Alkali Charge Uniformity Limits in Continuous Kraft Digesters

Continuous kraft digesters operate with a moving chip bed through which white liquor is distributed through multiple injection points, extraction screens, and circulation loops. Alkali charge uniformity in this context is not a single measured quantity but a spatial and temporal property of effective alkali concentration—defined as sodium hydroxide plus one-half sodium sulfide equivalent on an oven-dry wood basis—within the impregnation, cooking, and wash zones. In single-vessel hydraulic digesters, the upper section serves as an impregnation zone, while the lower section carries out bulk delignification; in two-vessel hydraulic systems, a separate impregnation vessel discharges preheated, alkali-impregnated chips through a high-pressure transfer line. Modified continuous cooking and Lo-Solids configurations superimpose counter-current washing and multi-point extraction onto the base hydraulic regime. Uniformity limits are derived from the observation that the rate of delignification is a first-order function of effective alkali concentration and a temperature-dependent function of H factor; therefore a local alkali deficit cannot be fully corrected by later alkali addition if condensed lignin structures have already formed. Mill operating data from Kamyr downflow, EMCC, and Lo-Solids digesters indicate that residual effective alkali at the extraction screens should remain above 8 g/L as sodium oxide for softwood furnishes, while kappa targets of 18–30 millilitres in kraft softwood are typically associated with cooking temperatures between 150 °C and 170 °C and white liquor total titratable alkali between 120 g/L and 160 g/L as sodium oxide. These figures represent industrial operating envelopes rather than universal limits, because chip species, chip thickness distribution, digester geometry, and sulfidity shift the acceptable range. Nevertheless, they establish the control band within which continuous digester alkali distribution must be held to limit kappa standard deviation and post-digester screen reject generation. Any deviation greater than approximately 0.5–1.0 percentage points of effective alkali charge can shift blow-line kappa beyond specification in a calibrated softwood line; the precise sensitivity depends on the local liquor-to-wood ratio and the degree of counter-current washing.

The physical origin of alkali nonuniformity lies in the competition between convective liquor dispersion and alkali consumption. Wood chips are non-uniform in thickness, moisture, and permeability; the chip bed in a continuous digester has a radial porosity distribution that is lower near the screen rings because finer particles accumulate there. Liquor introduced through central pipes and screen rings preferentially follows the path of least hydraulic resistance, so high-permeability channels receive a disproportionate fraction of the white liquor while compacted zones experience local starvation. The resulting concentration differences cannot be measured readily by a single extraction sample, because black liquor at an extraction screen is a mixture of liquid that has passed through the entire upstream bed, liquid that has short-circuited through annular gaps, and liquid that has been recycled from the circulation pump. Online conductivity and density sensors are therefore used in many mills to infer effective alkali, but these signals require calibration against titration data obtained according to TAPPI T 625 for white liquor and TAPPI T 650 for black liquor residual effective alkali. Deviation between the inferred value and the laboratory value is highest when the dissolved organic load in black liquor increases, because organic acids contribute to conductivity without contributing proportionally to effective alkali. A carefully planned sampling and calibration protocol, such as that outlined in TAPPI T 624, is therefore required to maintain measurement confidence. Full-scale measurements of radial concentration profiles inside the cooking zone are technically difficult to obtain; published data for local radial concentration profiles in operating digesters is limited due to the extreme pressure and temperature conditions and the inaccessibility of the interior chip bed. As a consequence, industrial uniformity is usually assessed indirectly through kappa distribution, screen rejects, residual alkali at multiple extraction locations, and pressure-drop trends across screens.

Chip moisture and chip size interact with alkali uniformity because the diffusion time for sodium hydroxide into the chip interior is proportional to the square of the chip thickness. Chips with thickness above 8 mm may leave the impregnation zone with a partially neutralized core even if the surrounding liquor has adequate effective alkali. Subsequent cooking causes the outer fiber wall to delignify faster than the core, creating kappa and strength gradients that are indistinguishable from local alkali deficits when measured at the blow line. Softwood mills therefore often screen chips using a chip thickness screen; the accepted oversize fraction depends on the digester residence time but is commonly kept below 5–10% by mass for continuous digesters. This chip quality boundary is not an alkali charge limit in itself, but it defines the maximum degree of alkali uniformity achievable under otherwise stable digester operation.

