Effective Alkali Operating Range 75 to 120 g/L in Kraft White Liquor

Effective alkali (EA) in kraft white liquor is defined as the concentration of sodium hydroxide plus one-half sodium sulphide, expressed as Na₂O in g/L. The operating range of 75 g/L to 120 g/L is not a direct cooking specification but rather a liquid-phase concentration envelope that interacts with chip moisture, liquor-to-wood ratio, digester hydraulics, and recovery-cycle capacity to determine the actual chemical dose delivered to the delignification front. White liquor drawn from the clarified storage tank at 75 g/L EA has a lower volumetric chemical density than a liquor at 120 g/L EA; therefore, for a fixed effective alkali charge on oven-dry wood, the lower concentration requires a larger white liquor flow. The resulting change in volumetric loading alters black liquor dry solids concentration, evaporator steam demand, and the sodium-sulphur balance across the causticizing plant. Analytical control of this window follows TAPPI T 624 and TAPPI T 625, which define sampling and titration procedures for white liquor total titratable alkali, active alkali, effective alkali, and sulphidity. In these methods, sodium hydroxide and sodium sulphide are differentiated through paired titrations and expressed on an Na₂O basis. Because the EA definition includes only half the sodium sulphide, sulphidity changes shift the relationship between EA and active alkali. A white liquor with an EA of 100 g/L and a sulphidity of 30% has a different chemical potential and different delignification kinetics than a white liquor with the same EA but a sulphidity of 35%. The 75 g/L to 120 g/L band exists because mills must balance these liquor chemistry variables against digester residence time, final kappa, pulp viscosity, brownstock washing efficiency, and the physical constraints of the lime kiln and recausticizing loop. The lower and upper boundaries are not arbitrary thresholds; they represent limits beyond which either delignification chemistry or recovery-cycle equipment imposes a measurable penalty.

The distinction between EA and active alkali is not merely analytical; it changes the interpretation of mill data. Active alkali is the sum of all sodium hydroxide and sodium sulphide expressed as Na₂O, while effective alkali discounts one-half of the sodium sulphide on the assumption that sulphide hydrolysis yields one hydroxide equivalent per two sodium sulphide molecules. For a white liquor with an EA of 100 g/L and a sulphidity of 30%, the active alkali is approximately 114 g/L as Na₂O because the sodium sulphide concentration is roughly 31 g/L as Na₂O and the sodium hydroxide concentration is roughly 83 g/L as Na₂O. These conversions are mill-specific and are verified by titration rather than calculated from a single online density meter, because dead-load species change the density–alkalinity relationship. The use of TAPPI T 625 differential titration is required when a digital control system relies on an inferred EA value. Online analyzers using conductivity or density provide a fast trend, but they must be calibrated against the laboratory titration at least once per shift to avoid drift caused by carbonate and sulphate variations.

Does the lower 75 g/L threshold reflect delignification kinetics or hydraulic capacity?

At EA concentrations near 75 g/L as Na₂O, the challenge is both volumetric and kinetic, but the binding constraint differs by mill configuration. A continuous digester with a softwood chip feed of 1,000 oven-dry tonnes/day and a target EA charge of 18% on oven-dry wood requires 2,400 m³/day of white liquor at 75 g/L, compared with 1,890 m³/day at 95 g/L and 1,500 m³/day at 120 g/L. This increase in white liquor flow directly raises the hydraulic load on the feed system, digester top separator, and extraction screens. It also reduces the black liquor dry solids concentration delivered to the recovery boiler when the wash dilution factor is held constant. Kinetic limitations at the lower end are equally specific. The bulk delignification stage in softwood kraft cooking follows an apparent rate that depends on hydroxide ion concentration, not solely on temperature and time. Published delignification studies using softwood furnish show that the residual effective alkali should not fall below 8 g/L to 10 g/L as Na₂O at the end of the bulk phase, and that below 5 g/L dissolved lignin and hemicellulose fragments can undergo condensation and deposit on fiber surfaces. A white liquor EA of 75 g/L leaves a smaller residual alkali margin than a higher-concentration liquor, particularly if the chip furnish contains an oversize chip fraction above 8 mm thick. Thick chips develop radial temperature and alkali gradients because hydroxide consumption occurs first at the chip surface, while the chip center remains below the bulk-liquor temperature and below the free-hydroxide concentration needed for β-aryl ether cleavage. Presteaming and chip compression reduce this heterogeneity, but they do not remove the stoichiometric requirement for free hydroxide at the reaction front. If the residual EA in the extraction screen drops below 5 g/L before the target H-factor is reached, the control system must either reduce the digester solids throughput, increase the white liquor flow, or accept a higher blow-line kappa. The lower 75 g/L limit is therefore not a single chemical threshold; it is a practical boundary at which hydraulic capacity and chip furnish heterogeneity begin to consume the remaining safety margin for residual alkali control.

