1. A plant-floor view: why the kraft process
Walking into a chemical pulp mill, the most eye-catching things are the tens-of-meters-tall vertical continuous digester, and beside it the roaring recovery boiler and rows of evaporators. The reason the kraft process (German for "strong") has thoroughly beaten the sulfite process over the past century to become the absolute mainstream of chemical pulping is not the cleverness of the cooking reaction itself, but its nearly zero-waste chemical-recovery loop — most of the alkali consumed in cooking can be recovered from the black liquor and recycled, so the larger the scale the cheaper and the greener it gets.
This episode breaks this main thread into four parts: first, the chemistry of the white liquor — what exactly dissolves the lignin in cooking; second, the three-stage delignification kinetics — why cooking must be temperature-controlled in stages; third, the H-factor, the key engineers use to compress wildly varying heating curves into one comparable number; and finally back to the recovery loop, to see how black liquor turns back into white liquor. Once these are understood, the settings of temperature, alkali charge and kappa number on the floor are no longer guesses but have a clear origin.
One caveat: the kraft process is extremely tolerant of raw materials — softwood, hardwood, bagasse and wheat straw can all be cooked, only the process window differs. All values below are given as common engineering ranges, and we keep the range wording throughout, never masking uncertainty with false precision — this is also basic discipline in field technical communication. By the way, precisely because kraft "can cook anything," it holds an overwhelming share of global chemical-pulp capacity; understanding it is almost equivalent to understanding the skeleton of the modern chemical-pulp industry.
2. White-liquor chemistry: the synergy of NaOH + Na₂S
The kraft process is named after its cooking liquor — the white liquor. Its active ingredients are a mixed strong-alkali system of sodium hydroxide (NaOH) and sodium sulfide (Na₂S). Mills conventionally convert all alkali quantities to sodium oxide (Na₂O) for accounting, so as to have a uniform basis across different sodium salts. The so-called effective alkali (EA) usually refers to the total of NaOH plus ½ Na₂S converted to Na₂O, with a typical range of 18–25% (based on oven-dry wood-chip mass). Note this "based on chip mass" basis; different mills sometimes report against "wood volume" or "oven-dry pulp," so when reading reports be sure to check the denominator.
Another core parameter is the sulfidity, defined as the ratio of sulfur in Na₂S to (NaOH sodium + Na₂S sodium), generally 25–40%. Why add sulfur to the alkali specifically? Because the HS⁻ ion penetrates wood chips better than OH⁻ alone and more selectively cleaves the α-aryl ether bonds in lignin. The result: at the same yield the end kappa number is lower, or at the same kappa number the yield is higher. This is especially pronounced for softwood. The cost is an extra sulfur cycle in the system — H₂S appears in the off-gas, requiring desulfurization and incineration; the environmental facilities cannot be omitted.
Before entering the digester, chips must first be prepared: chipping, screening and debarking, with the target of chip length 15–30 mm, thickness 3–8 mm and as high an acceptance rate as possible. Poor chip uniformity causes "undercooked" and "overcooked" chips to coexist in cooking — a hidden root cause of kappa-number fluctuation in many mills. The prepared chips and white liquor are heated in the digester to 155–175 °C and react under pressure for 1.5–4 hours (shorter for continuous, longer for batch cooking). After the reaction, the pulp together with the black liquor is flashed through the blow valve into the blow tank — this step both forcibly terminates the reaction and safely discharges the high-temperature, high-pressure stock.
On-site, white-liquor concentration is often reported in g/L (e.g., NaOH 100–120 g/L, Na₂S 25–40 g/L). When converting to Na₂O remember two factors: NaOH molar mass 40, Na₂O molar mass 62, so 1 g NaOH is equivalent to 0.775 g Na₂O; Na₂S molar mass 78, its sodium-to-Na₂O factor is about 0.808, while sulfidity looks only at the "sulfur-sodium" share. The DCS usually displays effective alkali (% Na₂O on chips) and sulfidity (%) directly, but the chemical-dosing operator must still be able to back-calculate the dosing rate of the chemical tank.
Raising sulfidity improves delignification selectivity, but too high (> 40%) increases sulfur loss and odor load; hardwood, having little lignin to begin with, often only needs the lower sulfidity limit of 25–30%.
3. Three-stage delignification kinetics
Viewed from the main line of lignin-removal rate, the entire cooking process can be very clearly cut into three stages — this is the most core engineering intuition of kraft kinetics, and also the starting point for all later "extended delignification" processes.
