Wood Chips for Kraft and Sulfite Pulping

(1)

Wood Chips for Kraft and Sulfite Pulping

Evaluation of Novel Forest-Industrial Drum- Chipping Technology

Jessica Gard Timmerfors

Doctoral Thesis, Department of Chemistry Umeå University, 2020

(2)

Responsible publisher under Swedish law: the Dean of the Faculty of Science and Technology

This work is protected by the Swedish Copyright Legislation (Act 1960:729) ISBN: 978-91-7855-234-4

Electronic version available at http://umu.diva-portal.org/

Tryck/Printed by: KBC Service Center, Umeå Umeå, Sweden, 2020

(3)

Allt stort som skedde i världen skedde först i någon människas fantasi.

Everything great that ever happened in this world happened first in somebody’s imagination.

- Astrid Lindgren, 1958, reception of the H C Andersen Award

(4)

(5)

Abstract

Wood chipping and the supply of high-quality wood chips are of critical importance for most forest-industrial processes. The quality of wood chips affects product yield, product quality, and processability.

Wood chips from a novel type of forest-industrial drum chipper, with a large drum and specially designed wood-chip channels, were evaluated with regard to wood chips for the Kraft and sulfite

processes. Wood chips from a full-scale demonstration version of the drum chipper and from a conventional disc chipper at a Kraft mill were compared. The average bulk density and the fractions of oversized and overthick wood chips were similar, but the

demonstration drum chipper produced 51% more large accept chips, 11% more total accept chips, and 74% less pin chips and fines. A pilot-scale drum chipper based on the new technology was used to produce short wood chips designed for acidic processes. When the drum velocity was 30-34 m/s and the average wood-chip length 21-22 mm, the fraction of pin chips and fines was 4.2% and the fraction of total accept was 89-90%. When the average wood-chip length was decreased to 17 mm, the fraction of pin chips and fines increased to 8.5% and the fraction of total accept decreased to 80-82%. The pilot drum chipper was used to investigate the influence of using different tree species (aspen, birch, pine, and spruce), processing of wood with different moisture content, and frozen wood. For hardwood (aspen and birch), the fraction of total accept reached ~90% when the average wood chip length was 17 mm. The pilot drum chipper was also used to generate wood chips of heartwood of pine for a comparison of 15 sulfite-process reaction conditions that differed with regard to impregnation and cooking procedures. The analyses included

absorption of liquid in a specially designed impregnation reactor, pulp yield, reject, viscosity, kappa number, brightness, fiber properties, and chemical composition as determined using compositional analysis based on two-step hydrolysis with sulfuric acid and pyrolysis-gas chromatography/mass spectrometry. The results reveal in detail how the individual wood constituents were affected by the different treatments, and demonstrate the benefits of using a pressurized impregnation step prior to sulfite cooking.

(7)

List of Abbreviations

α Clearance angle/Pulling angle β Sharpness angle/Knife angle λ Complementary angle

ε Spout angle/Cutting angle ε’’ Infeed angle

CHP Combined Heat and Power CTMP Chemo-Thermo-Mechanical Pulp

DW Dry Weight

EA Effective Alkali HMW High Moisture Wood

HW Heartwood

LMW Low Moisture Wood

m³fub Cubic meter solid volume excluding bark m³sk Cubic meter standing volume

MCS Multi Channel Sweden AB MMW Medium Moisture Wood MS Mass Spectrometry

p Probability (level of significance in Student's t-test) SCAN-CM Scandinavian Pulp, Paper and Board Testing Committee

test methods for chemical (C) and mechanical (M) pulps and wood chips

SEM Scanning Electron Microscopy SL Setting Length

SW Sapwood

TMP Thermo-Mechanical Pulp

(8)

List of Publications

This thesis is based on the following four papers. In the text, the papers are referred to by their Roman numerals.

Paper I Gard Timmerfors J, Sjölund T & Jönsson LJ. New drum-chipping technology for more uniform size distribution of wood chips. Holzforschung 2020, 74, 116-122.

Paper II Gard Timmerfors J & Jönsson LJ. Evaluation of novel drum chipper technology: pilot-scale production of short wood chips.

TAPPI J. 2019, 18, 585-592.

Paper III Gard Timmerfors J, Salahi H, Larsson SH, Sjölund T

& Jönsson LJ. The impact of using different wood qualities and wood species on chips produced using a novel type of pilot drum chipper.

Manuscript submitted to Nord. Pulp Pap. Res. J.

Paper IV Gard Timmerfors J, Gandla ML, Sjölund T & Jönsson LJ. Evaluation of chipping and impregnation of Scots pine heartwood with sulfite cooking liquor. Manuscript.

(9)

Enkel sammanfattning på svenska

När vi producerar massa till papper och kartong vill vi använda så mycket av vedråvaran som möjligt, dvs. ha ett högt utbyte när råvaran konverteras till produkt. Det är viktigt för att minimera transporter, råvaruåtgång och kemikalieåtgång, vilket bidrar till en ekonomiskt bärkraftig process. Ett av processens första steg är att sönderdela stockarna till flis. Hur jämn flisen är i storlek, dess fukthalt och dess packdensitet påverkar hur bra massa det blir.

Idag huggs nästan all flis som blir massa med skivhuggsteknik.

Skivhuggen uppvisar en hastighetsgradient, vilket leder till att en större andel av råvaran blir till pinnflis och spån. Pinnflis och spån har för små dimensioner för att fungera bra som råvara vid impregnering med kokkemikalier och vid massakok.

En ny typ av trumhugg har utvecklats. Trumman har inte den hastighetsgradient som skivhuggen har och borde teoretiskt sett kunna producera en mindre andel pinnflis och spån. För att utvärdera den nya tekniken så jämfördes en fullstor demonstrationshugg baserad på den nya trumhuggstekniken med en konventionell skivhugg på ett

sulfatmassabruk. För att utvärdera effekten av trummans hastighet och stockens inmatningsvinkel användes istället en pilothugg baserad på den nya tekniken. Med hjälp av pilothuggen undersöktes också hur fliskvalitén påverkades av stockarnas fukthalt, vilket trädslag det var och om veden var fryst eller tinad.

Den nya trumhuggstekniken skapade flis som visuellt liknade den från skivhuggen. För att få bra massa behöver en andel på ungefär 85% av flisen vara acceptflis, dvs. dess dimensioner ligger inom de storleksintervall som industrin har valt att definiera som acceptabel.

Den nya trumhuggen gav en andel på 85% acceptflis direkt efter huggning i demonstrationsskala, ett värde som brukar uppnås först efter sållning av flis. Framför allt minskade andelarna av de fraktioner som kan ses som förlust, pinnflis och spån. Dessa småfraktioner stör massaprocessen och även om de kan användas som bränsle innebär det en värdeförlust. Bra huggresultat gick också att få med andra inställningar och varierande råvara. När flis som var betydligt kortare än normalt producerades gick som väntat andelen acceptflis ner, dock fortfarande inom gränser som är industriellt acceptabla. Detta pekar på att den nya trumhuggstekniken kan vara speciellt fördelaktig för sura processer där kortare flis är att föredra.

