Our desalination brine could be a magnesium mine

Every day, the WEB plant at Balashi turns seawater into about 40,000 m³ of drinking water and sends roughly 60,000 m³ of concentrated brine back to the sea. A longer outfall past the reef islands was too expensive, so that brine goes into a nearshore lagoon. We call it waste. I spent the bulk of my PhD research, done at KU Leuven while teaching in the SISSTEM program at the University of Aruba, asking what else it could be.

My thesis, Valorization of Reverse Osmosis Brines in Small Islands, was defended in May 2024. To valorize is to turn a waste stream into something with VALUE. My main conclusion is that a desalination plant does not have to be just a water factory with a waste pipe. It can be a seawater refinery, where fresh water is one product among several.

Why Aruba is the right place to ask

Few places have more experience with desalination than we do. The first commercial plant on the island dates from around 1903, WEB has produced public water since 1932, and since 2015 all of our water comes from seawater reverse osmosis. Water sales bring in about US$54 million a year, and desalination uses roughly 13% of the island’s electricity.

We are also not alone. At least 28 Caribbean islands run seawater desalination plants, with about 650,000 m³ a day of installed capacity between them. Whatever works at Balashi can be copied from Curaçao to the Cayman Islands.

What I tested

I looked at different ways to get value out of the brine but then focused on four main possibilities: lithium, chlorine and caustic soda, energy, and magnesium.

Lithium was the one everybody asked about, and it was also the clearest NO. Our brine holds about 0.33 mg of lithium per liter. Even if we captured every atom from the whole stream, we would get about one tonne a year. That does not pay for a plant. I did build and test a new type of electrode, manganese oxide held in a Kevlar film, that could one day soak up lithium passively at many plants and be shipped to one central facility for processing.

Chlorine and caustic soda (sodium hydroxide) are made by splitting salt water with electricity. Aruba imports about US$2.7 million a year of these products, and the desalination plant itself uses caustic soda. Making them locally would cut the risk of shipping chlorine by sea. It makes sense, but it is a small local market, not an export.

Energy can come from mixing brine with treated wastewater. In pressure retarded osmosis, fresh water flows through a membrane into the salty side and the pressure it builds drives a turbine. I modeled about 0.25 kWh per m³ of mixed water. With the Bubali treatment plant 13 km away, the pipeline kills the economics and payback runs past 30 years. With a treatment plant next to the desalination plant, the payback could drop to about five years. The lesson for planners is simple: the next wastewater plant should sit near Balashi.

Magnesium was the clear winner. Our brine carries about 2,150 mg of magnesium per liter. In the lab at KU Leuven, working with an MSc student, we added caustic soda to brine-strength solutions and precipitated over 95% of the magnesium as magnesium hydroxide, Mg(OH)₂, within minutes. That is the first step of the Dow process, which supplied much of the world’s magnesium from seawater in Freeport, Texas, from 1948 until a storm wrecked the plant in 1998. The chemistry is proven. What is new is doing it with renewable power, on a brine stream that already exists.

The magnesium numbers

The arithmetic is short. 60,000 m³ a day at about 2.15 kg of magnesium per m³ is around 129 tonnes a day, or roughly 47,000 tonnes a year. World demand in 2021 was about 1.03 million tonnes, so the whole stream at Balashi holds close to 5% of it.

Where that magnesium comes from today matters more than the volume. In 2021 China supplied 88% of the world’s magnesium, mostly through the Pidgeon process: dolomite rock heated with ferrosilicon in coal-fired retorts, in batches of about 11 hours, with workers loading and emptying the tubes by hand. It emits about 28 kg of CO₂ for every kilogram of magnesium. Electrolytic plants running on hydropower come in at about 8.5 kg, and 5.3 kg when they recycle their chlorine. The footprint is set mostly by the electricity, so magnesium from brine and solar power belongs at the low end.

