—

A H A N D B O O K BY H Å K A N K A R L S S O N W I T H U P DAT E S FR O M
P E R - I VA R FR A N S S O N A N D H Å K A N Ö S T E R H O L M

Fiber Guide
Fiber analysis and process applications
in the pulp and paper industry
—
Fiber Guide
SECOND EDITION
Fiber analysis and process applications in
the pulp and paper industry

A handbook by Håkan Karlsson

with updates from Per-Ivar Fransson
and Håkan Österholm

—
Fiber Guide by Håkan Karlsson
Copyright © 2006 AB Lorentzen & Wettre, Box 4,
SE-164 93, KISTA, Sweden

This edition with updates from Per-Ivar Fransson and

Håkan Österholm published in 2021. All rights reserved. No
portion of this book may be reproduced in any form without
permission from the publisher. For permissions or additional
copies, contact: [email protected] or visit www.new.
abb.com/pulp-paper/abb-in-pulp-and-paper/literature

• Cover | Admind
• Microscope images | Joanna Hornatowska, RISE
• Illustrations | Timo Rinnevuo and others
• Printed at | Elanders, Sweden
• Printer | Elanders Sverige AB
• Price | € 180

ISBN | 91-631-7899-0
Second edition
 ≔ → ≡
3
Table of contents Next chapter Index

—
Table of contents

How to use this document

To easily navigate to different sections, this interactive book features a top navigation
on every page that allows you to jump back to the Table of Contents at any time.
You can also use the navigation bar to advance chapters, as needed. For specific topics,
jump right to the Index for more precise navigation. The document functions as a
normal PDF, so you can scroll pages as you would normally.
Happy reading!

05 Acknowledgements 76 – 77 Calculated properties

06 – 07 Background 78 – 82 Pulp properties

08 – 09 Introduction 83 – 89 Statistics and reports

10– 12 Wood parts 90 – 94 Sampling

12 – 18 Characteristics of wood 95 – 104 Impact of fibers on products

19 – 25 Fiber origins 105 – 112 Models

26 – 28 Fiber structure 113– 121 Applications of online

fiber analysis
29 – 41 Manufacturing processes –
an overview 122– 127 Cost of raw material and return
on investment
42 –49 Fiber/Process
128– 129 Reference testing
50 – 61 Traditional testing of pulp
130– 133 Bibliography
62 – 66 New technologies
134– 137 Index
67– 73 Measurements of fiber properties
138– 139 Abbreviations
74 – 75 Objects other than fiber
 ≔≔ ←
← →→ ≡≡
4
Table of
Table of contents
contents Previous chapter
Previous chapter Next chapter
Next chapter Index
Index
 ≔ ← → ≡
5
Table of contents Previous chapter Next chapter Index

—
1. Acknowledgements

A lot of my experience and knowledge comes

from STFI*, where I spent many, very active, and
inspiring years with research and development
work. Many results presented in this book are
based on work done there – Thank you STFI.

Thanks also to my colleagues at Fibertracker AB;

all of them have contributed with material, and
thanks to all partners who have been so engaged
in the introduction of the new technology.

A special thanks to Therese Jost, who did a great

job helping me to write this book. Finally, many
thanks to the great staff at Lorentzen & Wettre,
who have encouraged, supported, and helped me
to write this book.

Håkan Karlsson, 2006

*STFI (Svenska Träforskningsinstitutet, 1945-2008), Then became

Innventia AB (2009-2017), now RISE (Research Institutes of Sweden)
 ≔ ← → ≡
6
Table of contents Previous chapter Next chapter Index

—
2. Background

The first quality sensors for online purposes In the late 90’s, the company Fibertracker AB¹,
using modern electro-optical and digital introduced the first commercial product to the
processor components (photodiodes and a market – the STFI Fibermaster. Its standard
microprocessor) came in about 1975. One of features were measurements of length, width,
the first was an electro-optical shive analyzer, deformation, bendability, coarseness and fines,
suitable for the design and optimization of as well as detailed statistics about these
screen rooms. properties.

A few years later, systems for monitoring Some years later, Fibertracker AB was acquired
refining in mechanical pulping that combined by Lorentzen & Wettre and the fast development
measurements of optical fiber length, an optical of the online technique and new modules
shive analyzer and a freeness measurement were continued with the introduction of a new
developed. Ways of measuring average fiber benchtop laboratory unit for fiber analysis,
length and radius for online purposes were the L&W Fiber Tester².
invented during the 1980’s.
Historically, Lorentzen & Wettre supplied the pulp
All fiber analyzers at that time measured a and paper industry with a number of different
projected length, and people were not fully products to evaluate the quality of pulp, including
aware of the importance of fiber deformations. wood chips, classifiers, laboratory digesters,
But the industrial potential of these improved fiber fractionation units, handsheet machines,
measurements was very evident. Since lab refiners and freeness testers. More semi-
measurements of the distributions of fiber automated devices, such as the L&W Dynamic
properties were required in combination with Sheet Former (DSF), which produced a paper
improved length measurements, RISE (now sheet that closely resembled machine-made
Swedish Pulp and Paper Institute (STFI)) started paper, were also developed at Lorentzen & Wettre.
to develop the technology for this in the early
1990's.

1. F
 ibertracker AB was founded in 1998 by Håkan Karlsson and
Per-Ivar Fransson (both from STFI).
2. For clarity, L&W Fiber Tester Plus is the most recent product model
of the previous generations L&W Fiber Tester & STFI Fibermaster.
 ≔ ← → ≡
7
Table of contents Previous chapter Next chapter Index

Measurements of fiber properties are now widely

used online also for chemical pulps. Measurement
of fiber properties by optical techniques are now
a feature of standard laboratory equipment.

In 2011, Lorentzen & Wettre was acquired by ABB,

who is proud to continue this long heritage of
supporting the research and development of pulp
measurements, most recently introducing ABB’s
L&W Freeness Online and L&W Fiber Online
in 2017.

The purpose of this book is to support and

motivate people working in all areas of the pulp
and paper industry, by introducing new methods
for paper making. The book gives a general
overview, as well as detailed information, and
experience about fiber analysis, and possibilities
to use the new information in the pulp and paper
processes. It is specifically written for people
working with production of pulp and their
developers, customers and suppliers.

—
Learn more about the products mentioned in this book

GO TO L&W FREENESS
GO TO L&W PORTFOLIO GO TO L&W FIBER TESTER PLUS
AND FIBER ONLINE
 ≔ ← → ≡
8
Table of contents Previous chapter Next chapter Index

—
3. Introduction

The pulp that paper is made from, is more than a Fiber length was determined by fractionation
diffuse slurry of disintegrated wood. Pulp through screens with specified mesh sizes.
consists of fibers and each fiber is a fantastic Traditionally, for mechanical pulp (SGW, RMP,
basic building element. The cellulose fiber is TMP, CTMP, etc.), freeness, shives or mini-shives
strong and has a great bonding capacity of its content, linting propensity, fiber length, strength
own, and it can be used to obtain specific and optical properties were measured after
properties in a given paper grade. Paper should handsheets had been made.
be regarded as an engineered product and the
optimal use of the fiber is therefore of the All of these methods were laborious, time-
greatest economic importance. It is also a consuming, and prone to error as the pulp went
renewable raw material. through different operators and equipment. If a
mill wanted to determine the quality of its market
There are many different paper and board grades, pulp, it could conceivably take several days before
and therefore many different uses of fiber: the beating, handsheet making, and testing could
packaging and liquid carton board, newsprint, be completed. These methods do not lend
fine paper, corrugating medium, security paper, themselves to process optimization, and pulp
cigarette paper, tissue, towels, filters, building mills therefore had to rely upon the online sensors
elements, etc. Each grade has its own specific of the day, such as freeness and kappa number.
demands; the selection and treatment of fibers Development in the electronic and computer
should be optimized for the specific grade's technology sector have made new measurement
properties and processes. methods available for the pulp and paper
industry. New technology is used to achieve
For many years, pulp quality was determined by higher production, less discharge, improved
preparing handsheets and then performing paper quality (uniformity) and a better use of energy
testing in a “standard” way to assess the and raw material.
suitability of the pulp for sale or for internal use
within the integrated pulp and paper mill. Operator-independent testing provides better
Chemical pulp quality was, and in many locations accuracy in the measurements. Automatic
still is, determined by beating the pulp at methods allow for frequent testing and make it
specified intervals (PFI-mill) and then making possible to follow the process in a way that was
handsheets at each interval to evaluate the not achievable in the past with only manual tests.
development of the strength and optical Feedback (and feedforward) for process control
properties. and a final check of product quality before
delivery to the customer are also now possible.
 ≔ ← → ≡
9
Table of contents Previous chapter Next chapter Index

—
01
The new techniques enable mills to think and
Wood
Papermaking starts work in new ways. The development in fiber
earlier than in the
paper machine. The
quality measurements gives us a new language
most important cost to communicate quality – the fiber language.
item in the production
of pulp and paper is
This is key to more efficient papermaking
the cost of fibers. (Figure 01).
Pulping

Some people talk about a paradigm shift,

which means that people must now have a
broader view of what may seem like causal
process interactions. How are the processes
Fibers
connected, and why does trouble or deviations
from targets occur? In the past, it was usually
sufficient if each person managed his or her
own part of the process and if suppliers were the
Market pulp Chemical energy
subject matter experts for their products. Now,
Chemical fillers
generally speaking, everybody is required to
Recycled
energy know a lot about the whole system. Customers
expect the supplier to not only understand the
Broke customer’s process well, but also the problems
of the customer’s customer. Instead of customers
and suppliers, you can consider them members of
the same “departments”.
Paper machine

Understanding and speaking the fiber language

is a useful tool for communication of quality all
along the process chain.
— Paper
01
 ≔ ← → ≡
10
Table of contents Previous chapter Next chapter Index

—
4. Wood parts

—
02
Generally, a tree consists of three parts: a crown
Cross-sectional sketch composed of leaves and branches, a stem, and a
of a mature stem with
fiber annual rings and
root system. Figure 02 shows a cross-sectional
horizontal channels. sketch of a stem. It is a schematic diagram valid
— for both softwood and hardwood. The stem
03 consists of bark (inner and outer), cambium,
Cross-section of
hardwood with annual sapwood and heartwood. The bark is of no
rings, vessel cells and interest in papermaking.
horizontal ray cell
channels (cf. Figure 2.6)
The cambium is located between the bark and
the sapwood. It is a thin layer of cells where cell
growth takes place. The rate of growth varies
with the seasons and the growth place, giving
rise to the deposition of thin-walled fiber cells in
the spring and denser thick-walled fibers in the —
summer. 02

The cambium is dormant during the cooler

months of the year. This yearly growth cycle is
responsible for the annual rings phenomenon,
which also enables a tree to be aged by counting
the number of rings (Figures 02 and 03).

Fibers

Ray cells

Vessel

—
03
 ≔ ← → ≡
11
Table of contents Previous chapter Next chapter Index

— The sapwood begins within the cambium layer.

04
Detail of annual ring
It provides structural support for the tree crown,
acts as a food storage reservoir, and has an
—
05 important function of water conduction up from
Earlywood and latewood the roots. The inner heartwood functions only
fibers. The wider, brown
fiber is the earlywood. as a mechanical support. In most species, the
heartwood is much darker in color than the
sapwood. Differences in structure of the wood
capillaries make liquor penetration in the pulping
stage more difficult in the heartwood than in the
sapwood.

There are distinct differences between earlywood

(springwood) and latewood (summerwood) in —
04
regions where the climate has seasonal changes.
The earlywood has an open structure due to the
fact that during springtime the growth is very
intense and large amounts of water are
transported through the fibers, making the
structure very loose.

Earlywood fibers have a large diameter with a

thin fiber wall and tend to collapse in the pulping
process. Earlywood is succeeded by latewood,
which has thicker fiber walls because of the
slower growth during summer and fall. The
earlywood is the lighter part of the annual ring,
while the latewood is the darker part. These —
details can be seen in Figures 04 and 05 05

Latewood fibers in particular have to be

beaten to develop strength.
 ≔ ← → ≡
12
Table of contents Previous chapter Next chapter Index

—
5. Characteristics of wood

—
06
Wood for pulp making can be classified in two
A sketch exemplifying main groups: Softwoods (coniferous trees,
the structure of
softwood. Earlywood
such as pine, fir, spruce and hemlock) and
and latewood fibers Hardwoods (deciduous trees, such as birch,
are evident. A resin
channel can be seen at
aspen, eucalyptus, acacia and oak).
the top of the picture.

— 5.1. Softwood
07 Softwood (Figure 06) is composed of up to
Sketch of a wood
block exemplifying the 90% long, tapering cells called tracheids.
structure of hardwood. These function both as support and water
The open hole is a vessel.
The crossing channels transportation. The hole in the middle of the
include ray cells. tracheid is called lumen. The tracheids are
connected to each other through pores. Apart
from the tracheids, there are other cell types —
06
present, for example resin channels withholding
resin, and ray cells. These cells are principally
connected with lignin, forming a stable structure.

Ray cells are oriented in a radial direction from

the outside of the tree trunk towards its center.
They transport waste material (extractives)
towards the heartwood and may be used for
storage of various food substances. The ray
cells accomplish horizontal transport of liquids
across the annual rings. Fibers from softwoods
are long and strong. The strongest papers are
manufactured from chemical pulp made of
softwood. —
07
 ≔ ← → ≡
13
Table of contents Previous chapter Next chapter Index

5.2. Hardwood The biggest difference between hardwood and

Hardwoods have a more complex structure softwood fibers is that the hardwood fiber is
than softwoods, with different cells for water considerably shorter and thinner than the
transport and support. Elongated libriform fibers softwood fiber. Generally, hardwood contains
function as support and are thick-walled in more cellulose and hemi cellulose and less lignin
proportion to the diameter. Shorter, wider than softwood, while the proportion of
cells called “vessels” are responsible for extractives, i.e. resin, is higher.
water transportation (Figure 07).
5.3. Geographical differences
Hardwoods also contain a vertical parenchyma In Scandinavia, the growth period is short but
system and a horizontal ray parenchyma system. intense, particularly in the northern parts. It lasts
Hardwood fibers are shorter and thinner, giving a from May to September-October, and the growth
better paper formation than softwood fibers. is idle during wintertime. It takes about 75 years
Hardwood fibers give paper a smooth printing for a Scandinavian spruce or pine to be ready to
surface and high opacity. Because there is less harvest, compared to a pine in the southern U.S.,
lignin in hardwood, compared to softwood, it is where it only takes 25 years for the tree to be
also easier to bleach the pulp to high brightness. ready to harvest. This is due to the warm climate
all year round.
These qualities make the hardwood fibers
suitable for use in printing paper, although this It takes only seven years for a eucalyptus tree
type of paper generally consists of a blend of to be ready to harvest when grown on a
hardwood and softwood pulps to meet both plantation in Brazil.
the strength and the printing surface demands
of the customer.
 ≔ ← → ≡
14
Table of contents Previous chapter Next chapter Index

—
08
5.4. Plantations Three types of pulp raw material with different
Separating tree parts; Monoculture tree plantations continue spreading fiber properties are available:
the dark area is used
by the sawmill.
throughout the tropics and sub-tropics, and field • The top from older trees
trials with genetically modified trees aimed at • The lower, outer parts from older trees
producing higher yields of a more uniform type (sawmill chips)
of wood are in progress. • Young trees (thinnings)

The seedlings are produced in nurseries, where 5.5.1. Top and lower section of the stem
they undergo rigorous checking. The parameters Juvenile wood comes from the first years of
that are considered are: length, coloring, growth; it stretches to the top of the tree
and physiological state of the seedling and throughout its lifetime. There are fewer annual
consistency of the attached clod of earth. This rings at the top of the tree, and the cambium has
result in a homogeneous and controlled structure not yet reached its maturity stage, which means
of the wood produced at the plantations. Another the wood is thin-walled, slender, and with
advantage with plantations is the cost-effective relatively short fibers.
and fast transportation of wood to the mills.

Among the most significant disadvantages of

plantations are pests, diseases and fires. The
more homogeneous a plantation is, the higher
the risk of pests and diseases. These risks can
Toplog with low or
normally be controlled through proper planning medium coarseness is
suitable for linerboard
and management.
and tissue

5.5. Differences in fiber properties within a tree

There are differences in fiber properties both
between different tree species and within the
Dark area is sawlog
same tree species. This leads to heterogeneity
and time variations in the wood flow to the mill.
Area outside sawlog is
slabwood, which has
The usual way to process the tree is that the
high coarseness, long
lower part of the stem and middle section of the and wide fibers and is
suitable for sack kraft
stem go to sawmills, while the top section and
thinnings become raw material for pulp and paper
manufacturing (Figure 08). At sawmills, the outer —
layers of the stem become wood chips. 08
 ≔ ← → ≡
15
Table of contents Previous chapter Next chapter Index

—
09
The oldest parts of the tree are located in the Even inside one tree there is a big difference in
Cross-section of lower section of the stem. The number of annual properties for different parts; from pith to bark,
earlywood and
latewood fibers
rings is at a maximum, as are the number of cell from ground level upward (fiber length increases
divisions that have taken place in radial direction. from top to root) and differences due to growth
Fibers are longer in more mature wood. rate. There are also differences between adjacent
Therefore, there is a large variation in fiber cells and within the growth ring. Fibers from
dimensions, since the tree has both juvenile earlywood have a large diameter, a thinner fiber
and mature wood. wall and a shorter fiber length than latewood
(Figure 09).
5.5.2. Thinnings
A stand of trees, or a contiguous community of
mostly uniform trees, is often thinned several
times during its lifetime. The trees are often
about 30 years old at the first thinning. Wood
from these kinds of thinnings has a considerable —
proportion of juvenile wood, with thin-walled and 09
short fibers. Final cutting wood has, on the other
hand, a large proportion of thick-walled long Latewood fibers are thick walled and longer than
fibers. earlywood. Table 1 shows differences between
earlywood and latewood and amongst species.
5.5.3. Variations within a tree Fiber property models for different stands have
The variations within a tree species depend on been reported and can be used as tools to predict
geographical locations, stands, variations the fiber properties of logs. Effective sorting of
between trees within a stand, variations between pulpwood logs requires the detection of log
logs within a tree and variations within a log. diameter and number of growth rings, and also
the position of a log within a stem. The
An understanding of the principal differences in advantages of the application of fiber property
properties between and within trees can be used models are added value and cost savings, but
to predict the kind of fiber properties that one there are also additional costs for wood
can find in the above-mentioned assortment. procurement and mill handling.

—
Table 1: Fiber dimensions in pulp for some different trees.

Loblolly Pine (US) Swedish Pine

Earlywood Latewood Earlywood Latewood Eucalyptus Birch
Fiber length [mm] 3 3.5 2.7 3 1 1.1
Fiber width [μm] 45 35 35 25 16 22
Lumen diameter [μm] 32 12 30 10 10 16
 ≔ ← → ≡
16
Table of contents Previous chapter Next chapter Index

—
10
5.5.4. Sawmill chips
Cross-sectional sketch The parts of the log that are not used for making
of a stem and how its
divided into planks.
planks become chips. These chips are mostly
The outer parts are produced from the outer layers of the stem
processed into chips
for production of pulp.
(see Figure 10); therefore, the larger portion
of the sawmill chips is from sapwood.
—
11
Fiber dimensions The sapwood chips are proven to have a
depend on species and
origin of the wood. considerably faster impregnation rate than those
made of heartwood. Sawmill chips of pine are
stiffer than those made of spruce because they
have a thicker cell wall.

Since sawmill chips come from the outer part of

—
the tree, these fibers are longer and wider than 10
those from round-wood.

Spruce fibers are on average slightly longer and

Fiber width
thinner than pine fibers. Wood chips from Pine Sawmill chips
sawmills normally include longer and broader Thinnings Spruce
fibers than round-wood or thinnings. The
two-dimensional fiber distributions move as Fiber length

indicated in Figure 11, depending on the mix

—
of wood parts, pine and spruce. 11
 ≔ ← → ≡
17
Table of contents Previous chapter Next chapter Index

—
12
5.6. Other woodparts
Vessel cells from acacia,
eucalyptus, and birch
respectively. The first
5.6.1. Vessel cells 5.6.2. Schlereids (stone cells)
two samples show In softwoods, longitudinal tracheids function Schlereids are present in Nordic deciduous trees
similar vessel cells, as
they are both tropical
both for support and for water conduction. The and can cause problems in the papermaking
trees, compared cells have closed ends. Most hardwood species process. They are a strengthening element in the
to birch which is a
tempered zone tree.
contain short, wide, open-ended cells called tree and have thick, often lignified secondary
vessels, which function for water conduction, and walls. Schlereids are found in wood and bark and
—
13 closed-ended fibers that function for support. are called “stone cells” (Figure 13). They are
Stone cells Vessel elements vary in size and shape between roughly square and not elongated like a fiber.
and among species and are normally much wider They are usually equivalent in width to a shive of
than the fibers (Figure 12). Vessel cells are not several fibers. They can be mistaken for a shive.
wanted in the pulp because they cause problems, Schlereids are hard and highly lignified
such as vessel picking. In Nordic birch, about 9% supporting cells that in paper cause holes, tears
of the wood volume is vessel cells. The percentage and fisheyes because fibers do not bond to them.
of weight is much less because of the thin walls of Schlereids can be measured with a shive analyzer.
vessel cells.

—
13

—
12
 ≔ ← → ≡
18
Table of contents Previous chapter Next chapter Index

5.6.3. Ray cells (parenchyma cells) 5.6.4. Resin acids

The horizontal structure of conifers is composed As mill waters are recycled, the concentrations of
of narrow rays that are only one cell in width but wood extractives increase. Of special concern is
often several cells high. There are two specialized the resin acid because of its toxic nature and
types of ray cells: the ray parenchyma, present in detrimental effects on the pulp and paper making
all species; and ray tracheids, present in only process. Resin acids are released from wood
certain species. Extractives are produced by the during mechanical pulping processes, and high
ray parenchyma cells. Ray tracheids are dead cells. levels of these substances are associated with
Parenchyma cells affect the strength of pulp problems such as pitch deposits, reduced paper
sheets negatively. Generally, most hardwoods strength and decreased brightness. Resin acids
have shorter parenchyma cells than softwoods. are also known to be toxic to aquatic life. Pulp and
About 1–2% of the volume in Nordic softwood paper mills using softwoods as a furnish release
and about 5% of Nordic birch is estimated to be resin acids from wood chips into process waters,
ray cells. regardless of the pulping method.
 ≔ ← → ≡
19
Table of contents Previous chapter Next chapter Index

—
6. Fiber origins

—
14
Pulp fibers can be extracted from almost any 6.3. Birch
Twig of acacia vascular plant found in nature, but a high yield of Birch belongs to the hardwood category, and it
— fibers is necessary to give an economic profit. grows mainly in Europe, North America and
15 Today, pulp from virgin fibers is produced mostly temperate Asia. Their leaves may be toothed or
Twig of birch
from wood fibers (more than 90%). Depending on pointed with serrated edges (Figure 15). The
location and climate, pulpwood properties vary. fibers from birch are much shorter than those
The rest is produced from non-wood fibers like from pine and spruce. They are mainly used in
bagasse, straw and bamboo. chemical pulp production.

6.1. Acacia
Acacia is a fast-growing tree, which grows
naturally in Western and Northeast Africa. It is
raised on plantations for pulp production in e.g.
Asia. They often have many small oval leaves and
attach to the main stem by a leaf stalk called a
leaf petiole (Figure 14). The fibers are short and
suited for fine paper production.

6.2. Aspen
Aspen is a hardwood with brighter fibers than
birch. It grows all over the world, e.g. Europe,
Asia, North Africa and North America. Aspen is
used to produce both chemical and mechanical
pulp. The fibers are short with fine walls and
suited for fine paper. Mechanical pulp from aspen
is so bright that it can be used as a bonding layer — —
in coated fine paper. 14 15
 ≔ ← → ≡
20
Table of contents Previous chapter Next chapter Index

—
16
6.4. Eucalyptus
Twig of eucalyptus Eucalyptus is the world’s fastest growing tree.
— It is raised in plantations where the climate is
17 warm and humid, e.g. in South America, Southern
Twig of pine
Europe and Asia. Eucalyptus leaves are long and
—
18
pointed with smooth sides and a leathery texture
Example of dimensions (Figure 16). The eucalyptus fibers are short and
of fibers (softwood and
hardwood), vessel cells
suited for fine paper production.
and a ray cell. In practice
—
they all vary a lot in size
and must be described
6.5. Spruce 17
with distributions Spruce fibers are, like pine fibers, relatively dark
(see Ch. 16 "Statistics
and reports"). The
and long and grow all over the world. They are Spruce
smallest object to the mainly used in the production of mechanical pulp.
right might be a fibril.
Fibrils build up the
Mechanical pulp made from spruce has an ISO- Pine
layers of the fiber wall. brightness of 60–65%; a sufficient brightness
for newspaper. For the production of many other Birch
grades, the pulp needs to be bleached to achieve
the desired brightness. Birch
vessel cell

6.6. Pine
Eucalyptus
Different species of softwood pine grow all over
the world and can be identified by needles Eucalyptus
vessel cell
growing in clusters (Figure 17). Fibers from pine
are relatively dark and long. They are used mainly
Ray cell
in the sulphate process but also in the production
of mechanical pulp. The high content of resin Fibril
makes it difficult to use fibers from pine in the
sensitive sulphite process.
—
18

—
16
 ≔ ← → ≡
21
Table of contents Previous chapter Next chapter Index

—
Table 2: Properties of pulpwoods

Fiber length Fiber diameter Fiber wall Coarseness

Species (mm) (μm) thickness (μm) (µm/m)
North America
South Pine 3.6–4.9 35–45 5–10 230–300
Northeast region
Spruce, Pine, Fir 3.3–3.5 25–40
Northwest region
Fir, Hemlock 2.5–4.2 30–45
Redwood 6.1 50–65
Aspen 0.7–1.3 10–27 2–3 86
Birch 1.2–1.85 20–36
Beech 1.0–1.5 16–22 2–3.6 86–114
Oaks 1.40 14–22
Red Gum 1.70 20–40
Black alder 1.2 28
South America
Eucalyptus 0.9–1.0 14–16 4–4.5 80–90
Europe
Pine 2.5–3.6 25–39 2.9–4.0 160–200
Birch 1–1.5 16–22 2–3.6 86–114
Spruce 2.5–3.5 24–59 2.9–6.2
Asia
Acacia 0.9–1.0 14 75–90
Balsam 3.5 35–40
 ≔ ← → ≡
22
Table of contents Previous chapter Next chapter Index

6.7. Non-wood fibers • Hemp fibers are extremely long, which makes it
After wood, cotton is the second most important hard to use them to produce a homogeneous
fiber type. For example, linters, i.e. the fine, silky paper sheet. Therefore, they have to be cut to
fibers combed away from the collected cotton, the right size before pulp manufacturing.
are used in security paper. Annual plants Hemp is an annual crop; and the wood
containing very long and strong fibers like flax components are therefore younger than the
and ramie are used mainly in textiles, but they are average tree fibers used in papermaking. There
also found in cigarette and filter papers. Cotton is a very special use for the hemp pulp, which
must be extensively beaten to develop the partly explains the high price. Most of the mills
desired properties. that use hemp fibers are located in China and
India and produce moderate quality printing
Other non-wood fibers besides cotton include: and writing papers. In the western world, hemp
is used in specialty papers like cigarette paper,
• Bamboo is classified as a tropical grass with filter paper (for technical and scientific uses),
a high content of long fibers and is a widely coffee filters, tea bags and security papers.
used paper source. Bamboo pulp is an • Mulberry paper (known as Kozo in Japan) is
important raw material for strong paper in made from mulberry tree pulp. The fiber is
Asia. An interesting feature of bamboo is that obtained from the inner bark of the mulberry.
its fiber characteristics offer a wide range Mulberry contains long fibers, less lignin, high
depending on the species; some bamboos have pentosan and high alpha-cellulose and is
fibers similar to hardwoods, softwoods or an therefore a good raw material for papermaking.
intermediate between the two. However, this chemical composition depends
• Straw pulp is manufactured by a relatively on the age of the mulberry tree. Mulberry paper
simple alkali cooking procedure. In addition to is unique and is a good material for letters,
bast fibers, straw contains cells of different security paper, business cards, wrapping paper,
kinds of botanical character, of which many are hand-made flowers, etc. Because of its special
thin walled and short. characteristics—such as strength, ink
• Bagasse is sugarcane residue left after receptivity and stability—it can be used in many
extracting the juice and consists of short fibers. applications like paper for oil painting, filter
Its fibers are principally used as a raw material paper, paper clothes and tea bags. The main
for unbleached wallboard. pulping process applied in the production of
mulberry paper is the kraft process.
 ≔ ← → ≡
23
Table of contents Previous chapter Next chapter Index

6.8. Synthetic fibers • Influence of atmospheric conditions: Natural

It has been predicted that a widespread use of cellulose fibers are hygroscopic; i.e., they
synthetic fibers may eventually occur in the paper absorb moisture from the air and reach an
industry. An active interest has been evident in equilibrium that depends upon the relative
recent years, from both fiber producers and humidity. Therefore, the moisture content
paper manufacturers. A lot of specialty paper of paper changes with the atmospheric
products are currently being made from synthetic conditions. These changes cause swelling and
fibers. The advantages of synthetic or man-made shrinkage of fibers. Synthetic fibers are not
fibers for papermaking fibers are summarized susceptible to these changes and can be used
below: to produce dimensionally stable papers. Rayon
• Size: Natural cellulose fibers vary considerably swells and absorbs water, which needs to be
in size and shape, whereas synthetic fibers can considered when testing with
be made uniform with a selected length and rayon fibers.
diameter. Long fibers are necessary for
producing strong, durable papers. There are, While synthetic fibers offer many benefits, they
however, limitations to the length of synthetic can be more expensive than wood pulp. The
fibers that may be formed from a suspension cheapest man-made fiber (rayon) costs three-to-
in water because of their tendency to tangle six times as much as an equivalent amount of
and rope together. wood pulp, whereas most of the true synthetics,
• Heat and chemical resistance: Natural cellulose such as the polyamides (nylon), polyesters,
fibers have a limited resistance to chemical acrylics, and glass, cost 10–20 times as much.
attack and heat. Synthetic fiber papers can
be made resistant to strong acids and are This increase in cost does not prevent the use of
therefore useful for chemical filtration. Papers synthetic fibers, but it limits their use to special
can also be made of glass fibers. These papers products in which the extra qualities justify the
have a great resistance to both heat and additional cost.
chemicals.
—
Table 3: Properties of non-wood fibers

Species Fiber length (mm) Fiber diameter (µm)

Bamboo 2.8 15
Straw 1.5 13
Bagasse 1.7 20
Hemp 20 20
Cotton 20 20
Reed 1.2 12
Wheat 1.5 13
Rice 1.5 8.5
 ≔ ← → ≡
24
Table of contents Previous chapter Next chapter Index

—
19
6.9. Recycled pulp 6.10.2. Stickies
To the left: a shive of The definition of secondary fiber (recycled pulp) Stickies are troublesome contaminants found in
1.2 mm length (from
DIP pulp), and a fiber
is any fibrous mass that has undergone a recycled pulp. They originate from adhesives in
as comparison. To the manufacturing process and is being recycled as tapes and labels, self-sealing envelopes, hot
right: a fiber bundle that
has separated during
the raw material for another manufactured melts and latex used in the coating.
the pulping process but product. Fiber substance and strength are lost
“bundled up” again.
during each recycling step. It is generally The best way to deal with stickies includes
considered that a fiber can recirculate on an avoiding them (by selecting the source of pulp),
average of six-to-seven times, so it is necessary removing them during the deinking of
to sort and classify the material into suitable wastepaper, or use slotted screens. Since they
quality grades for an effective utilization of escape the mechanical cleaning and screening,
wastepaper. they end up in the paper machine, wire or in the
end product, causing machine downtime or
6.10. Contaminants downgraded products.There is a pronounced
There are many contaminants that can affect the problem in fine paper production, especially if the
papermaking process (Table 4 on the next page), machine is fast and/or has a twin-wire former.
but we’ll cover the most common.
6.10.3. Dirt
6.10.1. Shives Dirt is any foreign material in pulp. TAPPI defines
Shives are fiber bundles found in pulp and paper. dirt as “foreign matter in a sheet which, when
They derive from incomplete fiber separation examined by reflected (not transmitted) light, has
during the pulping process. Shives are an a marked contrasting color and has an equivalent
important factor that can trigger web breaks black area of 0.04 mm² (0.2 mm × 0.2 mm) or
and also reduce paper strength. more." Dirt content of pulp, and particularly of
recycled pulp, is important for its suitability to
make fine paper. The standard procedure for
measuring dirt content is laid out in TAPPI T213.

