastewater pH control Mg(OH)2 in dry solids, solids content, calcium, particle size, viscosity Martin Marietta Magnesia Specialties FloMag H data sheet, version dated 3/5/2025 (supplier document): Mg(OH)2 98.5% minimum of dry solids (typical 98.7%), dry solids 58.5% minimum (typical 60.9%), CaO typical 0.6%, median particle size 2.5 µm, viscosity 70 to 600 cP
Dead-burned magnesia for refractories MgO purity, boron as B2O3, lime to silica ratio, grain density Synthetic magnesia made with boron below 0.02% B2O3, ideally below 0.01%; commercial MgO grain targeted above 3,400 kg/m3 bulk density (IspatGuru technical reference on magnesia refractories)
Caustic-calcined magnesia, magnesium salts Reactivity, purity, consistency About 78% of US magnesium compound consumption goes to caustic-calcined magnesia, magnesium chloride, magnesium hydroxide and magnesium sulfates for environmental, de-icing, chemical and agricultural uses (USGS 2026)
Magnesium metal Made from MgO or magnesium chloride by smelting or electrolysis China produced about 950,000 t of the world’s 1.1 million t of primary metal in 2025 (USGS 2026, magnesium metal)

Compare the lime results above with the first row. A product at 61.7 to 91% Mg(OH)2 does not meet a 98.5% wastewater-grade specification, so a lime route on unsoftened brine is a non-starter for that market unless calcium is dealt with first.

Why boron decides the refractory market

Boron is the quiet specification killer. It forms a low-melting phase with lime that weakens magnesia refractories at temperature, which is why refractory grades cap it so tightly. Brine carries boron, and magnesium hydroxide picks it up: a 2004 study in Desalination used magnesium hydroxide precipitated from seawater to co-precipitate boron, removing up to 95% of it under the best conditions (Polat et al., 2004).

Worked example (hypothetical). Assumptions: brine at 2,540 mg/L magnesium (the Perth figure) and 5 mg/L boron (within the 4.3 to 5.7 mg/L range a 2025 review lists for desalination brine); all magnesium precipitated and calcined to MgO; worst case, all boron reports to the solid.

  • MgO per cubic metre of brine: 2,540 g × 40.30 / 24.31 = 4,212 g.

  • B2O3 per cubic metre: 5 g × 69.62 / 21.62 = 16.1 g.

  • B2O3 in the MgO: 16.1 / 4,212 = 0.38%, about 19 times the 0.02% refractory limit.

Real co-precipitation will be partial, but if more than about one-twentieth of the boron reaches the solid, the product fails the limit. If refractory magnesia is the target, boron removal has to be designed into the flowsheet and proven in testing. Wastewater-treatment grade, where the data sheet above sets no boron limit, is the easier first market.

What the reagent costs against what the product earns

Prices for magnesium hydroxide vary widely by source and region, and reporters rarely state the grade:

Product Price Period Source
Mg(OH)2, USA US$796 per tonne Q3 2025 IMARC, published 26 December 2025
Mg(OH)2, China US$490 per tonne Q3 2025 IMARC, as above
Mg(OH)2, France, India and Brazil US$751, US$750 and US$1,034 per tonne Q3 2025 IMARC, as above
Mg(OH)2, USA, delivered CFR Los Angeles about US$1,639 per tonne Q4 2025 ChemAnalyst
Magnesium metal, European free market US$2,500 per tonne (estimated annual average) 2025 USGS 2026, magnesium metal
Quicklime, US average value at plant US$260 per tonne (estimated) 2025 USGS 2026, lime
Caustic soda, USA about US$358 per tonne, basis not stated Q2 2026 IMARC caustic soda price index

Two reporters put US magnesium hydroxide about a factor of two apart in consecutive quarters, partly because the higher figure is a delivered price. For a business case, that spread is the point: get quotes from the buyers you would actually sell to, for the grade you can actually make.

The caustic break-even. Because 1.37 t of sodium hydroxide makes 1 t of Mg(OH)2, caustic alone costs as much as the product when caustic reaches US$357 per tonne against a US$490 product, or US$580 per tonne against a US$796 product. The US caustic figure above sits right at the lower break-even, before energy, filtration, drying, labour, capital and freight.

Worked example (hypothetical). Assumptions: a 10,000 m3 per day brine side-stream at 2,540 mg/L magnesium; 90% of the magnesium precipitated; caustic at US$400 per tonne (100% basis); quicklime at US$260 per tonne with 83% CaO; product valued at IMARC’s Q3 2025 China and US figures, US$490 to 796 per tonne.