Europe noticed. Magnesium metal is on the EU’s strategic raw materials list. The Critical Raw Materials Act sets a 2030 goal that no more than 65% of any strategic material should come from a single outside country. Aruba is part of the Kingdom of the Netherlands, has a brine stream, a disused refinery site and plenty of sun. That is an unusual combination.

There is a hard constraint, and it is energy. Refining magnesium metal takes about 14 to 25.6 kWh per kilogram. For 47,000 tonnes that means 660 to 1,200 GWh a year, while the whole island generated about 937 GWh in 2018. Turning all of our brine into metal would need about as much electricity as the whole island uses today, or more, most likely from offshore wind. So metal is a long-term goal, and the place to start is magnesium hydroxide.

Magnesium and cleaner cars

Magnesium is the lightest structural metal: 1.74 g/cm³, about a third lighter than aluminum and three quarters lighter than steel. According to the US Department of Energy, magnesium alloys can cut the weight of some parts by more than half, and every 10% cut in a car’s weight improves fuel economy by 6 to 8%. For electric cars, a Frost & Sullivan study for Messe Frankfurt puts the gain at 4 to 6% more range for the same 10%. Batteries make electric cars heavy, so every kilogram saved elsewhere means a smaller battery or a longer drive. Yet magnesium is still less than 0.5% of the weight of an average vehicle, mostly because of cost.

The catch is the production footprint. A 2015 life-cycle study in Environmental Science & Technology found that replacing steel with magnesium usually raises the emissions of building the car, and that the fuel saved over the car’s life usually, but not always, makes up for it. That calculation gets tougher for electric cars. As grids get cleaner, each kilogram saved on the road avoids less CO₂, so the CO₂ spent making the metal counts for more. An earlier study by Hakamada and colleagues adds that recycling magnesium parts is essential for the climate benefit to hold.

So the climate case for magnesium in cars depends on where the magnesium comes from. Coal-fired magnesium at 28 kg CO₂ per kilogram makes lightweighting a weak trade. Solar-powered magnesium from brine, made with a fraction of that, makes it a good one. That is where an island like ours could fit into the vehicle supply chain.

Cars are not the only market. Magnesium hydroxide is sold as a flame retardant, for neutralizing exhaust in ship scrubbers and in pharmaceuticals. Reactive magnesium oxide can be used in cements that absorb CO₂ as they harden, which is worth exploring with our local construction sector.

From thesis to project

I am now working with local and Dutch partners on the first step: a module on the former refinery terrain in San Nicolas that would take about a tenth of the brine and produce roughly 11,500 tonnes of magnesium hydroxide a year. It would make its own caustic soda with bipolar membrane electrodialysis, which splits salt water into acid and base with electricity, powered by dedicated solar. A side benefit is that the brine going back to the sea would be less salty.

I want to be honest about the money. A first module this size will not pay commercial investors on its own. It is a demonstration plant that needs grant support, and its job is to prove the technology at real scale so that the larger phases can be financed on normal terms. Our studies show that the economics improve with scale, and that is the reason to build the first step well.

What should happen next

Treat brine as a resource in policy. The water utility, the government and the environmental authority should agree that brine can be supplied, under clear rules, to a company that recovers materials from it.
Put the next wastewater plant near Balashi. Co-location makes energy recovery possible and closes the water loop on the island.
Start with magnesium hydroxide, not metal. It proves the process with energy we can supply and builds the track record that a metal refinery would need later.
Use it to train our people. The SISSTEM BSc and MSc programs can supply the engineers and researchers, so the knowledge stays on the island.

For a hundred years we have been world leaders in turning seawater into drinking water. The next step is to use the rest of what we pump out of the sea. The brine is already there every day, and the first module should be built before the decade is out.

The thesis was supervised by Prof. Bart Van der Bruggen (KU Leuven) and Dr. Nigel John (University of Aruba), within the EU-funded SISSTEM program. The main findings are also published in the Journal of Chemical Technology and Biotechnology.