(10)

Den nya flisningstekniken användes också för att producera flis från kärnved av tall. En stor andel av vedråvaran i Sverige består av tall, men det är utmanande att använda en stor andel tallkärnved i sulfitprocessen. Flisen användes till försök där 15 olika

reaktionsförhållanden jämfördes. Skillnaden mellan de 15 olika reaktionsförhållandena bestod i hur impregnering och sulfitkok utfördes. Resultatet utvärderades genom analys av upptaget av impregneringsvätska i en specialkonstruerad impregneringsreaktor, bestämning av utbytet av massa, rejekt, viskositet, kappatal, ljushet och fiberegenskaper, samt detaljerad analys av råvarans och de 15 produkternas kemiska sammansättning. Försöket gav detaljerad information om hur veden påverkas under olika reaktionsbetingelser och visar tydligt de positiva effekterna av att inkludera ett trycksatt impregneringssteg innan sulfitkoket.

(11)

1. Background

In a global context, around 3300 million m³sk of wood is used every year. The largest use of wood is as fuelwood in developing countries (Gellerstedt 2009a). Around 45% of all felled trees are used by the industry, mostly in developed countries (Gellerstedt 2009a).

The total felling in Sweden during the season 2016 amounted to 70.7 million m³fub. In addition, 7.8 million m³fub wood was imported and 0.7 m³fub exported (Skogsindustrierna 2017). In 2016, the Swedish forest industry used 36.3 million m³fub wood for the saw mills and 35.3 million m³fub as pulp wood. From the saw mills, 9.9 million m³fub wood was transferred to pulp production.

Sweden is the world's 3^rd largest exporter of forest-industrial products (pulp, paper, and sawn timber). In 2018, the export value of forest-industrial products amounted to 124 billion SEK. In 2016, pulp production in Sweden amounted to 11.3 million tons

(Skogsindustrierna 2017). In comparison, around 184 million tons of pulp were produced globally during the same year (FAO 2019).

The focus of this thesis is on the first steps in the production of pulp. In most forest-industrial processes, the wood logs are chipped after debarking. The yield and quality of wood chips are important for the industry. The cost for the feedstock is an important part of the operating costs for the mills, and the quality of the wood chips that are fed into the processes is critical for the function of digesters and for the quality of the pulp.

1.1 Feedstocks in the forest industry

The total forest area in the world is probably around 4 billion ha.

Around 5% of the total forest area consists of plantations. The countries with the largest plantations are China, USA, and India (Skogsindustrierna 2020; Henriksson et al. 2009). More than half of the total land area of Sweden (69%) is productive forest land. This area corresponds to 23.6 million ha(Skogsdata 2019).

Woody species can be divided into softwoods and hardwoods.

Softwood comes from gymnosperm trees, i.e. conifers, and hardwood comes from angiosperm trees, i.e. broad-leaved trees (Wiedenhoeft 2013). The wood of many softwoods is softer than that of many hardwoods, but this is not always true. Whereas there are obvious visual differences between conifers and broad-leaved trees, the

(12)

fundamental differences between softwood and hardwood are apparent also at a cellular level.

The boreal forest, or taiga, stretches eastward from the

Scandinavian peninsula to Russia and northern China, and further on to Canada and northern USA. It is dominated by relatively few species of conifers (Henriksson et al. 2009). The temperate forests south of the taiga consist of a more even mix of conifers and broad-leaved trees, and the forest is more rich in different tree species. Tropical rain forests contain a multitude of tree species, and they are dominated by broad-leaved trees. Nevertheless, some conifers are abundant also in southern areas, such as sub-tropical regions. In plantation forestry, both broad-leaved trees, such as eucalypts, and conifers, such as pines, are common.

The Forest Act from 1903 has protected Swedish forests through mandatory replanting of forests after felling. There is more forest in Sweden now compared to when collection of data started in the 1920s.

The standing volume has increased with around 80%. Around 25% of the forest in Sweden is not used as productive forest land, and 9% is officially protected (Skogsverige 2020). Protected areas include 30 national parks, 5000 nature reserves, and 8300 wildlife conservation areas (Skogsindustrierna 2020). There are two certificates for

sustainably forestry: FSC (Forest Stewardship Council), with focus on sustainable forests and forestry, and PEFC (Programme for the

Endorsement of Forest Certification), with focus on sustainability and trackability from forest to product (FSC 1993; PEFC 2017).

According to data from 2013 (Sveaskog 2020), roughly 70% of the Swedish forest land was certified through FSC and/or PEFC.

The standing volume in Swedish forests (2014-2018) consisted mainly of Norway spruce (40.4%), Scots pine (39.3%), and birch (12.5%) (Skogsdata 2019). Other common trees species include aspen (1.7%), alder (1.7%), lodgepole pine (1.3%), and oak (1.3%).

Wood logs from typical long-rotation Swedish forestry are sold as timber, pulp wood, or fuel wood. The price paid for timber is higher than the price paid for pulp wood, which in turn is higher than the price for fuel wood. Therefore, it is advantageous for forest owners to sell their wood as timber, or at least pulp wood, rather than as fuel wood, if it is possible. It is the quality of the wood logs that

determines this. The separation of wood logs into different qualities is

(13)

usually done in the forest after harvesting. Factors that affect wood log quality include tree species, dimensions, straightness, time of storage after felling, occurrence of cracks, residual protrusions left after removal of branches, occurrence of forest rot at end surfaces,

occurrence of root bones, and occurrence of impurities, such as gravel, in wood and bark (Biometria 2019; 2020).

Fuel wood is typically used in big heat and power plants, but to some extent by households. Around 300,000 Swedish households used wood burners or fireplaces for heating their homes

(Energimyndigheten 2003). Apart from fuel wood, a minor fraction of the felling residues, such as branches and tops, are also utilized for energy purposes.

1.2 Wood structure and chemistry

The tree is a woody material, which can be seen as having three main parts: the branches and twigs, the stem, and the root system (Henriksson et al. 2009). It is the stem that is used in pulp mills and biorefineries. Some industries use feedstocks from non-wood

materials. For example, paper from straw of wheat and rice is made in China, and cotton-based paper is used for making paper money.

Tree stems consist of a pith, wood (xylem), cambial zone, and bark.

However, the innermost part, the pith, is very small, and the bark is removed during the debarking process (Sjöström 1993).

The cambial zone (cambium) is a thin layer of cells outside the wood, but inside the inner bark. It consists of living cells and is the growing zone generating new tissue (Sjöström 1993).

Around 15% of the dry weight of the tree consists of bark. The bark can be divided into outer and inner bark (phloem) (Brännvall 2009b;

Sjöström 1993). The outer bark is dead tissue and serves as protection for the tree, whereas the inner bark transports water and nutrients. The composition of bark differs between species. The fibers in the bark consist mostly of cellulose, hemicellulose, and lignin, as fibers in the wood. The bark also contains cork cells, which die early and which resist water and gas, and parenchyma cells, which store nutrients.

Generally, the bark contains high fractions of extractives and minerals compared to the wood.

