Bayer Liquor Evaporation: Sodium Removal and Energy

Last updated: 2026-11-05

Bayer liquor evaporation is the duty that closes an alumina refinery’s water balance: for a 1 Mt/y refinery, a typical net duty is 100–150 t/h (about 440–660 US gpm) of water removed from spent liquor while caustic concentration is held near 200 g/L Na2O. Miss the balance and digestion capacity, oxalate control, and energy cost all degrade together.

Key Takeaways

• Evaporation exists to close the water balance: wash water and rainfall enter, only vapor and bleed leave.
• Spent liquor caustic of 180–260 g/L Na2O (typical range) preserves digestion capacity at the digesters.
• Sodalite and cancrinite scale grow fastest where temperature and caustic are both high — design for cleaning.
• Falling film suits cleaner liquor; forced circulation is preferred for oxalate-rich, high-scaling Bayer service.
• The worked example: a 1 Mt/y refinery carries a 140 t/h duty at roughly 1.0 GJ/t alumina.

Why Does Bayer Liquor Evaporation Anchor the Refinery Water Balance?

Every alumina refinery runs a closed caustic loop, and every liter of water that enters the loop — as residue wash water, hydrate wash water, gland water, or rain falling into open tanks and lakes — must leave it somewhere. Bayer liquor evaporation is a water-removal duty that holds the circulating caustic inventory at the concentration the digesters demand, and it is the only controllable water outlet apart from deliberate bleed. The chemistry gives the reason: digestion of gibbsite and boehmite needs sodium aluminate liquor of roughly 180–260 g/L Na2Ocaustic (typical industry range, site-specific), but washing residue and product hydrate dilutes the loop continuously. If wash water is not evaporated out, caustic strength falls, digestion conversion drops, and throughput follows it down.

Three named duties sit inside the umbrella term. Spent liquor is a cooled, depleted sodium aluminate stream that returns from precipitation carrying dissolved carbonate and oxalate, and evaporating it before it re-enters digestion is the standard duty. Pregnant liquor evaporation is a smaller, optional duty that concentrates green liquor ahead of precipitation to lift precipitation yield at cool-climate sites — the same equipment family, a different position in the flowsheet. Sodium salt control is the third duty: as water is removed, sparingly soluble sodium oxalate and sodium carbonate approach their solubility limits, and the train is deliberately operated so those salts crystallize where they can be removed, not where they do damage. Oxalate that stays dissolved poisons precipitation; carbonate accumulates and dilutes effective caustic; both are managed through the evaporator train and a causticization or salt-removal side loop.

The duty is bounded by the water balance. Countercurrent red mud washing returns most wash water to the circuit, but a net make-up stream always remains — and in high-rainfall locations it grows quickly. Refineries in tropical climates can carry evaporation duties two to three times those of equal-capacity arid-site plants, purely from rain and dilution. That is why the water-balance arithmetic comes first in any Bayer liquor evaporation project, and why the equipment belongs to the family described on our evaporator and crystallizer systems pillar page: the same principles that serve brine concentration serve hot caustic, with different materials and very different scaling chemistry. The full picture of where this duty sits in the metal’s flowsheet is covered on our aluminium industry applications page.

Which Parameters Govern Bayer Liquor Evaporation?

A short parameter list governs the duty, summarized below. Every value is a typical industry range — test work on your own liquor sets the final numbers, because gibbsite versus boehmite feeds, silica levels, and carbonate inventories shift the whole table.

All liquor analyses above are stated wet basis (grams per liter of liquor as pumped), with caustic values expressed as Na2O equivalent — the reporting convention used across Bayer operations. Where a plant reports on dry-solids or different-equivalent bases, numbers will not compare directly; always confirm the basis before benchmarking.