How Does the Impregnation Vessel Partition Acid Neutralization From Delignification to Buffer Alkali Demand?

In two-vessel hydraulic systems, the separate impregnation vessel operates at a temperature between 110 °C and 130 °C, a range selected to accelerate penetration of sodium hydroxide and sodium hydrosulfide into saturated steam-preheated chips while minimizing bulk delignification. During this stage, a significant fraction of the effective alkali is consumed by neutralization of acetyl groups and other acid groups released from hemicelluloses; measurements from mill audits indicate that the alkali consumed during impregnation can account for 20–40% of the total effective alkali demand in softwood pulping, depending on species and chip size. The remaining alkali enters the co-current and counter-current cooking zones where the temperature is raised to controlled cooking plateau conditions. A split of white liquor between the impregnation vessel, upper cooking circulation, and lower cooking circulation is maintained by flow controllers on each injection line. Uniformity breaks down when the impregnation vessel discharge consistency or transfer pressure fluctuates, because these changes alter the chip plug flow through the high-pressure feeder and introduce residence-time dispersion. In one class of production-scale failure, a partial plug in the top circulation screen causes the flow controller to increase white liquor to the lower cooking loop to maintain the extraction residual alkali, resulting in excessive peeling reactions in the upper zone and under-delignified chip cores in the lower zone. This type of compartment-level imbalance is not immediately visible in the blow-line kappa average, but it appears as an increase in shives and a reduction in pulp viscosity. Viscosity is commonly measured using ISO 5351-1:2010; strength properties are evaluated with ISO 1924-2:2020 or TAPPI T 494, depending on the mill specification. These test methods provide indirect confirmation of alkali maldistribution because localized under-cooking reduces fiber wall uniformity and increases weak fiber content.

The buffering effect of the impregnation vessel is not unlimited. If the temperature is allowed to drift above 135 °C, delignification begins before the acid-neutralization reactions are complete, and the alkali demand becomes highly non-linear because the two reaction families compete for the same hydroxide pool. The resulting local pH can fall below the threshold at which dissolved lignin begins to condense onto fiber surfaces, a condition that is difficult to reverse by subsequent alkali addition in the cooking zone. Published data for the exact pH threshold in black liquor is dependent on dissolved solids, temperature, and ionic strength, so no universal limit exists; however, industrial practice in softwood mills is to keep the impregnation extraction liquor pH above 12.5 as measured in a cooled, filtered sample. The more reliable diagnostic is the residual effective alkali at the end of the impregnation vessel, which is often maintained between 4 g/L and 8 g/L as sodium oxide for softwood two-vessel lines. When the residual effective alkali in the impregnation extraction falls below this band, the subsequent cooking zone receives chips with variable acid-neutralization state, and the variance in blow-line kappa increases disproportionately. This nonlinear relationship between low impregnation residual alkali and downstream kappa standard deviation is an established mill diagnostic. The stoichiometry of acetylation and neutralization reactions in softwood xylan is not constant; acetyl content in softwood ranges from about 1.0% to 1.8% on oven-dry wood, and each acetyl group hydrolyzed consumes one molar equivalent of sodium hydroxide. This consumption occurs before substantial lignin removal because the hydrolysis of acetyl moieties has a lower activation energy than the β-O-4 cleavage that dominates delignification. Consequently, the effective alkali concentration in the impregnation vessel falls rapidly during the first portion of the retention time, and the resulting concentration profile is steep. Uniformity in this region is managed by maintaining a large enough liquor-to-wood ratio and by distributing impregnation liquor through multiple points, but even then the residual effective alkali at the impregnation vessel extraction cannot be used as a standalone indicator of alkali uniformity because it represents a flow-weighted average of liquor that has passed through chips of different ages.

Operating Limits Recorded From Production-Scale Two-Vessel Hydraulic and Modified Continuous Cooking Systems

Production-scale control systems for continuous kraft digesters are commonly tuned around a set of correlated variables: effective alkali charge, sulfidity, white liquor TTA, residual effective alkali, liquor-to-wood ratio, H factor, extraction differential pressure, and circulation flow. The following table summarizes monitoring parameters that mill laboratories and distributed control systems use to maintain alkali uniformity. The standard designations are those most frequently referenced in mill operating procedures, although test method revisions may be updated by the issuing bodies.

