Top_Ten

�Paperi ja Puu – Paper and Timber Vol.89/No. 4/2007

Martin MacLeod

The top ten factors in kraft pulp yield

to determine the gross chemical composi-tion of the wood in use.

The chemical composition of wood is probably the primary variable in kraft pulp yield. Fig. 3 shows normal yields in conven-tional research-scale kraft pulping of spe-cies-pure chips to bleachable-grade kappa numbers versus their typical lignin contents in the wood. This relationship makes rea-sonable sense: the higher the lignin content – which will be mostly removed in pulping – the lower the pulp yield. It is remarkably accurate over a yield range of 42–55%: Pulp yield = - 0.69[Lignin] + 65.8 (r2 = 0.95). North American wood species are illustrat-ed in Fig. 3, but major commercial species elsewhere in the world will conform to this general picture.

AbstrAct

Kraft pulp yield depends on a plethora of fac-tors: the nature of the wood and the quality of the chips, the cooking recipe (especially the key independent variables – alkali charge, sulphidity, temperature, and kappa target), the pulping equipment, and so on. Here, the factors have been assembled into a “top ten” list, and are assessed in terms of relative importance, potential to influence yield val-ues, and contribution to practical knowledge of how pulp yields can be improved. the ten factors can be re-ordered at will, to rank the magnitude of the yield changes they can pro-duce, for example, or to see which factors have the highest potential for yield improve-ment at modest cost.

What are the principal factors affecting pulp yields in kraft mills? How comprehensive is our understanding of them? Are there practical ways to use exist-ing knowledge to improve yields?

To address these questions, here is a Top Ten list (Fig. 1) of the key factors to con-sider, followed by brief descriptions of why each is important, what the size of the yield gain might be, and how substantial and re-liable the information base is. The focus is on practical opportunities for yield gains in kraft mill operations, tying them to scientific knowledge of cause-and-effect relationships. The broad perspective is two-fold: how wood and chemistry interact in the kraft pulping process, and why uniformity of treatment (whether chemical or mechanical) matters. An anthology of papers on the subject of kraft pulp yield is also available /1/.

The ten factors have been assembled in the same order as fibrelines, i.e., from chips through pulping to bleaching. The order can be changed for different purposes, as will be obvious later when they are ranked for mag-nitude of potential yield gain and also for what is practical to do at modest cost.

Wood species

Wood, an organic raw material, consists of polysaccharides (cellulose and hemicellulos-es), lignin, and extractives. Their concentra-tions vary substantially among commercial wood species /2,3/: cellulose, approximately

40–50% of wood; hemicelluloses, 25–35%; lignin, 15–30%; extractives, 2–10%. The higher the polysaccharide content (especial-ly cellulose) and the lower the amounts of lignin and extractives, the higher will be the yield of pulp from wood. Aspen is a leading example – with lignin content often below 20% and (acetone) extractives below 3%, it cooks rapidly to the highest bleachable-grade kraft pulp yield in industrial practice, typically about 55% at kappa 12. Western red cedar, with an unusually high extractives content, is at the low end of the spectrum, providing a bleachable-grade pulp yield in the low 40s at kappa 30 /4/.

In commercial kraft pulping prac-tice worldwide, the typical yield range (unbleached pulp, in percent from wood) is about mid-40s to mid-50s for bleachable-grade hardwood pulps, and about 40–50 with softwoods (Fig. 2). We can widen the soft-wood range to about 60% by in-cluding linerboard basestock, the high-kappa end of the kraft pulp-ing spectrum. It is also possible to extend the lower limits of these ranges by invoking the use of saw-dust or fines (or decayed wood of any particle size).

Surprisingly for a worldwide industry which has been in busi-ness for many decades, there is no simple, fast, and cheap way

1 )noitisopmoc lacimehc( seiceps dooW

2 )serbif fo noitroporp( ymotana dooW

3 noitubirtsid ezis pihC

4 )ezis naht rehto( ytilauq pihC

5 )lanoitnevnoc( yrtsimehc gnipluP

6 yrtsimehc gniplup decnavda/deifidoM

7 smetsys retsegid lliM

8 gniplup dnoyeB

9 pihsnoitaler appak/dleiY

01 tsil hsiW

Fig. 1. These “top ten” factors in kraft pulp

yield are addressed in terms of their relative

importance, their potential magnitude, and their

reliability.

% ,dooW morf dleiY pluP tfarK

001080604020

sdoowdraH

sdoowtfoSedarg-elbahcaelB

1

Fig. 2. On a global basis, bleachable-grade kraft

pulp yields from hardwoods and softwoods fall

into these ranges. The softwood range can be

extended to 60% by including unbleached kraft

paper and linerboard grades.

WH/appak 51 tAWS/appak 03 ro

nepsA

hcriBhceeB

elpaM

eniP kcaJ

riF maslaB

kcolmeH E

radeC E

hcraL E

ecurpS

1

eniP yllolboL

04

54

05

55

06

5303520251

% ,dooW ni tnetnoC ningiL

Pul

p Y

ield

fro

m W

ood,

%

Fig. 3. There is a linear relationship between lignin content of

wood and probable yield of bleachable-grade kraft pulp.

