Sodium sulfate recovery from wastewater is technically routine and commercially hard. The chemistry cooperates: chill a concentrated sulfate liquor and Glauber’s salt (sodium sulfate decahydrate) crystallises out; boil it down above about 32 °C and the anhydrous salt forms instead. The difficulty is everything around that step. Recovered sodium sulfate (sodium sulphate in British usage) competes with mineral producers that ship product at 99% purity and above, and in most markets it sells for a few hundred US dollars a tonne.
This page sets out the recovery routes, how much a cooling step can actually pull out of a liquor, the purity buyers expect, and dated 2025 to 2026 prices, so you can judge whether a sulfate-rich stream is worth recovering or only worth managing.
Where recoverable sodium sulfate comes from
Sodium sulfate forms wherever sulfuric acid meets caustic soda, or sulfate salts meet sodium. The streams that most often raise the recovery question are:
Lithium hydroxide refineries that convert spodumene through a sulfate route. Our lithium refinery wastewater and ZLD guide covers that case in detail, including tonnages per tonne of product.
Battery materials, battery recycling and mining. VTT names the mining, battery and forest industries as producers of effluents high in sodium sulfate that cannot be discharged under their concentration limits.
Chemical plants under zero liquid discharge rules, such as the coal-to-ethylene glycol plants studied in China, where sulfate and chloride end up in the same concentrate.
Textile dyeing, which uses Glauber’s salt in the dyebath. One engineer asking for help on Eng-Tips described a textile effluent at about 3.5% sodium sulfate that had to meet zero liquid discharge.
Sulfate is rarely alone. Most streams also carry sodium chloride, and the ratio between the two decides which route can work. A liquor dominated by sulfate can make a saleable product. One with chloride and sulfate at similar levels usually ends as mixed salt for disposal unless the two are separated first.
The recovery routes and what each suits
| Route | How it works | Best suited to | Watch for |
|---|---|---|---|
| Cooling crystallisation (Glauber’s salt) | Below 32.4 °C, solubility falls more than tenfold toward 0 °C, so chilling a concentrated liquor crystallises the decahydrate | Sulfate-dominant liquors already close to saturation | Crystals are about 56% water by mass; chloride lowers the yield; selling anhydrous product means melting and drying the crystals |
| Evaporative crystallisation of anhydrous salt | Above 32.4 °C solubility is almost flat, so anhydrous sodium sulfate crystallises as water is boiled off in a multiple-effect or MVR crystalliser | Liquors where a dry, saleable product is the target | Energy per tonne; co-crystallisation of sodium chloride when chloride is high; scaling |
| Nanofiltration splitting, then crystallisation | Nanofiltration rejects divalent sulfate and passes most chloride, giving a sulfate-rich and a chloride-rich stream that are crystallised separately | Mixed sodium chloride and sulfate brines that would otherwise give mixed salt | Fouling and antiscalant carry-over; the split is never perfect, so each product still needs a purge |
| Freeze crystallisation | Cooling to about −2 °C crystallises the decahydrate; below that, ice forms too and leaves as clean water | Sulfate-rich saline process waters where heat is expensive | Decahydrate purity; sensitivity to chloride |
Four routes are in use or at demonstration scale: cooling crystallisation, evaporative crystallisation, nanofiltration splitting before crystallisation, and freeze crystallisation. The chloride content of the liquor decides which of them can work. The solubility figures behind the first two rows come from standard reference data: about 4.8 g per 100 mL of water at 0 °C, 28.1 g at 25 °C and a maximum of 49.7 g at 32.4 °C.
Published results show both what the routes can do and where they fall short:
Nanofiltration split. In a three-month pilot on coal-to-ethylene glycol wastewater, sulfate rejection stayed above 98% and chloride rejection below 10%. After evaporation and crystallisation the plant made sodium sulfate at 98.4% purity and sodium chloride at 98.9% (Fu et al., Journal of Water Process Engineering, 2025).
Freeze crystallisation. Four years of operation at a pipe freeze-crystallisation demonstration plant in South Africa raised decahydrate output from 3.5 to 9.1 t a month through pipe upgrades. Decahydrate purity measured by X-ray diffraction rose from 50% in 2021 to 84.9% in 2024, and the authors report that high sodium chloride in the feed delayed sodium sulfate crystallisation. They report about 330 kJ per kg for the freeze process against about 2,200 kJ per kg for distillation (More, Maree and Mahlangu, Water, 2025). These are the developers’ own plant results.