Two parameters deserve commentary. First, boiling point rise: because liquor is a concentrated caustic solution, it boils above the pure-water temperature at the same pressure, and the gap widens as caustic strengthens. With 8–12 degrees C of BPR in the concentrated effects, the useful temperature difference per effect shrinks, which caps practical effect count and shapes the choice between multi-effect and vapor recompression. Second, the caustic concentration limit itself. Pushing spent liquor well beyond ~260 g/L Na2Oc (typical range) raises viscosity, accelerates sodalite scaling dramatically, and pushes oxalate and carbonate past saturation earlier in the train — so the optimum is rarely the maximum. The energy consequences of these parameters are treated in depth in our evaporator energy consumption guide.

Falling Film or Forced Circulation for Bayer Liquor Evaporation?

Equipment selection resolves duty by duty, not by brochure. The core physics: falling-film evaporators carry liquor as a thin film down heated tubes, giving high heat-transfer coefficients, low liquor holdup, and gentle operation at small temperature differences — ideal for multi-effect trains. Forced-circulation evaporators pump liquor through tubes at high velocity and flash it in a separated chamber, giving lower coefficients but near-immunity to deposition, because supersaturated salts stay suspended rather than nucleating on the wall.

For relatively clean spent liquor with oxalate held below roughly 2 g/L, falling-film evaporation is preferred because its high heat-transfer coefficient minimizes heating area — and, as shown below, nickel inventory. For liquor near oxalate or carbonate supersaturation, forced-circulation evaporation is preferred because tube velocity, not surface area, controls deposition; the liquor simply does not dwell on a hot wall long enough to scale. Where the duty includes deliberate salt crystallization — a salt-cake removal effect, for example — forced circulation with a properly designed crystallizer body is the only sensible choice. The trade-offs are tabulated in our forced circulation vs falling film comparison, and the area arithmetic behind the first verdict is in our evaporator heat transfer area guide.

Multi-effect versus mechanical vapor recompression (MVR) is the second decision. A four-effect train delivers a steam economy of roughly 3.0–3.5 kg of water per kg of live steam (typical for caustic liquor, where BPR eats temperature difference faster than in salt brines). MVR replaces steam with compressor work, but the compressor must lift vapor through the full BPR plus the heat-transfer temperature difference — at 8–15 degrees C of BPR, that is a demanding lift, and the working principle and its limits are explained in our MVR working principle guide and the MVR vs multiple effect comparison. For cheap-steam sites with constrained power, multi-effect wins; for power-rich sites with expensive boiler fuel, MVR can win even at high BPR — provided the compressor is specified for the true lift, not the clean-water lift. Our brine concentrator sizing and selection guide covers the same sizing logic for high-BPR duties generally.

Materials of construction decide the project’s capital curve. Hot concentrated caustic attacks carbon steel (caustic embrittlement and general corrosion at elevated temperature), and austenitic stainless steels are not a general answer either; commercially pure nickel and nickel-clad steel are the standard materials for first-effect service above roughly 120 degrees C, with carbon steel acceptable in cooler, less concentrated sections within its temperature-concentration envelope (site-specific, to be confirmed by corrosion test work). Because nickel drives cost, the falling-film verdict above has a financial edge in clean service: less area means less clad surface. A complete alumina refinery equipment list for the train typically includes feed preheaters (plate or shell-and-tube, with cleanable surfaces), the effects themselves, flash coolers, surface or barometric condensers, the vacuum system, liquor circulation pumps, an oxalate/salt removal loop (settlers or filters on a controlled slipstream), and condensate monitoring — because a caustic leak into condensate shows up in a conductivity meter long before it shows up anywhere else. Both a triple-effect evaporation crystallization unit and an MVR evaporation crystallization unit can be configured for this duty with appropriate materials and salt-removal provisions.

Why Does Bayer Liquor Scale Evaporators So Fast?

