The effluent problem at a spodumene-to-hydroxide refinery is mostly one salt. Converting lithium sulphate to lithium hydroxide with caustic soda leaves sodium sulphate behind: at least 1.69 tonnes for every tonne of lithium hydroxide monohydrate by the chemistry alone, and between about 1.8 and 2.3 tonnes in the licences issued to Western Australian refineries. Zero liquid discharge for a lithium refinery is therefore less a water question than a salt question. Where do those tonnes go, in what form, and who will take them?
This page deals with refinery effluent. Getting lithium out of brine in the first place is a different problem, covered in our post on how lithium is extracted from brine.
Where the sodium sulphate comes from
Most hard-rock lithium hydroxide plants follow the same sequence. Spodumene concentrate is heated to convert it to beta spodumene, roasted with sulphuric acid and leached in water to give a lithium sulphate solution. After impurities are removed, caustic soda swaps the sulphate for hydroxide:
Li2SO4 + 2 NaOH gives 2 LiOH + Na2SO4
Lithium hydroxide is crystallised out, and the sodium sulphate has to be taken out of the liquor before it can be recycled. Albemarle’s compliance report for its Kemerton plant describes exactly this flowsheet: causticising with sodium sulphate decahydrate crystallisation, anhydrous sodium sulphate crystallisation with lithium removal, then evaporation and crystallisation of crude and pure lithium hydroxide (Kemerton compliance assessment report, 2022 to 2023).
Refineries that start from brine-derived lithium carbonate, or that make carbonate with soda ash, have different effluents, and the tonnages below do not apply to them directly. This page concentrates on the sulphate route, because that is where the approvals below record sodium sulphate tonnages.
How much sodium sulphate does a lithium hydroxide refinery make?
Between 1.69 and about 2.3 tonnes per tonne of hydroxide: the chemistry sets the floor, and the Western Australian approvals sit above it. One tonne of lithium hydroxide monohydrate (LiOH·H2O) carries 1.69 tonnes of sodium sulphate with it, using standard molar masses (41.96 for the monohydrate, 142.04 for sodium sulphate). Western Australia’s approvals allow for more:
| Refinery (Western Australia) | Lithium hydroxide allowed | Sodium sulphate allowed | Tonnes of sodium sulphate per tonne of hydroxide | Source |
|---|---|---|---|---|
| Tianqi Lithium, Kwinana (Train 1) | 24,000 t/yr (monohydrate) | 43,000 t/yr | 1.79 | DWER licence L2981/2025/1, issued 3 March 2026 |
| Albemarle, Kemerton | up to 100,000 t/yr | up to 200,000 t/yr | 2.00 | EPA WA, June 2018 |
| Covalent Lithium, Kwinana | up to 50,276 t/yr | up to 116,531 t/yr (anhydrous) | 2.32 | Ministerial Statement 1170, 15 July 2021 |
| Chemistry alone | 1 t (monohydrate) | 1.69 t | 1.69 | Stoichiometry |
Two cautions. These are approved maximums, not measured output, so they include design margin. And the gap above 1.69 is not explained in the approval documents. If you are designing or financing a refinery, ask the process licensor to account for its own figure line by line, because every extra tenth of a tonne is salt that needs a home. FEECO, an equipment supplier, quotes 2 to 2.5 tonnes per tonne of hydroxide in its own article, which overlaps the upper end of the licence range.
Glauber’s salt more than doubles the tonnage
Sodium sulphate comes out of cool solution as the decahydrate, Glauber’s salt (Na2SO4·10H2O). Its solubility collapses below about 32 °C: roughly 4.8 g per 100 mL of water at 0 °C against 49.7 g at 32.4 °C (reference data). That makes chilling the liquor a cheap way to pull sulphate out, which is why decahydrate crystallisation sits early in the Kemerton flowsheet.
The catch is the water of crystallisation. Every tonne of sodium sulphate leaves as 2.27 tonnes of Glauber’s salt, and 1.27 tonnes of that is water.
Chill or evaporate? Below about 32 °C the decahydrate is the solid that forms, so chilling makes Glauber’s salt; anhydrous sodium sulphate crystallises from hotter, evaporating liquor. Kemerton’s flowsheet uses both, chilling first and then anhydrous crystallisation with lithium removal. Which product you send off site depends on where the salt is going, as the example below shows.