(14)

1.2.1 Wood structure

The wood can be divided into heartwood and sapwood. The heartwood forms the inner part, outside the pith. It often identified as darker than the outer part of the wood, which is the sapwood. It is called sapwood as it distributes the sap through the tree.

In boreal and temperate forest land, the growth activity of the trees varies over the season. This results in growth rings in the wood. The springwood (earlywood) is formed during periods of rapid growth, whereas summerwood (latewood) is formed during later seasons.

Earlywood and latewood differ visually, but also with regard to mechanical properties and physical structure (Browning 1967;

Rydholm 1965). For example, there are typically large differences in density. Earlywood has a density of 250 kg/m³ and latewood has a density of around 750 kg/m³ (Hartler 1990).

There are significant differences in mechanical properties between species, but also between trees of the same species. Not even the wood from the same tree is of uniform quality. There is a mechanical

variation between the bottom and top of the stem, and also between the center and the periphery of the log (Twaddle 1997). This variation in wood quality and the angle between ring orientation and knife create differences in the relationship between wood chip thickness and length.

Most of the cells in the tree are elongated cells, oriented in the direction of the stem (Fengel & Wegner 1989; Henriksson et al.

2009). The cells transport and store liquids, nutrients, and resins. The name and structure of the cells differ between softwoods and

hardwoods (Henriksson et al. 2009). Softwood has a relative simple structure compared to hardwood. Softwood contains 90-95% tracheid cells, which provide mechanical strength and transportation of liquid, and 5-10% parenchyma cells, which are involved in transport and storage of nutrients (Fengel & Wegner 1989; Henriksson et al. 2009).

Hardwood consists mostly of libriform fibers, which provide

mechanical strength, vessels, which are involved in transport of liquid, and parenchyma cells, which are involved in transport and storage of nutrition. Hardwood typically has shorter wood fibers than softwood (Rydholm 1965).

(15)

1.2.2 Wood constituents and chemistry

The main constituent of fresh wood is water. The water can be free water in the cell wall and the lumen, and bound water in the cell wall.

The moisture content of the wood consists of the total amount of free and bound water. The moisture content also varies between heartwood and sapwood (Rowell 2013).

Dry softwood typically contains 40-45% cellulose, 25-30%

hemicelluloses, and 25-30% lignin. For hardwood, the corresponding values are 40-45% cellulose, 30-35% hemicelluloses, and 20-25%

lignin (Henriksson & Lennholm 2009).

Cellulose is a long linear polymer consisting of β-linked D-glucose units (Henriksson & Lennholm 2009). Cellulose typically has a high degree of polymerization, and may consist of up to 15,000 glucose units. This makes cellulose to one of the longest polymers in nature.

The secondary structure of cellulose is created by hydrogen bonds, which stabilize the chain, make it stiff, and form the basis of cellulose sheets (Henriksson & Lennholm 2009). The cellulose sheets are stacked next to each other and are held together with van der Waals bonds. The cellulose sheets form a long and thin fibril. The fibrils consist of many cellulose chains and can be up to 40 μm in length.

Hemicellulose chains are shorter than cellulose chains. They consist of around 200 units. The chains are branched and contain not only glucose units, but also other sugar units and sugar acid units (Teleman 2009). Common sugar units in hemicelluloses include D-glucose, D- mannose, D-galactose, D-xylose, and L-arabinose (Fig. 1). The most common hemicelluloses are galactoglucomannan, glucomannan, arabinoglucuronoxylan, arabinogalactan, and glucuronoxylan (Teleman 2009). Hemicelluloses are usually found between the cellulose fibers and form the bulk of the cell wall.

Pectins are sometimes classified as hemicelluloses, but usually not (Teleman 2009). The ability to gel comes from the pectins and they constitute only a few percent of the dry-matter of wood. Another common plant polysaccharide is starch, which consists of glucose units arranged as amylose (~20%) or amylopectin (~80%).

Lignin is a branched aromatic polymer consisting of phenylpropane units (Sjöström 1993; Ralph et al. 2004). Lignin is synthesized from monomeric precursors, referred to as monolignols. The most

important monolignols are p-coumaryl alcohol, coniferyl alcohol, and

(16)

sinapyl alcohol (Ralph et al. 2004; Henriksson 2009). p-Coumaryl alcohol, which lacks methoxy substituents, gives rise to p-

hydroxyphenyl (H) units in lignin. Coniferyl alcohol, which has one methoxy group, gives rise to guaiacyl (G) units. Sinapyl alcohol, which has two methoxy groups, give rise to syringyl (S) units. The composition of H, G, and S units differs depending on the biological origin of the lignin. Softwood lignin consists predominantly of G units, whereas hardwood contains SG-lignin. Lignin from grasses and herbs may contain a considerable fraction of H units.

Lignin has a more branched structure than cellulose and

hemicelluloses (Ralph et al. 2004). The phenylpropane subunits of lignin are commonly connected by ether bonds (b-O-4, a-O-4, and 5- O-4, g-O-a) or carbon-carbon bonds (b-5, 5-5, b-1, and b-b) (Ralph et al. 2004). The number of units in lignin is difficult to estimate, as preparation of lignin would typically disrupt its structure. It is a possibility that all lignin in a tree is one big molecule (Henriksson 2009).

The purpose of lignin is to give stiffness to the cell wall and glue the different wood cells together (Henriksson 2009). It also makes the wood hydrophobic and protects it against microbial degradation.

Extractives are small molecules in wood that can be extracted using various solvents (Björklund Jansson & Nilvebrant 2009). The content of wood extractives varies between softwood and hardwood, between different tree species, and between different trees of the same species.

Water-soluble extractives do not cause as much problems in pulp and paper production as lipophilic extractives (wood resin). Therefore, most work on wood extractives has had focus on resins (Björklund Jansson & Nilvebrant 2009).

Wood resins consist of fats, fatty acids, steryl esters, steroids, terpenoids, and waxes (Björklund Jansson & Nilvebrant 2009).

Extractives also contain phenolic constituents, such as stilbenes, lignans, hydrolyzable tannins, flavonoids, and condensed tannins (Sjöström 1993). Pinosylvin and pinosylvin monomethyl ether (Fig. 1)

(17)

are stilbenes and well-known constituents of heartwood of pine (Sixta 2006).

Fig. 1. Common monosaccharides derived from wood (A-E) and pinosylvins (F-G):

(A) glucose, (B) mannose, (C) galactose, (D) xylose, (E) arabinose, (F) pinosylvin, and (G) pinosylvin monomethyl ether.

1.3 Forest-industrial processes

In forest-industrial processes wood is utilized to produce sawn goods, fiber boards, pulp and paper, tall oil, specialty cellulose, lignosulfonates, and other bio-based commodities. In many cases, residual fractions, such as bark and partially degraded lignin in black liquor, are used for energy production. Pulp and paper processes are traditionally divided into mechanical and chemical pulping processes.