Hardwood operations typically tolerate lower EA than softwood operations because hardwood lignin has a higher syringyl-to-guaiacyl ratio and is more readily cleaved under alkaline conditions. A hardwood mill digesting aspen or eucalyptus at kappa 18 to 22 may operate with a white liquor EA near 75 g/L to 85 g/L, while a softwood mill targeting kappa 30 to 32 more commonly uses 90 g/L to 105 g/L. This difference is reflected in the chip column residence time and the digester temperature profile. Hardwood cooking temperatures are often 5°C to 10°C lower than softwood temperatures for the same H-factor because the lignin reactivity is higher. The lower EA threshold therefore applies differently across fiber furnishes, and a mill that swings between hardwood and softwood must change the EA setpoint not only by species but by chip size distribution and target kappa.

White liquor dead load modifies the lower threshold because total titratable alkali includes carbonate, sulphate, thiosulphate, sulphite, and chloride species that do not contribute to delignification to the same extent as sodium hydroxide. A liquor with an EA of 80 g/L and a TTA of 140 g/L contains a larger fraction of dead load than a liquor with an EA of 80 g/L and a TTA of 105 g/L. The dead load increases the ionic strength of the white liquor without increasing the effective hydroxide content, and it displaces liquor volume that could otherwise carry active pulping chemicals. High dead load also raises the boiling point rise in the black liquor evaporators, increases black liquor viscosity, and adds sodium and sulphur load to the recovery boiler without improving pulp yield or kappa. Recausticizing plant operation is particularly sensitive to this ratio. Pressure filters and lime mud washing systems that receive white liquor with excessive carbonate carry-over produce a lime mud with higher residual soda, and the lime kiln must then burn this soda from the mud, increasing fuel demand without adding process benefit. When a mill is constrained at the lower EA boundary, the combined effect of high dead load and low effective alkali is a larger volumetric liquor flow, a weaker black liquor to the recovery boiler, and a higher risk of alkali depletion in the digester extraction zones. This explains why mills do not routinely operate below 75 g/L unless they have compensated with lower chip throughput or higher sulphidity.

The effect of white liquor EA on evaporator and recovery boiler load is indirect but measurable. At a lower EA concentration, the larger white liquor volume to the digester produces a higher flow of weak black liquor to the multiple-effect evaporator. For a fixed black liquor dry solids target of 72% to 75% before the recovery boiler, the evaporator plant must remove more water per air-dry tonne of pulp, increasing steam demand. A mill operating at 75 g/L EA can see an increase in evaporator steam demand of 0.2% to 0.5% on a mill-wide basis compared with 95 g/L EA, depending on the wash dilution factor. This increase is not large in percentage terms but is continuous and compounds with any decline in washing efficiency. At 120 g/L EA, the lower volumetric flow reduces evaporator water load, but the higher residual alkali and higher black liquor sodium content can increase the viscosity of the concentrated black liquor. High-viscosity black liquor at the recovery boiler feed leads to poorer atomization in the black liquor spray nozzles and can reduce combustion stability if the dry solids exceed 75% or if the liquor viscosity at processing temperature exceeds design limits. The liquor droplet size from a recovery boiler spray plate is a function of viscosity and surface tension; at constant temperature, a more concentrated black liquor with higher sodium content shifts the droplet size distribution upward and can increase the carryover of unburned char particles.

When effective alkali is pushed to 120 g/L, which recovery-cycle constraints dominate?