Worked example (hypothetical). Assumptions: a refinery making 25,000 t/yr of lithium hydroxide monohydrate, sodium sulphate at 2.0 t per tonne (the middle of the licence range), 365 operating days.
Sodium sulphate made: 50,000 t/yr, or about 137 t/day.
Sent off site as Glauber’s salt: about 113,400 t/yr, or 311 t/day of wet crystal to store, load and dispose of.
Sent off site as anhydrous salt: 50,000 t/yr, but only after about 63,400 t/yr (174 t/day) of crystal water has been melted out and evaporated, usually with mechanical vapour recompression (MVR) or a dryer.
If the salt is going to a buyer, plan on the anhydrous step: Kemerton’s plan to sell its sodium sulphate rests on anhydrous crystallisation, drying and packaging. If it is going to disposal, compare the per-tonne cost of handling and landfilling the decahydrate with the energy cost of the anhydrous step. Our MVR evaporator guide explains what drives that energy cost.
How Western Australian refineries handle the effluent
The three approvals above describe three different end points for the liquid and the salt. Together they show that “zero liquid discharge” is a regulatory and site decision, not a single design.
| Refinery | What the approval says about wastewater and salts |
|---|---|
| Covalent, Kwinana | Wastewater of up to 252 ML/yr (about 690 m3/day) may go to the Sepia Depression Ocean Outfall Line at Cape Peron. Process-derived waste includes up to 9,479 t/yr of mixed salt material. No process-derived waste may go to landfill on the Swan Coastal Plain, and waste must leave site within 12 months (MS 1170). |
| Tianqi, Kwinana | Process water and wastewater must be reused in the refinery or taken off site to an authorised liquid waste facility, and must not go to stormwater. The emission points the licence lists are all air stacks (L2981/2025/1). |
| Albemarle, Kemerton | During commissioning, sodium sulphate went to the Sandy Ridge intractable waste facility. Albemarle states it intends to sell the material once the plant runs reliably and the product meets a consistent specification. Condensate replaces fresh reverse osmosis water for belt filter process water (2022 to 2023 compliance report). |
A coastal refinery with access to an approved outfall has options an inland plant does not. In Western Australia, refineries go through Environmental Protection Authority assessment and a Ministerial Statement, and the Department of Water and Environmental Regulation then licenses emissions, discharges and storage under Part V of the Environmental Protection Act 1986, as the Covalent statement notes. If you are planning a site elsewhere in Australia, our Australia page and brine management in Australia cover the wider picture.
When full ZLD makes sense, and when it does not
Full zero liquid discharge, meaning every liquid stream recycled or evaporated and every salt leaving as a solid, tends to be justified when:
there is no discharge route, because the site is inland, the licence rules it out, or outfall capacity is taken;
a buyer will take the sodium sulphate at a specification you can meet every month, not only on a good day;
the purge stream holds enough lithium to be worth recovering;
fresh water is scarce or expensive, so condensate has real value.
It is harder to justify when a permitted, well-monitored discharge exists and there is no salt market. Then ZLD turns a dilute liquid into a solid you still have to landfill, at the cost of the energy to dry it. Start from the end point, as our ZLD guide chapter on evaluating the disposal endpoint first explains, and then decide how much of the liquor to crystallise.
Design points that decide whether it works
Size the purge, not just the main salt. Potassium, chloride and other impurities build up in recycle loops and leave through a bleed that ends as mixed salt. Covalent’s approval allows up to 9,479 t/yr of it. In our view, mixed salt is usually the hardest residue to place; see our page on mixed salt disposal in ZLD.
Keep lithium out of the sodium sulphate. Lithium that leaves with the salt is lost product and a specification problem for the buyer. Kemerton’s flowsheet includes a lithium removal step in its anhydrous crystallisation for this reason.
Plan for commissioning salt. Off-specification salt during ramp-up needs an approved destination before start-up, as Kemerton’s commissioning period showed.
Agree the product specification before sizing the crystalliser. Purity, moisture and trace limits are set by the buyer, not the designer. Our chapter on purity, specification and qualification covers how to pin them down.
Treat condensate as your best water source. Evaporator and crystalliser condensate is low in salts and already on site.
Build the mass balance from measured liquor analyses. Our free brine mass balance calculator is a quick first check before a supplier sizes equipment, and our ZLD vendor proposal checklist helps compare the offers that follow.
Is there a market for the salt?