Industrial plants that convert lignocellulosic feedstocks, such as wood, to multiple products are referred to as biorefineries. One type of biorefining process that is currently in the focus of much research and development is the sugar-platform process. In a sugar-platform process, lignocellulosic polysaccharides, such as cellulose and hemicelluloses, are converted to sugars, typically by using cellulose- degrading enzymes. The sugars can then be converted to desirable

A B C

D

F G

E

(18)

products using microbial fermentation processes or through chemical catalysis. Wood and residual forest-industrial streams can also serve as basis for various thermochemical processes including combustion, gasification, pyrolysis, hydrothermal liquefaction (HTL), and

hydrothermal carbonization (HTC).

1.3.1 Saw mills

Sweden has about 140 saw mills (Skogsindustrierna 2020). At the saw mills, the wood logs are debarked and the wood is then sawed and dried.

A by-product from saw mills are wood chips that are produced from the outer part of the wood logs, the sapwood. This makes wood chips from saw mills, i.e. wood that mostly consists of the sapwood, slightly different from wood chips made from whole logs, which consist of a mixture of both sapwood and heartwood.

Saw dust is another by-product from saw mills. Saw dust is not suitable for pulp and paper production, but it is sometimes made into pellets that are used as a solid fuel. There are 58 pellet factories in Sweden producing pellets with a total energy content of 9.3 TWh per year (Bioenergitidningen 2020).

1.3.2 Heat and power plants

There are around 200 combined heat and power (CHP) plants in Sweden (Skogen 2019). In Sweden, low-quality wood logs and other residual woody biomass are common feedstocks for CHP plants.

Wood logs that are dry, damaged by rot, or have too small diameter cannot be used by saw mills or pulp mills and are instead sold to a lower price as fuel wood. If a pulp and paper mill does not have any bark boiler, it can sell bark and fine fractions to CHP plants.

One of the most important parameters for energy production from woody materials is the moisture content. Most plants work best with wood with a moisture content of 20-30% (Skogen 2019). The typical dimension of particles fed into CHP plants is usually up to 15-50 mm (Skogen 2019).

1.3.3 Mechanical pulping

The principal of mechanical pulping is to grind wood or wood chips to generate cellulose fibers. Whole wood logs are used to make

groundwood pulp, whereas in other mechanical pulping processes

(19)

wood chips are used to make refiner pulp (Brännvall 2009a).

Groundwood pulp was more common in the past, but it is still produced by some mills. As for other mechanical pulping processes, an advantage with the groundwood process is the high pulp yield, which is due to that all major organic constituents, cellulose, hemicelluloses, and lignin, remain in the pulp. A disadvantage with mechanical pulping is the high energy demand (Gorski et al. 2010).

Refiner pulp is made by grinding wood chips between refiner discs.

The mechanical pulp consists of fibers released from the woody material, but also of fines, smaller particles of broken fibers and other material from the fiber walls (Brännvall 2009a).

To make thermo-mechanical pulp, TMP, the wood chips are steamed before they are inserted into the refiner. To make chemo- thermo-mechanical pulp, CTMP, a relatively mild chemical and thermal treatment is carried out before the refining step (Brännvall 2009a).

1.3.4 Chemical pulping

Compared to mechanical pulping, chemical pulping provides a pulp with more flexible fibers (Brännvall 2009a). By degrading the lignin and a part of the hemicelluloses, the cellulose fibers can be released and form chemical pulp. When the lignin is partially degraded in chemical pulping, charged groups are introduced. This facilitates solubilization of lignin fragments, which then can be washed away (Brännvall 2009a) (Fig. 2).

Fig. 2. Delignification during chemical pulping. The upper path shows an alkaline process (Kraft pulping), whereas the lower path shows an acidic process (sulfite pulping).

(20)

Today there are two main chemical pulping processes (Sjöström 1993). The most common is the Kraft process, which was developed from soda pulping. The other one is the sulfite process, which was once very common but which has now decreased in importance. The first sulfite mill was installed in Sweden 1874. The Kraft process to a large extent replaced the soda process and then the sulfite process, in particular after the development of multistep bleaching in the 1930s.

Organic solvents can also be used for delignification during chemical pulping. In the organosolv process, ethanol, methanol, or peracetic acid are used in the cooking step (Brännvall 2009a). Even if this idea stems from the 1930s, it has not gone much further than laboratory and pilot scale, which is due to problems associated with solvent recovery (Sjöström 1993).

Other chemical processes that can be used for delignification and for breaking free the cellulose fibers include treatments with oxygen and steam explosion. Oxygen is a good delignification agent that is used for bleaching, but not for full-scale pulping processes (Sjöström 1993). To use hot steam at high pressure and then suddenly reduce the pressure to atmospheric conditions is referred to as steam explosion.

However, due to fiber damage steam explosion is not used for production of pulp (Sjöström 1993).

To make chemical pulp, wood chips are packed into a reactor.

There are several types of reactors, but only two main cooking procedures. These are based on batch-wise and continuous digesters.

The benefits of a batch reactor is a more reliable production, a more flexible production allowing changes in pulp quality, and a more efficient turpentine recovery. The most common batch reactor is a stationary vertical cylinder with a conical or spherical bottom. The reactor is filled with chips from the top and hot cooking liquid from the bottom, and is heated with a heat exchanger. Mills usually have more than one reactor and a normal size is 150-400 m³ (Brännvall 2009c). To stop the reaction when the desired delignification is reached, the hot cooking liquid is removed by pumping a cooler displacement liquid from the bottom (preventing the cooking liquid to vaporize in the woody material). The pulp is discharged and pumped from the pressurized digester to a flask tank (blow tank) with

atmospheric pressure (Fig. 3). After cooking, the fibers need to be

(21)

separated from each other. This is done mechanically in a defibrator (a hot-stock refiner/blow line refiner).

Fig. 3. Schematic figure showing forest-industrial processes in which wood is converted to pulp. The figure shows a wood log truck (A), a drum debarking system (B), a disc chipper (C), screening of wood chips (D), a wood chip pile (E), batch (F) and continuous (G) cooking systems, and washing and bleaching (H).

A

B

C

D

E

F

G H

(22)

The most common continuous digesters are tall and slim vertical flow digesters with production rates of 1,000-30,000 air-dry metric ton pulp/day (Brännvall 2009c). The process is continuous from the chip bin to the pulp blow. The process is arranged as a single digester or as two digesters depending on whether a separate vessel is needed for impregnation or not. The continuous digester schematically

depicted in Fig. 3 has a separate impregnation reactor. The wood chips are fed from the chip bin to a horizontal steaming vessel for pre- steaming (removal of gases from inside the wood chips) and are then fed into the top of the digester. To get a high pressure in the digester, a high-pressure feeder is used. The wood chips are then transported by a transportation liquid into the top of the digester, where a screw feeds the chips into the digester. The wood chips fall down into the digester onto the top of the chip column and then move continuously down through the digester. The cooking zone of the digester has a temperature of 160-170 °C (softwood) or 150-160 °C (hardwood).

The chips then enter a zone with washing liquid. After the passage through the digester, the fibers need to be defibrated. This occurs in the line defibrator, when the pulp is still under high pressure. The pulp is thereafter screened in a deshiving refiner and then washed.

Sometimes the mill has a second refiner after the wash. There are also continuous digesters that have blow units or blow tanks (Bryce

1980a). After cooking, the pulp can go through screening, washing, oxygen delignification, and bleaching steps depending on the desired pulp quality.

