Factors Limiting Effective Alkali Charge in the Causticizing Circuit

In the kraft recovery cycle, the term effective alkali charge in the causticizing circuit refers not to raw lime addition but to the concentration of hydroxide alkalinity actually generated from sodium carbonate and returned to white liquor after equilibrium, side reactions, filtration losses, and dead-load accumulation have been accounted for. A green liquor entering the slaker typically carries total titratable alkali in the range 120–160 g/L Na₂O, of which sodium carbonate may represent 70–85% of the alkali on an equimolar sodium basis, while sodium sulfide, sodium sulfate, and sodium thiosulfate make up the balance. The stoichiometric conversion of sodium carbonate by calcium hydroxide requires one mole of calcium hydroxide per mole of sodium carbonate, but industrial circuits consistently operate at lime charges of 0.90–1.10 mol CaO/mol Na₂CO₃ because conversion is limited by slaking kinetics, lime surface area, liquor composition, and thermodynamic equilibrium. The difference between the theoretical alkali charge and the measured hydroxide concentration after causticizing defines the efficiency loss, and that loss propagates directly into white liquor active alkali, lime mud loop inventory, kiln fuel demand, and sodium make-up requirement.

Two independent measurement streams define the limit in a mill: the residual sodium carbonate in clarified white liquor and the sodium oxide loss in the lime mud filter discharge. The first stream is normally measured by auto-titrator using TAPPI T 632 cm-03, with a residual carbonate range of 12–20 g/L Na₂O in a well-operated circuit; the second stream is measured on a dry mud composite and typically runs 0.5–2.0 kg Na₂O/t dry mud when displacement washing is adequate. Because the residual carbonate represents the unconverted fraction of the green liquor carbonate charge, each gram of Na₂O left as carbonate in white liquor displaces active hydroxide that could have been generated if the equilibrium and kinetic constraints allowed. The effective alkali charge is therefore best expressed as the ratio of sodium hydroxide to total sodium hydroxide plus sodium carbonate on a Na₂O basis, and this ratio rarely exceeds 85% in industrial causticizing, with 75–83% being the common working envelope.

Why Does Calcium Carbonate Solubility Product Cap the Carbonate Conversion?

At the molecular level, the causticizing reaction is a heterogeneous system in which dissolved sodium carbonate reacts with suspended calcium hydroxide to produce dissolved sodium hydroxide and solid calcium carbonate. The equilibrium constant is governed by the solubility product of calcium carbonate and the solubility product of calcium hydroxide, and the measured conversion is determined by the ratio of hydroxide ion activity squared to carbonate ion activity. Because the product sodium hydroxide is highly soluble, its accumulation in solution drives the reverse reaction, so the equilibrium efficiency decreases as the total alkali concentration increases. A green liquor TTA of 120 g/L Na₂O can support an equilibrium causticizing efficiency near 83–85% at 99–103°C, whereas the same circuit at 160 g/L Na₂O may plateau near 75–78%; published data for specific mill configurations is limited, but this inverse relationship is consistently observed in equilibrium calculations and mill audits. Temperature exerts a smaller and opposite kinetic effect: the reaction is mildly exothermic, so higher temperatures slightly lower the thermodynamic conversion, yet the industrial optimum remains above 95°C because reaction rate and particle mass transfer improve sufficiently to offset the equilibrium penalty.

The practical consequence of this equilibrium constraint is that a mill cannot simply increase lime charge to push carbonate conversion to complete reaction. Once the dissolved carbonate and hydroxide activities satisfy the equilibrium quotient, further calcium hydroxide remains as unreacted solid and contributes to lime mud calcium oxide, raising mud solids and kiln energy demand without producing additional sodium hydroxide. This explains why residual sodium carbonate in white liquor is not an arbitrary operating target but a thermodynamic floor at a given TTA, temperature, and sodium sulfide concentration. Sulfide reduces water activity and changes the ionic strength of the liquor, so high-sulfidity green liquors above 35% sulfidity often require either a lower TTA set point or a slightly higher lime-to-carbonate ratio to maintain the same effective alkali charge; published data for this specific interaction is limited.

Slaker temperature trajectory is the first measurable constraint that separates a highly efficient causticizing line from a carbonate-limited line. When quicklime is added to green liquor, calcium oxide slakes exothermically to calcium hydroxide, and a reactive mill lime will raise the slaker batch from the green liquor inlet temperature of 85–90°C to 99–105°C within 2–3 min. If the lime is overburned, underburned, or contaminated with uncalcined carbonate, the temperature rise may be delayed or may never exceed 90°C, and the resulting calcium hydroxide particles are coarse and dense, with a specific surface area too low to sustain the causticizing reaction within the available detention time. A slaker operating at 15–30 min total retention and an agitator power input of 0.5–1.0 kW/m³ can compensate for moderate reactivity loss, but no mechanical adjustment can restore surface area once the lime has slaked at an unfavorably low temperature.