� Paperi ja Puu – Paper and Timber Vol.89/No. 4/2007

Wood anatomyThe physical nature of wood also plays an important role in yield. Large differences exist among wood species, especially in percentage of “fibres” (the preferred cell type for papermaking) versus that of less desirable cells (e.g., ray parenchyma in soft-woods, vessel elements in hardwoods) /5/. This is compounded by large ranges in the principal wood fibre dimensions: length, diameter, and cell wall thickness /6/. For example, loblolly pine kraft pulp fibres can be five times longer than sugar maple fibres. Further, there are dimensional dif-ferences between earlywood and latewood, and between juvenile and mature wood. Of all of these, only fibre length distribution is routinely measured in the kraft pulping world.

From a papermaker’s perspective, a more appropriate concept might be the yield of papermaking fibres from wood. In this sense, different wood species offer very different potential yields. If only long, narrow fibres are sought, for example, then softwoods have a large advantage over hardwoods, in which wood anatomy is much more diverse (Fig. 4). But by acknowledging that hard-woods inevitably contain significant amounts of vessel elements, we can add them back since they are part of the pulp yield, bringing the hardwood cases much closer to the softwood ones. Still, there is a substantial amount of cell material in all woods that is not ideal from a papermaking standpoint.

We can generalize with the following observations:• The higher the percentage of long, nar-

row fibres (as opposed to any other cell types) in the wood raw material, the more uniform will be the pulping, en-

hancing the yield of pulp which is ideal for papermaking.

• The greater the range of wood cell types, the wider will be the dimensional ranges of length, width, and cell wall thickness in the raw material before pulping, and hence in the kraft pulp which is pro-duced.

• The anatomy of hardwoods is much more complex and – in some papermaking ways – adverse than that of softwoods.

Chip size distribution

In chip size, two things are clear – thick-ness is the principal dimension of concern in kraft pulping, and 2–8 mm thick chips are ideal /7/. Thickness distributions are routinely measured in chip classifiers, and modern chip thickness screening systems in mills are capable of controlling the thick-ness range reasonably well. Sadly, they of-ten don’t. Greater precision in chip making would help, whether during sawmilling operations or in log chipping. Undersized “chips”, although they pulp rapidly, carry a substantial yield penalty. With oversized chips, the danger is in generating rejects, inherently a penalty in mills producing bleachable-grade pulp whether the rejects are re-processed or are removed from the fibreline. If small wood particles can go to a dedicated, separate production line, and overthick chips are processed mechanically to make them more amenable to pulping, significant yield gains can be obtained when pulping only the properly-sized chips, on the order of 1–2%.

Fig. 5 illustrates two thickness distri-butions of same-species softwood chips on final delivery to two kraft digesters. The mill on the left achieves excellent control from a chip thickness screening plant with

disc screens and slicers. From pilot-plant pulping, these chips gave 46% pulp yield at 25 kappa when only the 2–8 mm frac-tion (containing 95% of the total mass) was cooked. Using mass fractions and reasonable assumptions to calculate the fractional yields shown in Fig. 5, the actual total yield from this chip furnish was 45.8%.

The older mill on the right had rudi-mentary chip screening and therefore a much broader thickness distribution. At 25 kappa, the penalties with the undersized and oversized fractions were more serious, bringing the total yield down to 45.1%. Note that with significantly less 2–8 mm material present, its fractional yield was ten percentage points lower.

A yield difference of 0.7% may seem rather small, but at a pulp production rate of 1000 tpd the older mill requires 12,000 t more wood (on an oven-dry basis) annu-ally. That can easily translate into a cost increase of a million dollars or more a year. The penalty will be worse when accounting for wasted volume in the digester occupied by overthick chips, higher alkali consump-tion, greater knotter rejects recycling costs, more shives going forward, less uniform pulp, and higher bleaching costs.

A chip thickness screening plant is a necessary part of a modern kraft pulp mill. But simply buying and installing such a plant is not enough – it must also be main-tained, tested periodically, adjusted, and improved.

Chip quality (other than chip size)

Many yield-related considerations fall into this category. In mixed-species chip fur-nishes, the proportions of the species, each with its own yield potential, will affect over-all pulp yield. Moisture content can in-

fluence yield values if green wood (rather than dry wood) is the basis for calculation; it can also affect the efficiency of pulping if the “reci-pe” changes (e.g., an unintentional change in alkali charge due to an un-seen change in wood moisture might penalize pulp yield). Mechanical damage to wood fibres can make them more susceptible to chemi-cal attack during pulping, lowering yield. Biological decay, bark, or the presence of biological knots and overthick chips in chip furnishes all impair pulp yield relative to fresh, sound wood of suitable thickness Any of these factors may represent only a small yield penalty; togeth-er, they may reduce pulp yield by 2–4%.