1.3.4.1 Kraft process

In a global context, the most common process is the Kraft process.

The cooking liquor in Kraft pulping, i.e. the white liquor, consists of sodium hydroxide and sodium sulfide. The active cooking chemicals are hydrogen sulfide ions (HS^-) and hydroxide ions (OH^-) (Brännvall 2009a). Hydrogen sulfide serves as the main delignification agent.

Reasons why the Kraft process is popular include that many wood species can be used, that the cooking time is short, and that the pulp has excellent strength (Bryce 1980a).

The Kraft process can be divided into two steps; an initial increase in temperature (impregnation) followed by a period with high

temperature. It is desirable that the cooking liquid penetrates the wood chips before the temperature reaches 140°C. The final cooking

temperature depends on the wood species, and the yield depends on

(23)

the temperature. The strength of the pulp seems not to be affected as long as the temperature is below 200°C (Bryce 1980a).

The most important parameters of the wood chips used for Kraft cooking is the thickness, length, and moisture content (Gullichsen 1992; Ressel 2006). The wood chips used for Kraft pulping are relatively long to minimize fiber shortening, which would result in weaker pulp. An even moisture content and an even thickness are important to get an even impregnation process.

1.3.4.2 Sulfite process

Until the 1950s most pulp was produced using the sulfite process, but the production of sulfite pulp has decreased as the Kraft process became more and more common (Sjöström 1993). The reason why the Kraft process became more common is two main disadvantages associated with sulfite pulping, namely that it is limited to fewer wood species and that the pulp is weaker compared to the pulp from the Kraft process (Bryce 1980b). In recent years, the focus area of the sulfite process has shifted from producing pulp and paper to producing dissolving pulp and lignosulfonates. This is due to the demands from textile and concrete manufacturing (Rødsrud et al.

2012; Sixta et al. 2013).

In acidic sulfite cooking, the active chemicals are sulfurous acid (H2SO3) and bisulfite ions (HSO3-) (Brännvall 2009a). The counter ion can be calcium, sodium, magnesium, or ammonium (Bryce 1980b;

Gellerstedt 2009b).

The sulfite system is based on two equilibria:

The equilibria are dependent on the temperature. At the temperature used for pulping (130-170°C), the pH will be higher than measured in room temperature (Sjöström 1993; Gellerstedt 2009b).

To obtain a high level of sulfonation of lignin, bisulfite ions (Fig. 2) need to be present in the liquor. If there are insufficient amounts of bisulfite, the woody material will turn dark, i.e. a black cook will occur, and there will be no efficient dissolution of the lignin

(Gellerstedt 2009b). By increased sulfonation (Fig. 2), the lignin will successively become more and more hydrophilic and water soluble.

SO2H2O + H2O(l) ⇌ HSO3-

(aq) + H3O⁺(aq) pKa=1.9 (1) HSO3-

(aq) + H2O(l) ⇌ SO32-

(aq)+ H3O⁺(aq) pKa=7.0 (2)

(24)

Most sulfite processes are based on batch digestion systems, but there are some mills that utilize continuous systems. The batch digesters are typically 70-350 m³ (Bryce 1980b).

Acidic sulfite pulping is performed with a cooking liquid in the pH interval 1.2-1.5 and there is an excess of free SO2 (Bryce 1980b). The initial temperature is often around 70-80°C to assure that the cooking chemicals have completely penetrated the wood chip before the temperature reaches 120°C. If the temperature is too high under acidic conditions, condensation of lignin will occur resulting in a black cook (Bryce 1980b).

Bisulfite pulping is a process where the liquid has equal fractions of free and bound SO2. The pH is in the range 3-5, and the duration is typically 5-7 h. The temperature increases faster than in more acidic sulfite cooking processes (Bryce 1980b).

Alkaline sulfite pulping, in which a combination of sodium sulfite and sodium hydroxide is used, results in low yields and a brightness that is similar to that of pulp from the Kraft process (Bryce 1980b).

However, the rate of pulping is rapid, and the pulps have high strength.

Multistage sulfite pulping is an approach designed to get the

benefits that are associated with different pH intervals (Bryce 1980b).

The first step could be neutral or alkaline, and the second step could be acidic. Alternatively, the first step could be acidic, and the second step neutral or more acidic (Bryce 1980b; Sixta 2006).

Dissolving pulp is a pulp with special properties that make it possible to use it for regenerated fibers that are used for textiles. The pulp needs to have a low content of hemicellulose, and a high content of alpha-cellulose. The final degree of polymerization of the cellulose and the viscosity need to be carefully controlled. The temperature of the process is usually higher than for other sulfite processes (150°C or higher) and cooking continues until low kappa numbers are achieved (Bryce 1980b).

The desired size of wood chips for sulfite cooking depends on the products. If the product is dissolving pulp that will be used to produce viscose, the strength of the pulp is not important. Fiber shortening caused by using short wood chips is then not a limitation. An average length of 19 mm has been determined as suitable for having a good

(25)

packing degree (Howard 1951). If the sulfite process is used for making paper, longer wood chips are needed for minimization of fiber shortening. The thickness of the wood chips is, however, not as critical as for the Kraft processes.

1.3.5 Sugar-platform processes

The fundamental difference between chemical pulping processes and sugar-platform processes based on lignocellulose is that in chemical pulping the cellulose is preserved in polymeric form, whereas in sugar-platform processes the cellulose is converted to sugar. Conversion of cellulose to glucose, i.e. saccharification of cellulose, is typically catalyzed by cellulose-degrading enzymes. The sugar is then utilized as substrate in a microbial fermentation process or in a chemical conversion process. Typical products would be bio- alcohols, such as ethanol or butanol, or bio-acids, such as lactic acid or succinic acid (E4tech et al. 2015).

Another difference is that chemical pulping primarily targets lignin, which is modified and at least partially degraded. In sugar-platform processes, too extensive lignin degradation is avoided and polymeric lignin, hydrolysis lignin, remains as a co-product (Ragauskas et al.

2014). Too high concentrations of lignin-degradation products, for example phenolic substances, can inhibit cellulose-degrading enzymes and microorganisms (Jönsson and Martín 2016).

Prior to saccharification of cellulose, a pretreatment is needed to make the cellulose accessible to cellulose-degrading enzymes (Arantes and Saddler 2011). There are many different types of pretreatment (Sun et al. 2016; Gandla et al. 2018), but the predominant technology is hydrothermal pretreatment, which is typically carried out at temperatures between 160°C and 240°C (Sun et al. 2016), and which primarily targets hemicelluloses (Gandla et al.

2018). To make it more efficient, hydrothermal pretreatment can be combined with steam explosion and addition of acids, for example sulfuric acid. Regardless of whether acid is added, the process will be acidic, as degradation of hemicelluloses will result in formation of carboxylic acids, for example acetic acid, formic acid, and levulinic acid (Jönsson and Martín 2016).