In practice, the tightest control limit is often the residual effective alkali at the extraction screens, because this variable integrates all upstream distribution errors. Mills with two-vessel systems typically monitor the difference between the upper extraction and lower extraction residual effective alkali values. When the lower extraction residual alkali is more than 3 g/L lower than the upper extraction value, a flow imbalance or channeling episode is suspected, and operators reduce production rate or adjust white liquor split before kappa variation becomes severe. This differential limit is site-specific and depends on the number of extraction points and the degree of counter-current flow; it is not a universal standard but has been reproduced in multiple mill audits as an early-warning threshold. The liquor-to-wood ratio also limits uniformity because at ratios below approximately 3.0, the volumetric capacity of the moving liquid phase is insufficient to homogenize local concentration differences generated by reaction depletion. At ratios above 5.0, the additional pumping energy and thermal load produce diminishing returns and can increase the rate of fines migration to the screens. The practical result is a narrow operating window that continuous digesters must respect if alkali uniformity is to be maintained under production variability.

White Liquor Manifolds, Circulation Screens, and Radial Flow Maldistribution

The white liquor manifold in a continuous kraft digester distributes reagent to multiple injection rings along the vertical axis. Branches supply the top circulation, upper cooking screens, lower cooking screens, and, in modified units, the counter-current wash circulation. Radial maldistribution occurs when a screen ring does not withdraw liquor evenly across its circumference. A common mechanism is the accumulation of fiber fines and extractives on the internal side of a screen plate, which raises local pressure drop and diverts withdrawal to the unobstructed segment. The circulating pump then returns heated liquor to the central pipe, where it may short-circuit through the chip bed along the same low-resistance channel. Temperature sensors arranged radially in the circulation line can detect a split of more than 2–3 °C between screen segments, indicating that a segment is partly plugged. Because the local alkali consumption rate depends on temperature through the Arrhenius relationship for delignification, a temperature difference of 2 °C can produce a measurable shift in local kappa. Mills therefore treat uneven screen-ring temperature as an actionable maintenance trigger, and scheduled outages include internal inspection of screen slot width and clearance against the chip bed. Screen slot openings in continuous digesters are commonly between 2 mm and 5 mm, depending on the furnish; smaller openings reduce fines carryover but increase the risk of screen plugging. Published data for slot-specific plugging frequencies across all digesters is limited because manufacturers classify internal inspection records; however, the operational pattern of rising extraction differential pressure and increasing radial temperature variation is well documented in mill reliability reports.

Corrective action for radial maldistribution usually includes increasing circulation flow, backflushing screens with hot white liquor, reducing extraction rate to decrease screen velocity, or changing the chip feed screen at the chip chute. A single-sentence operating rule for well-established continuous digester practice is that white liquor and extraction flow meters must be calibrated quarterly against master meters to maintain split-ratio accuracy.

In-line monitoring of alkali uniformity depends on the ability to distinguish transport-related concentration noise from true process drift. Conductivity probes installed in extraction lines respond to total dissolved ionic species, not specifically to hydroxide and hydrosulfide, so the signal includes contributions from carbonate, sulfate, thiosulfate, chloride, and dissolved organic acids. For this reason, a conductivity-based effective alkali estimate is considered reliable only after calibration against laboratory titration over the expected black liquor dry solids range. Coriolis flow meters are used for white liquor and black liquor service because they provide density and temperature in addition to mass flow, allowing the control system to calculate sodium oxide mass flows from volumetric additions. Density meters on white liquor lines are calibrated against TAPPI T 625 titration data to convert density to total titratable alkali concentration, but the correlation is affected by sulfidity and the presence of suspended solids. Periodic grab samples for sulfidity and TTA are therefore necessary, and the sampling points should be located downstream of mixing tees to avoid stratified samples. Automatic titrators reduce operator bias, but they require routine verification against primary standards and are subject to sulfide oxidation if sample handling is delayed. The sampling protocol in TAPPI T 624 specifies precautions for black liquor samples, including closed sample containers and minimized exposure to air. In addition, pH probes used in hot black liquor service have limited lifetime due to the alkaline sulfide environment and scaling; a redundant installation with automatic potassium chloride re-filled reference junctions is typical in mills that close-loop pH control. These instrumentation choices do not guarantee uniformity, but their absence removes the early warning capability required to detect a developing alkali imbalance before kappa and screen rejects are affected.