—
19
 ≔ ← → ≡
25
Table of contents Previous chapter Next chapter Index

—
Table 4: Common contaminants in recycled pulp.

Contaminant Originates from Types of in-mill processing difficulties

Cleaning system cannot handle adequately, fouls
Hot melts Adhesives and coatings equipment, causes defects in products
Polystyrene foam Blocks, beads, etc. used in packing Difficult to remove - sticks to rolls, indents sheets
Breaks into small pieces, hard to remove,
Dense plastic chips From blister packs and see-through packages causes “shiner” in product
Slows down pulper process,
Plastic films Laminated to paper causes product defects
Does not disperse in pulper,
Wet strength resins Papers treated with resins can cause spots in product
Rubber latex as adhesive, lining or coating, Degrades product,
Latex including flying pasters and rubber bands difficult to remove
Miscellaneous uses; splicing rolls, Sticks to wires and felts and deposits on wires,
Pressure sensitives case-sealing etc. can cause web breaks
Waxes Paper or paperboard laminates and coatings Fails to disperse in pulper, degrades product
Coagulates in pulping process, sticks to wires,
Asphalt Paper or board laminated or coated with asphalt causes black spots in product
Fibers Vegetable and synthetic fibers used for rope Causes product defects and web breaks
 ≔ ← → ≡
26
Table of contents Previous chapter Next chapter Index

—
7. Fiber structure

—
20
7.1. Structure of the wood fiber wall Cellulose is also very important to paper
Illustration of a cell wall Every fiber has a complex inner structure and properties because the attraction between
organization of fibers
a chemical composition that determines its cellulose molecules in different fiber surfaces is
mechanical properties. The fiber structure the principal source of fiber-to-fiber bonding
contains layered walls. Figure 20 shows the in paper.
different layers labelled: P (primary wall), S1
(secondary wall first layer), S2 (secondary wall Hemi cellulose is the second major component
second layer) and S3 (secondary wall third layer). in a wood fiber. It is also a polymer, and it can be
The hollow region in the center of the fiber is removed by mild chemical action. It is built up of
called the lumen. The region between the fibers branched molecule chains of glucose and other
is called the middle lamella and consists mainly monosaccharides. The molecule chains are much
of lignin. The illustration shows the specific shorter than in cellulose. Hemi celluloses are very
alignment of the cellulose fibrils in the secondary important in papermaking because they promote
walls. Fibrils are an amassment of cellulose the development of fiber-to-fiber bonding
molecules, and their orientation can influence through their influence on the ability of fibers
the characteristics of a pulp fiber. to take up water during processing and their
direct participation in the bonding.
7.2. Chemical composition
Wood mainly consists of three types of materials; S3 - Inner layer of the
secondary wall
cellulose, hemi cellulose and lignin (Table 5) . (0.1 µm)
The relative composition in the wood varies
S2 - Main layer of the
in different species of trees (Table 6). secondary wall
(1–5 µm)

Cellulose is the main component of the fiber. S1 - Outer layer of the

secondary wall
It is a (carbohydrate) straight-chain polymer (0.1–0.2 µm)
composed of glucose residues, and it is the
P - Primary wall
structural material out of which the fiber is built. (0.1–0.3 µm)
Cellulose is insoluble in most solvents, and it is
M - Middle lamella
resistant to the action of most chemicals except (0.1–1 µm)
—
strong acids. 20 L - Lumen
 ≔ ← → ≡
27
Table of contents Previous chapter Next chapter Index

—
Table 5: Distribution of chemical components in wood fiber wall. The middle lamella contains mainly lignin.

Wall layer Lignin (%) Cellulose (%) Hemi cellulose (%)

Middle lamella 90 0 10
Primary wall 70 10 20
Secondary wall, S1 40 35 25
Secondary wall, S2 15 55 30
Secondary wall, S3 10 55 35

—
Table 6: Average compositions of softwoods and hardwoods (%)

Species Cellulose Hemicellulose Lignin Extractives and others Vessels Ray cells
Softwoods 42 27 28 3 No Yes
Birch 45 30 20 5 Yes Yes

The third major component in a wood fiber is Besides cellulose, hemi cellulose and lignin,
lignin. Lignin is found between the fibers where it the wood also contains extractives (for example
serves as a binding agent. It also occurs within resin). These contaminants form unwanted dots
the fiber wall. Lignin is very complex and does not on the paper. Chemical pulping and bleaching
dissolve in water or other common solvents, but processes mostly remove these components.
it can be made soluble by chemical action. The However, trace amounts do survive and cause
purpose of all chemical pulping processes is to problems in pulp and paper mills. Examples of
dissolve and remove lignin. However, there is no problems are linting in offset printing and
practical method for complete lignin removal by missing dots in gravure printing.
pulping. The residual lignin gives the unbleached
paper a brownish tint, and if white paper is to be The compositions of softwoods and hardwoods
made, the remaining lignin must be removed by differ from each other. Hardwoods contain more
bleaching. Lignin prevents the formation of fiber- cellulose, skeletal polysaccharides, extractives
to-fiber bonds in paper, reducing its strength. and hemi celluloses than softwoods, and the
lignin content is lower.
 ≔ ← → ≡
28
Table of contents Previous chapter Next chapter Index

7.3. Degradation pattern 7.4. Importance of fiber properties

In chemical pulping, the middle lamella is Paper properties depend strongly on the
dissolved during cooking. The primary wall and structure of the fibers that build up the sheet.
the first layer of the secondary wall (S1) are also The manner in which fiber properties affect sheet
affected but may remain after the digester properties is described in Chapter 18 "Impact of
depending of the yield. Mechanical refining will fiber on products". The fiber length and other
initially remove parts of these layers. For lower fiber dimensions are influenced by the method
yields or after bleaching, these layers will be used to separate the fibers. To manufacture
removed. The fibrils are strictly oriented in the Stone Groundwood (SGW) for example, the fiber
secondary wall (S2) and, according to the is shortened considerably. Chemical pulp is
literature, the fibril angle (relative to the fiber cooked, and in this process, a large amount of
axis) is known to correlate well with mechanical substances other than cellulose is removed by
properties of the sheet. dissolution.

The fibril angle changes during the first years This leads to a decrease in the fiber wall
of the tree’s growth and becomes more constant thickness, resulting in collapsed fibers (two
with age. The non-oriented layers P and S1 opposite walls are pressed together and
function as a kind of bond for the fiber. When completely or partly eliminate the lumen) and a
these layers are removed, which happens greater strength in the paper. From a general
in chemical pulping after some mechanical point of view, long fibers enable good paper
treatment, the orientation of the fibrils in the strength properties.
S2-layer affects the fiber shape. Small changes
in the length of the fibrils affect the shape of The curvature of the fibers has an influence
the whole fiber. on paper properties. Increased curvature
contributes to a high stretch ability and tensile
In mechanical pulping, the lignin is softened at a energy absorption, which is important in sack
high temperature, which helps to separate the paper for example. Thick-walled fibers provide
fibers mechanically. Lignin rich parts will still be high bending stiffness but weak paper; thin-
a part of the fiber fractions. More information walled fibers provide low bending stiffness
about refining and degradation pattern is given but good paper strength.
in Chapter 9.
 ≔ ← → ≡
29
Table of contents Previous chapter Next chapter Index

—
8. Manufacturing processes – an overview

—
21
Fiber properties vary within single trees, between
Papermaking starts trees (depending on the growing situation) and
with the selection
of raw material
of course between different species. Selection
of wood is the start of the papermaking process.

In recent years, more attention has been paid to

this early part of the process (Figure 21).

8.1. Table of symbols

Following is a presentation of the symbols that
are used in the different process illustrations in
this chapter.

—
21
—
Table 7: to be referenced for process illustrations used in this chapter.

Recommended
Impregnation Refining Hydro cyclone Screen Dispersion Batch digester Continous digester Flotation sampling positions

Screw Storage Latency Bleaching Washing Disc filter O1 Pulper (Slushing) Chips Sampling point
 ≔ ← → ≡
30
Table of contents Previous chapter Next chapter Index

—
22
8.2. Log preparation importance but thickness and length are (Figure
Log preparation to Large amounts of wood are handled by the pulp 23). The thickness of the chips is most important
create chips in
a pulp mill.
mill, and the organization of the wood yard has a because it helps ensure that all the wood material
considerable impact on the final pulp quality. can be efficiently impregnated with cooking
—
23 Chip damage means reduced pulp quality, and chemicals. The average fiber length is reduced
Typical dimensions fiber losses have a severe effect on the economy after chipping and depends on the chip length.
of a wood chip.
of the mill. When the logs arrive at the mill, their Chipping is usually carried out in a chipper or
weight or volume is determined. If the logs are reducer chipper. Before pulping, the chips are
frozen, de-icing is required in order to facilitate screened to ensure a uniform size distribution.
barking. Pin chips and fines are discarded and used as
fuel. Oversized chips are reduced in size and
Sand and stones are removed in the de-icing step fed to the refiner or digester.
as well as after barking. This is done particularly
to reduce the wear on the chipper knives. For that
reason, a metal detector is also placed before
the chipper infeed to detect any metal scrap.
Thereafter the logs are reduced to smaller pieces
thickness
in the chipper and passed on to storage. 6 – 10mm

8.2.1. Chipping
length
To promote a fast and uniform release of the 15 – 40mm
lignin in wood, logs are cut into chips (see Figure
22). The width varies and is not of great —
23

Oversized

To digester
or refiner
—
22 Chips from sawmills
 ≔ ← → ≡
31
Table of contents Previous chapter Next chapter Index

8.3. Pulping processes However, since the lignin is retained in the pulp,
In a pulping process, wood is separated into papers from mechanical pulps tend to yellow; a
fibers. This can be achieved mechanically, reduction in brightness as a result of aging or
thermally, chemically or through a combination of exposure to light.
these treatments (Table 8).
Mechanical pulps are therefore used in short
8.3.1. Mechanical Pulping lifetime papers. Mechanical papers are weaker
The principle of mechanical pulp making is than those with chemical pulps, principally
to liberate wood fibers entirely by mechanical because the fibers are shorter and stiffer,
means. There are two fundamental mechanical resulting in less bonding points. Mechanical pulp
pulp manufacturing processes: the groundwood is mostly produced from softwood, and is mainly
and the refining processes. The pulps used in newspaper, weekly magazines, some
manufactured are called stone groundwood tissue products, and cardboard.
(SGW), refiner mechanical pulp (RMP) and
thermo-mechanical pulp (TMP). The mechanical 8.3.2. Online sampling
pulp manufacturing process is very high in energy A number of process illustrations are presented
consumption. in this chapter, with recommended marked
positions for automatic sampling equipment like
The advantage of mechanical pulp is the high ABB's L&W Fiber Online. Besides the marked
wood yield—over 95%. The pulp also has a high positions, manual sampling with benchtop
opacity, which means that thin papers with good equipment like L&W Fiber Tester Plus is
printing properties can be manufactured. recommended on each individual machine
(refiner, etc).
—
Table 8: General classification of pulping processes

Mechanical Hybrid Chemical

Pulping with combinations of chemical and
Pulping by mechanical energy mechanical treatments Pulping with chemicals and heat
High yield (90–95%) Intermediate yield (55–90%) Low yield (40–55%)
Includes a significant number of fines “Intermediate” pulp properties Long, pure fibers –
and medium fractions – Weak paper (some unique properties) Good strength
Softwood Softwood and hardwood Softwood and hardwood
Good print quality Poor print quality
Difficult bleaching Easy bleaching
Examples: Examples: Examples:
Stone groundwood NSSC
RMP BCTMP Kraft
TMP CTMP Sulphite
 ≔ ← → ≡
32
Table of contents Previous chapter Next chapter Index

—
24
8.3.3. Stone groundwood pulping and 8.3.4. Refiner mechanical pulping
SGW-process. pressurized groundwood pulping Refiner mechanical pulp (RMP), is manufactured
The pulp is screened in
holed or slotted screens
The oldest method of mechanical pulping is by the defibration of chips or sawdust in a disc
and in hydro-cyclones. the stone groundwood process (SGW), where a refiner. Softwood chips pass through a narrow
The rejects are treated
in a reject refiner. Fines
block of wood is pressed lengthwise against a opening between rotating metal discs that are
content, length/width roughened, revolving grindstone (with a diameter 1.5 m in diameter. The defibration was initially
distribution, shives
and freeness are of
of 1.5 m and a rotation speed of 240–300 rpm) carried out at atmospheric pressure but refining
interest to monitor. and the fibers are torn out and washed away from under increased steam pressure was later
the stone surface (Figure 24). The water also introduced.
cools the grindstone. While the principle is
simple, it may be hard to achieve a uniform The process usually involves two refining stages
quality. The roughness of the grindstones must in a series, producing a longer fibered pulp than
be carefully controlled, and the pressure against conventional groundwood. The pulp is stronger,
the grinding stone as well as the temperature and bulkier and contains fewer fines, but it
water flow must be checked. is somewhat darker in color than stone
groundwood. The mechanical properties
Pressurized groundwood (PGW) is produced by are also better.
a similar method where the grinder is pressurized
with steam. This improves the fiber separation 8.3.5. Thermomechanical pulping
and the pulp gets better physical properties than Thermomechanical pulp (TMP) is the most
SGW- pulp, and also less energy is required than common type of mechanical pulp and is a
in the SGW-process. modification of RMP. It involves steaming the
raw material for a short period of time prior to
and during refining (Figure 25 on the next page).
The steaming has a softening effect on the chips,
resulting in pulp with a greater percentage of
long fibers and fewer shives than RMP. These long
fibers produce a stronger pulp than either SGW or
RMP.

—
24
 ≔ ← → ≡
33
Table of contents Previous chapter Next chapter Index

—
25
8.3.6. Chemical pre-treatments If the pulps are bleached, they are called bleached
Simplified process A thermomechanical pulp can be modified with chemical mechanical pulp (BCMP) and bleached
illustration over the
TMP-process. Two
small amounts of chemicals, usually sodium chemical thermo mechanical pulp (BCTMP –
refining stages are sulphite (Na₂SO₄), prior to defibration in a refiner. Figure 26 on the next page).
used here (1, 2 or 3
stages can be used)
This pulp is called chemical mechanical pulp
followed by a latency (CMP) or chemical thermo mechanical pulp There are many non-integrated BCTMP mills
chest and screening
in two parallel lines.
(CTMP). Hardwoods such as birch and aspen can producing market pulp.
Rejects are treated also be refined with these methods.
in a reject refiner,
which is also followed
Refining is usually done at high consistency
by a latency chest. The additives usually result in an improvement (30–40%), but it has also becomes more and
Freeness, shives
and fiber length are
in pulp strength, while the opacity and light more common to include low consistency
commonly measured scattering ability decrease. refining in mechanical defibration systems.
online. Manual
sampling positions
are marked after the
individual refiners.

—
25
 ≔ ← → ≡
34
Table of contents Previous chapter Next chapter Index

—
26
8.3.7. Chemical pulping 8.3.8. Kraft process
Simplified process In chemical pulping, the objective is to degrade The most common method is the (alkaline) kraft
illustration over the
BCTMP process.
and dissolve away the lignin and to leave behind process, which can be used with all kinds of
The TMP process is most of the cellulose and hemi cellulose, i.e. the wood. Advantages of the process are high
combined with pre-
impregnation stages
fibers. This is achieved by treating the wood chips chemical recovery and high pulp strength.
and with bleaching after at an elevated temperature in a solution Chemical recovery is an essential part of making
washing. Brightness,
freeness, shives and
containing the pulping chemicals until a certain the kraft process economically feasible. The kraft
fiber length are standard degree of delignification is reached. process involves cooking the wood chips in a
measurements.
solution of sodium hydroxide (NaOH) and sodium
The process is referred to as cooking (Figures 27 sulphide (Na₂S). The cooking process softens and
and 28 on the next page). It is undesirable for the dissolves the lignin that holds the fibers together
cooking liquor to boil, and thereby generate in the wood.
vapor, so it is performed in a pressurized system.
After cooking, the pulping chemicals are As a result, the fibers are readily separated when
recovered. the pulp is expelled from the digester. Strong
paper products are manufactured from kraft
Almost all of the lignin is generally dissolved, but pulp, but the unbleached pulp has a dark brown
the amount of cellulose and hemi cellulose is also color. The yield of unbleached kraft pulp is 65–
reduced. The yield of chemical pulp is low relative 70% (used in bags and linerboard), bleachable
to mechanical pulping, usually 40–70%. kraft pulp has a yield of 47–50%, and bleached
The two main procedures are the (alkaline) kraft kraft pulp has the lowest yield, 43–45%, and is
process and the (acidic) sulphite process. used in white papers.

—
26
 ≔ ← → ≡
35
Table of contents Previous chapter Next chapter Index

—
27
Continuous digester,
unbleached pulp. The
pulp is washed and
screened after the
digester and different
kinds of mechanical
refining are often —
carried out in the pulp 27
mill. Kappa number,
shives content,
fiber properties
and freeness are of
interest to monitor.

—
28
Batch cooking. Five
batch digesters are here
followed by a buffer.
The pulp is screened
before entering the —
oxygen reactor and a 28
new buffer. A bleach
sequence is performed
before the final pulp
storage. Brightness,
shape factor, fiber width
and length, vessel cells
8.3.9. Sulphite process
(HW) are monitored. Sulphite pulps are lighter in color than kraft pulps Resinous softwoods are more difficult to handle.
and can be bleached more easily. The yield is in Sulphite pulping makes the lignin water-soluble.
the region of 48–51%, but the resulting paper is Due to its sensitivity to wood species, weaker
weaker than that made with kraft pulp. The pulp, greater difficulty in chemical recovery and
(acidic) sulphite process works well for long cooking time, sulphite pulping is not used
softwoods such as spruce, fir and hemlock, and much today.
hardwoods such as poplar, birch and eucalyptus.
 ≔ ← → ≡
36
Table of contents Previous chapter Next chapter Index

—
29
8.3.10. Advantages of the two major processes The NSSC process utilizes sodium sulphite
NSSC cooking. A for chemical pulping cooking liquor, which is buffered with sodium
continuous digester
combined with high
• Kraft process: carbonate to neutralize the organic acids
consistency refining. - Produces highest strength pulp liberated from the wood during cooking (Figure
The shives content
before refining can
- Handles wide variety of species 29). An optimal balance between cooking and
be as high as 12%. - Tolerates bark in the pulping process refining results in the least amount of fiber loss
Kappa number, shives
and fiber dimensions
• Sulphite process: for the best economy. The shives content is quite
are important - Produces bright pulp that is easy to bleach to high. As a prime example, NSSC- pulp is suitable
measurements.
full brightness for the center layer in corrugated containerboard.
- Produces higher yield of bleached pulp than
the kraft process 8.3.12. Dissolving pulp
- Produces pulp that is easier to refine for Dissolving pulp is a chemical pulp intended
papermaking applications primarily for the preparation of chemical
derivates of cellulose. It is utilized for chemical
8.3.11. Semi-chemical pulping conversion into products such as rayon,
Semi-chemical pulping combines chemical and cellophane, cellulose acetate, cellulose nitrate
mechanical pulping. The wood chips are partially and carboxymethyl cellulose. Dissolving pulp is
softened or digested with chemicals, and the rest made by either a modified kraft or a modified
of the pulping process is carried out mechanically, sulphite process.
most often in disc refiners.
The purpose is to produce fairly pure and
Semi-chemical methods include pulp yields in the uniform cellulose. Lignin and hemi celluloses are
range of 55–90%. The NSSC process (neutral considered to be contaminants and are therefore
sulphite, semi-chemical) is the most common removed. The major raw material is softwood, but
used semi-chemical process. It is mainly applied some hardwood is also used.
to hardwood chips.

—
29
 ≔ ← → ≡
37
Table of contents Previous chapter Next chapter Index

—
30
8.3.13. Recycling and De-inked Pulp (DIP) Thereafter, the pulp passes through screens and
A recycled fiber plant The paper industry is the sole user of recovered the ink particles are removed from the pulp
with its different
operational stages.
paper as a secondary raw material. According to through flotation. The principle for flotation is
Measurements of Fisher Solve, recycled fibers make up 57% of the to inject tiny air bubbles into the pulp suspension.
residual ink, brightness,
fines content, ash.
total fiber consumption in the pulp and paper The suspension is then allowed to flow slowly
industry. Wastepaper consists of different types through a basin; the air bubbles rise to the
of paper that has been collected after use, surface, and on their way, they will capture hydro-
i.e. corrugated fiberboard, magazines and phobic ink particles, which are gathered on the
newspapers. The different operating stages of surface and removed as foam. The separation is
a recycled pulping process are described below not completed in one step and it is therefore
(Figure 30). repeated. To remove the remaining ink-particles
and plastics, the pulp is heated and dispersed,
When recycled wastepaper arrives at the mill, it is and then bleached and washed.
sorted and cleaned. For example, if newsprint is
to be manufactured, the recycled paper has to be Recycled pulp fibers do not reach the same
separated into individual fibers and cleaned from strength as virgin pulp fibers, since they have
mechanical impurities, stickies and printing ink. been dried once before. Decent properties can be
The process consists of two steps: loosening the achieved through beating, but a larger number of
ink from the fiber surface and separating the ink fines is also released, hampering the drainage on
from the process water. The paper bales are fed the wire. The recovered paper is usually a mixture
to the slusher and tumbled in water. The paper is of various fiber types or paper grades. Recovered
torn and turned into pulp by moving slowly paper contains a certain number of contaminants
through the slusher. Chemicals needed to remove and other detrimental substances. Nevertheless,
the printing ink are added to the slusher. recycled fibers must meet defined quality criteria
for a given furnish, just as virgin fibers.

—
30
 ≔ ← → ≡
38
Table of contents Previous chapter Next chapter Index

8.3.14. Bleaching Bleached chemical pulps are found in a number of

Bleaching of pulp involves removal or paper products such as fine paper, board, tissue,
modification of the colored components in pulp liner, etc., and most of the chemical pulp
with the aim of increasing its brightness. produced is bleached. The bleaching process
Unbleached pulps display a wide range of results in a brighter pulp with higher cleanliness
brightness values. The sulphite process produces with respect to a low number of shives, dirt and
the brightest pulps (65, according to ISO extractives. The bleaching process also removes
brightness), whereas those produced by the kraft, bacteria, which is a necessity for food and liquid
soda, and semi chemical processes can be quite packaging board.
dark. Mechanical pulp brightness is mainly a
function of the wood species and condition; Bleaching is usually carried out in a step-wise
species are generally selected to provide sequence utilizing different chemicals and
brightness in the 50–65% range. conditions in each stage, with washing in
between the different stages. The generally used
8.3.15. Bleaching of chemical pulps chemical treatments are listed in Table 9 on the
There is an essential difference between the next page. Usually chlorination and extraction are
bleaching of chemical pulps and mechanical carried out in a sequence, where chlorinated
pulps. The bleaching of chemical pulps is usually a lignin compounds are first formed and then
multi-stage process where the lignin is oxidized, solubilized in a subsequent extraction stage.
decomposed and finally eliminated from the pulp
fibers. This results in less chromophores in the The intention is to delignify the pulp, since very
pulp. In the bleaching of mechanical pulps, the little brightening takes place through the CE
lignin is not decomposed and removed from the sequence. The oxygen stage is used mainly for
fiber; instead oxidizing or reducing chemicals delignification.
attack the chromophores in the lignin. A large
amount of lignin is therefore still present in the Sulphite pulps and hardwood kraft pulps are
pulp, but the pulp is brighter. bleached more easily than softwood kraft pulps
because of their lower lignin content. A
Unbleached chemical pulps, mainly kraft pulps, brightness level of 89–91% in a softwood kraft
have a dark brown color, as a result of the large pulp can be achieved with a bleaching of six
number of chromophores in the pulp, which are stages. A CEDED, CEHDED, or OCEDED sequence
mainly found in the residual lignin. Numerous can be employed. Lower brightness levels can be
paper products like printing and writing paper, achieved with fewer stages.
food board, etc. require a bright pulp, as the
requirements for a smooth surface with good
printability and other related properties are high.
 ≔ ← → ≡
39
Table of contents Previous chapter Next chapter Index

—
Table 9: Symbols for bleaching stages. 8.4. Paper manufacturing
Different pulps are used to optimize the
Bleaching stage Symbols Comment properties of the fiber mix fed into a paper
Treatment with acid A machine. If the paper machine has only one layer,
Phasing out due to various paper requirements still have to be
environmental concerns
over dioxin formation.
satisfied. Long fibers give strength. Short fibers
Not very selective, give better formation and a more even surface. It
carbohydrate is necessary to use broke. Other components can
Chlorine C degradation also occurs.
also be included in different situations. The
Chlorine dioxide D Expensive but selective.
quality of the different components and of the
Removes fragments of
lignin. The effluent is
blend determines the quality of the paper and
Alkaline extraction E highly coloured. also the runnability on the machine.
More selective than
Hypochlorite H elemental chlorine. 8.4.1. Refining
Least selective for lignin.
Refining, or beating, involves a mechanical action
Oxygen O Cheapest to use.
carried out in continuous conical or disk-type
Highly corrosive.
Becoming more common refiners, where the flow is parallel to the bar
with the elimination of crossings. The goal is to modify the pulp fibers in
Hydrogen peroxide P chlorine.
an optimal manner to meet the demands of the
Chelating stage Q
particular papermaking furnish. As a result of
Peracetic stage T refining, the fibers are more flexible and conform
Treatment with
more readily, and the surfaces are modified to
water W
improve the bonding. But refining also lowers the
Enzyme stage X
drainage capability of the pulp, reducing the
Dithionite Y
production rate and increasing energy
Ozone Z
consumption. It can also cut fibers.

Some paper properties are improved with

beating but others deteriorate. An optimal
balance has to be reached.
 ≔ ← → ≡
40
Table of contents Previous chapter Next chapter Index

—
31
8.4.2. Blend chest 8.4.4. Paper machine
Mixing and beating Figure 31 shows how pulp A and pulp B from A paper machine consists of a number of
of pulp to get certain
paper properties.
different storages are refined separately. After components with different functions: The
Brightness, blend refining the two components are mixed in the headbox is a pressurized flow box from which the
(fiber dimensions),
fines content, shape
blend chest. Broke from the paper machine is diluted pulp slurry is evenly spread out onto the
factor and freeness also recirculated and mixed in the blend chest. moving wire to form a web. Several headboxes
are of interest.
Different mixes are used for different grades; may be used. The cross-machine profile is
— the broke may come from earlier grades. controlled by sectionalized slice screws and/or
32
Paper machine wet by sectionalized dilution across the headbox.
end. Fines and fillers 8.4.3. Additives The speed of the water jet relative to the speed
are partly circulated,
depending on the Chemicals are added to the furnish for various of the wire orients the fibers more or less in the
retention. Shape reasons and they can be categorized as functional machine direction. The wire allows water to be
factor, fines, blend
(fiber dimensions) additives or control additives. Examples of drained and retains the fibers. Twin-wire
are of interest. functional additives are dyes, sizing agents, machines are common and several single wires
strength aids and fillers. Examples of other and headboxes can also be used on the same
functional additives are fillers like kaolin and clay. machine. Fines and fillers can pass through the
Sizing agents help to repel water, and strength wire and have to be bonded to fibers or flocs to
aids are polymers that bind to the fibers to be retained on the fiber web (more on this later
improve strength. Examples of fillers are kaolin in the book). Special chemical systems are used
clays and titanium dioxide. Examples of control for this purpose. Figure 32 shows the water
additives are drainage aids and retention aids. circulations. Fines and fillers build up in the
Drainage aids increase the rate of water removal circulations, depending on the retention.
and retention aids help to retain fines.

Pulp A

Mixed pulp
to PM Recirculation To purification

Filler
Pulp B White water
Thickstock

Broke from PM

— —
31 32
 ≔ ← → ≡
41
Table of contents Previous chapter Next chapter Index

This is a dynamic system, and this causes

problems when changes are made to a different
paper grade. The drainage properties are
important for the runnability of the machine.
The paper web enters the press section in order
to remove water that has not been removed on
the wire. Pressing is an energy-efficient way of
getting rid of the water. The structure of the web
influences the efficiency of the press section. The
remaining water has to be removed by drying the
paper on steam-heated hot cylinders, which uses
energy. In the dryers, the web can be more or less
restrained. Restrained drying often gives
a more uniform web and often better quality
but for some paper grades, free drying is an
advantage. The fiber properties also have an
influence.