Scale is the defining operational problem of the duty, and each scale type has a distinct symptom, cause, and fix. Seven items cover most of what a plant will meet:

1. Sodalite scale on hot heating surfaces. Symptom: overall heat-transfer coefficient on first-effect plates or tubes sags noticeably within two to four weeks of a cleaning. Cause: reactive silica plus dissolved aluminate plus strong caustic react on hot surfaces above roughly 100 degrees C to form sodalite, a dense sodium aluminosilicate that bonds tightly to the metal. Fix: control silica entering the evaporator train (feed clarification and good residue washing), keep surface temperature as low as the duty allows, and run disciplined cleaning cycles between effects.
2. Cancrinite scale in the hottest effects. Symptom: hard, layered scale that resists ordinary chemical cleaning, appearing in the highest-temperature effects first. Cause: sodalite gradually transforms to cancrinite — a harder sodium aluminosilicate — with time at temperature, especially with carbonate incorporated in the lattice. Fix: remove sodalite before it transforms; shorten cleaning intervals on hot effects rather than letting scale age in place.
3. Silica and quartz scale in preheaters. Symptom: glassy, hard deposits in feed preheaters ahead of the first effect. Cause: reactive silica supersaturation as liquor temperature rises. Fix: eliminate reactive silica upstream — silica reports to the evaporators when residue washing and clarification underperform, which is another reason the washing circuit and the evaporator train must be designed as one system.
4. Sodium oxalate fouling in cooler effects and flash chambers. Symptom: mats of fine needle crystals, falling liquor flow, and plugged distributors. Cause: as the train concentrates liquor, oxalate crosses its solubility limit and nucleates preferentially on surfaces and in low-velocity zones. Fix: hold liquor oxalate below the supersaturation window with a controlled bleed or destruction loop, seed-and-settle oxalate in a dedicated removal stage, and design distributors for rodding access.
5. Sodium carbonate scale late in the train. Symptom: carbonate monohydrate crystals in the concentrating effects and salt hints in condensate inspection reports. Cause: combined concentration and cooling push carbonate past saturation. Fix: manage the carbonate inventory through bleed and causticization, and place a salt-removal effect or crystallizer where supersaturation is predictable.
6. Corrosion-product fouling. Symptom: red-brown or black deposits and rising in-service corrosion rates downstream. Cause: carbon steel corroding in hot caustic upstream, sending oxide particles into the train. Fix: respect the temperature-concentration envelope for carbon steel, upgrade marginal sections to nickel-clad construction, and filter or settle incoming liquor.
7. Run-length decay as a system symptom. Symptom: each cleaning buys less run time than the last. Cause: cleaning intervals mismatched to actual scale kinetics, or one uncontrolled driver (silica spike, oxalate drift) feeding all the others. Fix: trend the heat-transfer coefficient per effect, schedule cleaning from the trend rather than the calendar, and treat a shortening run length as a circuit-wide alarm, not an evaporator problem. Practical techniques for each of these appear in our evaporator scaling control resource.

Worked Example: Sizing the Evaporation Duty for a 1 Mt/y Refinery

The water balance, not the evaporator catalog, decides the duty. Six steps take a project from plant size to evaporation train specification — the arithmetic below is a synthetic worked example with every assumption stated.

1. Fix the production and time basis. A 1 Mt/y alumina refinery (anonymized, synthetic basis) at 8,000 operating hours per year gives 125 t/h of alumina. All downstream numbers scale with this basis.
2. Tally water inlets to the caustic circuit. Residue wash water make-up (net of filtrate return) 140 t/h; precipitation and cooling circuit make-up 45 t/h; rainfall and uncontrolled dilution at a tropical site 30 t/h. Total water in: 215 t/h.
3. Tally non-evaporative water outlets. Water leaving with washed residue 55 t/h; water leaving with wet product hydrate 12 t/h; miscellaneous losses (gland water, trap losses) 8 t/h. Total: 75 t/h.
4. Subtract to get the net evaporation duty. 215 − 75 = 140 t/h of water to be evaporated — 1.12 t of water per tonne of alumina, on the high side of the typical 0.4–1.2 t/t industry span because of the rainfall allowance. An arid-site plant with the same flows might carry 60–80 t/h.
5. Distribute the duty and compute the steam draw. A four-effect train with a steam economy of 3.2 kg water per kg steam (typical for caustic liquor, BPR-limited) draws 140 ÷ 3.2 ≈ 44 t/h of live steam.
6. Convert to specific energy and check the salt loads. At 2.9 GJ per tonne of live steam (typical saturated-steam enthalpy basis), the train consumes about 127 GJ/h, or roughly 1.0 GJ per tonne of alumina — against a total refinery fuel-and-steam demand typically quoted at 9–11 GJ/t alumina. Evaporation is thus a mid-single-digit percentage of refinery energy, but a controllable one, which is why energy benchmarking of the train is worth instrumenting separately.