seniP ,riF D

securpS

dna serbiFstnemele lessev

tnetnoc doow fo % ,serbiF gnikamrepaP

001080604020

2

hcrib etihW

nepsA

mugteewS

serbiFylno

Fig. 4. If yield is defined on the basis of suitable

papermaking fibres, softwoods have an

advantage due to wood anatomy. Whether vessel

elements are considered “suitable” makes a

large difference in the hardwood results.

ssam latot fo % ,ssenkcihT pihC

0

02

04

06

08

001

mm8>mm8-2mm2<

3.6 1.337.5

1.54

0

02

04

06

08

001

mm8>mm8-2mm2<

2.1 7.349.0

8.54

latoTdleiY

%

y/noillim 2.1$ ~ = doow erom y/t 000,21 ~ = %7.0 = dleiY

3

Fig. 5. Maximizing the 2–8 mm fraction of a chip

thickness distribution can significantly improve

pulp yield.


A comprehensive examination of pulp yield with respect to chip quality was part of hanging basket experiments in a mill trial to implement Paprilox® polysulphide-anthraquinone pulping of hardwood in conventional batch digesters at Domtar’s Espanola, ON, kraft mill /8/: Four aspects were measured (Fig. 6):• Reference Chips: The removal of all bark,

knots, decayed wood, and heartwood provided ideal chips for kraft pilot-plant pulping, accounting for a 3% yield ad-vantage over the mill’s normal chips. The reference chips were made from the stemwood of middle-aged white birch logs of uniform growth chosen at the Espanola mill, and their thickness range was 2–6 mm.

• Best Mill Chips: When only the 2–6 mm thick fraction of mill chips was used in pilot-plant experiments, a 0.5% yield gain was measured relative to whole mill chips, whether in kraft or PS-AQ pulp-ing. The mill chips had an average thick-ness classification of 11% < 2 mm, 59% 2-6 mm, and 30% >6 mm. Obviously, removing 41% of the raw material is

not a practical thing to do, but decreasing the over-thick fraction substantially would help. Since the trial, chip thickness screening and overthick chip crushing have been installed on the hard-wood side at Espanola.• Pilot-Plant Pulping: Due to good chip pre-steaming practice, ideal temperature control, and homogeneity of impregnation and cooking in small research digesters, greater uniformity of pulping resulted in a significant yield advantage (1.5%) regardless of whether reference chips or mill chips were cooked.• Wood Species: Species analysis of basket pulps from mill “birch” chips showed that they actually contained 24% maple on average. Taking maple as one-quar-ter of the mass, and assign-ing this fraction a 2% yield penalty from wood relative to white birch /4/, a 0.5% yield deficit was calculated.

Overall, the four factors illustrated here added up to a potential yield gain of 5.5%, whether associated with the kraft baseline yield or with

the PS-AQ yield. Achieving best perform-ance in all of these factors significantly im-proves pulp yield.

Conventional pulping chemistry

Among the primary independent variables of kraft pulping, high alkali charge, low sulfidity, high maximum temperature, and high lignin content in the wood are the most dangerous for inferior yield, poten-tially reducing the value by several per-centage points. By contrast, the higher the cellulose-to-hemicellulose ratio in the wood, the better. Lower extractives content is also desirable. Liquor-to-wood ratio can affect yield in that it has a strong influence on pulping rate, and therefore the time dur-ing which the polysaccharides (especially hemicelluloses) are degraded by alkaline attack. Hardwood lignin is chemically dif-ferent from softwood lignin, and accounts for part of the reason why hardwoods often have higher pulp yields (and faster deligni-fication rates).

How the main independent variables of kraft pulping affect kraft pulp yield is clearly explained in Kleppe’s classic paper “Kraft

Pulping” /9/. Higher alkali charge decreases pulp yield at a given kappa number, all other factors held constant, both with softwoods and hardwoods (Fig. 7). For every 1% in-crease in effective alkali charge (NaOH ba-sis) with softwoods, there is a 0.15% penalty in yield. The problem is three times worse with hardwoods, due mainly to the higher proportion of hemicelluloses (especially xylans) and their susceptibility to alkaline attack.

An independent example with kraft pulping of aspen to 15 kappa showed these results: total yield of 55.6% at 11% effec-tive alkali, 54.4 % yield at 13.5% EA, and 52.8% yield at 17% EA. Thus, an increase of 6% effective alkali led to a yield loss of 2.7%, just as predicted (i.e., 6 x 0.45%).

Although not particularly important in industrial kraft pulping (the majority of which is done at or above 30% sulphidity), how sulphidity affects yield is informative. Again from Kleppe /9/, with birch (at a ka-ppa target of 25), the yield plateau at 54% comes at 30% sulphidity. At 0% sulphidity, pulp yield is about 50% instead, a deficit of 4%; note that pulping rate is much slower as well. With pine at 55 kappa number, the 51% pulp yield plateau is at ~40% sulphid-

niaG dleiY pluP fo stnenopmoC

54

74

94

15

35

55

Scr

eene

d Y

ield

, %

8.64

ENILESAB TFARK)spihc llim ,steksab gnignaH(

3.84

5.1+yrtsimehC ssecorP QA-SP

3.15

0.3+spihC ecnerefeR

3.35

5.1+gnipluP tnalP-toliP8.155.0+spihC lliM tseB

8.355.0+seicepS dooW 5.5 +

4

hcriB etihW

Fig. 6. Many aspects of chip quality and pulping

practice offer substantial yield benefits,

including original wood quality, removal of

fines and oversized particles, and uniformity of

impregnation and cooking.