A fourth option avoids selling the salt at all. Bipolar membrane electrodialysis splits sodium sulfate back into sulfuric acid and caustic soda for reuse on site. VTT’s testing found electrodialysis one of the most feasible options for highly concentrated solutions, and in 2023 Finnish Minerals Group and Adven planned a plant-scale project backed by EUR 15 million from Business Finland (VTT). It suits sites that consume both reagents and can use them at the strength the cell delivers.
How much does cooling alone recover?
The answer depends far more on the liquor you start with than on the chiller. Glauber’s salt is 44.1% sodium sulfate and 55.9% water by mass, so every kilogram that crystallises also takes about 0.56 kg of water out of the liquor. That self-concentration helps, but only if the liquor is already strong.
Worked example (hypothetical). Assumptions: a pure sodium sulfate and water system, 100 kg of water in the liquor, cooled to 0 °C, equilibrium reached, solubility at 0 °C of 4.76 g per 100 g of water, crystals separated completely.
| Sodium sulfate in the liquor before cooling (g per 100 g water) | Glauber’s salt formed (kg) | Sodium sulfate recovered (kg) | Share recovered | Water leaving with the crystals (kg) |
|---|---|---|---|---|
| 28.1 (saturated at 25 °C) | 56.3 | 24.8 | 88% | 31.5 |
| 20 | 36.8 | 16.2 | 81% | 20.6 |
| 10 | 12.6 | 5.6 | 56% | 7.1 |
| 5 | 0.6 | 0.3 | 5% | 0.3 |
Method: the mass of decahydrate x solves (S minus 0.441x) divided by (100 minus 0.559x) equals 0.0476, where S is the starting sodium sulfate in kg.
Three things follow:
Concentrate first. A liquor at 5 g per 100 g of water, roughly what a textile effluent at a few percent sodium sulfate looks like, gives almost nothing on cooling. It needs reverse osmosis, electrodialysis or evaporation before a chiller makes sense.
Chloride makes it worse. These figures are a best case. The freeze plant authors found that high sodium chloride depresses the freezing point and delays decahydrate crystallisation, so a mixed liquor yields less than the table shows. Model your real liquor with a phase-equilibrium tool and confirm it on the bench before sizing anything (laboratory and pilot testing).
The water comes back. Removing 31.5 kg of water with the crystals takes load off the evaporator. But if the buyer wants anhydrous product, that water has to be driven off again when the crystals are melted and dried.
What buyers expect, and how recovered salt compares
Recovered salt is judged against natural and mineral producers. Two current producer specifications show the bar (both are supplier-published figures):
| Producer document | Sodium sulfate | Sodium chloride | Other values |
|---|---|---|---|
| Searles Valley Minerals, Trona, California, anhydrous technical grade (Bulletin 4100, June 2025) | 99.0% minimum (typical 99.0 to 99.7%) | 0.50% maximum | Sodium carbonate 0.50% maximum; iron typically 0 to 10 ppm; water insolubles typically 0 to 20 ppm |
| Saskatchewan Mining and Minerals, Chaplin, detergent grade (2025 averages, effective 1 January 2026) | 99.48% (99.0% minimum) | 0.20% | Iron 2.44 ppm; moisture 0.018%; water insolubles 0.01%; reflectance 90.6; meets the Food Chemicals Codex |
In China, industrial anhydrous sodium sulfate is graded under GB/T 6009-2014, implemented on 1 December 2014 and still listed as the current edition in November 2026, which divides the product into three classes. Ask a Chinese buyer which class and edition its contract cites.
Set the published recovery results against those numbers. The nanofiltration pilot’s 98.4% sodium sulfate is close, but still under the 99.0% minimum both producers sell against. The freeze plant’s 84.9% is a phase purity of the decahydrate, a different measure, and shows how much work remains between crystals and a detergent-grade product. In practice, salt recovered from wastewater usually needs washing, or re-dissolving and re-crystallising, before it meets a commodity specification. Purity is also not only about the sulfate assay. Buyers check:
chloride, the usual failure point for salt from mixed brines;
colour and whiteness, which organics and iron spoil;
metals, including anything the source process carries, such as nickel, cobalt or lithium from battery streams;
consistency, meaning the same result every month rather than on a good day;
waste status: in the EU, a salt made from a waste stream must meet end-of-waste and REACH requirements before it can be sold as a product, as our salt recovery overview explains.