8.4.5. Post-drying treatments

One example of a post-drying treatment is
calendering, which is the passage of paper
through one or more nips formed by a set of iron
rolls. The primary objectives of calendering are to
get a smooth surface, reduce sheet thickness to
the desired level and even out caliper variations.
Many printing paper grades require a more even
surface, which can be achieved by applying a
surface coating. This is done online or offline.
 ≔ ← → ≡
42
Table of contents Previous chapter Next chapter Index

—
9. Fiber/Process

—
33
Chemically separated fibers are flexible and have Refining causes deformation of the fibers. The
Chemical pulps give a high bonding potential. At the same time, they fiber deformations in mechanical pulps are called
better bonding, but
lower light scattering,
are not damaged and have kept their length in the “latency” because there is a latent strength
and lower yield than separating process. They make strong paper. associated with their removal. By heating the
mechanical pulps.
fibers up to 85°C for a period of time, the
— Why then separate fibers mechanically? Because curliness of the fibers is removed.
34
TMP fibers mechanical pulp has a high yield and a high light
(magnification 100×) scattering power. Refining increases both tensile The shape factor, also known as form factor,
index and light scattering power. This is not the is an important measure of pulp quality and
case for chemical pulps; refining increases tensile is a measure of straightness of the fibers
index but reduces light scattering power (projected length, divided by the true length).
(Figure 33). For mechanical pulp, it seems to be proportional
to temperature.

Mechanical pulps After removal of latency, the fibers are quite

Light scattering

straight and even when heated to a certain

temperature, the shape factor seems to stay at
Chemical pulps
that level. If the latency changes, the drainage
properties of the pulp change.

Tensile index

—
33

9.1. Mechanical pulps

Fibers from the Thermomechanical pulping (TMP)
process are stiff and wide with high lignin
content. A medium fraction and fines can be seen
in Figure 34. Ray cells can also be seen in the
picture; they are shorter and narrower than the
fibers. Other parts of the fines are derived from
the cell wall. Bigger flakes and more fibrillated —
material can also be seen. 34
 ≔ ← → ≡
43
Table of contents Previous chapter Next chapter Index

—
35
If the latency is removed before determining the Freeness is to a great extent a measure of the
CTMP fibers CSF (Canadian Standard Freeness) number, this fines content. The problem is that the drainage
(magnification 100×)
source of instability is removed (Reference to P52 resistance comes from ray cells, particles with
and P78). However, the CSF value will then be poor bonding potential, and fibrillar material with
different in the actual pulping process. Fiber good bonding potential. Another reported result
analysis makes it possible to measure from refining of TMP is split fibers, which appear
deformations very accurately. This makes it to be generated in first stage refining—especially
possible to avoid hot disintegration of the pulp in earlywood fibers. The amount of split fibers
before measuring the CSF value. has been reported to be up to 30% of the total. A
split fiber wall will yield a more flexible fiber. The
CSF still seems to be the most accepted stiffness of mechanical fibers is much higher than
measurement of quality for mechanical pulp, for chemical fibers and the reason why chemical
although greater accuracy, better repeatability fibers produce much stronger paper than
and less maintenance is desired by users. For a mechanical fibers.
simple product, with only one type of fiber having
few and similar grades, a standardized drainage Unseparated fibers (shives) are also present in
measurement may be sufficient. Today it is very mechanical pulps, and it is of great importance
common that different kinds of fibers are used to keep these under control. TMP is often made
and mixed to produce more advanced products, from spruce, which requires less refining energy
meaning that a simple measurement is of limited than pine. Thick-walled, Southern U.S. pine is also
value. used in the TMP process, generally with three-
stage refining.

Chemical thermo mechanical pulp (CTMP) fibers

are similar to TMP fibers, but they are one step
closer to chemical fibers with more narrow fibers
and less medium fraction and fines (Figure 35).
The fibers are deformed a lot in refining of
mechanical pulp, but the treatment at a high
temperature reduces the deformation to a large
extent, but not entirely.

—
35
 ≔ ← → ≡
44
Table of contents Previous chapter Next chapter Index

—
36
Figure 36 shows that two investigated refiners
Shape factor measured worked differently. After latency removal, the
on pulp taken directly
from two parallel
fibers were quite straight. The same effect with
refiners and dissolved respect to shape factor occurs in the reject
in cold water, and
on pulp taken later
refiner. Deformed fibers are found after the
in the process. refiner, when pulps are sampled and diluted in
— cold water. After latency removal with hot water,
37 the fibers are straightened out. Fiber length and
The average fiber
width decreases faster fiber width decrease with increased refining
after 3.3 MW because energy. The development of fiber width is shown
of fiber cutting.
in Figure 37; at 3.3 MW, both fiber length and
—
38
width decrease due to the cutting of fibers
A cut fiber (Figure 38) and the generation of smaller —
38
particles.

Shape factor CTMP 9.2. Unbleached kraft pulp

91
After latency chest Fiber deformation can occur when the pulp leaves
90
After washing the digester, but unbleached pulp fibers are
Shape factor (%)

89
88 generally still quite straight, although the shape
Refiner A
87 factor can drop after a limited amount of refining.
86 Further refining can straighten the fiber again.
85
Refiner A High Consistency (HC) refining is often used in
84
0.5 1 1.5 2 2.5 the production of sack kraft paper. It introduces
Sampling position on process deformations of the fiber, which leads to an
— improved stretch in the sheet with improved
36
Tensile Energy Absorption (TEA) as a result.
Fiber width reject refiner, CTMP
39
Low Consistency (LC) refining improves swelling
Average fiber width

37
35
of the fibers. Some of the fiber properties
33
achieved in HC-refining are lost in LC-refining.
31
29 There is always a balance (optimization) between
27 properties. Both HC refining and LC refining
25 reduce the number of shives significantly. Some
0 1 2 3 4
Electric power (MW)
primary fines can be seen in Figure 39 on the
—
next page. The purpose of refining is to improve
37 bonding strength without cutting the fibers too
much.
 ≔ ← → ≡
45
Table of contents Previous chapter Next chapter Index

—
39
9.3. Bleached kraft pulp 9.4. Softwood and hardwood fibers
Unbleached Refining energy needed to develop strength is Figure 40 shows bleached softwood fibers.
kraft pulp fibers
(magnification 50×)
different for individual fiber species. Hardwood Compared to the unbleached fibers (Figure 39)
fibers need less specific refining energy than we can see a lot of deformed fibers. Figure
—
40 softwood fibers. Bleached pulp always contains 41 shows softwood fibers at 100 times
Bleached softwood very deformed fibers. Mechanical treatment in magnification, where more details of
fibers from chemical
pulp (magnification 50×) combination with the chemical treatment seems deformations can be seen: both local
to cause these deformations. kinks and softer bends.
—
41
Bleached softwood The deformations are created in the process. The deformations in bleached hardwood pulps
fibers from chemical
pulp (magnification Some of the deformations seem to be reversible are similar, as can be seen in Figure 42. It is a
100×) and some irreversible. The reversible birch fiber, with typically long vessel elements
— deformations can be “healed” by a proper (the wide objects are vessel cells). Hardwood
42
Birch fibers with vessel
beating of the fibers. fibers are more sensitive to refining than
cells from chemical pulp softwood fibers.
(magnification 50×)

— —
39 41

— —
40 42
 ≔ ← → ≡
46
Table of contents Previous chapter Next chapter Index

—
43
9.5. After bleaching 9.6. After refining
Fibers at different The shape factor is a key quality factor for If a pulp is beaten in a PFI Laboratory Mill (used in
stages in the process
industrial chemical pulps. A high shape factor the laboratory for the beating of chemical pulps
— is in most cases positive for the paper strength under standardized condition), the shape factor
44
Deformations along the and means straight fibers. Fiber deformation is will improve significantly (Figure 45 on the next
pulping process line always found in chemical pulp. The deformations page). However, if the same pulp is beaten in an
increase through the process (Figure 43). industrial refiner, the result is seldom as good.
The improvement in shape factor also gives a
A main drop in the shape factor often occurs in corresponding improvement in tensile index.
the oxygen delignification stage, and the shape The difference in tensile index between PFI and
factor continues to decrease during bleaching. industrial beating is explained by the difference
The most striking feature of the case shown in in shape factor.
Figure 44 was that a considerable decrease of
shape factor was seen already before the O2 One way to characterize refining is by using the
stage. This phenomenon may be because wash specific edge load. It is easier to straighten the
liquid containing residual products from pre- fibers and obtain a straightening of the fibers
bleaching was washed out from the diffuser, similar to that in PFI refining using a lower-
causing the fibers to be more sensitive to specific edge load. PFI mill is known to have a
mechanical treatment in the process. fibrillation effect on the fibers. The difference
in effects of each type of refining can be seen
in Figures 46 and 47.
Fibers at different stages in the process
Shape factor in sulphate mill
91
Filter
90
Shape factor (%) 89

a. b. c. d. 88
Before O₂
87
a.Tree structure.
b.Middle lamella and parts of the outer layers (P, S1) are removed. 86
After O₂
c. A
 fter refining the fibers are fibrillated and starts to swell 85
(to some extent). After diffusor
84
d. A
 fter drying the fibers are partly collapsed. Fiber collapse After bleaching
83
also takes place earlier in the process.
Sampling position in process

— —
43 44
 ≔ ← → ≡
47
Table of contents Previous chapter Next chapter Index

— 90
45

Shape factor (%)

88
PFI beating represents
a kind of ideal beating, 86 PFI
showing the strength
84
potential for a pulp.
82 Industrial
— 80
46 90 100 110 120 130 140 150 160 170
Bleached sulphate,
PFI beaten WRV (%)
(magnification 100×)
Tensile stiffness index (kNm/g)

90
—
47 88
Bleached sulphate,
86 PFI
beaten in an
industrial refiner 84
(magnification 100×) Industrial
82
— 80 —
48 90 100 110 120 130 140 150 160 170 47
Shape factor vs.
Water Retention Value WRV (%)
(WRV) with Kraft
pulp (Mohlin, RISE). —
45 94

Shape factor (1.5–3 mm)

92

90 Unbleached kraft

88

86

84

82 Bleached kraft

80
100 120 140 160 180 200 220 240

WRV (%)
Conical, high prod. Disc ref, high prod.

Conical, low prod. Disc ref, low prod.

— —
46 48
 ≔ ← → ≡
48
Table of contents Previous chapter Next chapter Index

9.8. Screening of shives

9.7. Swelling
To evaluate a screen or a screening system, the
Modification of the fiber wall is the most
efficiency of shives removal relative to the reject
important effect of beating; the fiber becomes
rate of the screen should be compared with the
more flexible and more fibrillated. Swelling
long fiber yield relative to the reject rate of the
increases sheet strength. Hemi cellulose is
screen. A combined fiber and shive analyzer is an
considered to play a central role as a swelling
efficient tool in such evaluations.
factor in the fiber. Hemi-cellulose and amorphous
cellulose can be found between the crystalline
9.9. Fractionation of fibers
parts of the fiber wall. Because of its structure,
How can expensive fibers be better utilized? One
the fiber will swell mainly in the plane of the cross
way is to separate the long fiber fraction from the
section not along the fiber axis. The swelling is
pulp and use the long fiber fraction for
prevented by the primary wall, the outer
reinforcement purposes in one production line,
secondary wall, areas with string hydrogen bonds
where improved strength is important, and use
and lignin with a hydrophobic character, and a
the short fraction for fine paper production in
layered mix with the cellulose.
another line, where formation and surface
uniformity is more critical than strength.
Water is retained by the fiber bound to internal
and external surfaces (hydrogen bonds), in the
In a recycled pulp mill, the problem can be the
lumen, in pores in the cell wall and within capillary
opposite. The fines fraction may only reduce
forces on the fiber surface. Swelling decreases
production efficiency and not contribute to the
with increasing temperature.
strength of the paper. Then it can be optimal to
separate out the fines fraction and perhaps be
Partial disintegration of the fiber wall during
able to burn it. An online measurement system
beating allows the fiber to swell. The swelling ex-
can be used to optimize this process.
plains why reversible fiber deformations can be
healed during beating. The healing can be
Fractionation with hydrocyclones separates
measured as an increase in shape factor. A slight
particles with respect to surface area and
increase in fiber diameter is often seen as well.
density. High-density particles will go into the
This is similar to a balloon effect, when the fiber
rejects and particles with a large area will be
wall swells. But most of the swelling enlarges the
accepted. A very special application would be to
fiber wall towards the lumen. The swelling can
separate earlywood and latewood. Attempts have
also be local at a specific deformation point of
been made to do this kind of separation with
the fiber.
hydrocyclones.
 ≔ ← → ≡
49
Table of contents Previous chapter Next chapter Index

—
49
Earlywood fibers and latewood fibers have very 9.11. Wet web stretch
Fractionation scheme – different properties and different bendabilities In general, the most critical point for runnability
fine fraction represents
the earlywood fraction,
and cell-wall thicknesses. The accepts are rich in in a paper machine is after the press section,
and coarse fraction with flexible earlywood fibers (fine fraction) and where web breaks are most common. The wet
represents the
latewood fraction.
the rejects are rich in stiffer latewood fibers web stretch depends on the straightness of the
(coarse fraction), see Table 10. The process is fibers. It occurs before the paper has dried
—
50 illustrated in Figure 49. enough to form hydrogen bonds, which make the
Formation index is paper stay together when dry. Increased curliness
a measure of local
— means increased wet web stretch.
variations in the
Table 10: Results from fractionation
grammage of the sheet.
Increase of average
fiber length causes an Length Width Shape 9.12. Formation
increase of flocculation (mm) (μm) factor (%) Chemical long fibers are used as reinforcement
on the forming unit.
Fine fraction 2.66 36.0 85.7 pulp. However, their impact on the strength of the
Coarse fraction 2.85 35.7 85.3 sheet is not always completely clear. Formation
index is a measure of local variations in the
grammage of the sheet. Longer fibers cause more
or bigger flocs on the wire, and an uneven sheet
Fine may have a lower strength with the same
F2 fraction
conditions in the process (Figure 50).
Feed F1
Discarded However, flocculation can be reduced with a lower
F3
consistency during forming and improvements of
F4
the dewatering elements on the wire.
F5
Formation index

Coarse fraction
—
49

9.10. Fiber orientation

Fiber length
Do the fiber properties affect the fiber
orientation in the sheet? The fibers will always —
50
have a tendency to more or less line up in the
machine direction on a paper machine because
of speed differences between jet and wire.

This tendency may be less for shorter or more

deformed fibers.
 ≔ ← → ≡
50
Table of contents Previous chapter Next chapter Index

—
10. Traditional testing of pulp

Physical tests on pulp and paper are carried out ISO standards are used in association with
for various reasons, e.g. as a production control, international trade and are officially recognized
as a value indication of the paper properties around the world. As ISO standards become
between a seller and a buyer and, of course, for available, the corresponding SCAN-test methods
functional reasons. The main reason for pulp are being withdrawn and replaced with the ISO
characterization is to ensure a stable pulp quality, standards.
since different kinds of fibers have different
effects on the physical properties. Pulp 10.1. Pulp testing
characterization is also a tool for improving A lot of testing methods are in use to characterize
existing and developing new products of the pulps with respect to quality, processability, and
resultant paper. However, it is not possible to use suitability for various end uses. Many of these
results from the pulp characterization to predict test procedures provide useful behavioral
the properties of a paper made on a specific information, while some of the tests provide
paper machine. the means to predict behavior.

To ease the communication between researchers, 10.2. Pulp testing methods

producers and customers, standard test • Laboratory sheet forming
procedures have been agreed upon and issued by • Drainability
the pulp and paper industry. Standardized tests • Beater evaluation
make it possible to compare results in one • Kappa number
laboratory with results from other laboratories. • CED Viscosity
Some institutes responsible for standardization: • Bauer McNett Classifier
• Sommerville screening
• ISO — International Standardization • Water Retention Value
Organization • Zero-span tensile strength
• SCAN — Scandinavian Pulp, Paper, and
Board Testing Committee (Finland, Norway
and Sweden)
• TAPPI — Technical Association of the Pulp
and Paper Industry, USA
• APPITA — Australian Pulp and Paper
Industry Technical Association
 ≔ ← → ≡
51
Table of contents Previous chapter Next chapter Index

10.2.1. Laboratory sheet forming Comparatively, in industrial sheet forming, the

Many pulp properties are determined on hand- fibers align themselves to some extent in the
made laboratory sheets. A laboratory sheet direction of the machine, and there is a closed
former consists of an upper section, a wire screen white water system giving a good retention of
on which the sheet is formed, and a lower section fines. If a hand-sheet’s retention is required to be
for the drained water. The pulp is diluted to a low similar to that of commercial sheets from paper
concentration (<0.5 g/l), and a certain volume of machines, a Rapid Köthen laboratory sheet
the pulp slurry, sufficient to make one standard former could be useful. It is also well-suited for
hand-sheet, is transferred to the upper section of sheet forming of mechanical pulps. The Rapid
the sheet former. The very low concentration of Köthen laboratory sheet former simulates the
fibers results in a sheet with an exceptionally moving web when the water is drained from the
good formation with fibers uniformly distributed pulp stock and includes a closed white water
in the sheet. The wet sheets are pressed onto system accumulating the fines and improving the
drying plates and dried in a conditioned retention (ISO 5269-2:1998).
atmosphere at 23ºC and 50% humidity. The
sheets stick to the drying plate; shrinking is In order to get a sheet with fibers oriented in a
therefore prevented. certain direction, i.e. an anisotropic sheet, a more
complicated sheet former needs to be employed.
A number of different types of sheet-forming For example, the French sheet former (Formette
equipment for laboratory use exist. They differ dynamique) where the pulp slurry is sprayed from
from each other mainly in the concentration of a nozzle towards a cylindrical rotating wire,
the pulp slurry, sheet type (rectangular, square or creating a sheet where the fibers are more or
circular) and sheet dimensions, type of wire less aligned parallel to each other. There is also
screen etc. In general, sheets are made in an open a standard for determination of physical
white water system resulting in low retention of properties on laboratory sheets (ISO 5270:1988).
fines (ISO 5269-1:1998). The fibers are evenly and
randomly distributed, resulting in a so-called
isotropic sheet.
 ≔ ← → ≡
52
Table of contents Previous chapter Next chapter Index

—
51
10.2.2. Drainability The standard freeness test starts with one liter of
Schematic drawing of An important property with respect to pulp the pulp suspension (at a concentration of 0.3%)
a standard freeness
tester, for use in
processing and papermaking is the resistance of being poured into a drainage chamber (Figure
the laboratory. fibers to the flow of water. The most common 51:1). Thereafter, the bottom lid (Figure 51:2) is
methods of determining drainability are the opened. The air valve on the drainage chamber is
Canadian Standard Freeness (CSF) tester (ISO opened to start the drainage. The flow through
5267-2:2001, TAPPI T227) and the Schopper- the perforated plate decreases with increasing
Riegler test (ISO5267-1:1999). density of the fiber pad. The volume of the
sample passing through the side discharge
The freeness and Schopper-Riegler scales have an orifice (Figure 51:3) determines the freeness
inverse relationship. CSF is defined as the number value of the sample.
of milliliters of water collected from the side
opening of the standard tester when pulp drains There are a number of disadvantages in testing
through a perforated plate at 0.30% consistency freeness manually. To obtain reliable values, the
and 20ºC. Freeness measurements are widely sample has to be handled in the same way every
used as an indication of quality for mechanical time a test is made. The advantage of using an
pulps and as a measure of the degree of refining automatic freeness analyzer is that the sample is
for chemical pulps. The fines fraction is primarily handled the same way every time, both in the
responsible for changes in the drainage time. The preparation stage and in all the tests.
removal of the fines fraction from beaten pulps
can restore the original drainability, while the
pulp retains its beaten strength properties. This
1
is why there is criticism of the use of drainage
measurements as an index of pulp quality.

The purpose of measuring freeness is to simulate

2
the drainage rate on the wet end of the paper
machine. During beating or refining, the
proportion of fines in the pulp increases and the
fibers become more flexible. This results in a
lower drainage rate since the fibers are pressed
more tightly together on the wire. The freeness
value also gives an indication of how much the
pulp can be beaten before it reaches the limit of 3
the paper machine’s drainage capacity. A lower
freeness value through beating gives better —
bonding and more flexible fibers and a greater 51

strength.
 ≔ ← → ≡
53
Table of contents Previous chapter Next chapter Index

10.2.3. Beater evaluation Lignin is responsible for the yellowish-brown

Beating is a process to mechanically condition color of paper and is removed by bleaching.
the fibers for papermaking. In general, laboratory Bleaching, however, also reduces the mechanical
beating methods have a more selective action strength of the fiber. Therefore, the lignin content
than mill refiners and produce results that cannot must be well known, so that only a minimum
usually be duplicated in the mill. A number of amount of bleaching is used.
laboratory beating devices are in use around the
world. One common device is the PFI mill (ISO The standard laboratory test for kappa number
5264-2:2002). It utilizes a grooved roll that is off- (ISO 302:1981) is defined as follows: The number
center from a smooth trough. The roll and the of milliliters of 0.02 mol/l potassium perman-
bedplate rotate at a high speed but with different ganate solution consumed under the specified
peripheral velocities. This induces friction, conditions by 1 g of pulp (calculated on an oven
rubbing and crushing of the fibers to produce the dry basis). The results are corrected to a value
beating effect. The device does not require any corresponding to that obtained when 50% (mass/
calibration, since there is no metal-to-metal mass) of the permanganate is consumed in the
contact and no edges to wear. Another ad- test. Faster tests based on optical techniques are
vantage is that it only requires a small amount of available for online control.
pulp to carry out a complete evaluation. Another
standardized laboratory beater is the Valley 10.2.5. CED viscosity
beater (ISO 5264-1:1979). An indication of degree of polymerization of the
cellulose can be obtained by measuring the
10.2.4. Kappa number viscosity of a cellulose solution of known
Roughly, the Kappa number is an indication of the concentration. Cupriethylene diamine (CED) is
lignin content, or bleachability of pulp. A Kappa universally used as a cellulose solvent because of
number test is used in mill control work for two its ability to rapidly dissolve cellulose and
specific reasons: because of its chemical stability (ISO 5351-1:1981).
1. To indicate the degree of delignification
occurring during cooking 10.2.6. Bauer McNett Classifier
2. To indicate the chemical requirement for Fiber fractionation is a procedure that segregates
bleaching a blend of pulp fibers into different streams
based on some physical properties of the fibers
The two objectives are never combined in the (i.e. length or flexibility). The fibers of a particular
same test because of the large processing gap type can then be directed to the most suitable
between cooking and bleaching. process and product.
 ≔ ← → ≡
54
Table of contents Previous chapter Next chapter Index

The Bauer McNett classifier (BMN) is the most Another definition is to use, in a similar way, the
widespread of the fiber fractionation laboratory classes; R14, R28, R48, R100 and P200. Today,
devices, and it is used to assess the size classification of fiber length is usually achieved
distribution of fibers. The BMN operates by the with optical fiber analyzers (see chapter 9).
selective passage of fibers through a screen
mesh, and it separates the fibers mainly on the 10.2.7. Sommerville screening
basis of length. The device consists of a series of Laboratory screening of pulp can be done with
chambers (four or five), each with an increasingly the Sommerville method. The purpose of the
smaller screen mesh. Mesh is the number of method is to separate contaminants such as
threads/inch in a wire. shives in mechanical pulp, and stickies, plastics,
sand, metal pieces, and flakes in recycled fiber
A predetermined amount of pulp is poured into from pulp fibers for subsequent examination and
the first chamber at the start of a test. A constant quantification. Contaminants affect the paper
flow of water passes through this chamber and machine negatively and downgrade the final
the series of chambers that follows. Smaller product.
fibers are carried by the flow through the mesh
and to the following chamber to the sewer. At the The method uses a screening device, and the
end of the test period, the flow stops and the separation is based on size difference between
chambers are drained. Thereafter, the mass in fibers and contaminants. The screening device
each chamber is measured. Pulp with smaller consists of a rectangular screen box, a screen
fibers will have more mass in the chambers with plate with 150 µm slots (mini shives screens use
the smaller mesh sizes. BMN classification is smaller slots), a diaphragm chamber with an
considered to be repeatable within the same attached weir box, and a stand. A diluted pulp
laboratory, but to have less reproducibility sample (less than 1%) with known consistency is
between different laboratories. Length fractions added together with water to the screen box. The
of fibers are successively fractionated out. sample is screened and the amount of oven-dry
Definitions of the classes are as following: materials added to the device is calculated and
reported.
is the fraction that remains on the
R16 16-mesh wire (1190 μm) 10.2.8. Water Retention Value measurements
is the fraction that remains on the Water Retention Value (WRV) measurements are
30-mesh wire (595 μm), but passes
carried out according to ISO 23714:2019.
R30 through the 16-mesh wire
According to the definition, WRV is the ratio of
is the fraction that remains on the
50-mesh wire (297 μm), but passes the mass of water retained after centrifugation
R50 through the 30-mesh wire under specified conditions by a wet pulp sample
is the fraction that remains on the to the oven-dry mass of the same pulp sample.
200-mesh wire (74 μm), but passes
R200 through the 50-mesh wire
is the fraction passing through
P200 the 200-mesh wire
 ≔ ← → ≡
55
Table of contents Previous chapter Next chapter Index

The test pad of pulp fibers is formed in a Büchner For never dried pulp, typical values of retained
funnel connected to a suction flask. The suction water relative to the dry weight of the pad are:
is stopped when the dry matter content of the
test pad is between 5%–15%. The test pad is • Sulphite pulp: 258%
removed from the wire and placed in a former, • Sulphate pulp: 165%
and the complete holding unit with the test pad is • Refiner pulp: 148%
placed in a centrifuge bucket and centrifuged at a • Groundwood pulp: 128%
force of 3000 g for 15 minutes.
10.3. Paper testing
Immediately after the centrifuge stops, the test Since there are many different paper grades and
pad is transferred to a pre-weighed bottle and many properties of paper, there is a need for a
weighed. The bottle is placed in a drying oven and large variety of test methods. A number of
dried to constant mass at a temperature of 105°C properties are important for all grades, and
for at least fifteen hours. The weighing bottle is tests used to measure these properties have a
closed and allowed to cool in a desiccator (dry wide application. Test methods have also been
atmosphere) for at least one hour. The closed developed specifically for specialty grades
bottle is then weighed. and have a more limited application.

Water Retention Value (WRV) is calculated Measurements of paper properties are

according to: dependent on the instrument or equipment
used and on the details of the testing procedure.
7.1 The fact that paper is both viscoelastic and
hygroscopic makes testing more complicated.
m₁
WRV = -1 Given that paper has both plastic and elastic
m₂
qualities, any test which leads to deformation
where: or destruction of the sample will give results that
m₁ = mass of the centrifuged wet test pad are dependent on the rate of application of the
m₂ = mass of the dry test pad. force. Paper always seeks moisture equilibrium
with the surrounding air (it is a hygroscopic
WRV measured by centrifugation is used mostly material), and the properties of the paper are
for chemical pulps, where it has a good reproduc- therefore highly dependent on the relative
ibility. The reproducibility for mechanical pulp is humidity and humidity history. Types of
not so good. WRV is used as a measure of fiber properties include:
swelling. Other principles are also possible in 1. Structural properties
order to measure bound water, but centrifugation 2. Strength and stiffness properties
seems to be the proposed standard, recommen- 3. Surface properties
ded for chemical pulp. 4. Optical properties
5. Absorption properties
 ≔ ← → ≡
56
Table of contents Previous chapter Next chapter Index

10.3.1. Structural properties Most physical properties of paper undergo a

Structural properties include: change as a result of variation in moisture
content. Water has a negative effect on the
• Caliper (thickness) paper; it plasticizes the cellulose fiber, and
For thickness measurements, a thickness gauge relaxes and weakens the inter-fiber bonding.
is used in which the paper or a pad of sheets is The amount of water present in a sheet of paper
held under a given pressure between two plays an important role in the calendering,
parallel plates. The thickness determines how printing and converting process. Determining
bulky or dense a paper is for a given grammage. the moisture content of paper simply involves
Uniform caliper is important for reel building weighing a sample of the paper before and after
and subsequent printing. Variations in caliper drying.
can affect several basic properties, including
strength, optical properties and reel quality. • Grammage
Thickness is important for printing papers, The grammage (also called basis weight) is the
condenser paper, saturating papers etc. (TAPPI mass per unit area of the paper and board
T411, ISO 534, ISO 3034). expressed as g/m2. Determination includes
weighing a piece of paper with a known area.
• Density
Density is the mass per unit volume of the paper • Formation
or board calculated as the ratio between Formation is an indicator of how uniformly the
grammage and thickness of the material and fibers and fillers are distributed in the sheet.
is expressed as kg/m3. It is a very important Most of the paper properties depend on good
parameter of paper, particularly for printing formation, since a paper is as strong as its
papers.Bulk is the inverse of density. Sheet bulk weakest point. More weak, thin or thick spots
relates to a lot of different sheet properties, are found in a poorly formed sheet. This will
and a decrease in bulk makes the sheet affect properties like caliper, opacity, strength
smoother, glossier, less opaque, darker, lower in etc., and it will also affect the coating
strength etc. High bulk is desirable in absorbent capabilities and printing characteristics. The
papers, while lower bulk is preferred for tendency to flocculate depends on the type of
printing papers (ISO534:1998, TAPPI T500). fiber. In general, longer softwood fibers have a
greater tendency to flocculate than hardwood
• Moisture content fibers. There is no standard measure or unit for
Almost all grades contain some percentage of formation.
moisture. Depending on the relative humidity,
type of pulp, degree of refining, and chemicals,
the moisture in paper varies between 2–12%.
 ≔ ← → ≡
57
Table of contents Previous chapter Next chapter Index

• Air permeance 10.3.2. Strength properties*

In general, the air permeance is obtained by
measuring the air flow through a defined area Strength properties include:
under a known pressure drop. The unit for air
permeance is µm/Pa·s. This unit can be • Tensile strength
converted by calculation into other instrument- Tensile strength is determined by measuring
specific units e.g. Gurley seconds, Bendtsen the force required to break a narrow strip of
(ml/min) and Sheffield units. Air permeance test paper when both the length of the test strip,
methods are standardized in TAPPI T460, ISO and the rate of loading are closely specified. It
5636. The air permeance is influenced by e.g. is determined according to a standardized test
beating, pulp additives, sheet forming, and procedure. The tensile strength of a paper
pressing. depends on the fiber properties (see chapter
12). The results are also affected by the testing
By controlling the combination of these conditions. An increase in moisture content of
components, the desired air permeance level is the paper will decrease the tensile strength,
obtained. The air permeance has a close relation while increasing elongation. The tensile
to the porosity of the sheet, and thus the strength is highly dependent on the
absorption of printing inks. In sheet-fed directionality of the paper. (ISO 1924-1,
printing processes, the air permeance is crucial TAPPI T494, ISO 1924-2, TAPPI T404).
to the runnability. Heavyweight coated paper
must have a certain air permeance level in order • Zero-span tensile strength
to leave a passage for moisture in heat-set Another measurement of tensile strength is
printing processes. Air permeance that is too obtained by testing with a span as close as
low can cause blistering in the coating layer. possible to zero (or smallest distance possible
between the two clamps holding the sample in
place). By testing wet sheets, the effect of
bonds is reduced. The test is primarily used for
pulp testing to understand fiber strength. In
practice, it is dependent on the deformation of
the fibers.