Fig. 7. Alkali charge plays a major role in pulp yield

– the higher the charge, the lower the yield, due to

increased susceptibility of the polysaccharides to

alkaline degradation.

5

Tot

al Y

ield

, %

06

55

05

54

0907050301 011

rebmuN appaK

dexiMsdoowdraH

0.515.710.02

% ,AE

00108060402 021

rebmuN appaK

nrehtuoSeniP

0.51

5.71

0.02

% ,AE

dleiY pluP no egrahC ilaklA fo tceffE

06

55

05

54

Tot

al Y

ield

, %

45

25

05

84

05040302010

% ,ytidihpluS

65

)52 appak( hcriB

)55 appak( eniP

h2

h3

h1

h2h3

h1

dleiY pluP no ytidihpluS fo tceffE5

dleiY pluP no erutarepmeT fo tceffE5

A retsegiD B retsegiD

3.344.54 % ,dleiy latoT

1.2 % ,YT ni ecnereffiD

3.0 - spihc ot debircsA1.0 - AE ot debircsA

% ,xamT ot eud ssoL dleiY 7.1

Fig. 8. Sulphidity has a minor effect, providing

that it is at the plateau level of 30% or above

(this is true for the majority of kraft mills).

Fig. 9. Maximum temperature of cooking has a

major effect on pulp yield – although it speeds

up the delignification rate, it accelerates

polysaccharides degradation even more.


ity. At 0% sulphidity, pulp yield is 48%, a deficit of 3%. Again, the pulping rate de-creases significantly with lower sulphidity.

In both cases, then, pulp yield is direct-ly related to sulphidity, but not in a linear manner. Sulphidity needs to be at or above 30% for optimum yield and rate reasons.

The maximum temperature of pulping is also important for yield. In the case shown in Fig. 9, two chip furnishes from the same wood species were being delivered to two continuous digesters. They were pulped in a pilot-plant digester at process conditions taken from the two mill digesters (A: 18.5% effective alkali, 163°C maximum; B: 19.1% EA, 175°C max.). Case A had 86% 2-8 mm chips and 7% > 8 mm chips; Case B, 79% 2–8 mm chips and 14% > 8 mm chips.

The difference in total yield at kappa number 30 was 2.1%. When adjusted for the differences attributable to chip thickness distribution and applied effective alkali, the yield deficit due to the 12°C higher maxi-mum temperature in Digester B was 1.7%.

Modified pulping chemistry

The era of modified kraft pulping (origi-nally called extended delignification) which began in the 1980s was founded on chemi-cal principles intended to make kraft pulp-ing more selective for delignification over polysaccharide degradation. Combined with appropriate changes in mill digesters, some yield benefits have accrued. Liquor displace-ment batch systems can improve yield over conventional batch systems (as measured by hanging baskets) by 1–2% /10/. Continuous digesters with multiple white liquor inputs and black liquor extractions appear to of-fer a yield advantage – particularly with hardwoods – of up to 4% /11/. In general, however, evidence for a universal yield ben-efit with modified kraft pulping equipment is scanty.

Modifying kraft pulping with addi-tives (e.g., anthraquinone, or polysulfide, or both) can improve pulp yields by about 1–3%. The research knowledge is extensive and deep /12/, and both additives have been used for the past 30 years in mills scattered around the world. An obvious advantage with AQ is that it can work in all types of kraft digesters – no equipment changes are required. To achieve maximum benefits with AQ, its strategy of use needs to be based on optimizing all the key factors in kraft delignification, including alkali charge, sul-phidity, and kappa target. Fig. 10 shows an example /13/.

A recent implementation of PS-AQ pulping of hardwoods demonstrated that the change from kraft resulted in a yield gain

of about 2% whether measured by hanging baskets in the mill or in pilot-plant pulping using the chips and cooking liquors from the mill /8/ (see also Fig. 6).

Occasionally, an astonishing possibility emerges, such as alka-li sulphite-AQ pulping /14,15/. Although not in use industrial-ly because of its slow delignifica-tion rate and complex chemical recovery issues, AS-AQ pulping can provide yield gains of 5–10% (Fig. 11), depending on the sce-nario. No other industrially-feasi-ble process chemistry change can do better.

Mill digester systems

Digester equipment considera-tions can have a big influence on yield in kraft pulping. Especially important are the chip pre-steam-ing and liquor impregnation steps. Advanced batch and continuous digesters do an effective job of chip pre-steaming by providing enough contact time with atmospheric steam (15+ minutes), but most di-gester systems have either no delib-erate pre-steaming or not enough, even when it is a combination of atmospheric and low-pressure regimes. When air removal and water saturation of the inner void spaces in wood chips are inadequate, the result is a less-than-perfect liquid environment for pulping, leading to more heterogeneous delignification and inferior yield.