The same concentration that removes water concentrates sodium salts, so the salt loop is part of the duty, not an accessory. At 140 t/h of water removal, a train concentrating spent liquor from 200 to 240 g/L Na2Oc moves roughly 2,100 m3/h of circulating liquor (indicative); the oxalate and carbonate removal stages on a controlled slipstream must be sized to the salt inventory growth, with causticization converting carbonate bleed back to useful caustic. Understating this loop is the most common late-stage design correction we see.

What Does Field Data From Bayer Liquor Evaporation Trains Show?

Three anonymized operating snapshots (indicative values, synthetic-typical basis, to be confirmed by test work on site liquor) illustrate how the parameters above play out:

• Falling-film first effect, 1.1 Mt/y tropical refinery. Overall heat-transfer coefficient fell from about 1,300 W/m2·K clean to about 900 W/m2·K after twelve days on spent liquor with reactive silica near 0.3 g/L. Cleaning every 14 days restored design performance; stretching to 21 days cost measurable steam. Basis: vendor-supplied train, operator-reported coefficient trend, wet-basis liquor analyses.
• Forced-circulation effects on oxalate-rich liquor. A train handling liquor at 4–5 g/L oxalate ran 45–60 days between cleanings with tube velocity maintained near 2.5 m/s, versus under 20 days on a sister falling-film train before conversion. Basis: anonymized operator data, two-train comparison at one site, same liquor source.
• MVR trial on spent liquor at high BPR. With boiling point rise near 9 degrees C, a pilot MVR evaporator held specific power at 24–28 kWh per tonne of water evaporated with a compressor lift near 25 degrees C — technically workable, but the economics hinged entirely on local power price versus steam cost. Basis: anonymized pilot campaign on site liquor, instrumented trial.

One more field observation deserves a place in the design file: oxalate removal staged through a dedicated settling effect on a 200 m3/h slipstream recovered about 0.6 t/h of sodium oxalate cake at a 1.0 Mt/y plant (indicative) — salt that would otherwise have circulated until it fouled the coolest effects. If you are assembling an equipment list against a liquor analysis, send us the numbers and request a quote for train sizing and materials selection.

FAQ: Bayer Liquor Evaporation, Answered

What caustic concentration should spent liquor be evaporated to?

Most refineries hold spent liquor between 180 and 260 g/L Na2Ocaustic (typical range), with the set point chosen to match digestion liquor requirements minus the dilution the circuit will see. Higher concentration improves digestion capacity but accelerates sodalite scaling and brings oxalate and carbonate to saturation sooner, so the best value is usually somewhat below the maximum the digesters could tolerate, fixed by plant-wide optimization.

Why is boiling point rise so high in Bayer liquor?

Bayer liquor is a concentrated sodium aluminate-caustic solution: dissolved sodium salts lower the solvent’s effective vapor pressure, so the liquor boils several degrees above pure water at the same pressure. At 200–250 g/L Na2Oc, boiling point rise of 5–15 degrees C is typical, growing with concentration. This rise is lost from every effect’s driving temperature difference, which caps effect count and makes compressor design harder in vapor-recompression options.

Falling film or forced circulation for spent liquor evaporation?

For clean liquor with oxalate well below saturation, falling film is usually preferred for its heat-transfer efficiency and lower nickel inventory. For liquor near oxalate or carbonate supersaturation, forced circulation is preferred because high tube velocity keeps crystals suspended instead of letting them nucleate on hot walls. Many trains use falling film in early effects and forced circulation where supersaturation becomes predictable — a duty-by-duty split, not a brand loyalty.

