In the Bayer process, sodium hydroxide is circulated as spent liquor containing 200–260 g/L Na₂O to digest bauxite in high-pressure autoclaves or tube digesters at 145–265 °C and 3.5–7.0 MPa. Gibbsite-bearing bauxite requires residence times of 20–60 min at the lower temperature boundary, whereas boehmitic and diasporic ores require 60–120 min near the upper temperature boundary at the same caustic concentration. The feed slurry is controlled to a molar Na₂O/Al₂O₃ ratio of 1.5–2.5 in digestion, preventing excessive aluminium solubility while keeping iron oxide and titanium oxide inert. Fresh 50 wt% membrane-grade sodium hydroxide is added to the returning spent liquor to offset losses from red mud entrainment and sodium aluminium silicate precipitation; published Bayer operating data show make-up requirements from 40 kg to 150 kg NaOH per tonne alumina, with the higher values associated with diasporic bauxite and reactive silica contents above 4–6 wt%.

The principal process conflict is desilication. Reactive silica dissolving from clay minerals forms sodalite-type sodium aluminium silicates that consume free caustic and deposit on heat exchanger surfaces, particularly in shell-and-tube feed heaters and flash tank vapor lines. Desilication tanks are therefore installed upstream of digestion when reactive silica in bauxite exceeds 6–8 g/L in predicted pregnant liquor. These atmospheric tanks operate at 90–100 °C with 4–8 h residence time, allowing silica to precipitate as sodalite for removal with red mud. Scale control in tube digesters relies on maintaining a residual free caustic gradient from 220 g/L Na₂O at the inlet to 140–160 g/L Na₂O at the flash discharge, with periodic acid washing of heaters at intervals determined by heat transfer coefficient decay of more than 10%. Red mud washing in counter-current thickeners and horizontal disc filters recovers 95–98% of the entrained caustic; wash water addition is limited to a dilution factor below 1.5 m³/t dry residue to avoid excessive evaporation load.

What Limits Caustic Load in Kraft Pulp Digesters?

White liquor entering continuous digesters contains 85–110 g/L effective alkali expressed as Na₂O, distributed between sodium hydroxide and sodium sulfide at a sulfidity of 25–35%. Active alkali charge for bleachable softwood grades is 16–22% on oven-dry wood, with a liquor-to-wood ratio of 3.5:1–4.5:1. Chip impregnation proceeds at 110–130 °C before the cooking zone is raised to 150–170 °C; the H-factor is held between 1,800 and 2,200 to produce a kappa number of 20–25 for softwood market pulp measured by ISO 302:2015 or TAPPI T 236. Final reject content above 1.5–2.0% of screen feed indicates insufficient alkali penetration or uneven chip moisture, triggering adjustments to effective alkali and cooking temperature rather than extended residence time.

The caustic balance in the recovery cycle is closed through recausticizing. Green liquor containing sodium carbonate reacts with slaked lime in continuous causticizing trains at 85–90 °C, converting 75–85% of sodium carbonate to sodium hydroxide. Lime kiln operation at 1,100–1,200 °C regenerates calcium oxide from lime mud; the residual calcium carbonate content in burned lime is maintained below 5–8 wt% to avoid slaking bottlenecks. Sodium hydroxide make-up in a closed kraft mill is therefore limited to losses in dregs, grits, lime mud, and bleach plant spills, commonly below 15 kg NaOH per tonne unbleached pulp in modern softwood lines. Excessive effective alkali accelerates carbohydrate peeling and reduces screened yield, while low effective alkali raises kappa number variability and black liquor viscosity. Operators monitor residual effective alkali in black liquor at 5–12 g/L Na₂O after brown stock washing; values below 5 g/L shift delignification selectivity toward alkali-insoluble lignin condensation, which is irreversible during oxygen delignification.