Good impregnation is always a key to good kraft pulping. It needs to be long enough (usually 30+ minutes) and at a low enough temperature (120° ± 5°C) to ensure that the liquid-phase chemistry is ready to begin everywhere inside the chips when they are taken to delignification temperature. Fig. 12 illustrates results from kraft pilot-plant experiments on two softwood sawdust fur-nishes from a mill operating M&D digesters /16/. The M&D operations were simulated by combining the sawdust and cooking liq-uor in bombs and driving the temperature to 185°C as fast as possible (~ 10 minutes). Even when starting with tiny sawdust-sized wood particles, plenty of rejects were gener-ated. But when we used conventional kraft conditions designed for chips, including a 90-minute ramp of 1°C/min to cook-ing temperature for graceful impregnation, the rejects decreased by about two-thirds, meaning that the screened pulp yield rose by 2%. This case shows that extreme im-

pregnation conditions can carry a significant yield penalty.

Pilot-plant experiments have also shown that if chips are thoroughly pre-steamed and impregnation with white liquor is done with good temperature control, then bulk liquor circulation through the cooking chip col-umn inside a steam-jacketed 20L digester is not vital in producing kraft pulp of high yield and quality. Forced liquor circulation in mill digesters is a means to try to over-come temperature and chemical concen-tration gradients created during filling and impregnation. It is no surprise that the best liquor displacement batch digesters have the lowest measured kappa variability in-side them /10/.

The era of modified kraft pulping has fostered longer and slower delignification in continuous digesters and more effective impregnation in liquor displacement batch digesters. Both provide an inherent advan-tage in selectivity (although the main benefit seems to be better preservation of cellulose integrity).

Together, all of these factors can improve pulp yields by several percentage points.

Fig. 10. Optimal anthraquinone’s effectiveness

as a kraft pulping additive depends on the

strategy of use vis-à-vis other primary variables

such as alkali charge and sulphidity.

WS dexiM

enoniuqarhtnA htiw niaG dleiY

Tot

al Y

ield

appaK03

enilesab tfarK

03

QA ddAH ecudeR

03

3

QA ddAAA ecudeR

8.51 1.71 1.71 % ,AA01.0 01.0 0 % ,QA0532 0002 0532 rotcaf-H

8.44 5.34 6.24 % ,dleiY

2.2 9.0 0 % ,niaG dleiY

1 2

6

WS WH% ,)kcotsnworb( niag dleiY 016 *

6+01+ rebmun appaK6 5% ,)dehcaelb( niag dleiY

ssenthgirb dehcaelbnU5302 OSI % ,niag

lacimehc gnihcaelB02~ 03~ % ,esaercni noitpmusnoc

?drocer dlrow a 81 appak ta %56 fo dleiy latot :nepsa morF *

QA-etihpluS enilaklA htiw niaG dleiY6

Fig. 11. Alkaline sulphite-AQ pulping offers

astounding yield gains over kraft, but the

process is burdened by slow a delignification

rate and chemical recovery is complex.


Yield beyond pulping

Three main considerations apply here: the chemical selectivity of oxygen delignification and chlorine dioxide bleaching, the uni-formity of the fibrous pulp passing through the chemical operations, and any physical losses of fibres in the progression of opera-tions along a fibreline.

The yield losses accompanying oxygen delignification and ECF bleaching are much smaller than those in pulping, offering less opportunity to improve yield substantially by process changes. But attention is required to avoid unnecessary mechanical degrada-tion of pulp fibres through these areas of a mill’s fibreline so as not to lose yield solely due to “leakage” of fibrous debris. Also, any recycles of unacceptable fibrous ma-

terials need to be minimized – they are proof of inadequate upstream process conditions, they add to processing costs, and they make the pulp less uniform. Common examples are knotter rejects (especially from biological knots) being recycled to digesters /17/, and final screen rejects being re-fined and recycled in bleach-able-grade mills.

It is instructive to examine the yield/kappa relationships of pulping, oxygen deligni-fication, and ECF bleaching together. Fig. 13 provides a generic softwood case.

For kraft pulping, the slope of a softwood line to ~30 ka-ppa is 0.15 ± ~0.05; for hard-woods to ~15 kappa, the slope is the same. Both are straight lines. With softwoods, the line represents the bulk delignifica-tion phase starting from about 100 kappa (the high-yield end of kraft pulping), and is a fact which can’t be changed eas-ily. The kraft case in Figure 13 is for a softwood with a pulp yield of 47% at kappa 30.

With oxygen delignifica-tion, the slope is about 0.10,

and extends down to perhaps kappa 15 be-fore beginning a steeper fall /18/. With final lignin removal, in theory the slope is about 0.05; this is chemically close to what ECF bleaching actually does. In all three cases, lower slope means better selectivity during lignin removal, the right direction for yield enhancement.