1. Initial delignification (80–100 °C)
The temperature has not really risen yet and the lignin backbone is essentially untouched; this stage is mainly the dissolution of wood extractives (resins, fatty acids, terpenes), plus a small amount of easily hydrolyzed hemicellulose and some lignin–carbohydrate complex (LCC) being removed. The proportion of lignin removed here is small but critically important — it clears the path for the subsequent liquor to penetrate into the chip interior. The kappa number drops only slightly in this stage, but without it the later liquor cannot get in and cooking becomes uneven.
2. Bulk delignification (135–175 °C)
This is the main battlefield. When the temperature crosses about 135 °C, the β-aryl ether and α-aryl ether bonds in lignin break on a large scale under the combined action of strong alkali and HS⁻, lignin fragments into soluble pieces and diffuses out of the chips. This stage removes 80–90% of the total lignin and is the key window that decides the cooking end point (kappa number). The end kappa of softwood kraft typically lands at 25–32, while hardwood, having inherently lower lignin (18–24%), reaches only 14–20. The kappa number itself is approximately proportional to the residual lignin content in the chips and is the most important handover indicator between cooking and subsequent bleaching.
3. Residual delignification (after the high-yield end point)
Trying to push the kappa number lower, carbohydrate degradation accelerates sharply — the "peeling reaction" of cellulose and hemicellulose rises in rate faster than delignification as temperature increases. So modern cooking generally holds the end kappa at a "good-enough-to-bleach" level (e.g., around 25 for softwood), leaving residual delignification to downstream oxygen delignification. This is the basic idea of Extended Delignification: defend viscosity in the cooking stage, then lower kappa in the oxygen stage — overall preserving strength while saving bleaching chemicals.
4. The H-factor: compressing the temperature curve into one number
Engineers need a single metric to characterize "cooking intensity"; otherwise, with every heating curve different, there is no way to compare horizontally which batch is cooked deeper. The H-factor is the tool for this — it is a temperature–time integral:
T is the absolute temperature (K); the exponential term is normalized to 1 at 100 °C (373.15 K); 16113 = Eₐ / R, activation energy Eₐ ≈ 134 kJ/mol, R = 8.314 J/(mol·K). Physical meaning: the entire heating–holding curve is converted into an equivalent 100 °C cooking time; the larger H, the deeper the cooking.
The prototype of the exponential term is the Arrhenius reaction-rate constant k = A·exp(−Eₐ/RT). Taking the reference temperature as 100 °C and setting k = 1 there yields the simple form above. In other words, the H-factor is essentially the accumulated value of "equivalent 100 °C cooking time." The typical target H for softwood continuous cooking is ≈ 1500; to achieve extended delignification pushing kappa below 20, H must be pushed above 2000, but viscosity (representing cellulose degree of polymerization) drops faster and strength is compromised.
In actual production, the H-factor is not computed to be "looked at" but used as a control variable: the DCS samples cooking temperature in real time, integrates the formula online to obtain the instantaneous H, and terminates cooking when the target H is reached. It compresses "ever-changing temperature curves" into one comparable metric and is the basis for the long-term stable operation of continuous cooking. When raw material batches change or chip moisture fluctuates, the operator's adjustment target is often H rather than some fixed holding time — this is the "kinetic intuition" most used on site.
5. Cooking equipment: the vertical continuous digester
The mainstream equipment of modern kraft is the vertical continuous digester (Kamyr type, from the Metso/Valmet lineage). It is a tens-of-meters-tall steel vertical vessel; chips and liquor are fed from the top and slowly descend by gravity, discharging at the bottom. A single line can reach 2000–4000 t/d of oven-dry pulp. Compared with the batch digester (50–200 t per batch), continuous cooking has lower unit capital cost, lower energy consumption and more stable quality, making it the first choice for large mills.
The continuous digester is roughly divided into several zones internally: the top impregnation zone lets chips soak up the white liquor, the middle heating zone raises the stock to reaction temperature, the lower cooking zone completes the main reaction, and the bottom diffusion-washing zone displaces the spent liquor from the pulp with dilute black liquor before blowing. The whole line relies on level and temperature cascade control, and the DCS shows the real-time profile of temperature and alkali concentration in each zone.
The improvement directions in recent decades are highly consistent — "lower kappa, preserve viscosity." A few representative routes:
- Two-stage cooking / MCC (Modified Continuous Cooking): the white liquor is split into two streams added at different positions; the first stage uses low alkali and high temperature for rapid delignification, the second maintains mild conditions to avoid excessive residual-stage degradation, thus preserving viscosity at lower kappa.
- Low-solids cooking (Lo-Solids): the dissolved solids in the cooking liquor are pushed very low, reducing the reverse inhibition of dissolved lignin on the reaction, giving higher viscosity and better yield at the same kappa.
- Isothermal cooking and Rapid Displacement Heating (RDH): both are variants of the same idea, aiming to make chips heat more uniformly and the reaction more controllable.