Saponification Reactor Geometry and Caustic Stoichiometry

Continuous saponification of palm stearin, tallow, or coconut oil with 50 wt% sodium hydroxide is sized from the saponification value of the blended fat charge, determined by AOCS Cd 3-25 or ISO 3657:2020. Stoichiometric caustic demand is calculated as 0.713 kg NaOH per kg fat for every unit of saponification value expressed as mg KOH/g. High-shear loop reactors or scraped-surface reactors operate at 80–100 °C under atmospheric pressure, with a residence time of 10–30 min and a recycle ratio sufficient to generate local shear rates above 500 s⁻¹. Free alkali in neat soap is maintained at 0.05–0.15 wt% as NaOH; below this window unreacted triglycerides remain, while above it the soap phase viscosity increases sharply and glycerol separation deteriorates in the subsequent centrifugal extraction stage.

Sodium chloride at 0.5–2.0% of neat soap is added to control viscosity and drive glycerol into the spent lye phase. Spent lye containing 8–12% glycerol and 5–10% sodium chloride is sent to a recovery evaporator; overheating above 160 °C accelerates glycerol polymerisation and discolouration. The caustic charge is trimmed on-line by near-infrared free alkali analysers; a sustained deviation of ±0.03 wt% from the setpoint signals a metering pump fault or fat blend saponification value shift. Batch kettle saponification, by contrast, uses an initial caustic charge of 90–95% of theoretical, with the balance added after the soap reaches a thick paste stage to avoid foaming and boil-over in open kettles.

When cotton fabric is treated with 18–25 wt% sodium hydroxide at 15–20 °C under controlled warp tension, the cellulose I lattice converts to cellulose II. In a chain-type mercerizer, the fabric is wetted by a padder to a pickup of 80–100%, held under 2–5% stretch on a clip stenter, and then passed through counter-current washing stages that recover 90–95% of the caustic as weak lye for evaporation to 28–32 wt%. Residual alkali after neutralisation must fall within pH 6.5–7.5 when measured by AATCC Test Method 81-2016; incomplete neutralisation reduces reactive dye fixation and produces haloing. Swelling increases fibre cross-section by 15–25%, and the resulting changes in tensile strength are assessed by ASTM D5035-11(2019) for strip breaking force. The mercerising bath concentration must be maintained within ±0.5 wt%; a drift below 16 wt% produces no permanent mercerization, while a bath temperature above 25 °C reduces swelling uniformity in densely woven constructions.

Before mercerization, cotton goods are scoured with 2–4% sodium hydroxide on weight of fabric in batch kiers at 100 °C for 30–60 min, removing waxes, pectins, and seed coat fragments. Continuous scouring ranges use 4–6% caustic solution applied by saturator and steamed at 100–102 °C for 15–30 min. The same alkali recovery system is used, though the weak caustic stream contains higher suspended solids and requires screening through 100–150 μm rotary screens before evaporation. If residual wax content is not reduced below 0.3–0.5 wt% on fabric, measured by solvent extraction per AATCC Test Method 97-2019, downstream mercerization wetting is uneven and caustic penetration in high-thread-count constructions becomes the dominant defect.

When Caustic Soda Becomes a Precipitation Reagent in Mixed-Metal Wastewater

Acid-bearing wastewater is neutralised with 25 wt% or 50 wt% sodium hydroxide metered through a pH-controlled injection loop. Iron(III) precipitates as hydrous oxide between pH 3.5 and 4.5; copper, zinc, and nickel require staged elevations to pH 8.0–9.0, 8.5–9.5, and 9.5–10.5, respectively. Reaction tanks are sized for 20–30 min retention at peak flow, followed by lamella clarifiers designed for 1.5–2.5 m/h hydraulic loading, and sludge dewatering through plate-and-frame filter presses at 0.6–0.8 MPa feed pressure. Inline pH analysis follows ASTM D1293-18 with automatic temperature compensation; electrode fouling in mixed-metal streams is managed by sequential acid and alkaline cleaning cycles every 4–8 h.