Several aspects of yield/kappa relation-ships need to be remembered:• There are non-linear conse-

quences for yield when either pulping or oxygen delignifi-cation is taken below its prac-tical kappa limit where the selectivity for lignin removal is lost.

• The yield gap widens in fa-vour of oxygen delignifica-tion over pulping as kappa number decreases.

• Raising the kappa target of pulping lifts the whole pic-ture to higher yield, notwith-standing the higher cost of removing residual lignin later in the process line.

Yield/kappa relationshipThe typical yield/kappa relationship for kraft pulping (as illustrated in Fig. 13) requires some caveats. There is, of course, a yield in-tercept which is strongly related to wood spe-cies, chip size, and pulping conditions. The straight line represents the bulk delignifica-tion phase, which covers almost the whole kappa range of commercial kraft pulping from high-kappa linerboard base stock to bleachable-grades.

Fig. 14 amplifies the meaning of a spe-cific yield/kappa relationship. This is a spruce/pine/fir case in which pilot-plant kraft pulping of 2–8 mm thick chips was done at five H-factors (the highest one was duplicated). Because the fibre liberation point with softwoods is at about kappa 40, screened yield equals total yield at all but the highest kappa level. Three linear regressions can be calculated:• For all six total yield values, total yield =

0.12(kappa) + 41.3 r2 = 0.95• For the highest four yields, total yield =

0.11(kappa) + 42.0 r2 = 0.94• For the lowest three yields, total yield =

0.22(kappa) + 38.9 r2 = 0.98This demonstrates that where you stop

kraft pulping has a significant effect on pulp yield. For bleachable grades, the idea is to aim for the end of the bulk delignification phase without falling into the residual phase. Being seduced by ever lower kappa numbers prior to oxygen delignification or bleaching has its price!

With hardwoods, only bleachable-grade pulp is made, and the entire kappa range is about 12–18, so there is much less room for unintentional overpulping. The use of an excessive alkali charge is the greater risk.

At the low-kappa end, the onset of the residual delignification phase will begin to increase the slope rapidly, sacrificing yield

noitangerpmI retteB htiw stcejeR rewoL

tsudwas WS

Scr

een

Rej

ects

, %

0

5.0

0.1

5.1

0.2

5.2

0.3

5.3

0.4

tfarKD&M

noisreV

tfarKlanoitnevnoC

noisreV

QA-tfarKD&M

noisreV

QA-tfarKlanoitnevnoC

noisreV

YS %8.1+YS %0.2+

A B

7

8

gnipluP tfarKstniop 4 tsehgih rof RL

r 1.04 + appak32.0=YT 2 99.0=

gnipluP dnoyeB pihsnoitaleR appaK/dleiY

040302

rebmuN appaK

04

24

44

64

84

05

01

Pul

p Y

ield

, %

noitacifingileD negyxOstniop 5 tsehgih rof RL

r 2.44 + appak90.0=YT 2 99.0=

)ylno ningil( laciteroehTstniop 5 rof RL

r 2.54 + appak60.0=YT 2 99.0=

Fig. 12. Even with sawdust-sized wood particles,

inferior impregnation conditions lead to

excessive rejects; good pre-steaming and

conventional impregnation significantly reduce

rejects generation, translating it into higher

pulp yield.

Fig. 13. The yield/kappa lines of kraft

pulping, oxygen delignification, and ECF

bleaching have progressively lower slopes,

hence greater selectivity for lignin removal

over polysaccharide degradation. The kraft

pulping and oxygen delignification lines enter

danger zones below about kappa 20 and 15,

respectively.

deneercSdleiY

latoTdleiY

stniop 3 tsewol rof RLr 9.83 + appak22.0=YT 2 89.0=

eniL appaK/dleiY fo epolS

stniop YT 6 lla rof RLr 3.14 + appak21.0=YT 2 59.0=

9

51 55545352

rebmuN appaK

04

24

44

64

84

05

Pul

p Y

ield

, %

stniop 4 tsehgih rof RLr 0.24 + appak11.0=YT 2 49.0=

Fig. 14. When a typical yield/kappa line for kraft pulping of

a softwood is separated into parts, it becomes clear that

seeking kappa targets below the high 20s inevitably sacrifices

yield by entering the residual delignification phase.


despite the further slow decrease in kappa number. Because the residual lignin is more resistant to delignification while the polysac-charides continue to degrade, the selectivity of kraft pulping becomes progressively worse – the slope of the line becomes steeper.

This relationship is a crucial aspect of every kraft pulping scenario, and it should be known for every mill operation. Often, that is not the case. To obtain accurate num-bers, such information is determined in research-scale pulping. It should be done routinely when any significant changes are made in chip furnishes and cooking reci-pes, including any proposed use of pulping additives.

Wish list

Although industrial kraft pulping practice has changed slowly and incrementally over the years, it is always useful to imagine how it could be made better, and by how much. Figure 15 lists some possibilities, from the far-fetched to the practical:• Lignin-free trees: In Factor 1, Fig. 3,

the linear regression suggests that the lignin-free case has a Y-intercept of 66%,

far higher than any kraft pulp yield cur-rently obtained commercially.