* Source, private communication: Thomas Fürst (Oct. 2005).

 ≔ ← → ≡
58
Table of contents Previous chapter Next chapter Index

—
52
• Compressive strength • Z-directional strength
Tear/Tensile diagram Compression strength is defined as the Z-directional strength represents papers'
maximum compression force per unit width ability to resist tensile loading in a direction
that a test piece of paper or board can support perpendicular to the plane of a paper
until the onset of failure in a compression test. (z-direction). When the Z-directional strength
The actual test can be performed in several of the paper is exceeded, a break in the paper
different ways. The most common tests are the structure occurs in the sheet, but not at its
Short-Span Compression Test (SCT) described surface. The Z-directional strength is therefore
in ISO 9895, TAPPI T825, and the Ring Crush not equivalent to the surface strength or linting
Test (RCT) described in ISO 12192, TAPPI T818, tendency of the paper. There are two common
T822. By measuring the compression strength methods for measuring strength in the
of linerboard in cross direction, the stacking z-direction; the Z-directional tensile method
strength of a final corrugated box can be that measures the force needed to split a
predicted. sample, and the Scott-Bond method that
measures the energy needed to split a sample.
• Bursting strength
Bursting strength is the maximum pressure, • Tearing strength
which may be exerted perpendicularly to the Tearing strength is the force required to tear a
surface of a paper, or board, before rupture sheet of paper, under specified conditions,
occurs. (ISO 2758:1983, TAPPI T403). continuing from an initial cut. The tearing
strength is very dependent on the fiber
• Folding endurance orientation of the sheet. When a pulp is beaten
Folding endurance is the ability of a strip of (see Figure 52) the tensile index increases (a),
paper to resist breaking when folded under a whereas the tear index decreases, except for an
certain load. The folding strength is expressed initial increase. Differences in fiber quality may
directly as the number of double folds a paper change the position of the curve in the diagram
can withstand. Folding endurance is the ten- (b). Tear index is not considered to describe
based logarithm of the number of double folds. paper products very well, but it is still very
The folding strength value is very sensitive to commonly used in specifications. (ISO 1974,
local variations in the structure of a paper. TAPPI T414)
Consequently, the test results are very sensitive
to variations in testing conditions, i.e.
Tear index

variations in the relative humidity of the

b

surrounding air. (ISO 5626, TAPPI T423, T511)

а

Tensile index

—
52
 ≔ ← → ≡
59
Table of contents Previous chapter Next chapter Index

10.3.3. Surface properties Some methods of package handling rely on a

Surface properties include: certain amount of friction between cartons. It is
therefore important for the printer/converter to
• Surface strength control the amount of friction. The test to
The ability to resist a force that strives to pull determine the friction coefficient is empirical.
out fibers or fiber bundles from its surface is The quantity can be measured using an inclined
identified as the surface strength of paper. A plane test, and the result is then reported as the
low surface strength can cause linting and slide angle. It can also be measured on a
runnability problems. Methods for testing the horizontal bed tester. (TAPPI T815).
surface strength usually use high viscosity,
tacky oils, printing inks or adhesive tape. (TAPPI • Softness
T459). Softness is a subjective property, relating to
how velvet-like the paper is and how easily it
• Roughness or smoothness yields when crumbled. Softness relates to the
The purpose of a roughness or a smoothness feeling of softness when stroking the paper
measurement is to obtain a figure that surface gently with the fingertips.
describes the topography of the paper surface
in a simple way. The generally applied method 10.3.4. Stiffness properties
measures the ability of the paper surface to Stiffness relates to the elastic properties of a
prevent an air stream from flowing between material, and it refers to the resistance given by
the paper surface and a surface or an edge the material to an applied force. Two types of
(measuring head) pressed against it. The stiffness are usually measured: tensile stiffness
obtained result is an indicator of the paper’s and bending stiffness.
smoothness or roughness. Several test methods
exist e.g., Bendtsen, PPS, Bekk, and Sheffield. • Tensile stiffness (ISO 1924-3) is a very
The PPS-method is the most known worldwide, important property of a printing paper to resist
and it is used to determine surface roughness the stress of the tension during web printing
on most printing grades. (PPS method: ISO in order to avoid off-set color (or maintain
8791-4, TAPPI T555; Bendtsen: ISO 8791-2, TAPPI registers). A high tensile stiffness value in the
T816; Sheffield: ISO 8791-3, TAPPI T538; Bekk; liner layer is needed to obtain high bending
5627). stiffness in corrugated board.

• Friction • Bending stiffness represents the capability of

The resisting force that occurs between two paper and board to resist a bending force. (ISO
paper or board surfaces in contact, when the 2493, TAPPI T489, T558, T566, ISO 5628, T543)
surfaces are brought to slide relative to each
other, is called friction.
 ≔ ← → ≡
60
Table of contents Previous chapter Next chapter Index

10.3.5. Optical properties • Opacity

Optical properties include: Opacity is the ability of a material (i.e. paper) to
prevent the passage of light. It is important that
• Brightness a paper has enough opacity, to prevent printed
Brightness is the intrinsic luminous reflectance text from showing through on the reverse side
factor (R∞) measured around wavelength 457 of the paper. Opacity is important in printing
nm. The function of the term brightness is to be papers, book papers, etc. The opacity of paper
able to measure the effect of bleaching, and to is influenced by thickness, amount and kind of
measure the amount of colored substances in filler, degree of bleaching and coating etc.
pulp. Drying of pulp increases brightness, and (TAPPI T425)
pressing decreases it. Brightness may not add
much value to the “useful” properties of the • Color
paper, but it is the most important selling The color of paper, like that of other materials,
feature. (ISO 2470, TAPPI T452). depends in a complicated way on the characte-
ristics of the observer and a number of physical
• Light scattering factors, such as the spectral energy distribution
The borders between air and fibers provide of the illumination, the geometry of illumination
surfaces for the light to be reflected from. A and viewing, the nature and extent of the
sheet made of stiff, tube-like fibers scatters surrounding and the optical characteristics
more light than a sheet made of slender, of the paper itself. (ISO 5631)
collapsed fibers. Well-bound fibers provide
fewer surfaces for light scattering than fibers • Gloss
with fewer bondings. Gloss is the property of a sheet surface that is
responsible for its shiny or lustrous appearance.
The degree of cooking or bleaching of the pulp High gloss is desirable in high quality,
does not affect the ability of a sheet to scatter prestigious printed images. The tone range of
light, whereas beating, pressing or any a glossy image is wider than that of a matte
procedure changing the physical structure of surface. It is possible to print with more
the fibers and sheet, greatly influences the saturated colors and deeper blacks on a paper
scattering capacity. Both beating the pulp, and with high gloss surface. Specular reflectance
pressing a paper sheet, result in a higher degree often worsens the readability of a text;
of bonding and a denser sheet with less light therefore high gloss paper is undesirable in
scattering interfaces. applications such as textbooks. (ISO 8254-1,
TAPPI T480)
 ≔ ← → ≡
61
Table of contents Previous chapter Next chapter Index

—
Table 11: Typical brightness values for different pulps

Pulp ISO-brightness (%)

TMP ~ 60–70 %
Unbleached softwood kraft ~ 40 %
Bleached softwood kraft 90 %

10.3.6. Absorption properties

There are many methods available for testing
the absorption properties, and an important
requirement for all these tests is that the medium
used for absorption into the paper should
simulate as close as possible the end-use
product. The media used can be water, oils
of different composition and viscosities, or
printing inks of different types.

Both absorption capacity and rate of absorption

are key properties for fluff pulps and products,
and for tissue products. Absorption properties
include:

• Water absorbency
The most frequently used water absorption test
is the Cobb test, which measures the amount of
water absorbed by the sample during a given
time from an excess of water. This test is
primarily used as an indicator of the degree
of sizing, i.e. the water repellence of the paper.
(ISO 535, TAPPI T441)

• Oil absorbency
The absorbency of printing inks is often
measured with a printability test. As an indica-
tor of printability evaluation, the amount of ink
transferred from the printing plate to the paper
is used.
 ≔ ← → ≡
62
Table of contents Previous chapter Next chapter Index

—
11. New technologies

—
53
Traditional pulp testing is often characterized 11.1. Optical technique
Traditional testing of as time demanding, with several manual steps The optical technique is a fantastic tool. We can
pulp sheets, compared
with alternative testing
involved. The methods are not always follow the moon passing the sun (solar eclipse)
of fiber properties. representative of modern processes and with the same principle that is used to study a
(PLS-model, see
chapter 13)
products. New computer technology, electro- small cellulose fiber in a light microscope, or for
optical components, and modern signal automatic classification online with L&W Fiber
processing have made new types of Online (Figure 54). The visible light covers only a
measurements possible. limited part of the spectral range. Other parts of
the spectral range that are used to analyze fiber
These techniques are suitable for automatic quality will also be reviewed below.
measurements, which lead to fast, more
repeatable and operator-independent measuring
methods. Fiber properties
10 minutes

Pulp testing is traditionally done by making

laboratory sheets, which are later tested and
evaluated, as indicated in Figure 53. The pulp is Pulp
PLS-model Potential for
1 second paper quality
beaten in a defined way and with a given amount
of specific energy, or for a given time, to one or
several levels. Laboratory sheets are formed,
pressed, dried and cut to suitable test strips and Conventional sheet test
Sampling
finally measured in sheet testing devices. Consistency test
Disintegration
Forming of lab sheet
The time until a result is obtained can be from Pressing
3–16 hours. However, the pulp is made from fibers Drying
Cutting
and the fiber properties define the pulp Testing
properties. Direct analysis of the fibers based on
TOTAL: 3 to 16 hours
image analysis is the new way of assessing pulp
properties. The advantage is that more detailed —
information is measured, which may explain the 53

causes for deviations in pulp quality.

 ≔ ← → ≡
63
Table of contents Previous chapter Next chapter Index

—
54
Particle size affects the light scattering; Polarized light does not allow the imaging system
A large scale description for example, smaller particles increase the to see all the fines (ultrafine fibers), which means
of the measurement
principle in Fibermaster.
scattering. One effect of this is that mechanical that this technique is not suitable for mechanical
The moon (detected pulps scatter light better than chemical pulps. pulps since these pulps include a lot of fines.
particle) is passing
the sun (light source).
However, when the particle size is smaller than There can also be limitations when making
The image represents half the wavelength of the irradiating light, the measurements on synthetic and glass fibers.
the sensor signal.
light scattering starts to decrease, which means
— that scattering has a maximum at half the
55
If the dotted line radiation wavelength. This is important for
represents half the the selection of filler size. See Figure 55 for
wavelength of the light,
then the light scattering an illustration.
from particles with
different sizes,
which are hit by the 11.2. Polarized light
light, will follow the A phenomenon used in the early measurement
curve in the figure
of fiber length was the change in polarization of
light when it passes through crystalline cellulose.
This principle has also been used for consistency
sensors, since the impact of air and filler could be —
54
avoided.
Light scattering

Fillers and air bubbles do not change the

polarization. But drawbacks have been reported.
If the polarization does not change at a
compressed zone of the fiber (amorphous
cellulose), the fiber may be detected as two Particle diameter
—
fibers instead of one. 55
 ≔ ← → ≡
64
Table of contents Previous chapter Next chapter Index

—
56
11.3. Detailed analysis of fiber structures Absorption

Difference in spectrum between

0.35
NIR spectra showing Light microscopes are widely used in
the effect from water. 0.3

dry and wetted paper

Measurement of
laboratories. Different techniques exist, for
water content is just sample preparation, staining etc. This book does 0.25
one application.
not include these methods in any detail. But the 0.2

images in Chapter 6 are made with light 0.15

microscopy, and most of them are made with an 0.1
interference technique. Confocal techniques are
0
used to study three-dimensional structures. They 1000 1200 1400 1600 1800 2000 2200 2400
are still regarded as laboratory R&D instruments Wavelength (nm)
for studying details in well-prepared samples.
—
The scanning electron microscope (SEM) is widely 56
used to measure fiber properties. A modern
version, ESEM, can measure fibers without
Validation of the models is an important part.
sample preparation and with different
Longer NIR wavelengths will not penetrate very
humidities.
deeply into water. Another effect on the
propagation of radiation waves when they pass
In the NIR (near infrared) region, where the
through materials containing small particles is
wavelength is longer than the wavelength of
the interaction with particle size. In the visible
visible light, the alternating electromagnetic field
light area, a small particle size will change the
starts to interact with the electric field around
intensity of the color. This wavelength interaction
the molecules in the sample. The amplitude of the
with particle size will also appear at other
molecular movement increases and the energy of
wavelengths than those of visible light.
the NIR field is absorbed. The substance must
have molecular groups which can stretch or
11.4. Ultraviolet light
vibrate in at least one resonating over tone in the
Ultraviolet light (UV) can be used for studying
actual NIR wavelength range i.e. C-H, O-H and
smaller dimensions than tracheids because it has
N-H. The first tone lies in higher wavelengths
a wavelength shorter than of visible light. UV is
intervals (MIR etc), and that requires a more
used to measure lignin content on fibers with an
complicated technique. In that case, a method
automated optical technique.
using diffuse reflectance NIR is the best method
to define yield, kappa number, lignin content,
11.5. X-Ray technology uses very short
glucose, xylose and uronic acid during kraft
wavelengths
digestion of birch. Multivariate data evaluation is
With advanced laboratory tool for the radial
used.
determination of fiber properties in wood, such
as density (x-ray absorption), fiber width (image
Using NIR for chemical information and
analysis) and microfibril angle (x-ray
combining it with image analysis for physical
diffractometry). Fiber wall thickness, coarseness
fiber properties has been suggested.
and wood stiffness are calculated results.
The technique for the development of a
calibration model is described in Chapter 11.
 ≔ ← → ≡
65
Table of contents Previous chapter Next chapter Index

—
Wave length (nm):
57
The electromagnetic 20 170 400 800 2500 5000 1000000
spectrum
Visible light NIR
— Vacuum
X-Ray UV Mid IR Far IR Microwave
58 UV
The Great Wall of China
High Inner Overtones Molecular bond Molecular
Valence Low proton
protection shell Ionisation of molecular vibrations bond
electrons energy
energy electrons vibrations stretch bend rotation

—
57

The whole length of the fiber has to be in the

viewing field and at the same time the width of
the fiber has to be resolved. Detailed statistics
can be calculated from images with modern
image analysis. A special area of image analysis is
blob analysis, which for example is used in L&W
Fiber Tester Plus (current iteration of the STFI
Fibermaster).

High processing capacity is necessary for image

processing. The first Fibermaster system
— developed in 1992 used a pipeline processor for
58 image analysis, or more specifically blob analysis.
The image analysis functions were implemented
11.6. Image analysis in hardware modules. By connecting such
The Great Wall of China was 5000 km long, and modules in series, a number of image processing
legend says it can be seen from the moon (Figure functions could be performed with high speed.
58). But according to the astronauts Neil The technique was used by the military for the
Armstrong, Jim Lovell and Jim Irwin, contrary to detection and following of moving targets. In the
legend, it was not seen from the moon. Why? The first industrial version of Fibermaster, a new
ability to resolve an object is dependent not only compact digital image sensor with built-in image
on the length but, of course, also on the width of processing capacity was used. The images could
the object. The width compared to what can be be transferred and processed in parallel.
resolved by the eye or a camera sets the limit.
The thickness of the great wall of China was Today it is possible to use standard high
4.5–9.0 m. In fiber analyzers, objects significantly resolution digital cameras and standard personal
narrower than the pixel size will not be detected. computer technology for online fiber analysis.
The problem when measuring fibers with image
analysis is that fibers are much longer than they
are wide (up to 100 times longer).
 ≔ ← → ≡
66
Table of contents Previous chapter Next chapter Index

—
59
A strobe light catches an image (shadow) on size is 10×10 µm, a camera with 1000×1000 pixels
A digital image is built the digital sensor. First the grey scale image is covers 10×10 mm. A digitalized image of a
up of a number of pixels.
adjusted to compensate for the background. detected object can look like the image in
The background can be measured before each Figure 59.
measurement with all moving objects excluded.
The difference between the actual image and If we want to model the object with two
background removes the stationary background parameters, then a rectangle with a length
(anything that is stationary is filtered out). The and a width is a possible model. If the area and
next step is to set a greyscale threshold, where perimeter of the object are calculated, then the
all pixels with low intensity are set to zero and all length and width can be calculated from these for
with high intensity are set to unity. We get black the object and spatial filters applied on the grey
objects (blobs) on a white background. If the primary parameters. An approximation is to set
objects have holes, the holes must be detected the length equal to half the perimeter and then
and filled appropriately. Uneven contours can be calculate the width from the length and area. The
evened out by expanding and eroding the object. result will be similar to that shown in the figure.
Note that the resolution of length and width will
The skeleton can be found by thinning. A program be higher than the pixel size. A bigger object will
can detect the contour of the object and the fiber give a better resolution since more pixels are
ends. The area of the objects can be calculated. involved in the calculation. An alternative to
It is also possible to go back to the grey scale and calculating the perimeter is to calculate a center
calculate a grey value scale image. The grey scale line (skeleton) of the object for length.
of each pixel can be influenced by the greyscale
of the surrounding pixels. Other special functions
for fiber analysis are the detection of fiber-to-
fiber crossings and the detection and handling
of fibers crossing the border of the image.
Pixel
Object
Advanced filters like the fast fourier transform
(FFT) can be used, but these types of image Rectangular model

analysis take more time and are seldom used for

fast systems. A digital image is built up of pixels.
Objects which are significantly smaller than the —
size of one pixel cannot be detected. If the pixel 59
 ≔ ← → ≡
67
Table of contents Previous chapter Next chapter Index

—
12. Measurements of fiber properties

—
60
If we measure the lengths of 20,000 fibers, and
L&W Fiber Tester Plus the average fiber length is 2.5 mm, then we have
— measured a total length of 50 m of fibers in a
61 single analysis.
Measurements of length
and shape are shown
with two-dimensional 12.1. Measurement cell for image analysis
imaging technology.
This is explained of fibers
further in Table 12, Different measurement cells have been used to
on the next page.
analyze fibers in a suspension. A capillary tube
with a circular cross-section orients the fibers
in the flow direction and the fiber length is
measured in this direction. With this technique
it is natural to let all the fibers pass the capillary
once and to measure all fibers in the sample.
A drawback of this technique is that fiber
deformations and fiber length cannot be
measured independent of each other. The fiber
length will in fact be a projected fiber length. —
60

In a more modern fiber analyzer like L&W Fiber

Tester Plus (Figure 60), the highly diluted 3 mm gap between plates
0.5 mm gap between plates
suspension flows between two glass plates.
The distance between the glass plates is very
small when compared with other systems and Viewing angle
(Fibertester) Viewing angle
limits the possibility for the fibers to move in one
direction, but it allows the fibers to move freely in
the other two directions. True shape Too straight

True length Too short

Two-dimensional images allow us to measure
fiber length and deformations separately, if the —
61
fibers are well aligned in a plane (Figure 61).
 ≔ ← → ≡
68
Table of contents Previous chapter Next chapter Index

— —
62 Table 12: Comparison between different principles
An image from L&W for measurement cells. Glass plates with a short
Fiber Tester. Images
distance in between them combined with a two-
can be stored during
dimensional camera, create a true projection
measurement and
monitored afterwards of the fiber length and fiber deformations

Measurement cell Comment

Capillary tube: Alignment of
fibers only in one dimension Inaccurate projections
(along the capillary axis)
Glass plates with small gap Two-dimensional
(0.5–0.7 mm): Good alignment measurements —
of the fiber in two dimensions close to the truth 62
in parallel with the glass plates (Fibermaster)
Glass plates with larger gaps A typical image of eucalyptus pulp fibers from
(1–5 mm): Three dimensional Inaccurate projections
movement allowed
bleached chemical pulp can be seen in Figure 62.
Note that all the fibers in principle have a
curvature, or are kinked. A single vessel cell is
A three-dimensional appearance of the fibers and also seen in the image. The image is a greyscale
orientation across the image plane will cause an image, before compensation for the background.
error (Table 12 shows comparisons).
12.2. Fiber model
If the distance between the glass plates is greater A fiber has a complicated structure, which varies
than the fiber length, this error will be greater. with species, growing conditions, pulping
Low flow speed in the measurement cell gives technique and refining. In order to measure fiber
a laminar flow pattern. For very high flow rates, properties, we have to define typical properties
the flow will become turbulent. which are possible to measure. An obvious
parameter is fiber length. However, fiber length is
A problem with capillary cells and also with not simple to measure. One model can be to
very narrow gaps is that fibers can become stuck consider the fiber as a rectangle with a width and
there. Using a dynamic measurement gap solves a length.
this problem in L&W Fiber Tester Plus. The gap is
3 mm before measurement, decreases to 0.5 mm For example, in L&W Fiber Tester Plus, the area (A)
during measurement and increases again after and perimeter (P) are measured for each detected
measurement. object (fiber).
 ≔ ← → ≡
69
Table of contents Previous chapter Next chapter Index

—
63
Length (L) and width (W) are calculated from the 12.4. Fiber length
Fiber length for CTMP, following equations: Fiber length positively influences sheet strength,
chemical birch, and
chemical pine with fines
but it can also have a negative effect on sheet
(l<0.2 mm) excluded. A=L×W formation. Length is measured with a minimum
P = (2 × L) + (2 × W) impact of the degree of fiber deformation. This
means that fiber length is measured
Where: compensated for fiber deformation. After
A = measured area of detected object reporting average length, the most common
P = measured perimeter of detected object presentation of data is the length distribution,
L = calculated length of detected object which can be seen in Figure 63. If the limit for
W = calculated width of detected object fines is 0.2 mm, which is according to the
standard, the curve starts at 0.2 mm.
The length is about half the perimeter and the
width is then calculated from this length and the The international standards for measurement
area. All pixels in the image are used to calculate of fiber length are ISO 16065-1 and ISO 16065-2.
length and width averages for the object. These two methods are similar; the main
difference is that polarized light is used in version
Another way of defining length is to measure the 1, and non-polarized light is used in version 2. In
length of the center line along the fiber. Length, Figure 63 below, we can see length distributions
width and shape factor are measured for each for different kinds of fibers. Birch has very few
fiber in Fiber Tester and can be stored in a raw fibers above 1.5 mm.
data file. Round objects can be excluded.
CTMP of spruce has fibers as long as those of
Since length, width and shape factor are chemical pulp of pine, but spruce CTMP also has a
calculated for the same objects, it is possible lot of fines and cut fibers. The surface area below
to describe the statistics as, for example, two- each curve is the same. Each curve is built up of 75
dimensional length-to-width distributions or fiber length classifications. The ordinate-scale
even three-dimensional length-width-shape (Y axis) shows the proportion of the fiber length
distributions.The latter is more difficult to show (%) in each length class. Each length class is 0.1
in a diagram but two-dimensional distributions mm wide.
are a powerful tool.
14
Proportion of length in each

12
12.3. High resolution in fiber width
of 75 classes (%)

10
The fibre width is 20 μm and the length is 1000 8
μm. For an assumed pixel size of 10 × 10 μm, the 6
area contains 200 pixels. If the area changes 1 4

pixel, the width will have a corresponding change, 2

which means (although the fibre is only 2 pixels 0

0 0.4 0.7 1.0 1.3 1.6 1.9 2.2 2.5 2.8 3.1 3.4 3.7 4.0 4.3 4.6 5.0
wide) the possibility to resolve, changes will be in
CTMP Birch Pine Length (mm)
the order of 0.5% of the width, i.e. 0.1μm. —
63
 ≔ ← → ≡
70
Table of contents Previous chapter Next chapter Index

—
64
12.6. Fiber deformations

Width
The width of the A variation in shape factor between 81% and 85%
fiber varies along Length
the fiber axis.
can make a difference of 15 Nm/g in tensile index
— in unrefined samples from bleached Scandinavian
— 64
65 softwood market pulp (see Chapter 14 for more
Distribution of the fiber details). This difference remains after refining
20
Proportion of length in each of 75 classes (%)
width for different pulps
18 with constant energy even though the tensile
—
66 16 index level has increased due to fibrillation of
CTMP fibers are 14 the fiber surface.
straighter than
chemical fibers. 12
100% in shape factor 10 12.7. Shape factor
means a completely
straight fiber. 8 The shape factor (also called form factor) is an
6 important measure of pulp quality. Shape factor
4 is defined as the ratio of the maximum extension
2 length of the fiber (also called projected length,
0 which is close to the distance between the fiber
0 10 20 30 40 50 60 70 80 90 100 ends) to the true length of the fiber (along the
Width (µm) fiber contour) and is often presented as a
CTMP Birch Pine percentage.
—
65
S = 100 × l / L
Proportion of length in each of 75 classes (%)

18 l = projected length; L = true length;

16 S = shape factor
14
12 A high shape factor means straight fibers and
10 gives in most cases good mechanical properties
8 in the sheet. It is well correlated with tensile
6 strength and tensile stiffness. A gently treated
4 laboratory pulp has straight fibers, whereas there
2 are several process stages in a mill that are
0 potential curlers of fibers, like presses, mixers
50 55 60 65 70 75 80 85 90 95 100 etc.
Shape factor (%) Fibers with a shape factor below 50% are not
CTMP Birch Pine
—
66
 ≔ ← → ≡
71
Table of contents Previous chapter Next chapter Index

—
67
included in the statistics in L&W Fiber Tester Plus It is possible to save raw data to be able to look
Different kinds of because very few fibers have such a low shape at individual fibers and the computed kink angles
possible deformation
of the fiber wall
factor. CTMP fibers of softwood after latency afterwards. Kink and shape factor are often
treatment are straighter than softwood chemical correlated. All types of deformation are included
fibers. Nevertheless, softwood chemical fibers in the shape factor. The number of local
give a better sheet strength due to better deformations of the fiber may, on average, be one
bonding, since they are more flexible and contain per fiber, meaning that the fibers in general have
less lignin. weak points. These are probably important for
the fiber strength.
Curl is often used as an alternative to shape
factor. Curl is defined as:
Wrinkles

C = (L / l) - 1
Compression

12.8. Kink
Local deformations such as knees and wrinkles in Axial micro compression

the fibers are called kinks. They are detected as

changes in the direction of the main axis of the Knees
fibers within a limited distance of the fiber.
A direction on each side of a center point is
Twist
calculated. If the angle is above 20°, a kink is
recorded (Table 13 shows class by angle). Data
—
from kink measurements are: kink/mm, kink/ 67
fiber, kink> 60° per fiber, mean segment length
(average distances between the kinks), modified
kink index (20°) [Kibblewhite] and mean kink —
Table 13: Class weights for the kink index formula
angle. Detailed information about distributions
and actual position of kinks on each individual Class (i) Angle Weight
fiber is available.
1 10–20 1
2 20–50 2
One way to detect the effects of hidden
3 50–90 3
deformations is to treat the fibers chemically
4 90–180 4
or mechanically and measure the result on the
fibers. One such technique is to measure
dislocations and other weak points in spruce
pulp fibers involving hydrochloric-acid-induced
cleavage of the fibers and an analysis of length- Kink index = ∑ Weighti × Number of kinksi
weighted fiber length distributions. i=1 ..4 Total fiber length
 ≔ ← → ≡
72
Table of contents Previous chapter Next chapter Index

—
68 ent 2
Particles that are bigger than 7 µm are detected
Segm
A fiber with two in L&W Fiber Tester Plus (actual pixel resolution is
detected kinks will result Kink angle 2
in three undamaged Se
3.5 µm). A defined maximum length limit of
gm
segments of the fiber en Kink angle 1
200 µm is commonly used as a definition of fines.
t1 Seg
—
me
nt Filler particles are smaller than 5 µm and are not
3
69 detected with resolution limit set to 7 µm.
Shape factor correlates
—
well with the number
68
of kinks per fiber,
because all types of Fibers and fines according to ISO standard
deformations are 1.6 100
reflected in shape factor 90
1.4
Number of kinks/fiber

— 1.2 80
70 L<W
1 70
How different areas of
fines are classified. By 60

Width [um]
0.8
definition, no objects Fines P
0.6 50
can be in the L>W area. Fibers
At TAPPI settings of the 0.4 40
fines measurement,
0.2 30
the limit between fines
and fibers are at 100um 0 20
Fines S Threshold limit
instead of 200um
80 81 82 83 84 85 86 87 88 89 10

Shape factor 0

Undried Undried 0 50 100 150 200 250 300

Dried unrefined Dried unrefined Length [um]

— —
69 70

12.9. Fines 12.10. Fibrills

Fines often have a different impact on processes When beating the fibers, some thin parts of the
and products than the fibers. Primary fines are fiber wall are partly loosened. Those fibrils are at
available before beating and include ray cells. the limit of what is possible to see in a measuring
Primary fines have poor bonding properties. cell wide enough to let fibers flow through. By
Secondary fines are created during beating and filtering the image is it possible to see larger
they improve the strength of the sheet (Figure 70 fibrils (Figure 71). By measuring the fiber with and
shows classification of the two). without this filter is it possible to get a value of
the degree of fibrillation of each fiber. As a value,
Both types of fines have a negative impact on the the fibrils part of the area or perimeter of the
dewatering capacity on the paper machine. While whole fiber is used. By getting the value for an
beating softwood, a large part of the produced individual fiber, it is possible to study how
secondary fines are shorter than 100 µm and different fiber dimensions affect the fibrillation
thinner than 30 µm. When beating hardwood, the during the beating. When a mix of hardwood and
dimensions of the produced fines have a wider softwood is beaten, we often see fibrillation on
span in length (the variation in length of fines can the long, wide softwood fibers but not on the
be bigger compared to beating softwood fibers). thin, short hardwood fibers.
 ≔ ← → ≡
73
Table of contents Previous chapter Next chapter Index

—
71
12.11. Crill
A fibrillated In the pulp, there also is thinner fibrillar material,
softwood fiber
called crill. These particles can be a few 0.1 μm
— wide, and 100 µm long. Many of these particles
72
Fibrills contribute cannot be seen with normal visible light. In the
to fiber bonding crill add-on feature within Fiber Tester Plus, the
and closely relate to
strength properties attenuation that is dependent on light scattering
in the dry paper. is measured at different wavelengths of the light.

The light scattering is dependent on the size of

the particle or the structure that scatters and the
difference in index of refraction compared with
—
71 the medium around the particle.