Feedstocks for sugar-platform processes based on lignocellulose can be agricultural residues, energy crops, and wood. In either case, a diminution step is needed before the pretreatment. For wood, that

(26)

would typically imply wood chipping. As wood, and particularly softwood, is relatively recalcitrant compared to many other lignocellulosic feedstocks (Gandla et al. 2018), hydrothermal pretreatment would typically commence with an impregnation of wood chips with an acid.

1.4 Wood preparation

Regardless of whether it is a conventional pulp and paper process, a biorefinery process, or a CHP plant using wood, the primary step in the conversion of wood logs is mechanical disintegration of wood into chips of dimensions that are optimized to suite a particular type of process. Due to the large quantity of raw material that is processed (Section 1), it is important that the first steps are handled as

effectively as possible and with minimal losses. The handling of the wood logs will affect the processability, the end product, and the total production yield. Wood logs are often stored under variable weather conditions, both dry and wet, for several months. Long time storage can negatively affect the yield and the demand of bleaching

chemicals, depending on the process (Ressel 2006).

Wood handling is made in wood yards, and between the steps presented below there are also transportation and storage steps. The different steps are also showed in Fig. 3. Transport and storage differ between mill sites and are not in the focus of this thesis. Transport and storage may, however, affect wood chipping, as the average width of wood chips is often affected by cracking along the fiber direction.

1.4.1 Debarking

Debarking is not necessarily a step for CHP plants or for sugar- platform processes. It is, however, an important step in pulp

production. In saw mills, the logs are also debarked. However, in saw mills the bark is removed with a debarker that removes the bark from one log at a time. Saw mills only remove bark if they sell wood chips from the outer part of the wood to pulp mills.

For pulp mills, it is very important that bark does not enter the process. Bark has a large content of extractives, which can negatively affect the processes, and small pieces of bark can also result in dark spots on the finished products. The wear of the knives of wood

chippers is highly influenced by stones and sand, and small stones and sand are often embedded in the bark (Brännvall 2009b). However, for

(27)

the economy and for resource-efficient use of the feedstock it is important that wood loss during debarking is minimized.

The force that is needed to remove the bark depends on the tree species and on whether the bark is frozen. Debarking is typically carried out in a rotating drum (Fig. 3). The wood logs are fed into the drum and the bark is removed as the wood logs are rubbed against each other and, for some debarkers, against the inner surface of the drum (Ressel 2006; Brännvall 2009b).

1.4.2 Wood chipping

For most pulping processes, it is necessary to chip the debarked wood logs to wood chips of a dimension that is suitable for that particular type of process. The only exception is the groundwood process, in which the logs are ground against a rotating stone (Brännvall 2009b). For all other processes, wood chipping is

necessary even if it will damage the fibers causing lower strength and decreased pulp yield. Wood chips may also come to pulp mills from the saw mills.

Quality controls are needed to separate high-value wood chips from fuel-grade wood chips and wood pieces. If wood chips are too small, too dry, contain bark, or are too damaged, they are separated from the other wood chips and are used for combustion to generate energy.

To get a uniform chemical reaction, all fibers in the wood need to get their share of chemicals and heat. Deficiencies are shown as a higher degree of shives in the pulp (Uhmeier 1995; Hartler 1996). In the production of semi-chemical and chemicals pulps it is important to have a short impregnation time. Different processes vary with respect to sensitivity to uneven distribution of chemicals in the beginning of the impregnation, to uneven wood chip dimensions, and to the cooking temperature.

The dimensions of the wood chips are important for pulp production and for heat production. In pulping, good wood chip quality is commonly defined as chips that will give a uniform pulp, and which have qualities such as narrow distribution in size, bulk, mechanical properties, and moisture content (Hartler & Stade 1977;

Uhmeier 1995; Hartler 1996). For heat and power plants that utilize biomass, wood chips are not the only feedstock and the material that goes into the process is often referred to as particles.

(28)

With regard to pulp production, the impregnation process is affected by the properties of the impregnation liquid and the

dimensions of the wood chips in relation to the direction of the fibers (Hartler & Stade 1977). For a wood chip (Paper III, Fig. 1), the length is the dimension that has the same direction as the fibers, the thickness corresponds to the smallest dimension, and the width corresponds to the third dimension.

Common dimensions for wood chips are: length 20-30 mm,

thickness 3-8 mm, and width 15-30 mm (Brännvall 2009b). However, the dimensions of the wood chips depend on the chipping settings, wood species, temperature (frozen wood), and moisture content.

It is important that wood chips have sufficient length. Direct fiber shortening was noticeable for wood chips that were shorter than 20 mm, and below 15 mm the shortening was apparent (Hedenberg et al.

1999). There has been a trend that pulp mills utilize longer wood chips, and that leads to less compact wood chip columns in the digester (Hartler 1997). A less compact wood chip column gives a lower radial and axial filtration resistance for the liquid flowing through the column. When the wood chips are longer, the thickness increases as well, which gives larger fractions of overthick chips that need to be rechipped (Fig. 3) in order to avoid increasing the amounts of shives.

Wood properties as porosity and density also vary. Formation of individual chips depends on micro variation in wood properties occurring at the cell structural level (Twaddle 1997).

1.4.2.1 Wood chippers

The chipping process strongly affects the quality of the wood chips.

The settings of the knife angles and the distance to the wear plate (i.e.

the T dimension) result in chips of a certain length (Paper II, Fig. 1).

The settings can be changed during maintenance to reduce the influence of wood variation, for example the difference between summer wood and winter wood, but that happens only rarely due to the resulting loss in production time. Relevant knife angles are shown in Fig. 4 and are also presented in Papers I and II (Paper I, Fig. 2, and Paper II, Fig. 3). The knife penetrates a distance into the log,

depending on the T dimension, and the wood will experience both cutting and shearing forces. The further the knife penetrates, the higher is the build-up of shear force in the material. When the shear

(29)

force is big enough, the wood chips will be shaved off in the fiber direction, giving the wood chip its thickness (Brännvall 2009b). The thickness will vary, due to that the break occurs randomly. However, the average thickness of the wood chips increases almost linearly with increasing average length, in a relation determined by the

complementary angle (λ) (Fig. 4). Changes in this angle are due to changes in spout angle (ε) or in the sharpness angle (β) (Fig. 4). The spout angle influences the size distribution of the wood chips.

Increased spout angle gives increased fractions of small-sized chips, increased wood chip thickness (with constant length), and increased bulk density. A smaller spout angle is beneficial, but it will decrease the bulk density and reduce the limits for the maximum diameter of the logs. Increasing the spout angle from the commonly used value 30° to 40° or to 50° would increase cracking and decrease the percentage of accept chips, as it would increase the fraction of both overthick chips and pin chips (Hartler 1962; Hellström et al. 2011).

Fig. 4. Definitions of knife angles of wood chippers: ε, spout/cutting angle; α, clearance/pulling angle; β, sharpness/knife angle; λ, complementary angle.

In order to obtain the desired spout angle, the direction, the speed, and the angle of the incoming logs are of importance. A correct positioning of the logs decreases the formation of small-sized chips.

This is achieved by having sufficiently many knives on the disc or drum to make at least one knife engaged in the log all the time. A

(30)

narrow thickness distribution is achieved by having a low friction between the tool and the wood (Hartler 1996).