Item Caustic route Lime route
Magnesium precipitated 22.9 t per day 22.9 t per day
Mg(OH)2 produced (dry) 54.9 t per day 54.9 t per day, less pure
Reagent needed 75.2 t NaOH per day 63.5 t quicklime per day (52.7 t CaO)
Reagent cost about US$30,100 per day about US$16,500 per day
Product value at US$490 to 796 per tonne US$26,900 to 43,700 per day Lower, unless calcium is removed first

Read across the caustic column: at the low end of the price range the reagent alone costs more than the product earns. The lime column looks cheaper, but its product will not meet a 98.5% specification without the softening step, and softening has its own reagent and sludge cost. Magnesium recovery tends to work where the site can use the product itself, where caustic is made or priced cheaply on site, where calcium is removed anyway for another reason, or where a buyer pays for a premium grade.

Projects and players to know (status as of November 2026)

  • Established production: US companies recover magnesium compounds from seawater, well brines and lake brines (USGS 2026). US output of magnesium compounds was an estimated 400,000 t of MgO content in 2025.

  • Desalination brine: the US Government Accountability Office reported in July 2025 that one early project, in Saudi Arabia, had begun full-scale operation, producing magnesium to supplement drinking water rather than for industrial sale, and that scalability and economic viability have not yet been demonstrated (GAO-25-108484). The Saudi Water Authority’s 2025 report describes it as a nanofiltration magnesium supplementation plant at Shoaibah.

  • Magnesium metal from brine: USGS reports that in July 2025 a California-based company produced a pilot-scale metal sample that met a US defence purity target. In December 2025 TETRA Technologies and Magrathea Metals signed a term sheet for a joint venture to produce magnesium from TETRA’s Arkansas brine, with TETRA stating that Magrathea has Defense Production Act Title III funding for a first commercial phase (TETRA press release, 2 December 2025). These are company statements; no production capacity or date for magnesium was given.

  • Supply context: the only US primary magnesium smelter, in Utah, reportedly stopped production in 2022 and its operator filed for Chapter 11 protection in September 2025 (USGS 2026, magnesium metal). In the EU, magnesium is on the critical raw materials list and magnesium metal on the strategic raw materials list of the Critical Raw Materials Act, Regulation (EU) 2024/1252 (JRC Raw Materials Information System).

  • Laboratory routes: researchers at Institute of Science Tokyo reported in October 2026 a liquid tin process that recovered more than 99% of the magnesium the tin absorbed from artificial brine, in tests with about 8 g of tin (EurekAlert release). It has not been run continuously or on real plant brine.

When is magnesium recovery worth pursuing?

  • the brine is rich in magnesium, ideally concentrated or split by nanofiltration;

  • calcium and carbonate can be removed economically, or are removed anyway;

  • you have a defined grade and a buyer, or a use on your own site, such as neutralising acidic effluent;

  • caustic is cheap where you are, or a lime route can reach the grade after softening;

  • boron is either below the buyer’s limit or removable;

  • the residual brine after magnesium removal has a route, because it is now richer in sodium and, with lime, calcium, and if it is crystallised it can end up as mixed salt for disposal.

Site use, filtration and metal: other questions

Can we use the Mg(OH)2 on site? In our view it is often the best outlet. It avoids freight and packaging, the specification is yours to set, and it replaces a purchased neutralising chemical.

Why does our precipitate filter so badly? Fast, high-pH caustic addition makes fine, gel-like particles. Slower dosing, seeding with recycled solids and an ageing step usually help, but they need to be developed in bench and pilot tests (laboratory and pilot testing).

Is magnesium metal from brine realistic? Not as a first step for most sites. Metal needs high-purity feed and an electrolysis or smelting plant, and projects outside China were still at pilot or term-sheet stage at the end of 2025.

Plan the grade before the reactor

The product grade, the reagent and the buyer have to be chosen together, before anyone sizes a precipitation tank. Brine Consulting is an independent industrial water advisory firm in Amsterdam that sells no treatment equipment. For magnesium, our brine and resource recovery work starts from your brine analysis and the grade question: which product your calcium, carbonate and boron levels allow, which reagent reaches that grade at your prices, and which bench tests would prove it. Sites work with us remotely or through a defined engagement. If sulfate is also present in your stream, see our guide to sodium sulfate recovery from wastewater; for disposal of what remains, see brine disposal options.

Send us your brine analysis through the consultation page, or book a scoping call.

Written by the Brine Consulting advisory team. The Brine Consulting advisory practice includes Dr. A. Mohammed Farooque, a senior desalination and membrane technology expert with more than 30 years of experience, whose work includes developing multistage nanofiltration technology for magnesium enrichment of desalinated water, and Mahmoud Gomaa, a chemical engineer with a master’s degree and a sharp focus on crystallization processes and the recovery of high-value salts like potassium and magnesium. Meet our experts.

Recent Posts

Questions about your site's water or brine?

Tell us briefly what you are working on and one of our experts will get back to you.