40
Normally the scattering is highest when the
35
size of the particle is in the same order as the
wavelength of the scattered light. When choosing
Fibril perimeter (%)

30
25
wavelengths, light absorption of different
20
chemical groups needs to be taken in
15
consideration. In this case, we have chosen 365
10
and 850 µm. The ratio of the attenuations at the
5
two wavelengths is used as an indication on
0
the number of particles with a size in the 0.3-0.4
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 µm region.
Tensile strength (kN/m)
Kraft pulp CTMP Eucalyptus In the beating process, crill develops very similar
to both secondary fines and fibrils, hence are also
—
72 important for fiber bonding and strength
properties in the dry paper.
 ≔ ← → ≡
74
Table of contents Previous chapter Next chapter Index

—
13. Objects other than fiber

—
73
13.1. Vessels Eucalyptus
250
A typical collapsed vessel cell of eucalyptus

Width (um) (micrometer)

Each detected object
in an eucalyptus pulp
is marked with a dot.
can be of the size 200×200×20 µm. A typical 200

Objects narrower than hardwood fiber can be 800×20×20 µm. In the case 150
50 μm and longer than
0.2 mm are fibers.
of hardwood pulps, vessel cells can cause printing 100
Shorter are fines, and problems, and this can limit the use of hardwood
wider objects are 50
mainly vessel cells.
pulps for the production of certain paper grades.
0
Vessel cells eucalyptus The amount of vessel cells relative to the amount
(number vessel/100 0 500 1000 1500 2000 2500 3000
000 fibers)
of fibers, and the length/width distributions of
Length (µm)
the vessels are measured. —
73

The experience from L&W Fiber Tester Plus is that Fibers are defined as objects longer than 0.2 mm
the ratio of detected vessel area to the detected and never wider than 0.1 mm in the measurement.
fiber area in a sample is a good measure of Vessel cells are defined as objects wider than
percentage of vessels. Vessel cells are thin-walled a certain width, for example 0.1 mm. If a
and often collapsed, which makes their area comparison is to be made between different
interesting. The total volume of the fibers is a instruments, it is important to define detection
measure of the consistency. The ratio of number criteria in similar ways. When this is written, the
of vessels to number of fibers for the whole detection criteria for vessel cell detection are
sample (0.1 g) is also monitored. The size rather simple, but they work well for virgin
distribution of the counts is shown in a length/ eucalyptus and acacia.
width matrix, see example in Table 14. In this
case, only fibers wider than 100 µm were In more difficult applications, it is possible to use
measured. Another way of expressing the same the L&W Fiber Tester Plus to collect images with
thing is by the ratio of total vessel length to total vessel-cell-like objects and then manually check
fiber length. Note that these different ways of the nature of the objects. Since there are only
expressing the amount of vessel cells can give a few vessel cells per thousand fibers, longer
different results. It is not completely clear what is measurement times can be necessary in order to
the best definition of vessel cells, and that is the achieve good statistics for the measurement of
reason why alternatives are used. In the samples vessel cells.
used in Figure 73, the vessel cells were
concentrated with hydrocyclones.
 ≔ ← → ≡
75
Table of contents Previous chapter Next chapter Index

—
13.2. Shives Table 14: The table shows example of length/
Detected particles with a width greater than width distribution for vessel cells (acacia).

75 µm are measured as shives. 0.20 <W < 0.50 – 25 % 0%

0.15 <W < 0.20 24 % 15 % 4%
Shives have poor bonding ability to fibers and
0.10 <W < 0.15 33 % 0% 0%
they cause quality problems in the paper and in Width (mm)
the papermaking process. Short shives (< 1.5 mm) Length (mm 0.1 < L < 0.2 0.2 <L < 0.3 0.3 < L < 0.5
are considered to give linting problems. Large
shives (Length > 1.5 mm and width > 150 mm) are —
considered to increase the risk for web breaks. If Table 15 : The table shows example of length/width
the sums of the corresponding channels in Table distribution for shives (Red numbers = increased risk for
linting. Grey numbers = increased risk for web break).
15 are calculated, then signals for linting risk and
web-break risk can be monitored. 0.30 < W < 1.0 30 10 10
0.15 < W < 0.3 600 90 20
Different averages are reported such as number
0.075 < W < 0.15 2250 355 100
of shives per number of fibers, and total length Width (mm)
of shives in sample/total length of fibers in Length (mm 0.5 < L < 1.5 1.5 < L < 3 3 < L < 7.5
sample (length weighted), and number of
shives/g if the weight is known (in laboratory
samples). Total volume of shives in sample/total
volume of fibers in sample is recommended to
use in online applications.

If only a few shives are available in the sample,

the statistics can be improved by measuring at
somewhat higher consistency and by taking the
average at several measurements. For an online
system, it is recommended to use a sliding
average or an exponential filter to even out
the random noise.
 ≔ ← → ≡
76
Table of contents Previous chapter Next chapter Index

—
14. Calculated properties

—
74
14.1. Coarseness 14.2. Fiber-wall thickness
The fiber illustrated Coarseness is defined as weight per unit fiber An average fiber-wall thickness can be calculated
as a cylinder for the
calculation of wall
length. ABB’s L&W Fiber Tester Plus calculates from L&W Fiber Tester data. The density of the
thickness. A wall coarseness and reports coarseness on the fiber wall can differ between different species.
density has to be set.
standard laboratory report. Coarseness is For example, eucalyptus and pine do not
calculated as a constant multiplied by the number necessarily have the same wall structure. It is
of images included in the measurement and the possible to use different wall densities. The wall
dry fiber weight of the sample. This number is thickness influences the flexibility of the fibers.
divided by the total fiber length for the measure- Thin-walled fibers are the more flexible and
ment. The constant is the measurement volume collapse more easily than a thick-walled fiber.
for one image divided by the total volume for the More flexible fibers will give better bonding in the
whole internal circulation. sheet (boiled macaroni theory). A thick-walled
fiber needs more refining. The wall is only a few
C = constant× Nb× M/ΣL micrometers thick, which makes it difficult to
measure using normal light microscopy.
• Nb = number of images included in the
measurement
D
• M = measured weight of “dry” fibers fed into
the system T
• ΣL = sum of fiber length for the measurement R
—
• constant = the corresponding volume for one L
74
image per total volume in the circulation

T = D / 2 - √(D² / 4 - C / (π × R)
Note: The constant is in practice defined by
calibration.
Where:
• T = average wall thickness for the whole sample
Coarseness has commonly been used instead of
• C = coarseness (weight per length unit)
fiber density for historical reasons. It had not
• D = diameter (average fiber width)
been possible to measure fiber diameter in the
• R = density of fiber wall (must be defined by
past. An extension may be to calculate fiber
calibration)
density – a measure of weight per fiber volume!
• C and D are measured in Fiber Tester.
 ≔ ← → ≡
77
Table of contents Previous chapter Next chapter Index

14.3. Slenderness
Slenderness, defined as the ratio of length to
width, can be calculated from raw data and
weighted in the same way as other properties.
Volume weight can be a good choice.

The ratio increases for long and narrow fibers.

The amount of fibers in this corner of the length/
width distribution is an interesting fraction. The
length-to-width ratio for European beech is 37,
for aspen 61 and for spruce 95.
 ≔ ← → ≡
78
Table of contents Previous chapter Next chapter Index

—
15. Pulp properties

— 15.1. Freeness Both methods suffer from the fact that the
75
Regardless of the The Canadian standard freeness (CSF) method drainage of the pulp does not simulate the
similarities between is widely used for mechanical pulps. The scale drainage process in a modern paper machine.
CSF and SR, the test
results cannot generally range is from 20 to 700 ml. High freeness means Nevertheless, these methods are standard
be transformed easily de-watered pulp, i.e. pure water gives methods and in all trading of pulp or discussion
between the test
methods because a freeness value of 880–890 ml. For chemical of refining these traditional units are used.
of differences in the pulps, the Schopper-Riegler method (SR) is more
testing equipment.
This is a summary from common. The scale is opposite to that of CSF, so Automatic drainability testing systems, like the
several measurements, that 1000 ml corresponds to 0 SR-units, and a one in ABB’s L&W Freeness Online, are calibrated
as an indication
(Source: TIP 0809-01). volume of 0 ml corresponds to 100 SR-units. against CSF or SR, where freeness is calculated
The SR-method is not suitable for well-beaten from the dewatering of the pulp as a function of
hardwood pulps because these fibers can pass time.
through the wire without building up a sufficient
fiber bed. The drawbacks of manual freeness tests, such
as time-consuming sampling and consistency
Conversion from CFS to SR determination, combined with the very critical
90
handling of the entire test procedure, make the
80
laboratory testing of freeness one of the most
70
expensive and troublesome methods. It is quite
60 common that only a few measurements per
SR units

50 sample point can be done per day, and this makes

40 laboratory-determined freeness unsuitable for
30
process control. All these drawbacks are easily
20
overcome in an online, automatic test system like
10
the one in L&W Freeness Online.
0

D = diameter (average fiber width)

R = density of fiber wall (must be defined by calibration)

—
75
 ≔ ← → ≡
79
Table of contents Previous chapter Next chapter Index

—
76
The volume of the water is measured as a • The time when the flow F0 is reached is read
Dewatering curve function of time. The flow during the formation from the flow curve F(t). This flow corresponds
measured from
the balance
of the fiber pad is integrated; the volume flow to the flow that passes through the lower part
can be calculated. The volume signal is sampled of a manual freeness instrument (see Figure 78).
—
77 into the computer and is then adapted to a • The volume that, in a manual freeness sensor,
The flow from the mathematical function. The noise is reduced in would have flowed through the bottom pipe is
wire calculated from
the drainage curve. this way, which is important when the signal in subtracted from the total volume V(tk). The
—
the next step is to be differentiated. volume in the tank of the freeness module when
78 the measurement starts (Vk) is also subtracted.
The funnel in the
freeness sensor is V(t) The freeness value is calculated as follows:
in practice a flow
sensor. In Freeness
Online the integrated
Freeness = V(t ) - F × t - V
V(tk)
flow is measured
by the draining of
measurement tank as
a function of time.
F=V’(t)

tk t
—
76
F₀

15.1.1. Calculation of freeness tk t

—
• Volume of sample (water+fiber) draining 77
through the screen is measured as a function
of time (V(t)) . Freeness tester Freeness Online

• A mathematical function is adapted to V(t) to

obtain the filtering effect of the weight signal.
(see Figure 76) F F

• The drained water as a function of time is

recalculated (V(t) is differentiated) to a flow
rate curve (F) as a function of time (see
F₀
Figure 77).
—
78
 ≔ ← → ≡
80
Table of contents Previous chapter Next chapter Index

— 15.2. Near infrared (NIR) As described in chapter 8, the NIR spectrum

79
A typical NIR The NIR spectra are used for strength predictions depends a lot on water. Measurement on a dry
spectrum for a paper in combination with fiber distributions. Water is sheet makes the method more independent of
(wavelength in nm).
always present in a paper sheet. It affects the the water. The formation of the sheet is critical.
spectrum close to 950–980 nm and 1920–1940 The thickness of the sheet is also important.
nm (OH).
— 15.3. Optical properties
Table 16: Path length in water for NIR The optical measurements performed early in
Wavelength (nm) Path length(nm)
the process can be used to predict the same
properties of the final product. Then, the
1190 19.9
necessary actions can be taken immediately upon
1450 0.9
deviations observed in the pulp line. The time
1940 0.2
delay in the papermaking process is avoided and
the paper manufactured is always less likely to be
Cellulose affects the spectrum at 1490, 2102, and
off specification in terms of optical properties.
1818 nm; hemi cellulose affects the spectrum at
The pulp furnish is optimized regarding
1982 nm; and lignin at several wavelengths for
bleaching, optical brighteners and color dyes. The
example 1680 nm.
savings connected are obvious.

0.5525

0.4815

0.4105

0.3396
Absorbance

0.2686

0.1977

0.1267

0.0558

-0.0152

-0.0862
1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
Wavelength
—
79
 ≔ ← → ≡
81
Table of contents Previous chapter Next chapter Index

—
80
15.3.1. Brightness The same parameters are used in the pulping
Part of the spectrum The brightness parameter is used to monitor process to supervise the possible dosage of dyes
included in the
measurement of
the bleaching process. Brightness together with used to maintain the desired shade, whether
brightness with an fluorescence is used to supervise the dosage of white or colored. The recycled pulp requires
effective wavelength
of 457 nm.
OBA (optical brighteners). Brightness is the most particular supervision in this respect as the
commonly used parameter in the sales specifi- incoming recycled pulp may vary in terms of color.
—
81 cation of paper and paperboard products. See Then the representative information collected
Color matching Figure 80. early in the process is of great value.
functions.

—
measurement of brightness

82 120
2.0
Power distribution for

CIE L * a * b color space. 100

Power distribution
80 1.5

60 1.0
40
0.5
20
0 0
400 500 600 700 400 500 600 700
Wavelength (nm) Wavelength (nm)
— —
80 81

15.3.2. Color L* = 100

The colorimetric response of the human eye can
be described by means of the color matching
functions according to Figure 81. – a* +b*

In case of a spectrophotometer, the tristimulus

values X, Y, Z are calculated by spectral addition
c*
of the product of the color matching function,
h*
the illuminant function, and the reflectance
factor obtained by the measurement. All other
colorimetric parameters are then calculated
from the tristimulus values. – b* +a*

The colorimetric parameters commonly used

are L*a*b* at specified illumination. The L*a*b* L* = 0
parameters constitute a color space with a grey-
—
scale axis (L), a red-green axis (a), and a yellow- 82
blue axis (b) in a three-dimensional color system
(Figure 82). These parameters are used to specify
the color of the final paper product.
 ≔ ← → ≡
82
Table of contents Previous chapter Next chapter Index

—
83
15.3.3. Whiteness The black ink absorbs light to a much greater
Development of The parameter whiteness is used to describe extent than the cellulose fiber. This absorption
residual ink content
at different process
the perceived whiteness of a finished sheet. occurs in particular in the range between 900–
stages in a DIP plant. Whiteness is, for example, influenced by 1000 nm. Thus, by measuring this absorption at
bleaching, coloring and OBA. Though whiteness different flotation stages, the performance of the
may be considered as the most understandable flotation process can be supervised. When
of the optical parameters, it is a fairly new necessary, the recycled pulp can be additionally
parameter in paper product specification and is bleached.
rapidly gaining more and more interest. It is used
particularly for the OBA- containing paper grades. Measurement of residual ink is standardized and
the parameter measured is referred to as L&W
1000
Elrepho ERIC (estimated residual ink content).
900
The method is based on the Kubelka-Munk
800
theory, which means that the scattering
700
coefficient and the grammage must be
600
measured. The absorption coefficient and
Res. ink

500
the ERIC 950 value can then be calculated. A
400
simplified method is suggested for online
300
applications, where the scattering coefficient
200
is kept fixed at e.g. 55. Then the absorption
100
coefficient follows from a reflectance
0
Day 1 Day 2 Day 3 Day 4 Day 5
measurement on the dry pulp sample. This
simplified method should work well for process
Stage 1 Stage 3 Stage 5 control purposes.
Stage 2 Stage 4

— As an example, the residual ink was measured

83 over five days at a plant for newsprint. Figure 83
shows how the amount of residual ink varies in
15.3.4. Residual ink the incoming pulp. There is a steady increase of
One task in the flotation of recycled pulp is to residual ink in stage 2 for four days. However on
detach the black ink particles remaining from the the fifth day the process was adjusted and the
dissolved, printed sheet. removal of residual ink was significantly
improved.
 ≔ ← → ≡
83
Table of contents Previous chapter Next chapter Index

—
16. Statistics and reports

—
84
16.1. Gaussian density function Arithmetic mean value
Normal distribution; The normal distribution, called the Gaussian
σ = standard deviation,
Σi xi
density function, is the most important x=
μ = mean value. n
distribution function in mathematical statistics
(see below). It is widely used to describe • n = number of values
variations in signals. It describes random errors • x = arithmetic mean
in repeated similar measurements of the same
object. The probability for a result from a Example:
measurement to fall within certain limits can be Calculate the average value for 5, 9, 7, 7, 4
described. x = (5+9+7+7+4) / 5 = 6.4
(x-μ)²
1
f(x) = ×e 2σ²
Limitations:
σ 2π
• Does not predict uniformity of product very well
0.4
• Extreme values have large influence

0.3 Standard deviation

σ = Σ (xi-x)²/(n-1)
0.2

Example:
0.1

Value (xi - x) (xi - x)²

0 6 -1 1
10 μ-3σ μ-2σ μ-σ μ μ+σ μ+2σ μ+3σ 20
4 -3 9
— x
8 1 1
84
12 5 25
Confidence intervals: -5 25
• 68% of the observations are within +/-1
standard deviation (σ) from the mean value • x = arithmetic mean =
• 95% of the observations are within 2 standard (6+4+8+12+2+10) / 6 = 42 / 6 = 7
deviations from the mean value • Σ (xi - x)2 = 70 (n-1) = 5 σ=+/-3.7
• 99.7% of the observations are within 3 standard
deviations from the mean value
 ≔ ← → ≡
84
Table of contents Previous chapter Next chapter Index

16.2. Confidence interval 16.4. How many fibers have to be measured for
When a trial series is carried out and the value of statistical significance?
a certain property is estimated, it may be For the calculation of length, the number depends
necessary to indicate the uncertainty of the on the length distribution. If we say that the
estimation. This is usually done by indicating the length distribution for softwood has a normal
limits within which the true estimated property distribution (which is not completely true but is
would be with a certain indicated probability. an acceptable approximation) and has an average
L= 2 mm and σ = 1 mm, then σp for the calculated
x +/- tp× σ / n average will be σ / √n, where n is the number of
counted fibers. If we count 10000 fibers, then
Where the coefficient tp is derived from a σp= σ/ 100. In our example, this is equal to
reference table. The standard deviation σp, of the 0.01 mm.
mean representing 68% probability for the
calculated mean value based on n observations, We need the statistics to achieve accuracy in the
where the standard deviation of the measured average values. If we test 625 fibers 'a high count
property itself is σ, is given by: for manual testing (microscope)' then σ will be
0.04 mm. Accordingly, it is difficult to reach the
σp = σ / n same repeatability manually with a microscope
that can be reached with an automatic optical
16.3. Poisson distribution analyzer. In the optical device, we can easily
Detection of shive elements or vessel cells in fiber increase the count to 40000 fibers and decrease
images can be regarded as single rare events. The the statistical scatter (σ) to 0.005 mm. When the
detections are random and independent of each distribution is divided into classes, the scatter in
other. These types of events are regarded as each class will, of course, be higher than that of
Poisson distributed. If µ = mean value for the average, depending on the number of counts
detection, and σ² = the variance, then the in each specific class.
standard deviation for the average is:
There is a limitation for how many fibers are
σ= µ recommended to count. At some point, other
sources of error, such as sample handling, will set
If 625 shives are found, then the standard the limit for variations in the sample.
deviation will be 25.
 ≔ ← → ≡
85
Table of contents Previous chapter Next chapter Index

16.5. Weighting of data Weighting that is more in favor of large objects

In the following we will describe how certain will increase the average or increase the influence
weighting functions can be used. Each measured of higher parts of the distribution.
value is multiplied by a weighting factor before
the sum of the different values is calculated. Mass weighting is often a reference, when
comparing with screened fractions and this is
There are large differences in size between fiber one reason for using weights. Another reason is
objects. Fines are very different from fibers, and that if we want good statistics for the fibers with
the manner in which fine objects affect the less impact of fines, weighting is a way for
average of certain properties depends a lot on improvement.
the technique used for weighting the data. We
often define fines as elements shorter than 0.2 Weighting is a non-linear operation. If you
mm and exclude them from calculations of fibers use multivariate data analysis, it can be
with a length above 0.2 mm. But you will also find recommended to test different weighting
a lot of small objects above this limit. functions to improve the models, since the
mathematics in MVDA are linear and do not
Following weightings are used: cover non-linear operations.
• 1, which is equal to arithmetic weighting by
number Example A
• Length-weighted, weighted with respect to We shall calculate statistics for the synthetic
length fiber distribution shown in figure 85. Data are
• Length× length-weighted, sometimes called presented as mean values and distributions.
mass weighted (which is misleading, since Length-weighted data are more commonly used.
optical methods cannot measure mass but are If the data for fiber properties are to be used to
based on the fact that fiber length often estimate some sheet properties at a certain sheet
correlates to wall thickness of the fibers), grammage, then mass-weighted data should be
weighted with respect to length² better. The closest you can come to mass with an
• Area-weighted, weighted with respect optical system is to use area or volume as
to length × width weightings. If you think wall thickness is
• Volume-weighted, weighted with respect to proportional to fiber length, then length × width
length × width² × length might be a good weighting.
 ≔ ← → ≡
86
Table of contents Previous chapter Next chapter Index

Weighting of data means to study how much of a This is the same expression as in [Formula A] but
certain property is carried by certain fibers. the weighting factor 1 is substituted with the
volume of each particle. The volume is calculated
In the simplest case, the weighting is the number in the conventional way from length and width.
of particles. We then have a length distribution
where the value of each class is the number of The corresponding calculation based on the
particles in this class, or is expressed as the distribution is identical with [Formula B] but the
share of the particles by number in that class. class- value is now the sum of the volumes
The average will be: instead of the number of particles in the class.
n
Formula A Σ li = Σ 1×li
l = i=1 Data can be weighted with any property in the
n Σi 1 same way, but the results should have a physical
meaning so that they can be interpreted.
which is the arithmetic mean value. This value can
also be calculated approximately from data Data from L&W Fiber Tester Plus are often
classified in a distribution: presented as length-weighted (xl) or arithmetic
averages (xa) of length, width and shape factor.
Formula B
l = Σ(class valuei × class centeri)/ Σ class value
Σi li×xi
[14.10] xl =
Suppose that we wish to describe the result of a Σi li
fractionation where the reference is the true
Σi xi
weighted classes representing the distribution. [14.11] xa =
n
This means that the value for each class should be
how much absolute mass then is in each class. For
simplicity, assume the particles to be massive Data used in the example (Figure 85) were:
with given constant density.
Number of objects Length (mm) Width (mm)
If so, we can use volume weighting of the data. 10 2 0.2
The volume-weighted mean fiber length will be: 15 1 0.15
30 0.3 0.1
n
Σi vi×li 60 0.1 0.05
l = Σi vi
• Arithmetic mean length = 0.43 mm
• Length weighted mean length = 1.17 mm
• Volume weighted mean length = 1.58 mm
Length has been used as a standard weighting
partly because it is a better approximation to
mass than an arithmetic value and partly for
historical reasons. It was not possible to measure
width with the early instruments for length
measurement.
 ≔ ← → ≡
87
Table of contents Previous chapter Next chapter Index

0.6
— 0.6
85
Example
0.5
B

Share of total fibers

Example of weighting of A mixture
0.5 of CTMP, chemical birch and chemical

Share of total fibers

results. All distributions 0.4
describe the same
pine
0.4 fibers was analyzed in the L&W Fiber Tester.
0.3
sample (above). Values
0.3 in Table 17 were reported. The first column
Different weightings 0.2
highlight different
represents an arithmetic mean value, which
0.2
regions of the data. means
0.1 that the weighting factor is 1.
0.1
— 0
0
Table017: Class0.5 1
weights for the kink1.5index formula
2 2.5
0 0.5 1 1.5 2 2.5
Fiber length (mm)
Fiber length (mm)
0.7
Weighting
0.5
factor 1 L L×W V L×L
0.7 0.5 length
Mean 0.999 1.555 1.772 1.924 2.259
0.6
Share of total volume

0.4

Share of total length

0.6 Mean
0.4 width 23.9 26.9 31.8 37.0 30.6
Share of total volume

0.5

Share of total length

0.5 0.3 shape factor
Mean 91.3 90.0 89.6 89.5 88.4
0.4
0.4 0.3
0.3 0.2
0.3 0.2
0.2
0.2 The
0.1 other columns represent different weighting
0.1
0.1 factors;
0.1 length, area, volume and the square of
0 0
00 0.5 1 1.5 2 2.5 the 0length.
0 Length-
0.5 weighted
1 is traditionally
1.5 2 used.
2.5
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
Fiber length (mm) Fiber length (mm)
Fiber length (mm) Volume is used in theFiber
L&W Fiber
length (mm) Tester, when it is
important to get close to the mass. L × L are used
0.6
by older devices on the market to get close to the
0.5 mass because measurements of width did not
Share of total fibers

0.4 exist or were not very reliable. For calculations of

0.3 blends in the L&W Fiber Tester, volume is often
0.2 used. Volume-weighted proportions of fiber
species can be translated to mass-weighted
0.1
proportions.
0
0 0.5 1 1.5 2 2.5

Fiber length (mm)

0.5

0.4
Share of total length

0.3

0.2

0.1

0
1 1.5 2 2.5 0 0.5 1 1.5 2 2.5

Fiber length (mm) Fiber length (mm)

—
85
 ≔ ← → ≡
88
Table of contents Previous chapter Next chapter Index

—
86
The length distributions with 5 length classes for Since length, width and shape factor are
Fiber length distributed the actual mix of fibers (CTMP, HW and SW) are measured on the same objects, it is possible
into 5 length classes
with 5 different
shown in Figure 86 with different weightings. As to study multi- dimensional distributions. The
weightings. was exemplified in the synthetic example A, the easiest is to monitor two components with the
— same pattern is shown here based on data from third as a color-coded surface falling into
87 Fiber Tester. Different parts of the distribution different classes of the other two.
Volume weighted
distributions of are highlighted with different weightings.
fiber width. As can be seen by how the proportion of volume
— Figure 87 shows volume-weighted width (Figure 88) is distributed depending on length
88
Data from Fiber
distributions of SW, HW and CTMP. CTMP is well and width, it is clear that length and width do not
Tester divided into separated from SW with volume-weightings. The correlate. If we look at shape factor, we can see
15 length classes
and 15 width classes.
lignin still remains in CTMP fibers and thus the that it depends on length but not on width.
Two-dimensional fibers are clearly wider than chemical fibers.
presentations give
an extra dimension
to the data analysis.
0.7
Shape factor depends 40

Proportion of length
on the fiber length. 0.6

of 50 classes (%)v
Observe the red colored 30
0.5
area for bendability.
Proportion

For the shape and 0.4

20
bendability figures,
0.3
the color indicates the
value, and the lines 0.2 10
show the proportion 0.1
of the volumes. 0
0 0 1 2 3 4 5
1 2 3 4 5
Length (mm)
Weighting function
(1 Arithmetic, 2 Length, 3 Area, 4 Volume, 5 Length × Length)

Series 1 0,2-0,5 mm Series 2 0,5-1,5 mm Series 3 1,5-3,0 mm

Series 4 3,0-4,5 mm Series 5 4,5-7,5 mm
—
Proportion of volume (%/class)
86
70 4.2
10
3.8
Proportion of volume in each

9 60
3.4
8 50 2.9
of 50 classes (%)

2.5
Width (μm)

7 40
6 2.1
30 1.7
5
20 1.2
4 0.8
3 10
0.4
2 0 0
1 0 1 2 3 4 5
0 Length (mm)
Width (µm) —
88
Softwood Hardwood CTMP
—
87
 ≔ ← → ≡
89
Table of contents Previous chapter Next chapter Index

—
89
16.6. Decoupling of impact of length on For hardwoods, the 0.5–1.5 mm interval is a
Shape factor depends other fiber properties representative interval. For softwoods, the 1.5–
on fiber length;
chemical pulp.
By looking at different fiber properties in 3.0 mm interval is representative and often used.
separate length classes, the impact of length
on these properties can be minimized. The Deformations can also be measured as local
following classes are often used (Table 18). deformations of the fibers. In this case, the
dependence on length is a minor problem. In the
Shape factor, as it is defined, is correlated with case of chemical pulps, temperature and pH have
fiber length. In order to separate changes in small effects on shape factor within reasonable
shape factor from changes in length, it is limits. The flow speed in the measurement cell is
recommended to study the shape factor in selected to have a minor effect on shape factor.
different length intervals. Typical intervals The width of the fiber seems to have a very small
used can be seen in Table 18. impact on the shape factor (Figure 89).

— 95
Table 18: Division into basic length classes 90
85
Length (mm) Proportion Width Shape

Shape factor (%)

80
0.2–0.5 P1 W1 S1 75
0.5–1.5 P2 W2 S2 70
1.5–3.0 P3 W3 S3 65
3.0–4.5 P4 W4 S4 60

>4.5 P5 W5 55
50
S1 S2 S3 S4 S5

0.2–0.5 0.5–1.5 1.5–3.0 3.0–4.5 >4.5

—
89
 ≔ ← → ≡
90
Table of contents Previous chapter Next chapter Index

—
17. Sampling

—
90
17.1. Representative samples to check 17.2. Sampling theorem
In this illustration, A production delivered to customers For correct sampling of an analog time-varying
represents one sample
per 3,000 ton pulp,
One company introduced on-line measurements signal to a discrete signal, certain conditions of
B represents three of pulp quality for frequent checks of the the signal have to be fulfilled. It should be
samples per 1,500 tons,
and C represents 100
production of market pulp. Manual laboratory possible to recreate the analog signal without
samples per 1,000 tons. hand-sheet testing is time-consuming. In the false frequencies popping up. If the analog signal
example in Figure 90 (A), the standard sheet- includes time variations faster than half the
testing procedure with sampling, beating, sheet sampling frequency, false frequencies will occur
forming and standard tests was applied every (Nyqvist theorem). The way to handle this
second day. The large square represents two days potential problem is to filter the analog signal
production of 3,000 tons of pulp. The dot before sampling. The low pass filter should be
represents one sample taken from the designed to cut away frequencies greater than
production. half of the sampling frequency.