The target effluent pH after clarification is 8.5–9.0 before discharge or biological treatment, as higher pH causes zinc hydroxide resolubilisation above pH 10.5 and increases scaling on downstream pipe walls. Precipitation performance is verified by filtered metal concentrations using ISO 11885 inductively coupled plasma optical emission spectrometry; typical compliance limits are set at 0.5 mg/L copper and 2.0 mg/L zinc in the filtered sample. Caustic soda used in drinking water applications must meet AWWA B501-19 and be certified to NSF/ANSI/CAN 60. The compliance checklist in Table 1 summarises the minimum specifications.

Chlorine gas is absorbed into a recirculating stream of 20–22 wt% sodium hydroxide in a packed tower or liquid jet eductor to produce sodium hypochlorite. The reaction releases 103 kJ per mole of chlorine, so the recirculation loop passes through a plate-and-frame heat exchanger or shell-and-tube cooler to maintain liquor temperature below 30 °C. Final product at 12.5–15% trade chlorine contains 5–15 g/L excess sodium hydroxide and a pH above 12.5, meeting ANSI/AWWA B300-18 for commercial sodium hypochlorite. Excess caustic suppresses chlorate formation, but accumulation of iron, nickel, or copper above 0.5 mg/L catalyses oxygen evolution and reduces available chlorine half-life. Storage tanks are fabricated from high-density polyethylene or fibre-reinforced plastic, with UV-shielding and venting for oxygen; carbon steel and stainless steel fittings are omitted because they accelerate localised decomposition.

The process is operated with continuous caustic feed and chlorine flow cascade control, maintaining the overflow stream at pH 12.8–13.2. A drop below pH 12.0 shifts the equilibrium toward hypochlorous acid and permits chlorate formation; a rise above pH 13.5 increases carryover of unreacted caustic and shortens shelf life. Decomposition follows a temperature-dependent kinetic profile: at 15 °C, a 12.5% trade solution may retain 90–95% of initial available chlorine after 30 days, while at 30 °C the retention may fall to 70–80% over the same period. These figures are dependent on nickel contamination below 0.1 mg/L and on sealed storage without repeated partial withdrawals; published data for specific tank configurations is limited, and on-site jar testing is required for shelf-life commitments.

How Does Water in Sodium Methoxide Generation Affect Biodiesel Transesterification?

Solid sodium hydroxide flake at 98–99 wt% purity is dissolved in anhydrous methanol to generate sodium methoxide for base-catalysed transesterification. The equilibrium reaction produces water; water in the methanol stream above 0.2–0.3 wt% reverses the equilibrium toward sodium hydroxide and prematurely saponifies triglycerides in the downstream reactor. The resulting sodium soaps increase phase separation time in centrifugal separators and raise free glycerol above the EN 14214 limit of 0.02 wt%. The dissolution is exothermic and is controlled below 45 °C in a 316L stainless steel static mixer with a methanol recycle cooling loop. Catalyst dosing is set to 0.6–1.0 wt% sodium methoxide solution on oil charge; for an 800 kg/h continuous biodiesel line, this corresponds to roughly 5–8 L/h of 30 wt% methoxide solution.

Water is purged from methanol by a molecular sieve column or by distillation over a 3 Å sieve bed to below 0.1 wt% before catalyst preparation. In-line Karl Fischer analysers per ISO 12937 monitor methanol water content; a deviation above 0.15 wt% triggers diversion of the methoxide stream and a caustic neutralisation flush. The biodiesel phase after separation is washed with water acidified to pH 2–3 with citric or phosphoric acid, then dried under vacuum at 80–90 °C; residual soap and total glycerol are verified by EN 14103 and EN 14105, respectively. This alkali route is incompatible with high free fatty acid feedstocks above 1.0 wt% unless pre-esterification reduces the acid value below 2 mg KOH/g, because each free fatty acid consumes catalyst on a stoichiometric basis.