Effective Alkali Charge Control in Kraft Pulping Delignification

Effective alkali charge in kraft pulping is defined as sodium hydroxide plus one-half sodium sulfide, expressed as Na₂O, and its control determines whether the cook terminates at the specified kappa number with acceptable yield, viscosity, and downstream oxygen-stage reactivity. White liquor delivered from the recausticizing plant is characterized by active alkali, effective alkali, sulfidity, and total titratable alkali, with manual titration performed in accordance with TAPPI/ANSI T 624 cm-23; pulp kappa number is determined by ISO 302:2015 or TAPPI T 236 cm-23, and limiting viscosity number is measured by ISO 5351:2010. The alkali charge is not consumed by lignin alone: hydrolysis of acetyl groups on galactoglucomannan releases acetic acid, 4-O-methylglucuronic acid is cleaved from xylan, and aliphatic carboxylic acids generated during peeling and stopping reactions consume sodium hydroxide rapidly in the first 30–60 minutes of the cook. As a result, an industrial softwood cook charged at 17–21% effective alkali on oven-dried wood may show only 8–12 g/L residual effective alkali at blow when the kappa number is held between 28 and 32; hardwood furnishes typically require 13–17% effective alkali for kappa 16–20 under similar sulfidity and H-factor constraints. The distinction between charge and residual alkali matters because low residual effective alkali at the end of the bulk phase is associated with lignin condensation, poor oxygen delignification selectivity, and increased carryover to the recovery cycle.

Alkali Charge as a Dynamic Process Variable Rather Than a Fixed Recipe Parameter

Control of effective alkali charge on a modern fiberline requires a reconciled mass balance of dry wood mass, white liquor strength, liquor-to-wood ratio, sulfidity, and target H-factor. Chip moisture variation is usually the largest disturbance: a chip pile moisture change of 2 percentage points at 50% moisture content alters dry fiber mass by more than 4% relative, which translates directly into an alkali charge error of the same magnitude if wet-mass feed rate is held constant. Chip moisture is measured by continuous near-infrared analyzers or by oven-dry checks according to ISO 638-1:2022, and the feedforward charge calculation is updated on intervals of 1–5 minutes. White liquor effective alkali is measured by automatic titrators or by density and conductivity analyzers, but conductivity is not specific to hydroxide; it is also affected by carbonate, sulfate, thiosulfate, and temperature. Therefore, density and conductivity signals are calibrated against TAPPI/ANSI T 624 cm-23 titrations at intervals of 4–8 h from recausticizing plant sample taps. The liquor-to-wood ratio is maintained between 3.5:1 and 5.0:1 in continuous digesters, and between 3.0:1 and 4.5:1 in displacement batch digesters; in each case, increasing the ratio improves liquor mixing but increases steam consumption in the evaporators. A high liquor circulation rate in the cooking zone—commonly equivalent to 4–6 digester volumes per hour—reduces radial alkali concentration gradients, but excessive circulation can carry degraded lignin fragments into the early cooking zone and accelerate condensation. The target H-factor is derived from the kraft delignification activation energy, which is reported in published kinetic studies to lie between 125 kJ mol⁻¹ and 140 kJ mol⁻¹ for the bulk phase. At a cooking temperature of 170 °C, a deviation of 1 °C changes the relative delignification rate by approximately 7–9%, so temperature loops are typically tuned to hold the digester heating zone within ±0.5 °C and the cooking zone within ±1.0 °C.

Displacement batch systems, including Rapid Displacement Heating configurations, distribute effective alkali across black liquor fills, high-temperature black liquor displacement, and fresh white liquor addition. In a typical sequence, presteamed chips are first immersed in warm black liquor containing residual alkali; the liquor is then displaced with hot black liquor and finally with white liquor. If the residual effective alkali content of the black liquor is not measured and subtracted from the fresh white liquor charge, the total effective alkali dose can exceed the intended level by 1–3 percentage points on oven-dried wood, which increases carbohydrate losses and reduces pulp viscosity. These systems operate with a high degree of countercurrent liquor reuse, so dead load components such as sodium carbonate, sodium sulfate, and oxidized sulfur species accumulate in the black liquor and displace effective alkali. The control consequence is that charge calculations must use residual effective alkali, not total titratable alkali, because carbonate and sulfate titrate as alkali but have limited delignification activity. In mills with displacement batch digesters, blow-line kappa and residual alkali from each cook are used to adjust the white liquor charge on the next cook, creating a slow feedback loop that can oscillate if chip moisture changes more rapidly than the controller sample interval.

What Measurement Frequency Is Needed for Reliable Effective Alkali Control?