Not an easy one. A 2024 report by Chemical and Engineering News (C&EN, reprinted by Vinachem) said Albemarle planned to sell 200,000 t of sodium sulphate from an Australian refinery to Chinese detergent makers, that most battery materials are made in China, where sodium sulphate is sold as a raw material, and that outside China there is no comparable outlet. Liquid detergents, which are preferred in many newer battery markets, do not use it. The same report cited a 2022 RWTH Aachen and Roland Berger projection that battery material plants could be discharging almost 6 million t of sodium sulphate by the end of the decade, nearly eight times the 2022 level.
The alternatives in that report each carry their own risk:
Salt splitting by electrolysis turns sodium sulphate back into sulphuric acid and caustic soda, but the products come out dilute and concentrating them is costly.
Conversion to potassium sulphate fertiliser by reaction with potassium chloride. C&EN reported that Cinis Fertilizer had started a 100,000 t/yr plant in Sweden using sulphate from the Northvolt battery plant. That supply stopped, the plant never reached its target rate, and Cinis filed for bankruptcy in January 2026 (Norran, 22 January 2026).
Sulphate-free refining routes, such as Metso’s soda ash process, avoid the salt rather than treat it.
In our view, each of these is a project with its own offtake and scale-up risk, not a ready answer. Our guide to sodium sulphate recovery from wastewater goes deeper into recovery routes and product quality.
Classification, recycling and monitoring
Is refinery sodium sulphate hazardous? Sodium sulphate itself is not especially toxic; the C&EN report describes the problem as one of volume. How a particular batch is classified depends on what else it carries, which is why commissioning salt and mixed salt need their own disposal checks.
Can we recycle the sodium sulphate inside the refinery? Partly. Salt splitting can regenerate acid and caustic, but at low strength. Whether that fits depends on how much dilute reagent your flowsheet can absorb.
Does a brine-based plant face the same problem? Not in the same form. Brine plants produce mother liquors whose main salts depend on the brine, and their water questions sit mostly upstream. See our pages on lithium brine in Argentina and lithium brine in Chile.
What should a refinery’s discharge monitoring cover? The pollutants your ore and reagents can actually carry, not only salts. In April 2026 Inside Climate News reported that Tesla’s lithium refinery in Robstown, Texas, discharges an average of about 231,000 gallons a day (about 870 m3) into a drainage district ditch under a state permit. A sample commissioned by the drainage district, taken in the ditch rather than at the outfall, found traces of hexavalent chromium and arsenic (the arsenic below the federal drinking water limit) and levels of lithium, strontium and vanadium that the district’s attorney called abnormally high. The state regulator, TCEQ, said the plant complies with its permit and that a ditch sample is not appropriate for permitting decisions; its own February sampling covered dissolved solids, chlorides, sulphates and other basics but not heavy metals, and the permit does not monitor lithium. Tesla said it remains in full compliance and routinely tests its discharge (Inside Climate News, 21 April 2026). Whatever the outcome there, the lesson for any refinery is to build metals and lithium into the monitoring plan from the start, so the data exists before anyone asks. For water at the vehicle plants further down the chain, see our page on automotive plant water and wastewater treatment.
Settle the salt route before the crystallisers are ordered
Once the crystallisers are on order, the salt route is largely fixed. Brine Consulting, an independent industrial water advisory firm in Amsterdam, sells no crystallisers, evaporators or other treatment equipment, so our view on decahydrate versus anhydrous, outfall versus full ZLD, or sale versus disposal is not tied to a supply contract. We work with refinery teams and lenders remotely or through a defined engagement: checking the licensor’s sodium sulphate figure line by line, screening endpoints, and reviewing brine and resource recovery and water treatment and ZLD consulting scope. Book a call about your salt balance, share your project details or email christos@brineconsulting.com.
Written by the Brine Consulting advisory team. The Brine Consulting advisory practice includes Paolo Mattiuzza, Executive Advisor, Brine Management & ZLD Systems, who specialises in brine management, zero liquid discharge, minimum liquid discharge and industrial resource recovery, with more than 20 years of international leadership experience in the environmental and industrial B2B sectors, and Prof. Basel Abusharkh, a senior water technology expert, industrial innovator and ZLD specialist whose work includes membrane-based technologies for recovering lithium, potassium chloride, sodium chloride and other valuable resources from seawater and industrial brines. Meet our experts.