Hydrated lime particle size distribution after slaking determines the available interfacial area for the dissolution of calcium hydroxide and the nucleation of calcium carbonate. In a reactive lime, the slaked slurry has a median particle diameter below 20–30 µm, measured by laser diffraction per ISO 13320:2020, whereas a poorly slaking lime may produce a bimodal distribution with a coarse fraction above 150 µm that settles in the slaker classifier and is discharged as grit. The coarse fraction removes available calcium oxide from the causticizing reaction, and the grit flow must be purged because it otherwise accumulates in the causticizer train and reduces effective working volume. The loss is not limited to the grit stream: large calcium hydroxide particles dissolve too slowly to counteract the reverse reaction, so the final clarified white liquor shows higher residual carbonate at the same lime charge.

Lime Quality, Grit Fraction, and Available CaO on the Filtration Floor

Available CaO in kiln-produced lime typically falls between 85–92% when the lime mud feed is well washed and the kiln is operated with a front-end temperature below 1,100°C, but excursions may drop available CaO below 80% during fuel switches, refractory degradation, or lime mud moisture spikes. The unavailable fraction consists of uncalcined calcium carbonate, calcium sulfate, magnesium oxide, and refractory silicates, all of which behave as inert solids in the slaker. A drop in available CaO from 90% to 82% does not merely require an additional 8–10% lime charge to deliver the same molar calcium oxide; it also increases the total mud solids load by the entire mass of the inert fraction, raising the filtration resistance and washing dilution demand. Lime is normally dosed by a belt feeder or screw conveyor slaved to green liquor flow and carbonate content, and the dosing ratio is reset after each available CaO titration performed to ASTM C25-19 or TAPPI T 617.

Grit and unburned carbonate do not merely dilute available CaO; they consume slaker volume and alter the rheology of the slaked lime slurry so that the causticizing reaction is starved of mixing energy in the lower sections of the vessel. The grit fraction above 150 µm is routinely 5–15% of lime mass in mills using a single-pass slaker with a rake classifier, but values above 20% are associated with visible short-circuiting and with a loss of causticizing efficiency of 2–5 percentage points at constant lime charge. The reason is that large particles provide relatively little reactive area and occupy volume that would otherwise contain sub-20 µm calcium hydroxide particles. The standard practice is to monitor the slaker grit discharge and to adjust the slaker water-to-lime ratio, but if the lime source itself has a high grit tendency the circuit can only compensate by increasing lime dosage, which further increases mud load.

Sodium sulfide, sodium sulfate, and sodium thiosulfate coexist with carbonate in green liquor, and their concentrations alter the effective alkali charge even though they do not participate directly in the causticizing reaction. Sodium sulfide behaves as a strong electrolyte that contributes to ionic strength and suppresses the activity coefficient of carbonate, which shifts the equilibrium toward lower conversion at a given TTA. Sulfate and thiosulfate are dead-load species that consume liquor volume and increase TTA without contributing alkalinity; when the TTA accountant uses a fixed TTA set point, a rise in sodium sulfate from 2 g/L Na₂O to 5 g/L Na₂O reduces the sodium carbonate available for conversion and therefore lowers the effective alkali charge. This is a common cause of declining white liquor active alkali during periods of high black liquor carryover or oxidized white liquor sulfidity changes, and it requires a purge increase rather than a lime charge adjustment.

Dregs that bypass the green liquor clarifier and enter the slaker do not generally consume calcium hydroxide in stoichiometric amounts, but they exert a particulate and surface-passivating effect that is disproportionate to their mass. Green liquor dregs are typically maintained below 50–100 mg/L total suspended solids; excursions above 200 mg/L lead to deposition on the slaker classifier, blinding of lime mud filter cloth, and the formation of mixed dreg–calcium carbonate flocs that trap sodium hydroxide. The trapped sodium hydroxide is then carried with the lime mud to the kiln, where it causes sodium vaporization and ring formation, and the effective alkali charge drops even when the clarified white liquor titration remains stable. The mechanism is physical entrainment rather than chemical consumption, which is why effective alkali charge audits include a lime mud sodium balance and not only a liquor residual carbonate analysis.

When Lime Mud Washing Lacks Sufficient Displacement Ratio

On a precoat lime mud filter or a belt filter, the displacement washing step is often the largest single point of sodium oxide loss in the causticizing loop. The filter cake is formed at a thickness of 10–20 mm on a rotary drum with a differential pressure of 200–400 mbar, and wash water is applied at a ratio of 1.5–2.5 m³/t dry mud in a well-adjusted installation. If the wash ratio falls below 1.0 m³/t dry mud or the spray bars become plugged, residual sodium in the discharged lime mud rises from the target of 0.5–1.0 kg Na₂O/t dry mud to 2–4 kg Na₂O/t dry mud. That sodium is lost from the liquor cycle and must be replaced with purchased sodium hydroxide or sodium carbonate, reducing the net effective alkali charge available to the digester even though the causticizing reaction itself is unchanged.