Reliable effective alkali control requires sampling and analysis intervals that are short compared with the disturbance time constant. Continuous digesters with hydraulic residence times of 3–6 h impose a dead time from white liquor addition to blow-line kappa measurement that makes feedback-only control sluggish. In-line analyzers based on automatic potentiometric titration of filtered liquor can provide cycle times of 5–15 minutes for white liquor, but black liquor analyzers face additional delay from high dissolved solids, fiber, and scale. Manual titrations according to TAPPI/ANSI T 624 cm-23 every 8–12 h are adequate only for slow recausticizing inventory disturbances; they cannot resolve a slaker upset that changes effective alkali by 5 g/L as Na₂O within 30 minutes. For this reason, many control architectures separate fast feedforward from slow feedback. Fast feedforward uses density, conductivity, chip moisture, and white liquor flow to correct the alkali charge before the disturbance reaches the digester heating zone. Slow feedback uses blow-line kappa and residual effective alkali to trim the charge setpoint over the next 1–2 h. Sample conditioning systems for black liquor are designed with high-velocity recirculation, self-cleaning filtration, and dilution to reduce fouling. Analyzer availability is critical; a failed black liquor alkali analyzer removes the only direct measurement of whether the cook still contains sufficient hydroxide to prevent condensation during the residual phase.

Chip column permeability and alkali channeling introduce axial and radial concentration gradients in continuous digesters that cannot be detected by blow-line kappa alone. If extraction screens are fouled or the chip size distribution shifts toward fines, liquor may migrate preferentially through regions of high permeability, producing overtreated zones with low viscosity and undertreated zones with high kappa. In two-vessel continuous digesters with an impregnation vessel, the chip plug is transferred through high-pressure feeders and steam heater assemblies; the transfer circulation and in-line drainers are used to separate air, turpentine, and excess condensate. Circulation flows in the cooking zone are generally maintained between 4 and 6 digester volumes per hour. If the circulation rate falls below this band, the alkali concentration at the chip surface can drop below the threshold required for efficient bulk delignification, while the chip center remains above target kappa. If the circulation rate is too high, dissolved lignin and acid fragments are mixed into the upper cooking zone before they can be extracted, increasing alkali consumption and accelerating condensation. Pressure diffusers or atmospheric presteaming vessels reduce nonuniformity by separating chip heating from chemical impregnation, but they do not remove the need for zone-specific alkali split control. The alkali split between impregnation and bulk cooking is manipulated by adding white liquor to the transfer circulation or to the top of the cooking zone, while extracting black liquor to hold dissolved solids below 15–20 g/L in the early delignification phase.

Diffusion, Dissolution, and Alkali Uptake in the Impregnation Zone

Alkali penetration into wood chips is governed by diffusion of hydroxide and hydrosulfide ions through liquid-filled pores, and by chemical reaction with carboxylic acid groups and acetyl substituents. Softwood chips with a thickness greater than 8 mm can remain undercooked at the center because the rate of diffusion through the chip is slower than the rate of delignification at the surface, especially when the cooking temperature is advanced too rapidly. Pre-steaming at 90–110 °C for 10–15 minutes removes air and raises chip temperature, which increases the diffusivity of alkali and improves liquor penetration. The impregnation vessel is typically operated at 110–130 °C and atmospheric or slightly elevated pressure; this temperature is high enough to soften lignin and increase ion mobility, but low enough to avoid extensive carbohydrate degradation before alkali distribution is complete. During impregnation, a substantial fraction of the effective alkali charge is consumed by neutralization of wood acidity and cleavage of acetyl groups. This means that the residual effective alkali entering the bulk cooking zone is lower than the total charge, and the cook must be designed so that sufficient hydroxide remains for bulk delignification. If the impregnation temperature is raised too quickly or the alkali split is front-loaded excessively, the bulk phase may begin with high ionic strength and low residual hydroxide, which reduces selectivity.

Modified continuous cooking systems address alkali depletion by withdrawing black liquor from multiple extraction screens and adding white liquor at several points along the digester. The dissolved lignin concentration is lowered in the bulk cooking zone, the alkali concentration profile is flattened, and the final temperature can be reduced without sacrificing kappa. This is achieved by splitting the effective alkali charge into two or more additions: a first addition in the impregnation vessel or top cooking zone, and a later addition in the lower cooking zone. Such configurations can reduce the H-factor required for the same kappa, but they also make effective alkali control more sensitive to extraction screen plugging and circulation flow imbalance. If an extraction screen located between the alkali addition points fouls, the lower addition point may not receive sufficient liquor, and the final delignification stage will stall. Pressure diffusers located between cooking zones can improve liquor exchange and remove degradation products, but their operation depends on consistent chip column formation.