When Flash Tank Pressure Deviates by More Than 5 kPa From the Cooking Header Setpoint

In two-vessel hydraulic continuous digesters, the transfer line between the impregnation vessel and the cooking vessel includes a high-pressure transfer device and, in many layouts, a flash tank or pressure control valve that separates the cooking circuit from the impregnation circuit. The pressure in this transfer zone influences the flow of liquor that carries chips from the impregnation vessel to the top of the cooking vessel. If the flash tank pressure deviates by more than 5 kPa from the cooking header setpoint, the chip plug flow becomes unstable, causing momentary interruptions in the transfer line. These interruptions alter the local alkali environment because the white liquor that normally accompanies the chips into the top circulation zone is delayed or partially diverted. Mill incident reports describe a characteristic pattern: the upper cooking circulation temperature falls first, followed by a rapid rise in extraction residual effective alkali because the alkali charge is not being transported into the cooking zone, and then a rise in blow-line kappa after one residence time. The time delay between the pressure deviation and the kappa response depends on digester volume and production rate, but it is typically on the order of 30–90 minutes in full-scale softwood systems. This delay makes manual correction difficult, and therefore the transfer pressure control loop is often integrated into the digester advanced process control strategy with feedforward from the high-pressure feeder speed.

The operational boundary associated with transfer pressure is not primarily a reaction limit but a hydraulic stability limit. A pressure deviation of 5 kPa may appear small relative to the cooking pressure of 800–1100 kPa gauge in many softwood digesters, but it is sufficient to alter the flash steam balance and the net liquor flow in the transfer line. The control action on alkali charge must then compensate not by increasing the total white liquor flow but by restoring the pressure and the transfer liquor flow before the local concentration deviation propagates downward. When the pressure deviation persists, operators sometimes reduce the circulation pump speed to protect the transfer line and screens, but this action further reduces radial mixing and accelerates alkali nonuniformity. The correct response sequence includes checking the flash tank level, verifying the strainer differential pressure, and comparing the top circulation flow to its historical baseline before adjusting the white liquor split. This sequence is derived from operating experience on two-vessel lines and is commonly embedded in mill troubleshooting procedures.

Counter-current washing zones in modified continuous cooking and Lo-Solids digesters reintroduce wash filtrate and external washer filtrate at the bottom of the vessel. The residual effective alkali in these zones influences the extent of continued delignification during washing and the precipitation of dissolved lignin onto fibers when temperature and pH drop. Uniformity limits in this zone are therefore defined by the need to keep the residual effective alkali above the precipitation threshold while avoiding excessive alkali that would consume fiber strength. The residual effective alkali at the bottom extraction is commonly specified between 8 g/L and 12 g/L as sodium oxide for softwood lines; values below 8 g/L are associated with extractive and lignin deposits on extraction screens, higher screen differential pressure, and a rise in screen rejects. Kappa distribution across the final pulp is assessed by taking multiple samples from the blow line over a full residence time; a standard deviation above 1.5 kappa units for a target kappa of 25 is frequently used as an internal mill limit indicating poor alkali uniformity or chip quality variation. Laboratory methods for kappa include TAPPI T 236 and ISO 302:2015; the two methods use similar oxidation chemistry but differ in sample size and calculation procedures, so mills must not mix them when evaluating trend data. Screen plugging thresholds are pressure-differential based: an extraction screen pair is considered at risk when its differential pressure increases by more than 10–20 kPa above the clean-screen baseline, depending on the screen design and slot width. The threshold is validated against visual inspection during scheduled outages; published data for exact threshold values across all digester designs is limited because screen manufacturers provide proprietary design-specific curves.

The interaction between alkali uniformity and screen plugging is bidirectional. A plugged screen reduces local extraction, alters liquor distribution, and therefore reduces local alkali transport; the resulting low-pH region promotes precipitation that accelerates plugging. Conversely, a well-designed screen cleaning sequence can restore hydraulic uniformity before alkali distribution degrades. Continuous digesters with EMCC and Lo-Solids configurations therefore use an automated sequence of extraction screens and circulation screens, but the sequence must be tuned to avoid pushing fines into the chip bed. The tuning parameters include screen backflush duration, backflush flow, and the maximum allowed differential pressure before interlock. These values are not determined by a single standard; they are set during commissioning and reviewed after each outage based on screen condition and kappa variability data.