A factor often overlooked is the interaction between filter speed and green liquor dregs load during periods of high mud solids. If the filter speed is increased to maintain mud throughput while the dregs content rises, the cake formation time decreases and the applied wash water no longer displaces the interstitial mother liquor, producing a wetter cake with a higher sodium oxide concentration. The result is a false reading of causticizing efficiency: white liquor residual carbonate may be in specification while the measurable active alkali delivered to the pulping process is reduced because sodium hydroxide is leaving the circuit in the lime mud. Published data for specific filter configurations is limited, but the relationship between cake washing efficiency and the dimensionless displacement ratio is well established in solid–liquid separation design.

Control loops on the slaker and causticizer train operate on a time delay of 45–90 min between a lime feed change and the corresponding steady-state white liquor carbonate response, and this dead time is responsible for many oscillations in effective alkali charge. The slaker and three causticizing vessels typically provide a total retention time of 90–120 min, with each vessel agitated at 0.3–0.6 kW/m³ for the causticizers and 0.5–1.0 kW/m³ for the slaker, and the temperature in the last causticizer is held in the range 99–103°C. A control system that holds lime-to-carbonate ratio by feed-forward from green liquor flow and TTA can maintain stable residual carbonate if the available CaO is titrated frequently, but lime silo segregation, belt feeder bias, and slaker level control introduce batch-to-batch variance. When the measured white liquor residual carbonate deviates by more than 2 g/L Na₂O from the target, the correction should be made in lime-to-carbonate ratio steps of 0.02–0.04 mol CaO/mol Na₂CO₃, because larger corrections overliming the circuit and creating a free lime transient in the lime mud.

Non-process elements such as magnesium, aluminum, silicon, and phosphorus enter the causticizing circuit through wood furnish, make-up lime, and recycled mill streams, and their accumulation creates secondary reactions that consume calcium hydroxide or precipitate on the reactive surface. Magnesium hydroxide is highly insoluble and precipitates within the slaker, increasing mud solids without contributing to alkalinity; aluminum and silicon form sodium aluminosilicate scale that coats the lime particles and reduces their dissolution rate. Phosphorus can precipitate as calcium phosphate, which consumes a portion of the available calcium and therefore represents a chemical loss of lime that is not recovered in the kiln. The specific effect on effective alkali charge depends on the concentration ratio of non-process element to green liquor TTA, and published data for this specific configuration is limited; however, mills with high phosphorus input have reported a lime demand increase of 3–8% at a constant causticizing efficiency target.

Overliming Above 160 g/L Na₂O TTA Produces Free Lime in the Mud

When the causticizing circuit operates above a TTA of 160 g/L Na₂O and the lime charge is simultaneously pushed above 1.10 mol CaO/mol Na₂CO₃, the circuit enters a region where overliming no longer improves effective alkali charge and instead produces a lime mud containing free calcium hydroxide. Free lime in the mud raises the pH of the filter filtrate and increases the risk of scale on the white liquor clarifier and downstream piping. The excess calcium hydroxide also competes with the desired calcium carbonate precipitation and produces a lime mud with a higher specific filtration resistance, so the vacuum filter must be slowed or the pressure differential increased beyond the normal 200–400 mbar. The net effect is a loss of production capacity and a higher sodium oxide loss, which can cancel any gain in carbonate conversion.

In lines producing bleached grades, the tightest operational boundary is often white liquor turbidity rather than carbonate conversion. A poorly causticized white liquor with residual calcium hydroxide, dregs carryover, or fine calcium carbonate particles above 20–50 mg/L can create scaling in the digester heater and impair oxygen delignification. The white liquor clarifier is typically designed for a rise rate of 0.6–1.0 m/h, and if the lime mud solids concentration entering the clarifier exceeds 35–40 wt% because of overliming, the rise rate cannot be maintained, and turbidity breaks through. This forces a reduction in lime charge, which then lowers the effective alkali charge to keep the circuit within the physical separation capability of the clarifier.

Accumulation of sodium sulfate, sodium thiosulfate, and potassium salts in the liquor cycle has a delayed but measurable effect on effective alkali charge because these salts compete for water and reduce the solubility of calcium hydroxide in the slaker. A mill that operates its recovery boiler with high black liquor firing and minimal electrostatic precipitator catch may see dead-load concentrations climb over several months, and the first visible symptom is a rise in residual carbonate at the same lime-to-carbonate ratio. The corrective action is a liquor purge, not a lime charge increase, because adding lime to a dead-load-limited circuit increases mud solids and kiln energy input without increasing hydroxide production. Published data for this specific configuration is limited, but the inverse correlation between dead-load concentration and causticizing efficiency is widely documented in recovery cycle mass balances.