When the Digester Kappa Band Narrows Below ±1.5 Units, What Changes in Control Logic?

Operators attempting to hold softwood blow-line kappa within ±1.5 kappa units of a 30 target cannot rely on manual adjustments based on one composite sample per blow. At 170 °C, a 0.5 °C cooking temperature error for 1 h produces a relative delignification rate shift of approximately 3–4%, which is sufficient to move kappa by more than 1 kappa unit in the final phase. The control logic must therefore include automatic H-factor target adjustment after each kappa measurement, with limit constraints that prevent over-temperature excursions into the residual phase where carbohydrate losses become severe. A model-predictive controller that includes a soft sensor for blow-line kappa from residual alkali, digester temperature, and chip feed can reduce variance more effectively than a proportional-integral loop on kappa alone because it accounts for the dead time and multivariable interactions. When the kappa band is tight, the white liquor charge is not simply reduced to lower kappa; instead, the temperature profile and alkali split are moved together so that the final residual alkali remains above 5 g/L as Na₂O. Attempts to reduce kappa by increasing H-factor without increasing alkali charge often produce unacceptable viscosity loss. Severe viscosity loss is usually measured as a limiting viscosity number below 800 mL/g, and may result in reduced pulp yield and poor runnability in the fiber line.

Anthraquinone and polysulfide are the two established additives that modify effective alkali demand in kraft pulping. Anthraquinone acts as a redox catalyst by oxidizing the reducing end of polysaccharides and being reduced to anthrahydroquinone, which then transfers electrons to quinone methide intermediates in lignin, promoting β-O-4 cleavage and stabilizing lignin fragments against condensation. Industrial addition levels are typically 0.03–0.10% on oven-dried wood; at these levels, the effective alkali charge can sometimes be reduced by 0.5–1.5 percentage points at constant kappa, but the effect is nonlinear with wood species and cooking time. Anthraquinone dispersions must not be acidified or blended with strong oxidants because agglomeration and loss of redox activity can occur. Polysulfide, generated by oxidation of white liquor sulfur species, increases sulfidity locally and promotes carbohydrate end-group stabilization, allowing a higher alkali charge or longer cook without excessive yield loss. Polysulfide sulfur is typically generated at concentrations corresponding to 2–4 g/L sulfur in enriched white liquor; its stability declines rapidly above 100–110 °C, so it is most effective when added to the impregnation stage before the main temperature rise. The interaction between polysulfide and anthraquinone is not always additive, and published data for highly sulfidic commercial cooking configurations is limited.

Residual Alkali and Viscosity Boundaries That Limit Downstream Oxygen Delignification

After the digester, residual effective alkali, kappa number, and pulp viscosity determine the operating window of the oxygen delignification stage. Oxygen delignification is typically carried out at 80–110 °C, oxygen pressure 0.6–1.0 MPa, and alkali charges of 1.5–3.5% NaOH on oven-dried pulp, with magnesium sulfate addition to protect pulp strength. If the digester blow-line kappa is above target and residual alkali is low, the oxygen stage may require excessive alkali and temperature, which can reduce viscosity below the bleach plant acceptance limit. If the digester residual alkali is high, it may be carried into the oxygen stage and reduce the required alkali addition, but it can also increase pulp extractive content and soap carryover. The control strategy therefore extends beyond the digester: blow-line kappa analyzers, residual alkali titrations, and in-line oxygen-stage pH sensors are used to adjust the oxygen alkali charge. The limiting viscosity number measured by ISO 5351:2010 is used as a boundary condition rather than a target; a pulp with high kappa and low viscosity has likely experienced an alkali profile imbalance that overcooked the carbohydrate fraction during the bulk phase. If the digester effective alkali charge is increased to compensate for low sulfidity or high chip fines, the viscosity response should be monitored at the next composite sample point. The oxygen stage can remove 40–60% of the remaining lignin under optimized conditions, but it cannot correct a digester that produces uneven kappa with low residual alkali.

Extended delignification cooking methods rely on a descending temperature profile near the end of the cook. If the effective alkali charge is increased too late in the residual phase, the cook may not gain additional kappa reduction but will accelerate carbohydrate degradation and increase dissolved solids load to the recovery boiler. The final residual effective alkali target therefore must be balanced against evaporator and recovery boiler dead load: sodium carbonate, sulfate, and thiosulfate accumulate in the black liquor cycle and reduce the apparent effective alkali available to the digester. Liquor inventory analysis using mill balance data is required to determine whether a low residual alkali value is caused by insufficient hydroxide or by excessive dead load.