Pushing white liquor EA to 120 g/L as Na₂O shifts the bottleneck from the digester hydraulic system to the recausticizing plant and lime kiln. The causticizing reaction between sodium carbonate and calcium hydroxide is equilibrium-limited, and the equilibrium conversion to sodium hydroxide decreases as the product hydroxide concentration increases. Recausticizing plant operating data reported by equipment suppliers show causticizing efficiency of approximately 85% when the white liquor TTA is near 100 g/L as Na₂O, falling to 76% to 81% when the TTA exceeds 140 g/L as Na₂O. Producing a 120 g/L EA white liquor therefore requires either a higher excess lime charge in the slaker, a longer causticizer residence time, or an adjustment to the white liquor clarifier underflow to prevent sodium carbonate accumulation. The lime mud filtration step is affected in parallel. Free hydroxide in the white liquor stabilises fine calcium carbonate particles in the lime mud slurry, reducing filtration rate on a rotary vacuum precoat filter or a disc filter. In a pressure filter installation, this appears as higher lime mud moisture after the final wash, which increases the specific heat demand of the lime kiln. A lime kiln designed for 2.2 GJ/tonne CaO may shift toward 2.5 GJ/tonne CaO when lime mud moisture rises above 35% water by mass. The digester metallurgy also requires attention at the upper EA boundary. Carbon steel digester vessels and extraction screens in continuous service have known stress corrosion cracking susceptibility in alkaline sulphide liquors at temperatures above 150°C when the EA concentration exceeds approximately 100 g/L as Na₂O; many mills specify stress-relief of welded zones or use corrosion-resistant overlay on extraction screens and circulation piping. The concentrated liquor also creates a steeper osmotic potential difference at the chip interface. If chip moisture is below 35% on a wet basis, the 120 g/L liquor can retard the initial movement of water and hydroxide into the chip interior unless the chip column is presteamed and pressurized before white liquor contact. Presteaming at 0.6 MPa to 0.8 MPa displaces air from the chip lumens and condenses steam within the capillary structure, but it does not fully equalise the alkali concentration gradient. The benefit of operating at 120 g/L is the reduced white liquor volumetric demand, which lowers digester feed system hydraulic load and reduces the steam consumed in heating the cooking liquor. This benefit is real but must be weighed against the recovery-cycle losses and the increased risk of local alkali concentration overshoot in the chip mass.

Mid-range EA, sulphidity, and the modified H-factor control logic

In the middle of the operating range, between 90 g/L and 105 g/L EA as Na₂O, the control system can use the modified H-factor with reasonable accuracy because the alkali concentration does not become the dominant rate-limiting variable. The modified H-factor integrates the relative reaction rate over the cooking cycle using an apparent activation energy of approximately 130 kJ/mol to 140 kJ/mol. For a softwood cooking temperature of 160°C to 170°C, a chip column retention time of 4 h to 6 h, and a sulphidity of 30% to 35%, a white liquor EA of 95 g/L supports a blow-line kappa of 30 to 32 at an H-factor of 1,000 to 1,300. The sensitivity of this relationship to sulphidity is substantial. Sodium sulphide accelerates delignification by cleaving β-aryl ether bonds and by reducing the formation of alkali-stable condensed structures; the EA definition includes half the sodium sulphide concentration because only one hydroxide equivalent is produced from the sulphide hydrolysis equilibrium, but the kinetic promotion effect is not fully linear. Published cooking studies using Scandinavian softwood indicate that reducing sulphidity from 30% to 25% can require an H-factor increase of approximately 10% to 15% to maintain the same kappa, while raising sulphidity to 40% shifts the constraint to the recovery boiler sulphur balance and the causticizing plant. The mid-range EA window is also the region where the transition from bulk to residual delignification can be controlled by adjusting the extraction screen temperature and the white liquor split between impregnation and cooking zones. A two-vessel continuous digester receives white liquor at multiple points: impregnation, co-current cooking, and possibly counter-current washing at the digester bottom. The distribution of EA across these injection points determines the alkali profile along the chip column. If too much EA is injected at the impregnation zone with a mid-range concentration, the initial pH rises rapidly and the chip surface lignin is solubilised early, while the chip center receives insufficient chemical. If too little is injected, the chip center may not reach the critical hydroxide concentration until the temperature has already reached the bulk cooking plateau, producing a higher fraction of brownstock rejects. In-digester extraction screens provide the only direct measurement of residual EA before the blow line, and the dead-time from a white liquor flow or EA setpoint change to the observed residual response is typically 2 h to 4 h in a continuous digester. This dead-time forces the control logic to use a combination of feed-forward adjustment based on white liquor EA and liquor flow, plus a slow trim on the digester temperature setpoint. A mid-range EA concentration reduces the volumetric sensitivity of this loop and provides enough residual alkali margin to absorb chip moisture and thickness disturbances without exceeding the lime kiln and recovery boiler penalties seen at the upper end.