The use of disc-chipping technology for making wood chips for mechanical pulping has been studied previously. For example, by changing the spout angle and thereby using more energy for the chipping, more cracks were created and by decreasing the wood chip size energy could be saved in the refining step (Hellström et al. 2011).

It is not only the dimension that is an important parameter.

Damaged ends and cracking also influence impregnation, due to that the damage zones are easier accessible for the chemicals (Brännvall 2009b).

The cutting speed, also defined as laps of the knife disc or drum, also influences the dimensions of the wood chips. Too low speed will decrease the capacity of the chipper and decrease the wood chip production rate. A study of a newly installed chipper showed that decreasing the cutting speed (the speed of the knife disc) with 25%

reduced the fraction of small-sized material (pin chips and fines) to half of that of the original value (Bergman 1987).

At lower temperatures the wood is more brittle, and the influence of having a high cutting speed will be more pronounced. When the temperature decreases, the fractions of pin chips and fines will increase, and the fraction of wood chips that are too large will

decrease. At temperatures lower than 0 °C, the moisture content is the most important factor that influences the mechanical behavior due to the presence of frozen water in the log. For higher temperatures, the basic density is the primary parameter rather than the moisture content (Hartler & Stade 1977; Hartler 1996; Hernándes et al. 2014).

1.4.2.1.1 Disc chippers

There are two types of disc chippers; smaller mobile chippers that are used at sites in the forest, and bigger chippers at heating plants and pulp mills. The mobile chippers are traditionally used to produce wood chips for the energy market. The focus of this work is big chippers that are found at mill sites and that chip more than one wood log at the time. The main part of the feedstock from forestry is

converted to wood chips using energy-consuming conventional disc- chipping technology (Brännvall 2009b).

(31)

A disc chipper consists of a rotating disc with 10-16 radial knife holders and knives. The log is fed endwise onto the disc through a slot. Different disc chippers have different designs. This could be an angled disc with horizontal infeeding, chippers that are fed

horizontally, as in Fig. 5, or the disc could be radial and the feeding in an angle and the wood logs would fall into the slot through gravity feed (Brännvall 2009b).

Fig. 5. Schematic views of (A) a disc chipper, and (B) the velocity gradient across the disc (courtesy of Multi Channel Sweden AB).

There are certain drawbacks with disc-chipping technology. This has to do with the placement of the wood raw material, the lack of flexibility to handle wood of variable quality (such as summer and winter wood, different wood species, variations in wood quality caused by growth in different habitats, and variable log dimensions), and large demands for installed power to drive the wood chippers. The knife-angle settings of today's disc chippers cannot be changed during operation, and stopping the chipper to change the settings to make them optimal for the feedstock and for the process takes valuable time.

The design of the disc chipper results in a cutting gradient over the disc. The cutting speed towards the center is lower, which generates too large wood chips. Towards the periphery a higher speed is obtained, generating pin chips and saw dust. This results in lower product yield from the feedstock, which increases the cost and the environmental impact.

1.4.2.1.2 Drum chippers

In the past, drum chippers have mainly been used to produce chips for bioenergy power plants. These chippers are smaller and mobile and can be used in the forest, close to the raw material. The chippers

A B

(32)

are usually used as tractor implements or placed on trucks. The drum is usually stronger than small disc chippers, and is not only used for logs both also wood residues. The biggest problem with the drum chippers is the uniformity of the dimensions of the chips. This is due to that a large T dimension is needed to prevent wood chips from entering the middle of the drum or the pocket next to the knives (Fig.

6B). When wood logs with large diameters are chipped, changes in spout angle, ε, also affect the uniformity (Paper I, Fig. 1).

Fig. 6. Schematic views of (A) the novel drum chipper (courtesy of Multi Channel Sweden AB), and (B) an example of a drum of a conventional drum chipper.

Drum chippers have the same velocity over the knives and the same distance between the knives, similar T dimension as traditional disc chippers, and the same angle independently of where on the knife the logs are chipped. On disc chippers, the parameters change along the knives. That is not a problem when only one wood log is chipped, but when several wood logs are chipped at the same time there will be differences between wood logs that are close to the periphery of the disc and the wood logs that are close to the center of the disc. This problem will increase when the diameter of the disc increases.

The company MCS (Örnsköldsvik, Sweden) has developed a new type of drum chipper that potentially offers several advantages compared to conventional disc chippers. The new drum-chipper design also differs from that of conventional drum chippers. The MCS drum chipper has specially designed wood chip channels around the knife holder (Fig. 6A) to facilitate the use of drum dimensions over 2 m. The new drum chipper does not transport the wood chips from the

(33)

middle of the drum or have big pockets next to the knives resulting in a large distance between knives and wear plate. The knife-to-

anvil distance and the knife-to-wear-plate distance (the T dimension) are similar to those of traditional disc chippers, and the wood chips will fall down into a collecting vessel due to centrifugal forces and gravitation. The channels around the knife holders lead the wood chips out from the drum and not back to the knives (Fig. 6A).

The size of the drum is important for making the wood chips more uniform in size and make them similar to wood chips produced using conventional disc chippers. The use of smaller drum chippers can result in formation of boat-formed wood chips and in large variability with regard to length and thickness, which is due to the distance between the wear plate and the knife. The size of the drum will also affect the spout angle (ε). If a small drum is used, there will be a noticeable difference between the angles (ε) with regard to the upper part and the lower part of the wood log (Paper I, Fig. 1). The new drum chipper has more knives than conventional chippers, and the invariable distance between the knives gives a more uniform chip size.

It is the knives that feed the drum with wood logs, as the knives will pull the wood logs closer to the drum.

The new wood chipper is flexible, as it allows the infeed angle settings to be adjusted during running, which permits optimized chipping without process interruption. The construction has potential to provide a more even wood chip quality, minimization of the reject fraction, increased yield, and energy savings. As the new chipper has the same cutting speed across the knives, it has potential to generate smaller fractions of chips that are too large, and less pin chips and fines. This would result in a better use of the precious wood feedstock.

1.4.2.1.3 Other chippers

In saw mills, wood chips are produced from the outer parts of the timber, while the interior is utilized for production of sawn goods. A reducer, a disc with knives, is used to produce the wood chips.

Hand-cut laboratory chips are used in some studies. Laboratory chips cannot be compared to chips produced with industrial-style wood chippers. The laboratory wood chips have no damaged ends or cracks. Impregnation is affected by the cracking of the wood chips, and laboratory-cut wood chips for Kraft processes need to be much thinner than wood chips produced in a normal way (Brännvall 2009b).

(34)

1.4.3 Screening of wood chips

All wood chips that are produced by the wood chippers are not used in the digesters. The fractions with too big wood chips and too small wood chips and particles can create problems in the production of pulp. Therefore, the wood chips need to be screened before use. There are different kinds of screening equipment, but they typically consist of plates moving above holes. The plates can vibrate and rotate depending on the model

Oversized wood chips (diameter > 45 mm) are usually rechipped using a chip cleaver, or cracks are introduced in them using a chip conditioner, before they go back into the production. The fractions with too small material are sorted out, but in some mills the pin chips are fed into the process in controlled amounts. In some mills, the pin chips are sorted away together with the fines. Thereafter, they are sold for their energy value or used as fuel in the bark boiler.