• Extractives-free trees: The same general argument applies. Because there is no great business in by-products from ex-tractives any more, it would be nice to avoid dealing with extractives at all.

• Hardwoods without vessel elements: The wood would be denser, providing higher pulp yield per unit volume of digester space, and the pulp would be more uniform, allowing improvements in stock refining, papermaking, coating, and printing.

• Chip thickness screening: Most CTS plants don’t come close to their origi-nal specifications for segregating and controlling chip dimensions, nor work consistently well in cold-weather loca-tions. Overthick chip processing spans the range from very good to abysmal /17/.

• Working knowledge: Training of digester operators is not as good as it should be (especially in North America). There is usually no certification of personal knowledge of the chemistry of pulping, so digesters tend to be treated foremost as mechanical entities. Is this satisfactory for the operation of chemically complex systems worth upwards of $100 million that produce tens of billions of dollars worth of pulp per year? Standards are much stricter in many other lines of work, including regular continuing ed-ucation plus re-testing. Why not in our business?Having assembled this Top Ten list for

kraft pulping yield, it is possible to rank the factors in a variety of ways. Fig. 16 does this based on magnitude of yield gain. For example, a bleachable-grade kraft swing mill could gain 14% going from the lowest softwood yield to the highest hardwood one (Figs. 2 and 3). No mill has the wood basket to do this. But in the northern boreal for-est zone, a 7–8% yield gain is routine when going from spruces to aspen. The same is true in hardwood mills going from maples to aspen.

Alkaline sulphite-AQ pulping has been done industrially, but only briefly and confined to two mills. In the right circumstances, its use in linerboard pro-duction could be interesting from a yield perspective. Unfortunately, slow pulping rate and complex chemical recovery are serious hurdles to overcome.

Most of the opportunities in Fig. 16 provide yield gains of 3% or less – not so exciting, perhaps, but feasible and operating in some mills. In fact, there

Fig. 15. Substantial yield improvements would

come from all of these items. While the first

three remain intractable, the last two are

possible today.

seert eerf-ningiL

seert eerf-sevitcartxE

stnemele lessev on htiw sdoowdraH

snoitacificeps ot mrofrep hcihw stnalp STC)stidua raluger eviecer dna(

gniplup tfark fo egdelwonk gnikrow lacitcarPsrotarepo retsegid rof noitacifilauq a yrtsimehc

tsiL hsiW trohS A01

seiceps dooW%41WH ot WS%8 WS ot WS%7 WH ot WH

tfark lanoitnevnoc .sv QA-SA%6 WS%01WH

tfark lanoitnevnoc .sv QA-SP%3 WS%3 WH

noitacifingiled negyxo ddA %2

noitangerpmi evorpmIytimrofinu gnikooc dna %2

egnahC fo edutingaM

2

7

6

8

1

6

srotcaF

Fig. 16. When ranked according to magnitude of

potential yield gain, the top ten factors emerge

in this order. Very few options offer individual

gains above 3%.

esahp noitacifingiled laudiser fo tuo yatS

tnalp STC morf ecnamrofrep lluf teG

erutxim a ni seiceps tseb rof ezimitpO

xamT ,S ,AE rof epicer gniplup ezimitpO

QA ddA

semiger noitangerpmi ,gnimaets-erp evorpmI

tsoC tsedoM tA oD oT lacitcarP

9

3

2

5

6

7

1

srotcaF

are a lot of opportunities which can de-liver 1–3% yield gains: additives such as anthraquinone and polysulphide, moving to advanced modes of digester operation, oxygen delignification (especially with a higher kappa target after pulping), and close attention to the quality of chips being fed to a digester. It is also good to have a strong command of existing knowledge and apply it to the technical details of good kraft pulp-ing practice.

Enhanced yields can also come from bet-ter chip making and dimensional control, improved pre-steaming and impregnation practices, cooking at lower temperatures for longer times wherever possible, mini-mization of rejects from pulping (and the re-processing of them), efficient fibre spill collection, and tight process control of oxy-gen delignification and bleaching. Research demonstrates that impressive, cumulative yield gains are possible.

Finally, Fig. 17 is an attempt at reality – what can you do in a kraft mill to improve pulp yield at modest cost with the equip-ment you have today? The items are listed in order of increasing cost:• Get out – and stay out – of the residual

delignification phase.• Make your CTS plant perform to maxi-

mize the 2–8 (or 9 or 10) mm thick fraction. Minimize the fines going to pulping, and deal effectively with the (small) fraction of overthick material. Buy or make chips with a narrower dis-tribution of thickness.

• Push continually to increase your best species for yield. Know the real num-bers by species from R&D work done on your wood sources.

• Make sure that your alkali charge and maximum temperature of cooking don’t creep too high, or your sulfidity too low. Process creep can occur over the long term, and current process targets may lose their connections to the original reasons for change.

Fig. 17. When ranked according to what is practical

to do at a modest cost, the top ten factors offer

plenty of opportunities for improvement.