Batch digester operations respond differently to the EA range than continuous digesters because the entire white liquor charge is mixed with the chip charge at the start of the cook. In a batch digester with a liquor-to-wood ratio of 3.5 L/kg and a target EA charge of 18%, the initial EA in the free liquor is approximately 51 g/L if the white liquor EA is 75 g/L and the displacement ratio is not considered, but the local concentration decreases as the liquor penetrates the chips. A continuous digester maintains a more stable EA profile along the chip column because fresh white liquor is introduced at multiple points, and the co-current and counter-current flows provide a chemical gradient. This fundamental difference explains why the acceptable white liquor EA range is broader for continuous digesters than for batch digesters. Batch systems at the lower end of the range require extended chip impregnation and longer circulation times to compensate for the rapid depletion of free hydroxide, while continuous systems can adjust the extraction screen temperature and white liquor split to preserve residual EA in the bulk and residual phases. At the upper end of the range, batch digesters experience a higher initial pH spike, which can lead to early carbohydrate degradation if the heating rate is not controlled. Published data for this specific configuration is limited; however, the directional effect is consistent across mill types.

Residual alkali at the blow line is not a fixed value across the 75–120 g/L EA range

Residual effective alkali measured at the blow line is the best direct indicator of whether the white liquor EA setpoint is appropriate for the digester conditions, but the target residual is not constant. At a white liquor EA of 75 g/L, a softwood digester targeting kappa 30 to 32 typically carries a blow-line residual of 5 g/L to 8 g/L as Na₂O. This leaves a narrow margin above the 5 g/L threshold at which dissolved lignin and hemicellulose can condense onto fiber surfaces. At a white liquor EA of 120 g/L, the same kappa target is commonly achieved with a residual EA of 12 g/L to 18 g/L as Na₂O, because the high incoming concentration is deliberately underutilized to avoid overcooking and yield loss. The high residual is not chemically inert in the brownstock washing system. It increases the sodium content of the weak black liquor and therefore increases the evaporator load and recovery boiler smelt load. A brownstock washing train with a washing efficiency of 90% to 95% can lose 5 kg to 10 kg Na₂O per air-dry tonne of pulp, and this loss rises when the blow-line residual is kept at the upper end to protect pulp quality. The residual alkali management strategy must also account for the fact that bicarbonate and carbonate ions are generated during the later stages of cooking and are titrated as part of the residual alkali measurement, but they do not provide the same high-pH buffering as free hydroxide. A titration result of 10 g/L residual EA can therefore represent different free-hydroxide availabilities depending on the carbonate accumulation in the liquor. This is why mills use both residual EA and blow-line pH or free-hydroxide concentration, with TAPPI T 625 providing the differential titration basis. The relationship between incoming white liquor EA and residual EA across the 75 g/L to 120 g/L window is nonlinear because the liquor-to-wood ratio, chip porosity, cooking time, and kappa target all change the alkali consumption per tonne of pulp. A mill cannot simply scale the residual target with the incoming EA; it must determine the target empirically from the digester extraction screen profile, blow-line kappa trend, and brownstock oxygen-stage response.

The white liquor EA setpoint also propagates into the oxygen delignification stage through the residual alkali carried over from the digester and brownstock washing. At medium consistency oxygen delignification with a pulp consistency of 10% to 14% and a sodium hydroxide charge of 1.5% to 2.5% on oven-dry pulp, the initial pH is strongly affected by carryover of digester residual alkali. If the digester residual EA is held above 18 g/L because the white liquor EA is at the upper end of the range, the oxygen-stage initial pH can exceed 11.5 unless the alkali charge is reduced, and the resulting carbohydrate degradation decreases pulp viscosity. ISO 302:2015 kappa testing across the oxygen stage often shows a kappa reduction of 40% to 50% when the digester residual EA is controlled within the mid-range window, but the reduction can fall below 35% when the incoming residual alkali is excessive and the oxygen-stage chemical dose is not compensated. The digester, washing, oxygen delignification, recovery boiler, recausticizing plant, and lime kiln are coupled by the same sodium and sulphur balances that define the 75 g/L to 120 g/L effective alkali operating range. A change in white liquor EA is therefore not a local digester adjustment; it is a disturbance to the entire kraft chemical recovery cycle that must be managed through coordinated control of liquor flow, temperature, residual alkali, and causticizing efficiency.