1.5 Impregnation

To get a uniform chemical reaction with the wood, all fibers in the wood need to get their share of chemicals and heat. This step, the impregnation, is vital for achieving uniform pulp, low proportions of reject and shives, and a product of high quality. Different processes exhibit different sensitivity with regard the need for a thorough impregnation. The sulfite process is typically more sensitive than the Kraft process (Rydholm 1965; Hartler 1990).

Impregnation is not only about penetration of wood chips by pulping chemicals, as it also depends on outward diffusion of

entrapped air and gases, and dissolved organic matter. Impregnation is also affected by inward diffusion during cooking (Rydholm 1965).

Penetration of wood chips by liquid is determined by the character of the capillary cavities of the wood, the presence of trapped air, and the pressure gradient of the penetrating liquid. The diffusion is determined by the diffusion surface that is specific for different wood species, and characteristic of the ions and the concentration gradient of the diffusing agent (Hartler 1990).

Different tree species and parts of the same tree exhibit differences with regard to the capillary structure. Thus, there are differences between the capillaries of softwood and hardwood, sapwood and heartwood, and earlywood and latewood. Reaction wood is also

(35)

different compared to normal wood. Normal pulpwood consists of 50- 70% of void spaces filled with water and gases (Rydholm 1965). The difference between heartwood and sapwood will affect not only impregnation, but also the packing degree. Due to differences with regard to moisture content and wood properties, the wood chips will be affected differently by the cooking process.

Theoretically, the use of very thin wood chips, i.e. 0.2-1 mm, would be advantageous for alkaline processes, but this has not been possible to implement in large-scale operations. In most mills, the normal wood chip dimension is a length and width of 15-25 mm and a thickness of 2-5 mm (Rydholm 1965).

The flow of liquid in completely soaked wood can be described with the Poiseuille equation:

𝑉

𝑡 = 𝑘 ∗𝑛 ∗ 𝑟⁽∗ ∆𝑝 𝑙 ∗ 𝜂

...where the volume, V, flows through n capillaries of the radius (r) and length (l) at the time t if a pressure differential Δp is maintained and the viscosity of the liquid is η. An increase in the temperature of the liquid will increase the flow rate in proportion to the decrease in viscosity. Increasing pressure differential will also result in increased flow rate (Rydholm 1965).

Alkaline liquids, such as cooking liquid of Kraft pulping, cause a swelling of the wood structure. This results in almost equal diffusion rate in all dimensions. Neutral and acidic liquids diffuse more rapidly in the longitudinal than in the transversal direction of the chips. The diffusion rate for neutral and acidic liquids in the longitudinal direction would be about half compared to that of water, and in the transverse direction 3-6% compared to that of water (Rydholm 1965).

This is due to the total cross-sectional area of the capillaries that control diffusion. The cross-sectional area of the longitudinal direction is half of the total area. For the radial and transversal directions, the rate of diffusion increases up to 40% of that of water if the pH is in the range 12.8-13.5, whereas the diffusion in the longitudinal direction would still be 50% of that of water (Rydholm 1965).

Thickness is the most important wood chip dimension in a Kraft process. This is mostly due to the difference in dimensions, as the

(36)

length and width of the wood chips are 5-10 times greater than the thickness. When the distance is shorter and the diffusion rate is almost the same, impregnation in the width and length directions can be seen as insignificant compared to impregnation in the thickness direction (Rydholm 1965).

The initial impregnation can be described as penetration of air-filled wood by liquid. Trapped air results in uneven impregnation rate.

Proper gas diffusion is important with regard to both entrapped gases and gases that are produced in the chemical reactions. One method for air removal is to soak the wood chips in water. However, soaking large volumes of wood chips in water requires a lot of time and a lot of clean water, which makes implementation of soaking in industrial scale challenging. Using large amounts of water also has a negative influence on recovery of chemicals and it is not good for the heat balance of the pulp mill. Another way to remove air is through vacuumation, which is based on a pressure gradient. The drawback is that the low pressure needed to remove the trapped air is impractical to use in a normal digester. A third method is replacement of air with gases that are soluble in the penetrating liquid. This method is

commonly used and a lot of research has been made on pre-steaming of wood chips with water vapor (Rydholm 1965).

Impregnation of saw dust differs from impregnation of wood chips and therefore requires the use of special techniques. When saw dust is in contact with a liquid, a compact saw-dust matrix that limits the impregnation will form. The matrix limits the flow of liquid and sometimes completely prevents impregnation. The saw dust will consume more effective alkali than wood chips, which also results in a lower yield. The impregnation time required to get the same EA concentration in saw dust is twice as long as that of wood chips (Korpinen et al. 2008).

When wood logs are chipped, the material becomes compressed, which creates a plastic deformation. This deformation is manifested as cracks, structural damage, or micro changes in the fiber wall. For a single wood chip, one of the sides is subjected to compression and damage, but the other side is more or less intact. This mechanical effect seems to make the wood more sensitive to subsequent chemical damages. The damages make it easier for the cooking chemicals to impregnate the chips and the reaction rate is increased.

(37)

Cracking with a chip sizer (chip cleaver) is efficient to improve impregnation. Impregnation of cracked and sized overthick wood chips was found to be faster compared to accept chips (Määttänen &

Tikka 2008a).

1.5.1 Methods for studying impregnation

Different methods are used to study impregnation. The

impregnation is mostly influenced by the pH of the impregnation liquid, the dimensions of the wood chips, and the degree of cracking.

A sinkage test is used to study impregnation by measuring the time it takes for a wood block to sink into a liquid. However, the method has poor reproducibility and accuracy (Malkov et al. 2003).

Another possibility is the so called "uptake-method". Here, the treated wood chip sample is compared to an untreated or dried sample by comparing the weight. However, this method also has poor

reproducibility and accuracy (Malkov et al. 2003).

Another approach is to use image analysis by analyzing and

digitizing photographs of sliced frozen chips after impregnation. Only a rough estimation of the penetration degree can be achieved with this method, because of interference between diffusion and penetration (Malkov et al. 2003).

Specially designed penetration clamps have been used to study the permeability of different wood specimens. A wood block is clamped in a special cell, and water or air is being forced through the wood chip sample (Malkov et al. 2003).

The wood chips can also be hung in a quartz spiral balance. The method has been used to measure the flow of a liquid into a single wood chip on continuous basis, the influence of steaming time,

hydrostatic pressure, moisture content, cooking liquid penetration, and penetration rates (Woods 1956; Aurell et al. 1958; Malkov et al.

2003).

The way liquid penetrates wood can also be studied by using a microtome (Zanuttini et al. 2000). Other techniques include scanning electron microcopy (SEM), staining and precipitation techniques, radioactive tracer technique, and nuclear magnetic resonance (NMR).

A SEM equipped for energy dispersive X-ray analysis was used to determine the sodium content in a study of impregnation of wood by sodium sulfite (Bengtsson & Simonson 1988). Confocal laser

Wood Chips for Kraft and Sulfite Pulping