The daily operation of the continuous digester hinges on the word "stable." On the DCS one must watch several profiles: whether the top impregnation-zone temperature is up to mark, whether the temperature difference between heating and cooking zones is reasonable, and whether the black-liquor solids at the bottom diffusion washer are out of control. Once chip moisture or density fluctuates, the cooking temperature curve drifts, and the H-factor loop automatically extends the holding to compensate — but over-compensation damages viscosity. So what the operator truly manages is "keep the heat load as low as possible while H is met" — this is not solved by one knob but is the result of cascade regulation across impregnation, heating and washing. For new lines, besides the Kamyr type there are newer variants such as Compact Cooking; the core remains this logic of "stage-controlled alkali, viscosity-preserving kappa reduction."
6. The recovery loop: the lifeline of the kraft process
After cooking the pulp carries black liquor — an alkaline aqueous solution of dissolved lignin, hemicellulose degradation products, and inorganic chemicals (sodium, sulfur), dark in color and high in organic concentration. Direct discharge would both pollute and waste chemicals, so the kraft process must be equipped with chemical recovery. The black liquor is first concentrated by multi-effect evaporators to 65–80% solids and sent to the recovery boiler; in the boiler the organics burn to supply heat and by-produce steam for power generation, while the inorganic sodium salts become a melt (mainly Na₂CO₃ and Na₂S) that becomes green liquor after water quenching.
To the green liquor lime Ca(OH)₂ is added for causticizing: Na₂CO₃ + Ca(OH)₂ → 2 NaOH + CaCO₃↓. The resulting CaCO₃ is burned back to quicklime in the lime-recovery kiln and recycled, while the produced white liquor (NaOH + a little unreacted Na₂S) returns to the cooking section. The inorganic chemical recovery rate of the entire loop is >97%, and it is precisely this that makes the kraft process nearly unbeatable under economies of scale — the chemicals are almost free to buy, and part of the energy is self-supplied.
7. Deep dive: yield, energy consumption and black-liquor heating value
Yield: kraft pulp yield is 45–55% for softwood, slightly higher for hardwood, versus the sulfite process at 50–65% (higher yield but poor raw-material adaptability and heavy pollution). Each 1-percentage-point gain in yield is real, hard-cash wood savings.
Cooking steam: about 1.5–2.5 GJ/t pulp, lower end for continuous cooking; batch cooking trends to the upper end due to large per-batch heat loss. Steam consumption is strongly correlated with holding time and the H-factor setting.
Black-liquor heating value: the heating value of black-liquor solids is about 13–15 GJ/t solids (varying with lignin content) and is the source for the recovery boiler to achieve steam/power self-sufficiency. Softwood black liquor, higher in lignin, has a higher heating value and better energy self-sufficiency.
Wood factor: as noted in the briefs, 1 t of bleached softwood kraft needs about 4.5–5.0 m³ of wood (at density 400 kg/m³), eucalyptus pulp about 3.5–4.0 m³ — the difference behind this is precisely the lower softwood yield and smaller chip density.
Putting these numbers together, the kraft process's competitiveness is clear: its cooking-stage yield is not the highest, but it recovers almost all the chemicals and energy through recovery, so total cost is leveled out or even lower. This is also why kraft exceeds 90% of global chemical-pulp capacity — it may not be the most "wood-saving" process, but it is the most "closed" one.
8. The trade-off between batch and continuous cooking
Although continuous cooking is the mainstream for large mills, the batch digester still exists widely, especially in mills with small-to-medium capacity, multiple raw-material switches, or a need for flexible process adjustment. The trade-off between the two is essentially "scale vs. flexibility." Batch cooking is 50–200 t per batch, each kettle independently heating–holding–blowing; switching raw material or kappa target only requires changing the next batch's settings, with almost zero switching cost; the downside is large heat loss and steam consumption toward the upper end (close to 2.5 GJ/t), and quality fluctuation follows the batch. Continuous cooking amortizes unit cost through long-term steady state and is more stable in quality, but once raw material is switched, the waste-pulp volume during the transition is not small.
Whether batch or continuous, the prerequisite for cooking uniformity is the penetration of liquor into the chip interior. After chips are heated, the cooking liquor must diffuse centripetally against the air and moisture inside the chip, delivering alkali to every cell wall. Incomplete penetration produces an "undercooked core" — overcooked outer layer, undercooked core, and a widened kappa distribution. Engineering ensures this with three points: first, uniform chip size (length 15–30 mm, thickness 3–8 mm), so thin stock penetrates fast; second, sufficient liquor-to-wood ratio (liquor volume to oven-dry chip mass, commonly on the order of 3–5 m³/t) to fully surround the chips; third, enough "impregnation time" in the heating stage so penetration completes before the main reaction. Henderson's classic conclusion also points out that in the early cooking stage liquor penetration is often slower than the chemical reaction itself and is one of the rate-limiting steps.