An automated laboratory analyzer is easier In the pulp and paper industry, the processes are
to operate. Measurements from such an often slow because of the number and capacity of
analyzer were taken once every shift. With six the mixing chests. Automatic measurements for
measurements on the same production volume, fiber quality control with a 5–15 minute cycle time
we get more representative measurements as are often enough, but there are examples where
shown in Figure 90 (B). With automatic sampling the fastest possible measurements are required.
from an online fiber analyzer, such as ABB’s L&W
Fiber Online, quality can be measured once every Figure 91 illustrates how to select sampling
5-10 minutes, which is 100-200 measurements frequency. The same signal is sampled five and
per 1,000 tons, as shown in Figure 90 (C). three times respectively. With five samples, the
frequency is sufficient to describe the actual
signal (process) reasonably well. This is not the
case with the other example with three samples.
Here the curve is not well represented by the
sampling. It is not possible to recover a signal
A B C
—
which is similar to the initial curve.
90
 ≔ ← → ≡
91
Table of contents Previous chapter Next chapter Index

—
91
17.3. Master sample Actual trend before sampling Estimated trend from samples
Illustration of the One user of the L&W Fiber Tester noticed large

Property

Property
problem of sampling and
the risk of information
variations (scatter) in the data for samples of
loss. It is important to waste paper from the yard (waste stock).
have many samples to
be able to follow and
However, there was a large variation between
Samples taken often enough
understand the process. the waste paper reels. represents actual trends well.
More samples per unit
time will always give a
Actual trend before sampling Estimated trend from samples
better representation In order to obtain a representative average,
of the reality.
the user had to take a lot of different manual

Property

Property
samples from different reels. He then mixed all
the samples, disintegrated and mixed well and a
representative sample was taken from this mix to Samples taken too seldom,
will give false trend
the L&W Fiber Tester.
information.
—
91
In this case, this was a necessary step to obtain
reliable data from the L&W Fiber Tester. These
averaged values could then be correlated with the
runnability of the paper machine. Representative
samples are needed in all steps of the analysis.
 ≔ ← → ≡
92
Table of contents Previous chapter Next chapter Index

17.4. Manual preparation of samples for fiber Alt 2:

measurement • Take out 0.1 g by weighing, and then dilute with
Most fiber analyzers are accurate and have good water to about 100 ml
repeatability. However, a critical requirement is • Disintegrate the sample in a mixer, and fill it up
the correct preparation of the samples. You can to 200 ml with water
easily introduce variations by a few percent when
you take a smaller part out of a larger sample. Alt 3:
Since laboratory fiber analyzers use small • Take out 1 g dry sample by weighing, and then
volumes, it can be practical to prepare large dilute to about 300 ml
volumes and then take new representative • Disintegrate the sample with a hand mixer
samples from these. Stirring immediately before (avoid sharp edges in the mixer)
sampling is very important, since fibers have • Dilute to 1 liter. Transfer 100 ml, corresponding
different sedimentation speeds. Here are some to 0.1 g dry fibers, to a bottle and fill it up to
guidelines on how to prepare samples before you 200 ml
analyze in the L&W Fiber Tester:
The way in which samples should be prepared
Wet samples to the fiber analyzer: depends on the purpose of the analysis. However,
• Measure the consistency of the sample no matter what sample preparation method is
• Take out an amount that corresponds to 2 g of used, care must be taken to avoid altering the
dry fibers, and disintegrate it in 2 liters of water fundamental properties of the fibers to ensure
in a standard disintegrator (10 minutes or that false or erroneous results are not recorded.
30,000 revolutions) In general terms, it is desirable to achieve a stable
• Transfer a volume that corresponds to 0.1 g dry state for the suspension. The sample should not
fibers into a bottle, and fill it up to 200 ml with change properties during the analysis. ABB’s L&W
water Fiber Tester usually makes double tests. If the
sample is not stable, the double test will detect a
Dry samples: deviation and signs of an impending deviation
will also be monitored. For mechanical pulps, the
Alt 1: shape factor or latency will change with
• Pre-soak 2 g of the sample in water for at least temperature. To stabilize a sample before a
4 hours freeness test, hot disintegration is sometimes
• Disintegrate (re-slush) the sample in 2 liters of used, but if the deformations of the fibers are to
water, at 30,000 revolutions, in a standard be investigated, the sample preparation should of
disintegrator course avoid changing the property to be studied.
• Transfer 100 ml of this volume, corresponding It has been observed that fiber deformations can
to 0.1 g dry fibers, to a bottle and fill up to 200 also change with time in chemical pulps. For
ml with water example, the time between refining and
measurement of the deformations can play a role.
This will also have an impact on how pulps are
stored in the process before the paper machine.
 ≔ ← → ≡
93
Table of contents Previous chapter Next chapter Index

—
92
17.5. On-line samplers After the sample has reached the analyzer, the
Sampling from the Online samplers are available for low consistency tubing is flushed with clean water to be prepared
process, the principle
operation of the
pulps and medium consistency pulps. It is for the next sample extraction from the process.
L&W Pulp Sampler. possible also to develop samplers for high
consistency pulp. In mechanical pulping,
individual samplers for high consistency refiners
would make it possible to automatically follow
and optimize each refiner separately. Samplers
should take a representative sample from the
process. The analyzer must be able to handle
different distances to the sampling points and
account for transportation time of sample to the
1. Flushing (cleaning) the sampler with water
analyzer. It is important to capture the sample
from a pipe where pulp is transported upwards to
ensure that there is enough pulp in the pipe.
Sampling of processes is a difficult issue,
especially when small amounts of samples are
used. See Figure 92.

If particles are left in the sampling chamber, they

will be taken back into the process. At 2, the fresh
2. Taking a sample
water supply is switched off and the sampling
chamber is filled with the fiber suspension. At 3,
the sampling chamber is pulled out, and the
sample is isolated from the process pipe. Finally,
the fresh water supply is switched on again, and
the fibers pass into the sampling line to Fiber
Online. The fresh water enters the sampling
chamber as a water jet through a small hole
placed at the sample outlet. When the sample has
3. Pushing the sample with water
passed into the sampling line, it is already diluted
—
and it can therefore be flushed with pressurized 92
water, thus allowing for long distances between
the sampler and instrument.

If the consistency or the pressure in the process

pipe is low, several strokes of the sampler are
required in order to get the appropriate amount
of fibers to the analyzer for reliable results.
 ≔ ← → ≡
94
Table of contents Previous chapter Next chapter Index

—
93
17.6. An online fiber quality / pulp analyzer
L&W Fiber Online L&W Fiber Online measures fiber properties
with 4 positions for
the input of samples
online via a sampling device placed in the
from the process. manufacturing chain (an installation on a pipe for
the production of bleached chemical pulp is
shown in Figure 93).

Up to eight samplers can be used for automatic

sampling of the process. The instrument is also
equipped with an input for manual samples. The
—
online system and the laboratory analyzer are 93
built with the exact same components for
capturing images of the fibers. This makes it easy
for the laboratory to follow up calibration.

The measurement technology with its special

sampling system enables totally automated and
frequent analysis of pulp quality throughout the
manufacturing process.
 ≔ ← → ≡
95
Table of contents Previous chapter Next chapter Index

—
18. Impact of fibers on products

—
94
Paper can be described as a stratified structure It is thus possible to measure some properties
From tree to composed of fibers and fines. Each gram of paper at the end of the chain together with specific
end product
contains many millions of fibers. In addition to properties of the product.
the fibrous material, many paper grades also
contain large amounts of fillers. The fibers’ This may help to develop an understanding of
flexibility and the presence of fines increase the what is needed from the fibers. Measurement of
bonding surface and improve the bonding ability fiber properties makes it possible to connect the
between the fibers. Length, fiber deformations, fiber properties directly to the end-use.
cross-sectional properties, fiber surface and Measurements online early in the process chain
chemical composition are all important for paper are made in order to control and optimize the
quality. fiber properties.

Paper properties are dependent on the fiber All paper products have their own specifications
species used, on the pulp manufacturing process and special tests to verify whether the product
and on the paper machine. The final paper fulfills the requirements. An important thing to
properties can be established by controlling the know is which fiber properties are the most
fibers in the pulp, i.e. by making sure that they suitable in each case. Since these may be difficult
have the right quality. A paper with desired to define, tests on laboratory-made sheets are
properties can be obtained by choosing the raw used to test pulp properties. A lot of knowledge is
material and the treatment of the fibers. available regarding the properties of hand sheets.
Therefore, it is of interest to relate fiber
The quality of the end product is determined early properties to hand sheet properties as one step.
in the process chain. How can quality be It is also an advantage if we can describe pulp
communicated through the chain of processes? quality with a few basic fiber parameters.
It is possible to analyze the fiber quality in a Information about sheet properties described in
product by disintegration of the paper and also following sections of chapter 18.
by disintegration of the end product.

Pulping process Paper machine Converting

Paper Product
Wood Fibre quality
properties properties

—
94
 ≔ ← → ≡
96
Table of contents Previous chapter Next chapter Index

—
95
Let's examine the principal effects of fiber By using different temperatures, the pulps were
Shape factor in the sheet properties on traditional properties of hand given dissimilar degrees of fiber deformations
and in the suspension
sheets. and thereby different fiber shapes.
—
96
Only the straight fiber
18.1 Reinforcement About 0.5% of the fibers in the suspension were
will carry load initially Fiber length is an important property of pulp, and dyed with black ink to enable the shape in the
—
longer fibers generally improve the strength sheet to be studied with an image analysis
97 properties up to a certain point. Limited bonding system. The fiber shape obtained from L&W Fiber
The fibers are bonded
in networks.
of the fiber in the network will limit the possibility Tester Plus was roughly the same as the shape
for the fiber to carry a load at the ends of the factor in the sheet. The main work was to show
fibers. With longer fibers, the bonding will be less theoretically by FEM-simulation of fiber networks
critical. Very long fibers are more easily entangled that curled fibers lowered the tensile stiffness of
with each other, giving an uneven fiber the sheet.
distribution in the sheet, i.e. the sheet will have a
poor formation. 18.1.3. Stretched and free segments
A stretched fiber segment between two bonding
18.1.2. Shape factor in suspensions and in the points will carry load in the sheet. A non-
sheet stretched segment will carry load first when the
A comparison was made between the mean fiber sheet has been sufficiently stretched (figure 96).
shape factor in four pulps, and the mean fiber
shape factor measured on laboratory sheets Since the fiber is bonded in a network (figure 97),
made from these pulps, using an image analysis only parts of the fibers are free segments. Some
system. To make the laboratory sheets, a CTMP kinds of paper are very dense, while others have
from a laboratory refiner was used. The pulp had an open structure (like tissue). Pulp for tissue is
a high latency (highly deformed fibers) and was often not beaten very much, and this leads to the
divided into four separate categories. Each open structure. In this structure, fiber flexibility is
category was treated at different temperatures more important for strength since fewer fines are
(20, 55, 65 and 85°C) during the disintegration available for bonding.
according to SCAN-M 10:76.
STFI Image analyzing system, D/L (%)

—
96

— L&W Fiber Tester Plus, D/L (%) —

95 97
 ≔ ← → ≡
97
Table of contents Previous chapter Next chapter Index

18.1.4. Individual fiber strength 18.1.8. Secondary Fines

Wet zero-span test makes it possible to measure Secondary fines are produced during beating and
the fiber strength separated from other effects. give strength to the sheet. Modification of the
The zero-span tensile strength value is intended fiber wall leads to an increased uptake of water
to be an indicator of the average strength of by the wall and also to a more flexible wall.
individual fibers. However, it is dependent on Measurements like Canadian Standard Freeness
fiber deformations. It is also suggested that the (CSF), Schopper-Riegler (SR), and Water Retention
deformations themselves are more important Value (WRV) correlate with both these effects and
than the fiber strength in a non- deformed fiber. with bonding.
The local deformations of fibers may be potential
weak points of the fibers. 18.1.9. Coarseness
It is believed that high fiber wall thickness gives
18.1.5. Fiber width high fiber strength, which will provide a high tear
For certain pulps made from wood from a single index. To calculate fiber wall thickness,
species fiber, width does not correlate with fiber coarseness can be measured if we have input of
length and wall thickness. However, in a mix of the fiber weight. Coarseness is defined as weight
different pulps, it may correlate. A low fiber width per unit fiber length, and it seems reasonable
will give a sheet a more even surface and better that the diameter of the fiber will also play a role.
formation. Sometimes it can be of value to calculate an
overall fiber density, and this is done by dividing
18.1.6. Fibrillation the total mass of the measured fibers by the total
During the beating, a thin part of the fiber wall volume of the measured fibers instead of dividing
partly becomes loose. This is part of the by the total fiber length. This can be extended to
secondary fines that still are still connected to calculate an average fiber wall thickness as well.
fibers. The function in the sheet is the same as If an average wall thickness is to be calculated, a
for the secondary fines mentioned below. fiber wall density has to be set.

18.1.7. Primary Fines This fiber wall density may be different between
Primary fines give low strength because of their different species. A thin fiber wall increases
low bonding potential. Opacity and bulk are paper strength for a certain grammage, and a
affected. Drainage is affected negatively. Ray thick fiber wall gives bulk to the sheet.
cells are primary fines.
 ≔ ← → ≡
98
Table of contents Previous chapter Next chapter Index

18.1.10. Bonding strength 18.1.13. Vessel cells

The strength with which fibers adhere to each Vessel cells cause linting problems in the printing
other, together with fiber strength, contributes press. Bad bonding properties of the pulp
to the pulp strength. Both the bonded area and increase the risk of linting. Bad bonding between
the specific bonding strength between fibers fibers and vessel cells can also cause linting
affects the bonding strength. More flexible fibers problems in the paper machine.
and more secondary fines both increase the
bonding area. The sheet density and the light 18.1.14. Kink
scattering ability of a sheet may thus be used to Local deformations of the fibers are called kinks.
predict the bonded area. The fibers are more A kink can be a weak point of the fiber, since the
densely packed in the sheet, when there are a lot strength of the fiber is obviously unevenly
of contact points between the fibers. Increased distributed along the fiber axis. Kink correlates
density and decreased light scattering ability well with shape factor in most cases, since local
denote more bonded area. The bonding strength deformations are included in shape factor.
can be defined as the strength in the z-direction. Deformed fibers give a more elastic paper (a
Bonding is increased by beating. Hemi cellulose is good property when manufacturing sack paper) if
important for bonding. the paper is freely dried.

18.1.11. Freeness 18.1.15. Shives

Beating increases bonding and decreases Shives lead to paper weaknesses, as they cause
drainability at the same time. Freeness sensors web breaks.
are also sensitive to small particles that reduce
drainability, but do not necessarily increase 18.2. Tensile stiffness and tensile strength
strength. WRV is another measure of how well Pulps collected from different positions in
water is bound to the fiber. It is intended to different mills were analyzed with respect to
simulate the water content after the press shape factor in the L&W Fiber Tester; tests were
section in a paper machine. WRV is often said to also performed on handsheets. Tensile index
correlate better with strength than standard depends a lot on the deformation of the fibers.
drainage measurements (CSF etc). However, The samples with the highest tensile index were
freeness measurements are the most widely used laboratory-made samples. Note that these fibers
measurements in stock preparation testing. were very straight from the beginning and strong
already without beating.
18.1.12. Flexibility
Flexible fibers are believed to give stronger Tensile strength is used in pulp testing as a
sheets and increase the surface available for general characteristic for the capability of
bonding. bonding between fibers.
 ≔ ← → ≡
99
Table of contents Previous chapter Next chapter Index

—
98 9.0 18.2.2. Tear strength

Tensile stiffness index (kNm/g)

Shape factor measured
8.0
The tear strength depends on fiber length, fiber
on never-dried pulp in
non-integrated pulp
strength, cross-section properties, degree of
7.0
mill correlates very bonding between fibers, and degree of
well with the Tensile 6.0
Stiffness Index [RISE].
orientation of the fibers in the paper. Longer and
5.0 stronger fibers have higher tear strength. The
4.0 tear index increases in the early phase of beating
but then decreases with beating.
3.0
80 82 84 86 88 90 92 94 —
Brightness Table 20: Effect on the paper sheet of fiber properties
Shape factor with all other parameters kept constant
Fully bleached Unbleached Oxygen bleached
— Fiber property Tensile Tear Bust Formation
98 Increase in fiber
length + + + -
— Decrease in shape
Table 19: Light scattering ability for factor - + - +/-
different types of species Increase in % fines + - + +
Decrease in
Pine Birch Eucalyptus coarseness + - + +
Length (mm) 3.0 1.1 0.9
Width (µm) 30 18–22 12–16 18.3. Single layer products
Wall thickness (µm) 8 3 2–3
Number of fibers/mg 2000 8000 16000 18.3.1. Printing papers
S (m2/g) at tensile Newsprint furnishes and magazine papers for
index 50 kNm/kg 29 34 41
printed media belong to the category of wood-
containing paper where the life cycle of the
18.2.1. Light scattering printed product is quite limited. As a result, it is
For slender and thin-walled fibers, the surface important to have a high pulp yield. Newsprint
area/ weight ratio is greater than that for broad furnish consists of a mixture of mechanical pulp–
and thick fibers. This provides a larger surface groundwood or TMP (usually spruce) and lightly
area, both for bonding and for light scattering. refined chemical pulp. Sometimes a small amount
Increased beating lowers the light scattering in of filler is added. The mechanical pulp adds
the case of chemical pulp and increases the valuable properties to the newsprint, all of which
tensile index. Eucalyptus gives the highest light are related to printability, bulk, compressibility,
scattering at a given tensile index (Table 19). opacity and formation. Sometimes, the strength
properties of mechanical pulps are insufficient to
produce a sheet that runs well on the printing
press, and the pulp is therefore reinforced with
chemical pulp.
 ≔ ← → ≡
100
Table of contents Previous chapter Next chapter Index

The need for reinforcement pulp has decreased 18.3.2. Fine paper
with the development of paper machines and the Fine papers are white, uncoated printing and
properties of mechanical pulps. It is more and writing grades that contain less than 25%
more common to use recovered paper in a mechanical pulp in the furnish. The grammage
newsprint furnish and the waste paper then also ranges from 50 to 100 g/m2. A large item is A4
includes a certain amount of chemical fibers. The copy paper. Fine papers are made from a mixture
grammage of newsprint papers is 40–45 g/m2. of fully bleached softwood pulp, hardwood pulp
Important properties for printing papers are and fillers. Good sheet formation and a well-filled
listed below with an explanation of why they are surface are necessary attributes for fine paper
important. grades, while strength is not a limiting factor.
Therefore, sulphite, hardwood, and chemical
Surface strength – low surface strength will sawdust pulps are preferred over long-fibered
result in fibers being torn from the web surface, kraft as furnish. The pulp furnish for fine paper is
and this may lead to dusting. Linting can also usually given a light to moderate refining, and 10–
occur if the fibers are ripped off and stick to the 15% filler is added. Important properties for fine
printing press surface. Linting reduces the papers are listed below with an explanation of
availability of the printing machine, since it has to why they are important.
be shut down and cleaned.
Surface strength – low surface strength will
Fracture toughness – fracture strength is result in fibers being torn off the web surface,
important for the runnability of the printing and these may stick to surfaces within the
press. High fracture toughness gives the paper copying machine, causing machine failure.
web the ability to resist fracture from a crack in
the web. Dimensional stability – a paper should not curl
when toner is applied in a copying machine.
Opacity – it is not desirable to read the print on
the opposite side of the paper. Bending stiffness – the paper should not be like
a cloth and bend; it has to stay flat when held.
Smoothness – a smooth surface is a necessity for
good printing. A rough surface will lead to linting, Smoothness – a smooth surface is a necessity for
dusting and uneven print quality. good printing. A rough surface will lead to linting,
dusting and uneven print quality.

Picking
A problem with some hardwood species is vessel
picking. The phenomenon is illustrated in figure
101. Large vessel cells tend to fasten on the
printing plate. The problem is worse on fast
machines with a high-tack ink. A sheet with good
bonding potential may be one way for
improvement.
 ≔ ← → ≡
101
Table of contents Previous chapter Next chapter Index

—
99
LWC-papers must have good runnability and
Surface of newsprint good printability. Stable quality is regarded as
Environmental Scanning
Electron Microsope
important. LWC-papers consist of groundwood
(ESEM) image (150× pulp or thermomechanical pulp, even though they
magnification). A
lot of fines are seen
are quite different mechanical pulps.
between the fibers. Groundwood pulp is rich in fines, which gives
—
good printability to papers, but low strength.
100
The surface structure
of fine paper made
The conse­quence is that a large amount of
in ESEM, (150× chemical pulp has to be added to the furnish. The
magnification).
The wide bands are
long-fibered TMP is stronger and less chemical
vessel elements pulp needs to be added to the furnish. However,
probably from birch. — TMP gives poorer printabil­ity and has a higher
99
— specific energy consumption.
101
Picking - A vessel cell
on the surface of the Printing
Printing
plate plate
paper is poorly bonded
to the sheet and loosens Ink Ink

during printing.

Ink Ink Ink Ink

Coating
Coating LayerLayer

Base Base
sheetsheet

—
101

—
100

18.3.4. Lightweight Coated (LWC)

Lightweight coated (LWC) is a grade of
mechanical printing paper of low grammage,
primarily used as a publication paper where
surface quality and weight are important.
 ≔ ← → ≡
102
Table of contents Previous chapter Next chapter Index

18.3.5. Corrugating medium Air permeance – efficient sack filling requires

Corrugating medium is a lightweight board used that the product entering the sack displace the
for the fluted inner plies of corrugated box stock. air captured in the sack. Extreme permeance will
It is principally made from semi-chemical pulp result in dusting during filling, product wastage
and recycled fiber. High bending stiffness is and dirty sacks. Insufficient permeance will result
required for corrugated box construction, and the in stress to the sack during filling, and improper
fluted layer provides much of it by keeping the sack sizing.
two liner layers apart from each other. Mechanical
strength properties such as tensile, burst and Tear and Burst strength are also important for
tear are unimportant except in certain limited sack paper to prevent rupture in the paper bag
applications. Sheet finish and appearance can or sack when it is containing heavy objects.
also often be ignored. The ideal furnish for
corrugating medium is semi- chemical pulp 18.4. Stratified products
because of its high crush resistance, although
these pulps are notably weak with respect to the 18.4.1. Tissue
more common strength criteria. The term tissue covers a wide range of extremely
low-weight sheets. Sanitary tissue includes facial
18.3.6. Sack grades and bathroom tissues, paper towels, and paper
Sack grades are usually manufactured from well- napkins. Industrial tissues include condenser,
refined, unbleached softwood kraft pulp to meet carbonizing, and wrapping grades. The
the strength requirements. The pulp is refined in grammage for tissue paper is in the range of 15–
high consistency refining in order to introduce 25 g/m2. Tissue cannot be produced on a
microcompressions and crimps. Rosin sizing and/ conventional paper machine because of the low
or starch is added to the furnish to provide grammage of some products and the loose
additional internal strength. The grammage is structures of others. The raw material used can
between 60 g/m2 and 150 g/m2. Good sheet be TMP, recycled paper or bleached kraft pulp.
formation is extremely important for uniform CTMP in the form of fluff pulp is used for diapers
strength, which is why these machines are and sanitary napkins.
operated with very low consistency in the forming
section. For this grade, free drying is preferable, Important properties for tissue grades are
to allow the sheet to shrink during drying. This listed below with an explanation of why they
provides good tensile energy absorption (TEA) are important.
through good stretch.
Softness – the paper should be velvet-like.
An explanation of important properties for sack
papers are listed below. Smoothness – a rough surface with fibers
sticking up, giving the surface an uneven
TEA and Fracture Toughness – sack paper needs appearance, would feel rough against the
to withstand applied forces of large magnitude human skin. The surface must be smooth.
by having a good stretching ability. Curled fibers
with microcompressions are more stretchable.
 ≔ ← → ≡
103
Table of contents Previous chapter Next chapter Index

—
102
Ability to absorb water – tissue paper is most The main requirements of linerboard are a
The surface structure of commonly used for wiping up liquid media of high compression strength and a high burst
tissue from microscope,
(30x magnification).
some sort, usually water, and the capability to resistance. The fibers are strongly oriented in
Notice an open sheet absorb a relatively large volume is probably its the machine direction, since stiffness in the cross
structure compared
with the figures 99 and
most important property. direction comes from the corrugating medium
100 shown earlier in the final product. Important properties for
Wet strength – the intention of tissue is to be linerboard are listed below with an explanation
wetted; therefore it has to have sufficient wet of why they are important.
strength not to fall apart immediately after
becoming moist. Burst strength – is often measured for liner. If the
liner has low burst strength, the box will be weak,
Dry strength – a certain dry strength is necessary easy to penetrate and/or lose its structural
in the converting process. integrity.

Compression strength – boxes and cartons made

of kraft liner are often piled on each other and are
thereby subjected to compression forces.

Good appearance and printability on one surface

are also important. Using high-yield kraft pulp in
the bottom liner, and a lower yield, well-refined
kraft on the top liner can satisfy these
requirements.

18.6. Paperboard
Technically, all sheets with a thickness greater
than 0.3 mm are classed as paperboard, but there
— is no clear definition of paperboard in precise
102
terms. Boards can have a single-ply or multi-ply
structure with a grammage of 250 g/m² or higher.
18.5. Linerboard
A partial listing of paperboard grades is given on
Linerboard, the facing of corrugated fiberboard,
next page (table 21). Bleached or unbleached kraft
is a relatively lightweight board. Linerboard
pulp is used as well as mechanical pulps and
consists of different layers; a rather dark and
recirculated fibers. An explanation of important
coarse high-yield, unbleached kraft base sheet
properties for paperboard are listed below.
(bottom liner), and a lighter, cleaner, better
quality layer (top liner) with a good printing
Bending stiffness – packages have to be rigid
surface. The grammage is between 100 g/m2 and
and not crumble when stresses are applied.
400 g/m2. When the raw material is recycled
corrugated fiberboard, the paper grade is called
“testliner”.
 ≔ ← → ≡
104
Table of contents Previous chapter Next chapter Index

—
103
Smoothness – a smooth surface is necessary for Pigments

A SEM-image from good printing. The top layers of paperboard

a cross section of Hardwood
paperboard. Note the
boxes are often used to print information
open structure in the regarding the box contents, etc. Coating layer Softwood
middle of the sheet,
Even surface Secondary
the black coating layer
layer fines
and the tight bonded Other properties, such as opacity and bulk, are
sheet at the surface
close to the coating.
more dependent on the area of use. The main Strong layer

advantage of multi-ply forming is the ability to Bonding CTMP

—
104
utilize bulky low-grade waste materials (mostly layer

It is possible to news) in the inner plies of the board, where low

build up sheets with
advanced structures
strength and extraneous materials (ink, coatings, Bulk layer

to optimize quality. etc.) have little effect on the properties of the

sheet. Since several furnishes may be used on a
Bonding
single multi-ply machine, the stock preparation layer
system is generally more complex than that
Strong layer
utilized in papermaking. The top liner, the under
liner, the filler plies, and the bottom liner may all
—
require individual systems.
104

In a sheet construction like that in figure 104, the —

surface layer of hardwood ensures an even Table 21: A partial listing of paper boards
surface. The next strong layer is made of chemical
softwood with good bonding properties. A board with high demand on bending
stiffness. Must be very neutral with regard
Hardwood and softwood are often mixed in the to taste and smell. Made of virgin pulp.
surface layer. Bulk in the middle layer is achieved Different species used in different layers
by using chemical-mechanical pulp with wider to meet different demands on surface
Liquidboard and mid-layers. Usually coated top layer.
and stiffer fibers. A bonding layer is in this case
A board having at least two plies, the top
applied outside the middle layer in order to keep layer being of relatively better quality;
the layers together. usually 100% virgin pulp furnish. Testliner
Linerboard uses 100 % recycled paper.
A board used for food packaging with
single-ply or multiply construction, usually
Foodboard with 100% bleached virgin pulp furnish.
A multiply board used to make folding
Folding box boxes with virgin pulp used for the top ply
board (carton (liner); the remainder of the plies are made
board) from secondary fiber (recycled waste).
A multiply board made from 100%
Chipboard low-grade secondary fiber.
A board that will ultimately be
Baseboard coated or converted.
A multiply board made from 100%
low-grade secondary fiber that is used for
—
Gypsumboard the outer surfaces of plasterboard.
103
 ≔ ← → ≡
105
Table of contents Previous chapter Next chapter Index

—
19. Models

It is desirable to be able to make measurements 19.1. Mathematical and physical models

and to supervise fiber quality parameters online Linear models have limitations; the world is not
in the process. However, in some cases it is linear, but it can be regarded as linear within
preferable to know how the fiber properties will certain limits, and we have good tools to handle
affect laboratory sheet characteristics. linear models. Non-linear models are much more
difficult to handle. One way of solving this
Available knowledge often relates to laboratory problem is to introduce known or estimated non-
sheets, which show the potential paper linear functions before using the linear modelling
properties of the pulp. In the laboratory tools. When statistical data from fiber
procedure, a laboratory beater is often used, but distributions are used for linear modelling it can
the result is different from that achieved in an be an advantage to know that the weighting
industrial beater. It is difficult to make realistic functions used are often non-linear operations.
trials to model the actual paper properties from Different statistics to represent the fiber
pulp properties since the papermaking process properties will give different results. Tests using
involves a lot of complicated parameters. different types of weighting functions may be
one way of improving the models.
A sheet from a paper machine is of course
affected by the set-up parameters in the machine. The stability of the models will increase if
A laboratory sheet former does not produce a knowledge about the process/relationships can
sheet with the same properties as a paper be included in some way in the models. A formula
machine. based on an understanding of concrete physical
relations is always better than a purely
For these reasons, one technique for the mathematical correlation. I would classify the
development of models relates to laboratory method to calculate the amount of fiber
sheets, although with the mentioned limitations. components in blends described in this chapter
The other technique is to develop models directly as being of this latter type.
from process data and the testing of the machine
made paper. A limitation of all models is that they are valid
only within a limited range.
 ≔ ← → ≡
106
Table of contents Previous chapter Next chapter Index

—
105
19.2. Multivariate data analysis (MVDA) PLS modelling solves problems with correlated
PCA model. Black is Multivariate data analysis is an efficient tool for variables by using a few orthogonal latent
recycled pulp. Green is
unbeaten virgin pulp.
the development of linear models. The theory of variables or principal components describing the
Blue is beaten virgin multivariate calibration does not differ from that X and Y block to perform regressions and find a
pulp. Red is the mix
and orange is broke.
of univariate calibration except that many correlation between the descriptor (X) and
variables are used to describe the predictor responses (Y). In other words, principal
space. Modern instruments collect a large set of components are calculated for the X-matrix in the
data when performing an analysis, but the same way as PCA, except that PLS finds relations
interpretation and evaluation of these large between the structures in the X- and Y-matrices.
amounts of data is tedious and difficult.
Multivariate analysis can be applied on large data From an experimental set of data, half the data is
matrices, e.g. PCA (principal component analysis) often used to calculate of a model. The other half
and PLS (partial least square, or projections to of the data is then used to test the predictive
latent structures). capacity of the model.

PCA Model Kraftliner Results from a joint project many years ago
between what was then STFI and Fibertracker
10 with Södra Cell and Vinnova (Swedish agency for
innovation systems that provided financial
Score X 2* (8.36 %)

support) are shown in figure 106. Tests to build

0 linear models of traditional sheet properties
showed good possibilities by combining the
standard Fibermaster properties with some other
-10 measurements. The most important data from
-30 -20 -10 0 10 20
Fibermaster were the shape factor values. The
Score X 1* (52.4 %)
fiber properties were measured on never- dried
Distributions of fibre properties were used as input variables. pulp and were used to predict sheet properties
—
105 on paper made of dried pulp. In the figure below
(figure 106) one result is shown.
PCA is used to obtain an overview and a good
approximation of the main variance in the data. Pulp samples are collected from the process
All the K-variables (number of variables) and during a certain time, with representative
N-variables (number of observations) form a variations. Laboratory sheets are made and
K-dimensional space in which each sample is tested with standard sheet testing equipment. It
represented as a point. By PCA, all samples in the is critical that the standard testing equipment
original K-space are reduced to a new low works well with small variations and is well
dimensional (usually 2–4 dimensions or principal calibrated. If the reference is not correct, the
components, PC) plane, onto which the samples developed models will not work. The fiber
are projected. The contributions of the K-variables properties of the same samples are then
to the sample-oriented projection are visualized determined.
by a variable-related projection (loadings).
 ≔ ← → ≡
107
Table of contents Previous chapter Next chapter Index

—
106 Tensile index (Nm/g) The single distributions from length, width and
An input of Fibermaster shape factor from Fibermaster were used and
data, water retention
value and specific
were combined with NIR spectra from paper
beating energy sheets made of the same samples, beaten at 1000

Predicted
could predict tensile
index quite well.
revolutions in a PFI beater, in order to predict
tensile index. Reference samples were about 30
—
107
fully bleached pulp sheets taken after the dryer
An illustration of how from a chemical sulphate pulp mill. Conventional
a model is developed
offline and then
pulp sheet testing was performed. The data
used online for real 40 50 60 70 80 90 100 included natural variations in strength including
time predictions.
Measured noise. The samples were divided into two data
sets and the degree of explanation and
Refined 50 kWh/ton
the prediction capability was calculated; the
Refined 100 kWh/ton
results are shown in the following tables (22 and
—
106
23). The first unit using this technology was
installed in a non-integrated pulp mill in Sweden.
*Multi
Lab sheet tests Variate
MVDA* Data The samples were divided into two data sets and
off-line Analysis
Tests off-line
the degree of explanation (R2Y) and the
prediction capability (Q2) was calculated; the
results are shown in the following tables.