Our ZLD Guide chapter on purity, specification and qualification sets out how to agree a specification and qualify a product with a buyer.
What sodium sulfate sells for (2025 to 2026)
Prices differ by region by a factor of two or more, and most price reporters do not state the grade. Treat these as indications, not offers:
| Market | Price (US$ per tonne) | Period | Source |
|---|---|---|---|
| United States | 216 | Q2 2025 | IMARC price index, published 15 October 2025 |
| Germany | 360 | Q2 2025 | IMARC, as above |
| China | 460 | Q2 2025 | IMARC, as above |
| Brazil | 240 | Q2 2025 | IMARC, as above |
| India, ex-Bharuch, natural anhydrous detergent quality | about 150 to 165 | early 2026 | Tradeasia market note, 9 April 2026 (a chemical trader) |
To put that in scale: 10,000 t a year of on-spec salt at the US figure of US$216 per tonne is worth about US$2.2 million before drying, packing and freight. Freight matters at these prices, because a bulk salt worth a couple of hundred dollars a tonne cannot travel far before transport takes most of its value. In many projects the stronger part of the business case is the disposal cost avoided by not producing mixed salt, rather than the sales revenue. Our ZLD Guide chapter on markets, offtake and logistics covers how to test a buyer’s interest before committing to a flowsheet.
When is sodium sulfate recovery worth pursuing?
Sodium sulfate recovery tends to pay off when most of these hold:
sulfate is the dominant anion, or nanofiltration can make it dominant in one stream;
chloride, organics and metals can be held below a named buyer’s limits month after month;
a buyer within sensible freight distance has agreed a written specification and will take a trial lot;
the alternative, mixed salt to landfill, is expensive or restricted where you operate;
you can test the flowsheet on real liquor, not a synthetic recipe;
or you can use the salt internally, for example by splitting it back into acid and caustic.
If most of these fail, the realistic goal is usually a smaller, cleaner residue stream rather than a product.
Practical questions on Glauber’s salt and route choice
Can Glauber’s salt be sold as it comes out of the crystalliser? The producer specifications above are for anhydrous product. Glauber’s salt is about 56% water, so shipping it means paying freight on water, and it melts in its own water of crystallisation at about 32 °C. Most flowsheets that aim to sell the salt melt and dry it to anhydrous first.
Is cooling or evaporation cheaper? Cooling removes sulfate without boiling water, but it needs a concentrated feed and chilling capacity. Evaporation gives anhydrous product directly but costs heat or electricity for every tonne of water removed. Many plants use both: a cooling step to take most of the sulfate as decahydrate, then evaporative crystallisation to finish the product.
Where do we start? With a reliable mass balance of sulfate, chloride and minor ions across the site. Our free Brine Mass Balance Calculator is a quick first pass, and the Water Analysis Validator checks that the lab data behind it is internally consistent.
Talk to us before the crystallisers are specified
Whether a sulfate stream becomes a product, a reagent or a residue is decided early, when the flowsheet is still on paper. That is when an outside check is cheapest. Brine Consulting, an independent industrial water advisory firm in Amsterdam, sells no treatment equipment, so we have no stake in whether you crystallise, split or send the salt to disposal. Our brine and resource recovery work on sulfate streams covers the site mass balance, a screen of cooling, evaporation, nanofiltration and electrodialysis against your chloride level, and the specification a named buyer would accept. We work with sites remotely or through a defined engagement. Related reading: mineral recovery from desalination brine and the ZLD Guide on product-specific pathways.
To talk through a sulfate stream, book a scoping call, send us your project details or email christos@brineconsulting.com with a recent analysis.
Written by the Brine Consulting advisory team. The Brine Consulting advisory practice includes Paolo Mattiuzza, a senior brine management and ZLD specialist with more than 20 years of international experience in environmental technologies, industrial water treatment, and resource recovery, 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.