• Anthraquinone? It is probably the sim-plest quick fix for yield gain if you can afford it. Don’t waste it by adding too much, losing some of it in an early black liquor extraction, or failing to recognize trade-offs with other primary factors such as alkali charge, sulphidity, and kappa target.

• Do anything you can to improve chip pre-steaming. Optimize impregnation by ensuring that the ingredients you put in your digester are the best you can provide. Don’t exceed what the chemis-try can actually do.

• And if the opportunity comes, go to an advanced batch or continuous digester system and advanced oxygen delignifi-cation.Happy kraft pulping!

References1. Kraft Pulp Yield Anthology (CD-ROM), 100 pub-

lished papers, 1990–2001, TAPPI, Atlanta, GA. 2. Gullichsen, J.: Fiber Line Operations, in Chemical

Pulping, Volume 6A, Papermaking Science and Technology, J. Gullichsen and H. Paulapuro, eds., TAPPI/Finnish Paper Engineers’ Association, At-lanta/Helsinki, 1999, Chapter 2, p. A27–28.

3. Process Variables, in Alkaline Pulping, Volume 5, Pulp & Paper Manufacture Series, 3rd edition,

Grace, T.M., Leopold, B., and Malcolm, E.W., eds., Joint Textbook Committee of the Paper Industry, CPPA-TAPPI, Montreal/Atlanta, 1989, Chapter 5, p. 82.

4. MacLeod, J.M.: Kraft Pulping: Connecting Theory to Industrial Practice, Notes of PAPTAC Kraft Pulp-ing Course, Session 1, Pointe-Claire, QC, October 23–25, 2006 (Typical Yields of Kraft Pulps).

5. Hakkila, P.: Structure and Properties of Wood and Woody Biomass, Volume 2, Papermaking Science and Technology, J. Gullichsen and H. Paulapuro, eds., TAPPI/Finnish Paper Engineers’ Association, Atlanta/Helsinki, 1998, Chapter 4, p.143.

6. ibid., p.141–150.7. Process Variables, in Alkaline Pulping, Volume 5,

Pulp & Paper Manufacture Series, 3rd edition, Grace, T.M., Leopold, B., and Malcolm, E.W., eds., Joint Textbook Committee of the Paper Industry, CPPA-TAPPI, Montreal/Atlanta, 1989, Chapter 5, p. 90–96.

8. MacLeod, J.M., Radiotis, T., Uloth, V.C., Munro, F.C., Tench, L.: Basket cases IV: Higher yield with Paprilox® polysulphide-AQ pulping of hardwoods, new Tappi J 1(8):3 (2002).

9. Kleppe, P.J.: Kraft Pulping, Tappi J 53(1):35 (1970).

10. Tikka, P.O., Kovasin, K.K.: Displacement vs. con-ventional batch kraft pulping: delignification patterns and pulp strength delivery, Paperi ja Puu 72(8):773 (1990).

11. Lebel, D.J.: Continuous Digester Operations, Notes of PAPTAC Kraft Pulping Course, Session 3, Pointe-Claire, QC, October 23-25, 2006 (Lo-Solids® Pulping).

12. Anthraquinone Pulping: a TAPPI PRESS Anthol-

ogy of Published Papers, G. Goyal, ed., TAPPI, Atlanta, GA, 1997, 600 pages.

13. MacLeod, J.M.: Improving kraft pulp yield with anthraquinone and polysulphide: science and strat-egy, 2002 Kraft Pulp Yield Workshop Preprints, TAPPI, Atlanta, GA, Session 6, Paper 6-1.

14. MacLeod, J.M.: Alkaline Sulphite-Anthraquinone Pulps from Softwoods, J Pulp Paper Sci 13(2):J44 (1987).

15. MacLeod, J.M.: Alkaline sulphite-anthraquinone pulps from aspen, Tappi J 69(8):106 (1986).

16. MacLeod, J.M., Kingsland, K.A.: Kraft-AQ pulping of sawdust, Tappi J 73(1):191 (1990).

17. MacLeod, J.M., Dort, A., Young, J., Smith, D., Kreft, K., Tremblay, M.-A., Bissette, P.-A.: Crushing: Is this any way to treat overthick softwood chips for kraft pulping? Pulp Paper Can 106(2):44 (2005).

18. Gullichsen, J.: Fibre Line Operations, in Chemical Pulping, Volume 6A, Papermaking Science and Technology, J. Gullichsen and H. Paulapuro, eds., TAPPI/Finnish Paper Engineers’ Association, At-lanta/Helsinki, 1999, Chapter 2, p. A146.

Martin MacLeod is a teacher, writer, and technical consult-

ant on kraft pulping. He can be reached at: 150 sawmill

Private, Ottawa, ON K1V 2E1 canada; phone + 1 613 526-

4798; e-mail [email protected]. this paper was

adapted from a presentation at the tAPPI Growing Pulp

Yield from the Ground Up symposium, Atlanta, GA, May

17, 2006.

184x133mm

Documents

Top_Ten