Image analysis Fibermaster

Process —
Routine conditions
based on statistics:
quality
Table 22: Model result for the calculation of
snapshots of Average values,
predictor Paper Tensile Index at 1000- PFI revolutions.
fibers distributions properties

— R2Y Q2
107
NIR 0.76 0.33
Fibermaster 0.69 0.5
The models are then developed from these data
NIR + Fibermaster 0.85 0.56
sets using, for example, the multivariate
modelling technique. The developed model
—
consists of a number of factors which can then be Table 23: Model results for the calculation of specific
used to for example predict strength properties energy needed to reach a Tensile Index of 70 Nm/g.
direct from measured Fiber Tester data.
This type of prediction (PLS) has been described R2Y Q2
by one user. NIR 0.79 0.41
Fibermaster 0.70 0.54
NIR + Fibermaster 0.96 0.68
 ≔ ← → ≡
108
Table of contents Previous chapter Next chapter Index

—
108
19.3. Distributions and prediction of PINE – length/width distribution (% / class)
Upper; length/ fiber blends 70
width distribution of 2.8
chemical birch fibers.
Volume-weighted 2-dimensional distributions 60 2.5
Lower; length/width from L&W Fiber Tester Plus are used to calculate 2.2
50
distribution of CTMP 1.9
fiber mixes in recycled pulp. Mixing of up to four

Width (μm)
made from spruce. 40 1.6
The distribution is different pulps can be predicted. Averages and 30
1.3
volume weighted. 1.0
distributions are available for the fiber 20 0.7
— properties, weighted in different ways in L&W 10 0.4
109 0.1
Upper; length/
Fiber Tester Plus. Both length-weighted and 0
width distribution of volume- weighted “surface data” are stored in the 0 1 2 3 4 5
chemical pine. Lower;
measurements for a
following classes; proportion of length and Length (mm)
mix of all three pulps. volume respectively in 2-dimensional length/
The distribution is Measured value (%/class)
volume weighted.
width matrixes in 15×15 classes. Figure 108
70 3.1
(below) and 109 show examples of three pulps
— 60 2.8
110
used in production of paperboard. 2.5
50
Chemical pine and CTMP 2.1

Width (μm)
of softwood pulps were 40 1.8
weighted separately BIRCH-length/width distribution (%/class) 1.5
30
and mixed together. 1.1
70 9.9
The composition of the 20 0.8
mix was calculated. 60 8.9 0.5
10
7.9 0.1
50
6.8 0
Width (μm)

40 5.8 0 1 2 3 4 5
30 4.7
Length (mm)
3.7
20 —
2.6
10 109
1.6
0 0.5

0 1 2 3 4 5 Mixes of softwood/hardwood

Length (mm) 98
97
Procent SW measured (%)

96
SPRUCE-length/width distribution (%/class)
95
70 1.7 94
60 1.5
93
1.4
50 92
1.2
40 91
1.0
Width (μm)

0.8 90
30
91 92 93 94 95 96 97 98 99
0.6
20
0.2 Aimed procent SW [%]
10 0.1 Fiber tester Linear (fiber tester)
0 0 Fiber online Linear (fiber online)
0 1 2 3 4 5
Length (mm)
— —
108 110 100
ed amount of birch from

90
80
erTester PlusBlend

70
60
50
40
30
20
 91 92 93 94 95 96 97 98 99

Aimed procent SW [%]

≔ ← → ≡
109
Table of contents
Fiber tester Previous
Linear chapter
(fiber tester) Next chapter Index
Fiber online Linear (fiber online)

—
111 100
The maximum in width is above 40 µm for CTMP.

Estimated amount of birch from

Chemical birch and 90 For chemical pine the maximum is clearly below
CTMP softwood 80

FiberTester PlusBlend
pulps were weighted 70 40 µm. CTMP also has significantly more fines.
separately and 60 These differences make it possible to use these
mixed together. The 50
composition of the mix 40 distributions to calculate data for mixes of these
was calculated with 30 pulps. Calculated and measured amounts are
the L&W Fiber Tester 20
Plus Blend package. 10 compared in figures 108 and 109 on the previous
0
0 20 40 60 80 100
page.
Measured amount of chemical birch in mix with CTMP
Volume-weighted L/W-distributions were used
one third by weight of each component. The same
Prop. Birch Prop. CTMP references were used and the same densities. The
90 89.43 result is shown in Table 24.
90 88.98
—
75 74.75
Table 24: Calculation of mix three components
75 74.33
50 49.45 Birch Pine Spruce(CTMP)
50 51.31 Blended mix
25 26.52 (% of weight) 33.3 33.3 33.3
25 25.56 Calculated mix
(% of weight) 34.4 31.3 34.4
10 10.17
10 10.66
19.4. Separation of spruce and pine
— Since the distribution functions of spruce and
111
pine are similar, the separation possibilities with
Fiber Tester are limited.
Three fiber components were used in a board
mill. Chemical fibers of Nordic birch and pine
19.5. Bauer McNett model
were used together with CTMP made from
Length distributions presented by optical
softwood. The three distributions are very
analyzers give results which differ from a
different. Birch has a narrow distribution with
conventional Bauer McNett (BMN) classification,
respect to both length and width and the average
based on screening and weighting. This can be a
values are low for both. The other two have a wide
drawback when introducing the new technology.
distribution with respect to length. CTMP made
One way to show the potential of the new
from softwood has a wider distribution and the
technology is to model BMN fractions with the
results were compensated with the fiber density
optical technique. How can this be done? This
measured in Fiber Tester for each reference in
way of doing this has been suggested: Fiber
order to predict weight relations. Finally, a mix of
intervals corresponding to the BMN fractions are
all three pulps was made with in width and also a
defined, and a weight factor is defined for each
higher average.
interval.
 ≔ ← → ≡
110
Table of contents Previous chapter Next chapter Index

—
112
By using actual fiber length data, a set of fiber But it was also shown that quite good models
Tensile index (refined classes and a set of weight factors, the BMN based on fiber parameters from the wet pulp
market pulp) vs. shape
factor from never-
classes (R14, R28, R48, R100 and P200) can be could predict tensile index after drying, refining
dried pulp based on estimated with good precision. with constant specific energy, laboratory sheet
bleached softwood
market pulp from four
forming and standard sheet testing. The results
Scandinavian mills. The methods can probably be improved by verify that shape factor (form factor) measured
defining each length class in terms of a non- by Fiber Tester in itself is a key fiber property for
rectangular filter, which conforms to the analyzed the prediction of strength properties. It is also
BMN fractions. The weights could also be measured online in many pulp mills in
developed to be divided into more classes than Scandinavia.
the number of BMN classes. Volume-weighted
data could also be used. The only limitation will 100
be if different species with different fiber wall 90
dimensions are used. 80

Tensile index (Nm/g)

70
The blend module in ABB’s L&W Fiber Tester Plus 60
can estimate the composition in a mix of four 50

components, and can be used by defining four 40

BMN fractions, and use each fraction as a 30

reference (cf. fig. 108 and 109). 20

10

19.6. Strength predictions from fiber properties 81 82 83 84 85

The shape (form) factor was measured with Fiber- Shape factor (LW) never dried

Tester on wet pulp samples from different mills

Dried, refined E-W 100 kWh/t Dried, unrefined
on different occasions, and the tensile index was Dried, refined E-W 50 kWh/t
determined on refined (with an industry-like —
refiner) market pulp from the same samples. A 112

very strong correlation could be seen (figure 111).

It is interesting that the same slope seems to be 19.7. Modeling of fiber networks
valid for different refining levels (constant The ultimate model to relate fiber properties to
specific energy). paper sheet properties would simulate a sheet
built- up from fibers. This would be a physical
Without considering all other variations such as model. By feeding the fiber statistics into a three-
fiber length etc., we can see that, with the help of dimensional model of the sheet built up by fiber
the shape factor and the specific refining energy, models and by using FEM (finite element method)
tensile index can be predicted quite well (figure to calculate the properties of the sheet, all paper
112). properties could be estimated.
 ≔ ← → ≡
111
Table of contents Previous chapter Next chapter Index

—
113
Small steps have already been taken in this 1.2
Simulated fiber network futuristic direction. It has been shown with very
with curled fibers.
simple models of fibers and sheets (figure 113)

Number of kinks/mm
1.0
— combined with FEM calculations how deformed
114
Number of kinks
fibers with different distributions influence the
0.8
increases with tensile stiffness of a two-dimensional sheet.
decreasing viscosity.
However, very simple models have been used
0.6
so far.

0.4
700 800

Viscosity (dm³/kg)

Dried, unrefined

Undried

—
114

19.9. Mental models

A reflection is that we all actually use mental
models. We make up our own mental images of
processes, and from those images or mental
models we try to analyze a certain situation. An
example, regarding the beating of fibers, is
cooked and uncooked spaghetti. Raw spaghetti
— (unbeaten pulp) consists of stiff ribbons unwilling
113
to form sheets.

19.8. Viscosity
Results from RISE indicate that the number of
kinks per mm of fiber increases with decreasing
viscosity. Accordingly, shape factor will decrease
at lower viscosities. A direct correlation is always
a better model than a more complicated model
with many variables.
 ≔ ← → ≡
112
Table of contents Previous chapter Next chapter Index

—
115
Cooked spaghetti (“beaten” fibers) is flexible, and
The spaghetti test the pieces follow each other’s shapes and easily
– a mental model
form tight structures. The surface of the
spaghetti is also modified during cooking.

One way to test whether spaghetti is ready to eat

is to throw it against the ceiling and see if it
sticks there. If it stays, the spaghetti is ready. It is
difficult to get rid of this mental model
describing the development of fiber properties.

—
115

19.10. Reconciliation of data

In systems with large amounts of data, it might
be possible to describe all the variations with a
few independent variables. All parameters could
then be predicted on the basis of these limited
variables. This can also be used to reduce noise in
signals.
 ≔ ← → ≡
113
Table of contents Previous chapter Next chapter Index

—
20. Applications of online fiber analysis

20.1. Delivery check and control of HW- and SW- 20.2. Communication of pulp quality properties
chemical pulp for the market Traditional laboratory tests seldom predict paper
A main goal for quality improvement is to test and quality well, and more frequent tests of fiber
verify final pulp quality more frequently. It is also properties are now becoming more and more
of interest to improve the pulp quality and to common. Experience suggests that fiber
decrease the variations. By measuring at the fiber properties give more stable signals (less noise)
level instead of relying on laboratory paper than sheet properties. The purpose of a quality
testing, deviations can more easily be detected control project is to deliver a pulp that works well
earlier in the process. Certain quality variations for the customers.
are present already in the wood supply, while
others are generated in the process. For example, The most important quality properties of the
deviations in fiber dimensions (length, width) can fibers to work with in the pulp mills are:
often be due to the raw material. Different types • Fiber length
of wood, for example roundwood and saw mill • Fiber width
chips, have different fiber properties. Deviations • Fiber shape
in fiber strength and shape factor are more • Coarseness
related to the process. A faster detection of • Fiber strength
different types of deviations will shorten the • Fiber surface
action time for correction. Management of grade
changes. If a certain grade requires a mix of fiber Measurements of the first three parameters is
types, one application is to use ABB’s L&W Fiber fully automated and installed online in mills
Online to accurately control the mix. producing bleached chemical pulp; this is well
established in the mills today. The measurement
of other parameters is partly automated but is
not used online so far. Fiber surface, which is
correlated to drainage properties, is widely used.
Some optical properties are also widely used.
 ≔ ← → ≡
114
Table of contents Previous chapter Next chapter Index

—
116
20.3. Linking data between pulp and paper mills 60

Predicted refining energy (kWh/t)

Input of fiber data The usage of fiber data in quality control also
predicts refining energy 50
in the paper mill.
makes it possible to tighten the alignment
between pulp and paper manufacturers. The 40
influence of fiber quality in the customer’s end
30
product can be studied by linking the fiber data
to the paper production. 20

10
Results from a study of the relation between fiber
characteristics and paper production in a paper
0 10 20 30 40 50 60
mill are shown in figure 116. A multivariate model
was developed with fiber data as X-variables and Measured refining energy (kWh/t)

refining energy as Y-variables. The study shows —

116
that the monitoring of fiber data can predict the
refining energy needed in the paper mill to
20.5. Process control – Action limits
maintain constant paper quality.
Fiber deformation measured as shape factor is
a property that is to a great extent created in
20.4. Uniformity
the pulp mill. Since it is created in the pulp mill, it
In deliveries of market pulp, uniformity between
should also be possible to control it there. One
shipments is of great importance. The installation
example described here is to follow shape factor
of the same equipment for online fiber analysis in
frequently online. One user has, based on
all the mills of one company, followed by
experience, defined action limits. When the
calibration of the systems to deliver comparable
signals pass an action limit, something
numbers, has given unique possibilities for
unexpected has happened and the operators
process and quality control. Process development
should follow up what has happened and take
using a comparison of fiber properties among
corrective action.
the mills is one of the most studied areas. Quality
comparisons aiming for uniformity and
In the example in figure 117, the mill uses action
exchangeability of products between the mills are
limits for fiber properties. The example shown is
also made easier by the installation of accurate
from a period before the active use of action
and frequent quality measurements. The
limits. In this case, the shape factor dropped
possibility of achieving stability in the quality
below the action limit over two days. If the
increases with frequent control, and it will also
operator had reacted immediately to the online
result in a faster feedback to the production staff
measured change, two days of costly
and better possibilities of avoiding deliveries out
downgrading of the pulp could have been
of specification. This will pay off in confidence
avoided. The reason for the quality change was a
between supplier and customer.
drop out of certain fiber-protective chemicals in
the bleaching process. In this case, the action was
first taken after the offline testing two days later.
Similar situations happen from time to time.
 ≔ ← → ≡
115
Table of contents Previous chapter Next chapter Index

—
117
82 20.7.
82 Control of production flow during grade
Shape factor is one of 81.5 changes
81.5 in the pulping line
the most important
fiber properties for 81
Pulp
81
mills alternating between hardwood and
paper strength. Higher softwood use Fiber Online to control the
80.5 80.5
shape factor means
higher strength. The
transition. It takes up to eight hours for the pulp
80 80
shape factor decreased to pass through the pulp mill, so it is important
in the pulping process, 79.5 79.5
which was monitored
to detect when the change comes to the
79 79
by L&W Fiber Online. headbox of the drying machine. But the transition
78.5 78.5
from one fiber type to another can also be fast or
78 slow.
78 It is important to know when the old quality
goes out of range and when the new quality is
Normal production over action limit reached.B-grade production undertime
The intermediate action can
limit vary from one

to seven hours. During this time, a lot of pulp is

82 produced. Some mills use the standard data from
81.5 Fiber Online. An alternative is L&W Fiber Online
81 Blend software that calculates the mix based on
two-dimensional length/width distributions and
80.5
monitors the calculated mix.
80

79.5
When the initial part of the grade change reaches
79
the final buffer, the mixed pulp is stored in a
78.5 separate buffer. When the final phase of the
78 change is detected, the new grade can start. The
pulp stored during the grade change can then be
uction over action limit B-grade production under action limit mixed into the new grade controlled from fiber
— measurements and can be sold at full price with
117
the fiber length variations always within the
specified limits. The cycle time for measurement
20.6. Fractionation of pulps
with L&W Fiber Online during transition is 5-6
Different qualities of pulp can be achieved
minutes.
through the screening or fractionation of pulp
from one production line into two separate
streams. This technique is often used in a part of
the production to produce reinforcement pulp.
This process is monitored by using length and
width data from Fiber Online. Also, the sampling
frequency is of great importance. Figure 117
shows length measurements from one of the
mills, in a transition stage from non-screening to
screening and separation into two different pulp
qualities.
 ≔ ← → ≡
116
Table of contents Previous chapter Next chapter Index

—
118 Fractionation of pulp
Fractionation of pulp
Even if the SW and HW lines are completely
The screening process is 2.9
2.9
“clean”, a mix of broke creates an uncertainty in

(mm)
length(mm)
monitored well, by using
L&W Fiber Online data. 2.7 the mix after the blend chest. The same reasoning
2.7
can be applied for broke: Use of online

fiberlength
— 2.5
2.5
119
measurements can improve the production
2.3

weightedfiber
Fiber length during 2.3 strategy and better control the final mix of HW
campaigns with HW
and SW. Many samples
Lengthweighted 2.1
2.1 and SW.
(dots) during the 1.9
1.9
transitions mean that
the grade change has 1.7
An example of a grade change on a fine paper
Length

1.7
taken a longer time. machine is shown in figure 121. The HW/SW mix is
1.5
1.5 changed at the time of the grade change. This
—
120 mean that the HW/SW mix in the broke will also
The HW line suddenly
includes some SW pulp,
— Drying machine 1
Drying machine 1
Drying machine 2
Drying machine 2 change, although this latter change will be some-
118
which causes the final what delayed. In the example in the figure, the
blend to drop from
80% HW to 60% HW.
3
3
change in broke came back to the machine six
(mm)
length(mm)

hours later. After which, the changed mix in the

2.5
2.5
broke led to a new disturbance on the machine.
fiberlength

2
2
weightedfiber

1.5
Without online measurement of the mix, as in this
1.5
case, the machine tender has no knowledge of
Lengthweighted

1
1 this effect. It is not possible to calculate it. It has
0.5 to be measured.
0.5
Length

0
0
120
A M J J A S
A M J J A S

— 100
119
Proportion (%)

80

20.8. Broke fine paper 60

The process described in figure 31 on page 40 is 40

common for an integrated fine paper mill.
20
Separate hardwood (HW) and softwood (SW)
lines come into the stock preparation. When a 0
grade change occurs, long fibers may be mixed 2005- 2005- 2005- 2005- 2005- 2005- 2005- 2005-
into the HW line (figure 120). This causes a drop in 09-26 09-27 09-27 09-27 09-27 09-27 09-28 09-28
19:12 00:00 04:48 09:36 14:24 19:12 00:00 04:48
the amount of hardwood after mixing and
changes the formation of the paper on the paper % HW in mix %HW in HW-pos %SW in SW-pos
machine. This knowledge can be used to avoid —
such a situation by changing the production 120

strategy. One possibility is to control the mix

directly from the measurement.
 ≔ ← → ≡
117
Table of contents Previous chapter Next chapter Index

—
121
20.9. Other experiences Proportion of hardwood in the virgin fiber line and broke
The effect of the broke In one mill, it was found that during a cook the 60
from a grade change
in this case came 6
softwood suffered a significant reduction in

Proportion af hardwood (%)

55
hours after the actual normal expected length. Because the fiber
grade change.
properties were being monitored with a fiber 50

property analyzer, it was addressed quickly 45

without any problems arising. Normally, a
40
situation such as this would not have been found
until the pulp got to the paper machine and 35
possible breaks occurred or some other type of
30
upset occurred due to the abnormal softwood 12 15 18 21 00 03 06 09 12 15 18 21 00 03 06 09 12
fiber length. Time (hours)

Target Mixed stock Broke

Another example from the same mill was a pulp
“swing” in the pulp mill during a transition from —
121
SW to HW. Formation of the paper dropped
without “any apparent reason” – i.e. no changes
20.10. Optimization of beating
were intentionally made at the wet end or to the
For one mill, the main objective was to decrease
furnish.
costs by reducing softwood without sacrificing
machine runnability or sheet strength properties.
It was found that the average fiber length
An online fiber analysis system allows operators
of the HW increased significantly. This caused the
to make decisions based on data that are very
average length to the stuff-box to increase as well
precise, repeatable and only minutes old instead
and resulted in a poor formation. With an online
of being several hours old. Mills cannot afford to
unit, it is believed that the sample frequency and
add testing manpower to quickly get the amount
reaction time of the machine crews would be
and accuracy of data that an online system can
improved to the point that they would be able to
provide and still attain the goal of cost reduction
react to the increase in HW fiber length and
of materials. They would simply be adding costs
compensate for it by a reduction in SW addition.
in one area to reduce those in another. There is
nevertheless the initial cost of equipment and
yearly maintenance.
 ≔ ← → ≡
118
Table of contents Previous chapter Next chapter Index

There are “sweet spots” for fiber length, 20.12. Sack kraft
fibrillation, shape factor and WRV (freeness). The fiber analyzer shows how variations in the
Online monitoring of these parameters, with a raw material and pulp mill affect the paper
system that can measure accurately and often, machine. This has resulted in improvements of
allows the papermaker to maintain these critical the wood handling and improved strategies for
targets to avoid upsets on the paper machine and refining.
produce consistent sheet strength, while at the
same time optimizing BSKP. The optimal refiner At one mill, there was a problem with the tear
settings: index for a certain grade. Experience from tests
• Maintain desired fiber length average, do not with Fiber Online, and comparison with strength
cut fibers data from the reel combined with PCA (see page
- Track fibrillation 106) showed that the problem could be solved
• Track freeness to move around desired target with another refining strategy for that grade.
(this can also be done online)
• Optimize shape factor to the desired level 20.13. Coated printing grades
• The straightening effect on shape factor has a A non-integrated paper mill tried to minimize
time effect. If the beaten pulp is stored too long costs by using different fiber grades on the
(several hours) before being used for paper market. Pure market pulp and rejected reels from
making, the shape factor may drop again! other mills were used. Using ABB’s L&W Fiber
Online, they found that the fines content could
Online sampling locations (figure 31 on page 40): vary a lot between different suppliers. Certain
• At the output of the softwood refiner(s) incoming pulps had high fines levels, which
• At the output of the hardwood refiner(s) caused trouble on the paper machine. Certain
• At the machine chest (by monitoring the soft- suppliers could be avoided. The same mill also
wood and hardwood line, the difference would uses measurements of fines and fiber length to
be the broke) optimize refining. This mill has also detected
problems with vessel cells on the coating blade of
20.11. Security paper the on-machine coater from one pulp and they
In the production of security paper, good also use an integrated shives analyzer in their
formation is a necessity for good watermarks and fiber analysis system.
superior printing. Formation is strongly
dependent on fiber length. Many of these mills
have successfully installed online systems for
fiber length measurement. Long cotton fibers are
cut to suitable and constant average fiber length
to obtain a uniform formation.
 ≔ ← → ≡
119
Table of contents Previous chapter Next chapter Index

—
122
20.14. Prediction of fiber mix 20.15. Vessel cells in DIP
Fiber width is important for recycled fibers Recently it has been discovered that recycled pulp
for newsprint based
on recycled fiber.
A good correlation between some of the variables has a larger proportion of hardwood fiber species
achieved from the Fiber Online and the paper containing vessel cells. Fiber debris, ink, and
composition expressed as news and magazine shives must be recognized in relation to vessel
content was found in one case. However, the cells. These cells are well suited for identification
control potential was limited in this specific case. in image analysis systems for fibers.
The evaluation showed that fiber width was the
most important single fiber parameter in 20.16. Other measurements in a deinked pulp
predicting news content, especially class W4 (DIP) plant
(which means average width for length class Spectrometers for VIS- and UV-measurement on
3–4.5 mm). W3 (which includes length class 1.5–3 dry sheets can be used for the automatic control
mm) showed similar results. The mix could also be of brightness and estimated residual ink content
described with a model based on reference (ERIC 950 nm). It is also possible to use the VIS
volume-weighted -dimensional length/width spectrometer for the measurement of color
distributions of BSKP, BHKP, TMP and SGW. See (L*a*b*) in the DIP plant.
figure 122.
20.17. Control of mechanical pulps
36 TMP refiners have to be controlled at two levels.
35 First, the process has to be stabilized and
34 controlled with respect to consistency and
33
production rate. This is a fast control.
32
WI (µm)

31
On the next level, the specific energy level has to
30 be controlled. This is normally done by automatic
29 sampling systems. Freeness measured as CSF
28
R2 = 0.9655 seems to be well accepted in mechanical pulping
27
for this purpose. This has led some suppliers to
26
0 20 40 60 80 100
try to copy the old manual CSF unit in their
automatic systems, even though the method is
W4 (see table 18 p89)
not very modern. CSF is operator-dependent, but
News content (%) it is reported that it should be possible with the
— same operator and routines to reach a 95%
122
confidence interval of +/-4 ml (in practice this
interval is probably wider).
 ≔ ← → ≡
120
Table of contents Previous chapter Next chapter Index

—
123
It has been shown in a case study that modern Narrow slots (150–200 µm) remove most of the
One approach for closed loop control of a TMP plant can reduce the shives, which are then treated in the reject
control of the refining
of mechanical fibers is
variance in freeness by 80%. The closed loop handling (reject refiner). Nevertheless, it seems
to use two parameters control used weighted mean fiber length sampled that there is still a need for the measurement of
to describe quality.
F1 (reinforcement;
every 7.5 minutes. The second stage refiners and shives in all production of mechanical pulps.
mainly fiber length) the reject refiner were run in the same way. The Reject refiners often tend to cut fibers. It is a
and F2 (bonding;
mainly freeness).
first stage was consistency controlled based on a balance to kill shives but keep the fiber length. In
mathematical model including thermodynamics; mechanical pulping, the refiners work at a high
mass flow balance and water balance. The disc consistency (HC), but low consistency refining
clearance for each refiner was optimized on the (LC) is used in some processes today. One effect
basis of manual sampling and the offline is a reduction of shives.
measurement of mean fiber length for each
machine. A possible control strategy for screen- It is important to measure fiber length in order to
rooms in order to separate the control of shives prevent or minimize fiber cutting. Treatment of
from the control of freeness has been the fibers in refiners results in a reduction in the
investigated with good results. However, screen shives content and also deformed fibers. These
rooms are seldom designed for automatic properties should also be of interest as a mean of
control. The active control element is the reject monitoring refiner operation, but much is not
refiner. Another control parameter is the fillings reported. Shape factor and width are relatively
in the refiner. Plates get worn out and have to be new measurements with good accuracy. By
replaced or the pattern can be optimized. This combining length and width for fibers it is
can be measured as a balance between fiber possible to work with volume-weighted averages
length (F1) and freeness (F2) (see figure 123). and distributions in fiber analysis, and this seems
Specific energy control tries to keep the process to be an advantage for mechanical pulps. In fact,
within a certain target area according to the shape factor is a direct measurement of latency.
figure. This can be used to obtain information In order to measure latency, the sample has to be
about when it is time to change the plates. taken before the latency-removal chest, which
means it has to be taken in the blow line from
F1 each refiner. This would also allow optimization
of each refiner.
Target
area
A BCTMP mill with many grade changes needs
online measurements of freeness, shives, fiber
F2
length and brightness to measure final pulp
—
quality frequently. It is critical for profitability to
123 quickly arrive within the specifications of the new
grade. For process control purposes, earlier
Shives are detrimental objects in mechanical positions in the production line are of course of
pulp. Today, screening systems are quite effective great interest.
in reducing shives.
 ≔ ← → ≡
121
Table of contents Previous chapter Next chapter Index

20.18. Impact from raw material

In one case study, a mill trial with a supply of
different wood is described. The wood used at
the mill during the time period of the study was
spruce from seven geographical locations; five
from Sweden, one from the United Kingdom and
one from Russia. The amount of each chipped
wood category was tallied and recorded each day.
Pulp samples were gathered every second hour
from the disc filter chest for full hand sheet
quality testing. This was done for two mills. The
trial was evaluated using factor analysis.

Note: The study showed that 40% of the

variations in tear index were attributed to
variations in raw material.

The raw material is obviously a source of

variation in quality.
 ≔ ← → ≡
122
Table of contents Previous chapter Next chapter Index

—
21. Cost of raw material and return on investment

* S ource, private 21.1. Prices on the different market pulps* Another parameter that usually has an impact on
communication:
Johan Persson, Per The annual production of pulp worldwide is the pulp price is the dollar rate. If the dollar tends
Malmquist, Johan estimated to 475 million tonnes (Mt). This to be expensive, the whole world except for the
Lundvik, Hans
Göte Andersson, production can be divided into two major North American market does not buy as much
Bo Fredriksson segments of which 275 Mt are recycled fiber and pulp and this causes a drop in the pulp price.
(Oct. 2020).
200 Mt are virgin fiber. The majority of the virgin This is due to the fact that the pulp is purchased
fiber, approx. 155 Mt, is used in integrated mills, in US-dollar (USD) and sold in local currency. Yet
and approx. 45 Mt is sold as market pulp. The another factor that influences the pulp price is
prices of these different market pulps are the price of recycled fiber (DIP). When the price
described in this chapter. of DIP increases, the price of the BCKP pulps
increases about six week later, and when the
BCKP: Approximately 40 Mt, or 90% of all kraft price of DIP drops, the price of BCKP drops
market pulp that is sold on the global market is immediately.
manufactured according to the BCKP-process
(bleached chemical kraft pulp). The pulp is CTMP: It is the second largest kraft pulp segment.
divided in two segments; NBSK (northern The amount of CTMP is about 9–13 Mt. The prices
bleached softwood kraft pulp) and BHKP follow the movement of the BCKP market and the
(bleached hardwood kraft pulp). In Europe, the prices are about 75% of the price of NBSK.
prices are set from PIX (price index), which is
taken from the web site www.foex.fi. This SGW, PSGW, NSSC, and TMP: These pulps follow
homepage is working like a stock exchange for the normal rules for the market with regard to
these pulps. When a pulp deal is to be made, the supply and demand. The prices are negotiated
pulp producer uses the PIX pulp index from www. between customer and supplier without any
foex.fi for that day as a price guideline. Special index from outside.
deals for certain customers or specific dis-
counts are calculated from that PIX pulp index. DIP: This pulp also follows the supply and demand
Another marketplace for pulp trading is NYBOT rules. The prices have recently been high because
(New York Board of Trade). of the large demand from the Chinese market.
The prices of DIP are also dependent on political
The prices follow the marketing forces of supply decisions such as how countries decide to collect
and demand. If there is a lot of pulp in stock at the paper or changing laws in different directions.
mills, or a lot in storage in harbors, the prices DIP-pulps are not dependent on the dollar rate,
tend to decline. In contrast, the prices increase if since they are purchased in local currencies.
the stocks are small.
 ≔ ← → ≡
123
Table of contents Previous chapter Next chapter Index

—
124
21.2. Operational costs A return on investment (ROI) of 20% is also
Pulp price differences Figure 124 shows the price differences on the included in the diagram. The operating costs for a
in Europe. % difference
compared with NBSK European market for different pulps compared typical mill in Sweden and Finland may be around
1976–1996 by quarter.* with the highest valued pulp, NBSK. 400 USD/t of pulp. The dotted line at 650*
USD/t of pulp is shown as the long-term trend
Operating costs for NBSK are shown in figure price for NBSK. Finally, table 25 presents some
124 for a hypothetical case with a new pulp mill price information for pulp and paper products in
with a capacity of 500, 000 t/yr. The operating western Europe.
costs are divided into wood, other variable costs
(such as energy and chemicals), personnel, other
fixed costs, and distribution costs (to western
* Source RISI and Moody estimate 2020
Europe).