Can MVR be used for Bayer liquor evaporation?

Yes, with caveats. The compressor must lift vapor through the full boiling point rise plus the heat-transfer temperature difference — often 20–30 degrees C total — which raises compressor cost and power draw. Field indications of 24–28 kWh per tonne of water evaporated have been reported on spent liquor (anonymized pilot data). MVR suits power-rich, fuel-expensive sites; multi-effect remains the default where cheap steam is available.

How is sodium oxalate removed from Bayer liquor?

Three routes coexist. Evaporative: concentrating liquor past oxalate solubility in a dedicated effect, then settling or filtering the crystals from a slipstream. Chemical: oxalate destruction or precipitation reagents dosed into a side loop, chosen by project chemistry. Dilution: simply carrying more oxalate in a larger liquor inventory, which costs capital elsewhere. Most plants combine evaporative removal with a chemical trim loop, sized from the oxalate inventory growth rate.

What materials survive hot caustic evaporator service?

Commercially pure nickel and nickel-clad steel are the standard for first-effect surfaces above roughly 120 degrees C, where hot concentrated caustic attacks carbon steel and most stainless grades. Carbon steel remains acceptable in cooler, less concentrated sections within its temperature-concentration envelope, which must be confirmed for the specific liquor. Pump internals, distributors, and instruments need the same discipline — a carbon steel component in the wrong effect becomes a corrosion-product generator.

How often must Bayer evaporators be cleaned?

Clean-on-trend beats clean-on-calendar. Typical first-effect cleaning cycles run 10–21 days on falling-film service (indicative), forced-circulation effects often run 30–60 days, and the right interval is found by trending each effect’s heat-transfer coefficient and cleaning when performance loss exceeds the steam cost of continuing. A shortening run length usually signals an upstream driver — silica, oxalate, or carbonate drift — rather than a cleaning problem alone.

Conclusion

Bayer liquor evaporation is where a refinery’s water discipline becomes a process variable you can actually control. The water balance fixes the duty — 140 t/h in our worked 1 Mt/y example, against an industry span of roughly 0.4–1.2 t of water per tonne of alumina — while the caustic concentration set point, the oxalate and carbonate inventories, and the scale chemistry fix the equipment: falling film where liquor stays clean, forced circulation where supersaturation threatens, nickel where temperature and caustic combine, and a salt-removal loop designed as part of the train rather than bolted on afterward. None of these verdicts is exotic; all of them depend on your liquor analysis, your rainfall, and your steam-to-power price ratio, which is why test work and a stated-basis water balance come before any equipment order. If you are scoping a new train or troubleshooting an existing one, send us your liquor analysis and water balance basis, and request a quote for train sizing, materials selection, and a cleaning strategy matched to your scale chemistry — the difference between a 1.0 and a 1.4 GJ/t alumina evaporation account is decided at design time, not discovered on the steam meter.

References

• U.S. Geological Survey, Mineral Commodity Summaries — Bauxite and Alumina chapter (annual edition). usgs.gov
• SME Mineral Processing and Extractive Metallurgy Handbook, Society for Mining, Metallurgy and Exploration, Englewood, CO — alumina and Bayer process sections.
• CSIRO Mineral Resources — publications on Bayer process liquor chemistry, sodalite and cancrinite scale formation, and residue washing. csiro.au
• TMS Light Metals, Alumina and Bauxite sessions (multiple years) — peer-reviewed papers on spent liquor evaporation, evaporator scale management, and salt cake removal.
• Wills, B.A. and Finch, J.A., Wills’ Mineral Processing Technology, 8th ed., Butterworth-Heinemann — Bayer flowsheet context.
• Kirk-Othmer Encyclopedia of Chemical Technology — “Alumina” chapter (alumina production and liquor circuits).
• Journal literature on sodium oxalate removal and causticization in Bayer circuits, described generically per Light Metals and hydrometallurgy journal proceedings.

Changelog

• 2026-11-05: Initial publication.