The blow step deserves an extra word. The reacted stock is at 155–175 °C and high pressure; flashing through the blow valve into the atmospheric blow tank, this process is itself an "explosive" fiber separation that improves the subsequent pulp freeness. But the blow also carries away a large amount of heat; modern mills generally install blow heat recovery (e.g., a heat-recovery system using flash steam for chip preheating or black-liquor evaporation) to bring this energy back into the system. This is especially true for batch cooking — each blow is a concentrated steam release with higher recovery value.
There is also a frequently overlooked detail: determination of the cooking end point. Continuous cooking relies on the H-factor loop, while batch cooking mostly terminates by holding time or sampled kappa after heating ends. Either way, online/offline kappa detection is the hard handover indicator to downstream sections — only when the cooking stage holds kappa steady can the chemical consumption of downstream oxygen delignification and bleaching be controlled. This is also why we repeatedly stressed "good enough to bleach": each point kappa is lowered, the extra alkali and heat consumed in the cooking stage and the yield lost must be saved back in the bleaching stage to be worthwhile; overall it requires system optimization rather than single-stage extremism.
9. Deep dive: the kappa-number handover account with downstream
The kappa number is approximately 0.13–0.15 times the residual lignin content in the pulp (the coefficient varies slightly by raw material), so it is the only common "lignin inventory" language among the cooking, oxygen-delignification and bleaching sections. The cooking stage pushes kappa from the raw material's 80–100 down to 25 (softwood) / 14–20 (hardwood); oxygen delignification then drops it a further 35–50%; the final bleaching stage only has to deal with the residual small amount of lignin.
From the cost chain: each kappa point lowered in the cooking stage consumes more alkali and loses more yield (carbohydrate degradation); but the downstream bleaching stage consumes less ClO₂ or H₂O₂. The optimal cooking kappa is therefore not a fixed number but drifts dynamically with chemical prices, energy prices and environmental constraints — this is exactly the economic motive of Extended Delignification for moving "kappa reduction" to the oxygen stage: oxygen-stage chemicals are cheap and do not lose yield.
Field experience: if softwood kraft cooking kappa is reduced from 30 to 22, yield may lose 1–2 percentage points, but ClO₂ usage in the oxygen and bleaching stages drops in sync, and overall on large lines it usually nets a gain. This is also why modern new lines generally pull cooking kappa toward 20–25.
There are two more easily overlooked handover indicators: first, viscosity (commonly measured by CPA or ISO method, units dm³/kg or mPa·s), which represents the cellulose degree of polymerization and is the "ceiling" of strength potential; the deeper the cooking, the more viscosity drops, so "lower kappa, preserve viscosity" is essentially defending this viscosity ceiling. Second, remaining alkali (the free NaOH+Na₂S in the black liquor at the cooking end point) — too low remaining alkali means the reaction is "starved" and carbohydrate degradation runs out of control, while too high wastes alkali and adds to the recovery load; the typical end-point remaining alkali is controlled in the 5–15 g/L range (as Na₂O). Looking at kappa, viscosity and remaining alkali together is what truly counts as understanding a batch of cooking.
Key points of this episode (quick-number memory)
- White-liquor effective alkali 18–25% (Na₂O on oven-dry chips), sulfidity 25–40%; adding sulfur improves delignification selectivity.
- Cooking temperature 155–175 °C; end kappa 25–32 softwood, 14–20 hardwood — good enough to bleach.
- Three delignification stages: initial 80–100 °C, bulk 135–175 °C (removes 80–90%), residual stage with accelerated carbohydrate degradation.
- H-factor ≈ 1500 (typical softwood), activation energy Eₐ ≈ 134 kJ/mol, essentially equivalent 100 °C cooking time.
- Continuous digester single line 2000–4000 t/d; MCC / Lo-Solids both "lower kappa, preserve viscosity."
- Recovery-loop chemical recovery rate >97%: black-liquor evaporation 65–80% → recovery boiler → causticizing → white liquor.
- Cooking steam 1.5–2.5 GJ/t, kraft yield 45–55%, black-liquor heating value 13–15 GJ/t solids.
Chemical pulp trades "dissolving lignin" for yield and strength, but mechanical pulp takes another path: almost no fiber loss, using a refiner to mechanically tear wood into pulp, with yield as high as 85–96%, at the cost of staggering power consumption. Next episode we go from SGW stone groundwood all the way to CTMP, taking apart why TMP's two-stage refining can preserve long fibers, why specific energy (kWh/t) is the lifeblood of this line, and how the three elements of refiner gap, consistency and energy staging are tuned. The energy-for-strength account will be calculated for you next episode.