Average +/- standard deviation

-6.8 +/- 4.7

Southern sw kraft
-5.3 +/- 5.8
Scand. birch kraft
-7.0 +/- 6.6
Brazilian euca kraft
-11.0 +/- 5.8
US southern mixed hw
-3.5 +/- 6.2
Bl. sw sulphite
-12.2 +/- 7.9
Bl. hw sulphite
-16.2 +/- 6.9
Unbl. sulphite
-12.2 +/- 7.4
Unbl. sw kraft
-30.3 +/- 6.4
Dry bl. mech.
-22.6 +/- 5.4
BCTMP
-29.7 +/- 9.1
Unbl. CTMP
-50 -40 -30 -20 -10 0 10 20 30
-%-

Maximum Average Minimum

—
124
 ≔ ← → ≡
124
Table of contents Previous chapter Next chapter Index

— —
125
It follows from the table on previous page , that
Table 25: Price levels of pulp and paper
Sales price requirements products (CIF) in Western Europe, quarter II. the main pulp and paper products are sold at
for new hypothetical
pulp mills (NBSK) with a
$ are translated to € by a factor of 0.9.* price levels of€650 - €1100 (CIF western Europe)*.
capacity of 500 000 t/yr As a guideline, the operating costs may be around
Grade Price (€) 65–75% of this price and the variable costs may
BSKP 426–480 be 40–60% of the sales price, of course with large
BSKP, Southern Pine 444–462 variations.
BHKP, Eucalyptus 373
BHKP, Birch 373–408 The economic return from the use of the fiber
RCP, Mixed grades 10 analyzer instrument may be related to one or
RCP, ONP 62
more of the following factors:
• Lower operating costs
RCP, OCC 42
• Higher product price due to improved (higher)
RCP, CPO 185
product quality
RCP, White woodfree shavings 140
• A more saleable product due to a higher or a
Newsprint, 48.8 g/m2 485–500
more uniform product quality
SC Paper, 60 g/m2 605–650
LWC Paper, 60 g/m2 740–765
Assuming that the return is based on an
Uncoated woodfree, 80 g/m2, sheets 700–800
investment cost of €200000 and that the
Unbleached kraftliner, 175 g/m2 420–470
payback time should be two years or less, a
409–435
payback of about €100000 per year
SC Fluting 330–360 corresponding to the sales of 131–268 tonnes of
Testliner 275–310 product per year is required, see table 26 below.
* Source: Jaako Pöyry (1998). * Source: Fastmarket RISI and Moody’s estimates (2020).

—
1000 Table 26 Payback requirement of 100 000 €
USD 650/t for different levels of product price.
800
A. B.
600 Product price Product price
– USD/t –

1100 €/t 650 €/t

400
Number of tonnes for a sales
of 100 000 € 288 131
200
Number of tonnes for
0 operating costs of 100 000 € 348 170
US Canada Chile Sweden US Finland Canada Number of tonnes for
South East West West variable costs of 100 000 € 532 362

ROI 20 % Personnel
Distr. to W. Europe Other variable costs
Other fixed costs Wood

—
125
 ≔ ← → ≡
125
Table of contents Previous chapter Next chapter Index

When considering the payback, different analyses 21.3. Example from a 350 000 t/yr
will have to be made for three different market pulp mill
situations: The mill specifies a minimum average fiber length.
If the fiber length is below this value, the pulp will
1. For a mill producing about 350000 t/yr, the be downgraded and sold at a lower price to a less
operating costs must be reduced by about demanding customer. If such downgrading can
0.33 €/ton. be avoided by using an online fiber analyzer for
2. For a mill producing about 350000 t/yr, an control actions/control strategies, the payback
increase in product price of 0.33 €/t would yield can be calculated as follows. The control action
the required payback. However, this figure is can be to use a buffer for the temporary buffering
too small to be considered in price negotiations of out- of-specifications production and later in-
and it is more probable that the higher quality mixing. A strategy can be purely technical, but it
makes the product more saleable (higher sales can also be a bonus system to avoid bad
volume) as in alternative 3. situations (process upsets, etc.). The important
3. For a product in the upper price range, the thing is to measure the result frequently.
pay- back would be satisfactory for additional
sales of about 300 t/yr, whereas for the 21.4. Using fiber analyzer data instead of
products in the lower price range, additional strength indexes
sales of about 600 t/yr would be required. Production of 20 t/h means 480 t/day. A price of
480 €/t means 230 400 €/day. Downgrading may
These payback levels are only 0.1% of the overall be in the range of 3–9%. A 6% reduction of the
figures for a mill. price leads to losses of 13824 €/day.

However more specific calculations are asked for A conservative estimation of 20 days
in each application. Thus, improved flexibility, downgrading gives losses of approx. 245769 €.
higher and/or more uniform quality, new
strategies for stock preparation (fiber mix and/or Larger savings per year are possible in the long
broke handling) and paper machine operation run with other quality profiles than are made
(fractionation and multi- layered formation, and today. If the tear index of a pulp is not as
higher production thanks to fewer web breaks) important as previously stated, and instead a
will have to be considered more in terms of their specification of shape factor will be better for the
inherent benefits in different respects when customer’s product, then the ability to change
evaluating online applications. yield and chemical dosage in the process may
provide savings of another order of magnitude
If the quality of the raw material (fibers) or the than the investment cost for a fiber quality
paper quality and/or production rate could be analyzer!
improved by using the fiber analyzer, the cost of
the instrument would easily be balanced by these
positive factors.
 ≔ ← → ≡
126
Table of contents Previous chapter Next chapter Index

21.5. Investment calculations 21.7. What can we do about these causes?

In an R&D environment, you can invest in new Early detection of the problem will reduce the
equipment if it will help to generate new negative effect.
knowledge. In industrial production, the • We can avoid unnecessary long
generation of knowledge about your process, production stops
quality and problems is of course just as • We can avoid running out of specification
important. However, for investment in new longer than necessary
equipment in the mill, it is usually not sufficient
to merely have the opinion that a new system will If we increase our knowledge about the causes of
help. You are required to show a return on the these problems we can make improvements by
investment within a certain time. For changing operating strategy.
measurement equipment, this time may be from
six months to two years—corresponding to an 21.8. How does this lead to cost reductions?
“interest rate” of 50–200%! • A non-uniform pulp/stock has more weak points
or periods than a uniform pulp/stock. If a defect
One way to do this is to list possible occurs in the web, the probability of a web
improvements and then estimate how these break is occasionally greater in a non-uniform
improvements will affect the costs. In practice, sheet. If the uniformity increases, one risk
cost reduction seems to be easier to calculate or factor for web breaks is reduced. Estimate a
at least to communicate than expanding percentage of how much you think this can
business. reduce the frequency of the web breaks.
• The risk for out-of-specification product will
21.6. How can these savings be achieved? decrease. Set a percentage of how much you
Let us take one example – more uniform pulp. think you can reduce culled paper.
Causes for non-uniform pulp or stock can be: • Variation in pulp quality may be the reason for
• Variations in broke handling from grade adding chemicals for compensation. Cost of
changes or web breaks chemicals may be reduced with fewer
• A temporary unknown fiber mix incoming variations.
• Fiber mix partly out of control • Customer claims because of out-of-
• Process upsets in the pulp mill specification product: What is the economic
• Process upsets in the stock preparations value if you avoid only one customer claim? It
• Process breakdown in the pulp mill can mean that you can reduce the risk of losing
• Process breakdown in the wood handling a customer.
• Variations in incoming raw material
 ≔ ← → ≡
127
Table of contents Previous chapter Next chapter Index

21.9. Other examples The following possibilities for reductions are

• Increased demands from the paper mills for often used in calculations:
frequent quality check on pulp production.
• The shape factor affects tensile index up to 15 SAVINGS (€/t)
units. To increase tensile, you have to add Optimistic Realistic Pessimistic
fibers, which will increase your costs. Reduction in
• Better selection of wood. Cost for raw material customer claims
is up to 50% for a pulp mill Reduction in
machine breaks
• Smaller losses due to grade changes and
Reduction in
broke handling culled paper
• Loss of customers due to the delivery of poor Tighter basis
quality (out of specification) weight limits
• Quality monitoring of a given pulp Reduction in
• Replacement of one commercial pulp by another additives usage
Fiber
• Positioning a competitor’s pulp
substitution
• Early prediction of physical paper properties Refiner energy
allow actions to be taken in time savings
• Control of fractionation Total savings
• Refiner control
• Control of screening
 ≔ ← → ≡
128
Table of contents Previous chapter Next chapter Index

—
22. Reference testing

22.1. Resolution 22.4. Reproducibility

The resolution of an instrument determines the This is defined as the consensus of results from
smallest change in an observed property that can tests from the same samples, on different
be sensed by the instrument. For ABB’s L&W instruments, and at different times. This is often
Tester Plus, the resolution sometimes is even called accuracy. The example that is most often
higher than the pixel size, which has been used is firing guns against a target. If the hits are
explained in earlier chapters. well placed together, then the repeatability is
good. If the hits are gathered around the middle
22.2. Measurement range of the target, then the accuracy is good.
The measurement range can be defined as the
range in which the instrument can measure within 22.5. Reference instrument
specified maximum errors. It can also be defined If a better method or instrument is available, this
as within which range the instrument has can be used as a reference. However, regarding
sensitivity. For your stereo, you usually have the – new fiber analyzers, the situation is that no better
3dB points defined for the spectral range of an method exists. Stable operation over time is
amplifier or loudspeakers. This represents the often followed up by tests of stable reference
point where the signal is 50% of what it is in the material. For fiber analyzers, synthetic fibers
middle of the spectral range. (rayon) are used. Round robin tests are also
organized to compare different instruments.
22.3. Repeatability
This is defined as the consensus of results from
tests from the same samples, on the same
instrument, and with same operator involved, if
the method is operator dependent. In process
control applications, repeatability is most
important. Repeatability tests are often defined
for acceptance tests of sensors. For example,
define standard deviation for 20 equal samples
tested on the equipment (see also chapter 10,
about statistics).
 ≔ ← → ≡
129
Table of contents Previous chapter Next chapter Index

RISE sends out samples every quarter to users of Results within WL OK

L&W Fiber Tester Plus as a subscription service.
Results outside WL Check next sample!
They send rayon fibers to be used for weekly tests Results outside WL at next
by users and also bleached hardwood and sample Run an extra test!
softwood fiber samples for more accurate tests. Contact responsible for
Statistics are distributed and the users can see if Results still outside WL instruments

their instrument is different from the others. Results outside AL Run a second sample

Periodic tests are always necessary to follow up if Contact responsible for

1. Still outside AL instruments
calibration has to be done.
2. Outside WG Run a second sample
Contact responsible for
One user of L&W Fiber Online uses all three Still outside WG instruments
reference samples mentioned. They check fiber
length, fiber width and shape factor. For each
Warning limits and action limits are set by
property they have set both a warning limit (WL)
experience for each property or by
and an action limit (AL). Actions that need to be
recommendation from the supplier.
taken are:
 ≔ ← → ≡
130
Table of contents Previous chapter Next chapter Index

—
23. Bibliography

Ander, P. & Daniel, G. Characterization and Hagedorn, A. et al. Optimizing Machine Efficiency
Comparison of Industrial and Laboratory Pulp Through Fiber Quality Management. TAPPI 2006,
Fibers using HCl, Cellulase and FiberMaster Papermakers and Coating and Graphics Arts
Analyses. Nordic Pulp and Paper Research Journal Conference.
Vol. 20 No. 1/2005.
Hallgren, H. & Fredlund, M. Formningsstudier på
Björk, E., Granlöf, L. & Mohlin, U-B. Pulp quality FEX; Effekt av formningsprincip,
with STFI Fibermaster. PUB 8, RISE, 2002. retentionsmedelsatsning och massaval på
pappersegenskaper vid finpapperstillverkning.
Bådenlid, R. et al. Sätt vid framställning av massa RISE C 175. 1989.
samt produkt av sättet. Patent application No. SE
0001419-1, Stora Enso R&D, Karlstad, Sweden. Hill, J. Process understanding profits from sensor
and control developments. Int. Mechanical
Dillner, B. Fiber Property Development – Tailoring Pulping Conference, Oslo, 1993.
Softwood Pulps for Different Products. SPCI,
2002, Stockholm. Hill, J., Karlsson, H. & Ostman T. Process design
Fellers, C. & Norman, B. Pappersteknik. KTH, 1998. and control of screen-rooms. Pulp & Paper Can.
Forgacs, O.L., Robertsson, A.A. & Mason, S.G. 83(5), T125–T129, May 1982. Horn, R.A. &
The Hydrodynamic Behaviour of Paper-Making Setterholm, V.C. Fiber morphology and new
Fibers. Pulp and Paper Mag. Can. 59(5), 117 (1958). crops. 1990

Fralic, G., Mozaraffi, S., Flynn, G. Mill-wide Jangmalm, A. Network modelling of the elastic
Advanced Quality Control for the Production of properties of paper. Licentiate Thesis no 52,
Newsprint. Strand W.C., IMPC 2001. Department of Solid Mechanics, KTH, Stockholm,
Sweden,1996.
Fransson, P-I. et al. Mätningar med STFI
FiberMaster i ett kartongbruk. RISE, TF 74, 1997. Johnson, R.A. & Wichern, D.W. Applied
multivariate statistical analysis (third edition).
Granlöf, L. et al. Vessel elements in hardwood Prentice Hall, ISBN 0-13-041773-4, 1992.
pulps – appearance and measurements. RISE
report, PUB 14, 2003. Joris, G. & Roux, J.C. High yield computerised
refiner machine. Current and future technologies
of refining, Vol. 2, 1991.
 ≔ ← → ≡
131
Table of contents Previous chapter Next chapter Index

Joutsimo, O. et al. The influence of fiber Karlsson, H.I. & Pettersson, J.G.T. Method and
ultrastructureal damage on pulp and paper apparatus for determination of refiner
properties. APPITA 2006, Melbourne, Australia. mechanical pulp properties. Patent; US 5491340,
1994.Karlsson, H. et al.
Karlsson, H. Improvements with L&W STFI
Fibermaster. PAPTAC 2004, Montreal, Canada. Image Analysis for Quality Monitoring of Recycled
Fibers. Control Systems, Stockholm, Sweden,
Karlsson, H. New techniques for measurements of 2002.
fiber properties including vessel cells and mix of
fiber species. APPITA 2006, Melbourne, Australia. Karlsson, T. Utveckling och systemintegrering av
mjukvara för detekt- ering och klassificering i
Karlsson, H., Fransson, P-I. & Mohlin U-B. STFI realtid av lokala defekter hos långsmala objekt
Fibermaster. SPCI 1999, Stockholm, Sweden. från bildsensor. TRITA-NA-E03120, Nada, KTH,
2003.
Karlsson, H. & Fransson, P-I. FiberMaster ger Kibblewhite, R.P. & Brookes, D. Factors which
pappersmakaren nya muskler. Sv. Papperstidning, influence the wet web strength of commercial
97:10, 26 (1994). pulps. Appita 28(4), January 1975.

Karlsson, H. & Fransson, P-I. Method and Krantz, T. Landvinningar under 20 års mät- och
apparatus for producing and testing a fiber bed. styrteknisk utveckling vid RISE. Sv.
Patent; US 6, 884, 320 B2, 2005. Papperstidning/nordisk Cellullosa nr 3, 1995 (99),
pp. 36-38.
Karlsson, H. & Fransson P-I. Filtering device.
Patent; US 6, 520, 343, 2003. Lindblad, L. & Fürst, T. The Ultrasonic Measuring
Technology on Paper and Board. AB Lorentzen-
Karlsson, H., Fransson, P-I. & Kastre L. Method wettre (2001).
and apparatus for measuring fiber flexibility.
Patent US 5, 331, 405, 1994. Lindström, J. & Karlsson, H. Optimizing the fiber
process by using new on-line analyzer and
Karlsson, H., Johansson, B-L. & Jung, E. experiences from pulp mill.
Experiences with computer control, based on Tappi Japan, October 13–15, 2004, Hokkai-do,
optical sensors for pulp quality, of a two stage JAPAN.
TMP-plant. Control Systems Conference, Halifax,
1980.
 ≔ ← → ≡
132
Table of contents Previous chapter Next chapter Index

Lundqvist, S-O. & Evans, R. Illustration of wood Page, D.H. The beating of chemical pulps – the
and fiber measurements with Silviscan. RISE action and the effects. 9th Fundamental Research
Report, ART 15, 2004. Symposium, Fundamentals of Papermaking,
Cambridge, Mech. Eng. Publ. Ltd, London.
Markström, H. Testing Methods and Instruments
for Corrugated Board. AB Lorentzen-Wettre Page, D.H.et al. Curl, crimps, kinks and
(2005). microcompressions in pulp fibers – their origin,
measurement and significance. Papermaking raw
Mohlin, U-B., Dahlbom, J., Hornatowska, J. Fiber material conference, 1985, pp. 183.
deformation and sheet strength. Tappi Journal,
Vol. 79, No. 6, June 1996. Paper and paper making. CEPATEC AB, Markaryd-
binder (2005). Pauler, N. Paper Optics. AB
Mohlin, U-B. & Miller, J. Industrial refining – Lorentzen-Wettre (2002).Pettersson, T. &
effects of refining conditions on fiber properties. Karlsson, H. Method for determining the average
Proceedings Refining 95, paper 04, PIRA, Atlanta, radius and/or the average length of particles
March 1995. carried by a flow medium. Patent; US 4529309,
1985.
Mohlin U-B. Market SBK and refining response.
Refining and Mechanical Pulping, Pira, Barcelona, Pöhler, T. & Heikkurinen, A. Amount and character
Spain, March 2005. of splits in fiber wall caused by disc refining. KCL,
Espoo, Finland.
Mohlin, U-B. & Olofsson, K. Formfaktor och
böjlighet hos barrvedssulfatmassa – inverkan av Olofsson, K. Mätmetoder för fiberrs
vårvedsandel, HC-behandling och kappatal. RISE, fibrilleringsgrad. RISE, TF 56, 1997.
TF 97, 1997.
Olsson, R.J. et al. Multivariate characterization of
Mohlin, U-B. Pulp evaluation for modern chemical and physical descriptors in pulp using
papermaking. NIRR. Tappi Journal, October 1995.
6th International Conference on New Available
Technologies, June, 1999, Stockholm. Ring, G.J.F. & Bacon, A.J. Multiple component
analysis of fiber length distributions. University
Mortens, H. & Naes T. Multivariate calibration. of Wisconsin, SP Department of Paper Science,
Wiley, New York, 1989. Paavilainen, L. The USA.
possibility of fractionating softwood sulfate pulp
according to cell wall thickness. Appita Scotch, M. Optimizing Bleached Softwood Kraft
45(1992):5, 319-326. usage on a Paper Machine by using properties
obtained through online fiber analysis. TAPPI
Conference, Milwaukee, WI, May, 2005.
 ≔ ← → ≡
133
Table of contents Previous chapter Next chapter Index

SIMCA-P Manual. Umetrics, Umeå, Sweden, 1995.

Smook, G.A. Handbook for pulp and paper
technologists. (1982). Storebråten, S.
Karakterisering av massa och fiberr.
Sv. Papperstidning No. 6, p. 125–130, 1999.

Strand, B.C. Analyzing the effects of raw material

variation on mechanical pulp properties using
integrated factor networks. Pacific Simulation,
Moscow, Idaho, USA.Turunen, M. et al.
Comparison of fiber morphology analyzers.
Appita Annual Conference, 2003, Melbourne,
Australia.

Tremainis, A. Advanced traditional methods of

analysis of fiber surface layers. La Chimica e l’In-
dustria, No. 2, March, 2006.

Yan, H. et al. Fiber length effect on fiber suspen-

sion flocculation and sheet forming. Nordic Pulp
and Paper Research Journal, Vol. 21, No. 1/2006.
 ≔ ← → ≡
134
Table of contents Previous chapter Next chapter Index

—
24. Index

A Chemical pulping, 27, 34

Absorption properties, 61 Chemical thermo-mechanical pulp, 33
Acacia, 19 Chips, 16, 30
Additives, 40 CMP, 33
APPITA, 50 Coarseness, 76
Arithmetic mean value, 56 Color, 60
Aspen, 19 Collapsed fibers, 28
Confocal techniques, 64
B Confidence intervals, 83
Bagasse, 19 Coniferous tree, 12
Bamboo, 22 Contaminants, 24
Barking, 30 Continuous digester, 35
Basis weight (see Grammage, 56) Corrugating medium, 102
Batch cooking, 35 Cotton, 22
Bauer McNett classifier, 50 Crill, 73
BCTMP, 33 CSF, 78
Beater evaluation, 50 CTMP, 33
Bending stiffness, 59
Birch, 10, 15 D
Bleached chemical mechanical pulp, 33 Deciduous tree, 12
Bleaching, 38 Degradation pattern, 28
Blend chest, 40 De-icing, 30
Bonding strength, 98 Density, 56
Brightness, 38, 60 DIP, 37
Bursting strength, 58 Dirt, 24
Dissolving pulp, 36
C Drainability, 52
Caliper (thickness), 56
Cambium, 10 E
Canadian Standard Freeness, 43 Earlywood, 11
CED viscosity, 50 Eucalyptus, 20
Cellulose, 8, 26 Extractives, 12
Chemical mechanical pulp, 33
Chemical pre-treatments, 33
 ≔ ← → ≡
135
Table of contents Previous chapter Next chapter Index

F Hemi cellulose, 26
Fiber debris, 119 Hemp, 22
Fiber deformations, 70
Fiber dimensions, 15 I
Fiber length, 69 Image processing, 65
Fiber model, 68 ISO, 50
Fiber orientation, 49
Fiber properties, 28, 67 K
Fiber wall, 26 Kappa number, 53
Fiber wall thickness, 76 Kink, 71
Fiber width, 97 Kozo, 22
Fibril, 72 Kraft process, 34
Fibril angle, 28
Fibrillation, 46, 72, 97 L
Filler, 40 Laboratory sheet forming, 50
Fine paper, 100 Latency, 42
Fines, 72, 97 Latewood, 11
Fisheyes, 17 Light scattering, 60
Flexibility, 98 Lignin, 27
Folding endurance, 58 Linear models, 105
Formation, 49 Linerboard, 103
Form factor, 42, 70 Log preparation, 30
Fractionation, 48 Lumen, 12
Friction, 59
M
G Measurement range, 128
Geographical differences, 13 Mechanical pulping, 31
Gloss, 60 Mesh, 54
Grammage, 56 Middle lamella, 26
Mini shives, 54
H Models, 105
Hardwood, 13 Moisture content, 23, 56
Headbox, 40 Mulberry, 22
Heartwood, 10
 ≔ ← → ≡
136
Table of contents Previous chapter Next chapter Index

N Refiner mechanical pulp, 31

NIR, 64, 80 Refining, 39
NSSC, 36 Repeatability, 128
Nylon, 23 Reproducibility, 128
Nyqvist theorem, 90 Reinforcement, 96
Residual ink, 82
O Resin, 12
Oil absorbency, 61 Resin acids, 18
Opacity, 60 RMP, 32
Optical properties, 60 Rosin, 102
Optical technique, 62 Roughness or smoothness, 59

P S
Paperboard, 103 Sack grades, 102
Paper machine, 40 Sampling, 93
Paper manufacture, 23 Sampling theorem, 90
Paper testing, 55 Sapwood, 10
Parenchyma, 13 Sawmill chips, 16
PFI mill, 46 SCAN, 50
PGW, 32 Scanning electron microscope, 64
Picking, 100 Schlereids, 17
Pine, 16 Screening, 48, 54
Pixel, 65 Secondary fines, 72, 97
Plantations, 14 Secondary wall, 17
Poisson distribution, 84 SEM, 64
Post-drying treatments, 41 SGW, 32
Press section, 44 Shape factor, 70
Pressurized groundwood pulp, 32 Shives, 24
Primary fines, 97 Schopper-Riegler, 52
Primary wall, 26 Single layer products, 99
Printing papers, 99 Slenderness, 77
Pulp testing, 50 Softness, 59
Pulping processes, 18 Softwood, 12
Sommerville, 50
R Springwood, 11
Rapid Köthen, 51 Spruce, 20
Ray cells, 12, 18 Standard deviation, 83
Ray tracheids, 18 Stickies, 24
Rayon, 23 Stiffness properties, 59
Recycled pulp, 24 Stone cells, 17
 ≔ ← → ≡
137
Table of contents Previous chapter Next chapter Index

Stone groundwood, 28
Stratified products, 102
Straw, 19
Strength properties, 57
Structural properties, 55
Summerwood, 11
Surface properties, 59
Swelling, 48
Synthetic fibers, 23

T
TAPPI, 50
Tear strength, 99
Tensile stiffness, 57
Tensile strength, 55
Thermo mechanical pulp, 32
Thinnings, 14
Tissue, 102
TMP, 32
Tracheids, 12

U
Uniformity, 114
Ultraviolet light, 64

V
Vessels, 13
Vessel cells, 17
VIS, 119
Viscosity, 50

W
Water Retention Value, 54
Whiteness, 82

X, Y, Z
X-ray technology, 64
Z-directional strength, 58
Zero-span tensile strength, 50
 ≔ ← ≡
138
Table of contents Previous chapter Index

—
25. Abbreviations

A Acid (symbol for bleaching stage) IR Infrared light. Longer wavelengths than VIS
ADt Air dry ton of pulp; ton of pulp at 90% dryness ISO International standardisation organisation
Australian pulp and paper industry technical LC Low consistency
APPITA association
LWC Light weight coated paper (mechanical grade)
BCKP Bleached chemical kraft pulp
MC Medium consistency; 5–15% consistency of pulp
BCMP Bleached chemical mechanical pulp
Mt Million tonnes
BCTMP Bleached chemical thermo-mechanical pulp
MVDA Multivariate data analysis
BHKP Bleached hardwood kraft pulp
NBSK Northern bleached softwood kraft pulp
Bauer McNett, fractionation equipment for
Near infrared light. Longer wavelengths than VIS
BMN laboratories
NIR and IR
BSKP Bleached softwood kraft pulp
NSSC Neutral sulphite semi-chemical
C Chlorine (symbol for bleaching stage)
NYBOT New York board of trade
CCD Charge coupled cevice
O Oxygen (symbol for bleaching stage)
CED Cupriethylene diamine
OBA Optical brighteners
Cost, insurance and freight (trade term requiring
OCC Old corrugated containers; grade of RCP
the seller to arrange for the carriage of goods by
sea to a port of destination, and provide the OMG Old magazines, a recovered paper assortment
buyer with the documents necessary to obtain ONP Old newspapers, a recovered paper assortment
CIF the goods from the carrier)
OW Office waste; grade of RCP
CMP Chemical mechanical pulp
P Hydrogen peroxide (symbol for bleaching stage)
CPO Computer print-outs; grade of RCP
PAPTAC Pulp and paper technical association of Canada
CSF Canadian standard freeness
PCA Principal component analysis
CTMP Chemical thermo mechanical pulp
Papir- og fiberinstituttet AS (Paper and fiber
D Chlorine dioxide (symbol for bleaching stage) PFI research institute, Norway)
DIP Deinked pulp (from recycled fiber, RCF) PGW Pressurized groundwood
E Alkaline extraction (symbol for bleaching stage) PIX Price index
EU European union Partial least square, or projections to latent
FEM Finite element method PLS structures

Experimental paper machine installed in the PM Paper machine

FEX 1980s at RISE, Stockholm, Sweden PQM Pulp-quality-monitor instrument
FFT Fast fourier transform Q Chelating stage (symbol for bleaching stage)
H Hypochlorite (symbol for bleaching stage)
Q2 Prediction capacity of models
HC High consistency
R2Y Explanation capacity of models
HW Hardwood
R&D Research and development
 ≔ ← ≡
139
Table of contents Previous chapter Index

Recycled fiber, i.e. pulp made Recycled fiber, i.e. pulp made
RCF from recovered paper RCF from recovered paper
RCP Recovered paper, raw material for recycled fiber WRV Water retention value
RMP Refiner mechanical pulp WFC Wood-free coated (paper); coated fine paper
Rest of EU, the European Union (apart from VIS Visible light
RoEU Denmark, Finalnd and Sweden and Scandinavia).
X Enzyme stage (symbol for bleaching stage)
ROI Return on investment
Y Dithionite (symbol for bleaching stage)
S1 Secondary wall-layer 1
S2 Secondary wall-layer 2 yr Year

S3 Secondary wall-layer 3 Z Ozone (symbol for bleaching stage)

SC Super calendered paper or semi-chemical pulp

Svenska Cellulosa AB (pulp and paper company
SCA with headquarters in Sweden)
Scandinavian pulp, paper and board testing
SCAN committee (Finland, Norway and Sweden)
SEK Swedish krona; Swedish currency
SEM Scanning electron microscope
Stone groundwood pulp; a type of
SGW mechanical pulp
Schopper-Riegler. See CSF. SR normally used in
SR papermills using chemical pulps
Pulp & paper research institute of Sweden,
STFI nowadays RISE (Research in Sweden Softwood)
T Peracetic stage (symbol for bleaching stage)
t Tonne; 1000 kg
Technical association of the pulp and paper
TAPPI industry, USA
TEA Tensile energy absorption
Thermo-mechanical pulp; a type of
TMP mechanical pulp
US United States (of America
USD US Dollar; US currency
Ultraviolet light. Shorter light
UV wavelength than VIS
W Water (symbol for bleaching stage)
—
ABB AB / Lorentzen & Wettre
P.O. Box 4
SE-16493 Kista
Sweden
Tel: +46 8 477 90 00

abb.com/pulpandpaper

© Copyright 2021 ABB. All rights reserved.

Specifications subject to change without notice.