Advanced Training Course on Deinking - Dispersion and Kneading

7/27/2019 Advanced Training Course on Deinking - Dispersion and Kneading

1/38

8th

CTP/PTS Advanced Training Course on Deinking Grenoble, May 29-30-31, 2007

1

PAST, PRESENT AND FUTURE OF DISPERSION AND KNEADING

SAURABH KUMAR, BENJAMIN FABRY, BRUNO CARRE, ALAIN COCHAUX,

FRANOIS JULIEN SAINT AMAND AND GERARD GALLAND

1. INTRODUCTION............................................................................................................ 32. PRINCIPLES OF DISPERSION AND KNEADING EQUIPEMENTS............................... 4

2.1. High Speed Dispersion............................................................................................ 52.1.1. Principle ........................................................................................................... 52.1.2. Theoretical approach........................................................................................ 72.1.3. Miles and May theory applied to dispersing...................................................... 82.1.4. CTP approach.................................................................................................. 9

2.2.

Low Speed Kneading .............................................................................................10

3. OPERATING CONDITIONS..........................................................................................114. DISPERSION................................................................................................................12

4.1. Asphalt...................................................................................................................124.2. Hot Melt Contaminants...........................................................................................12

4.2.1. Waxed papers and boards ..............................................................................134.2.2. Hot melt glues for book bindings and container sealing...................................13

4.3. Stickies...................................................................................................................144.4. Residual ink and specks.........................................................................................15

5. INK FRAGMENTATION AND DETACHMENT ..............................................................175.1. Hot dispersion between 2 deinking stages .............................................................18

5.1.1. Low speed kneader between 2 deinking stages ..............................................18

5.1.2. High-speed disperser between 2 deinking stages ...........................................195.1.3. Comparison of high-speed disperser and low speed kneader between 2deinking stages .............................................................................................................22

5.2. Dispersion before deinking.....................................................................................235.3. Two hot dispersion stages......................................................................................26

6. BLEACHING .................................................................................................................286.1. Use of high-speed disperser as mixer ....................................................................286.2. Bleaching in low-speed kneader and high-speed disperser....................................28

6.2.1. Peroxide bleaching..........................................................................................286.2.2. Reducing bleaching.........................................................................................29

6.3. Destroying catalase................................................................................................297. MICROBIOLOGICAL DECONTAMINATION.................................................................30

8. FIBRE PROPERTIES....................................................................................................318.1. High-speed dispersers ...........................................................................................318.2. Low-speed kneaders..............................................................................................338.3. Comparison of the effects of low-speed kneader and high-speed disperser on fibreproperties..........................................................................................................................33

9. Summary.......................................................................................................................3410. References ................................................................................................................35


2/38


3/38

8th


3

1. INTRODUCTION

Early work on the dispersion of contaminants began in 1946 by a group of American paper mills withthe intent of recycling kraft bitumen paper. It led to the development of a number of processes. Themost common thermo-mechanical processes, or asphalt dispersion, consisted of high temperature(150 C) treatment in a device such as a disk type refiner [1][2]. The pressurized disk refinerprocesses have experienced considerable industrial development since the 1960's in contrast to thechemical or purely thermal processes [3] which have remained largely unused.

In mills recycling packaging papers and boards, hot dispersion is used to disperse thermofusiblecontaminants, such as waxes, hot melt and bitumen in order to avoid problems of spots (especially onhot plates during the converting of liners) and sticking between sheets. Dispersion or/and kneadingequipment is found in most modern mills recycling packaging papers. It may be applied to the wholepulp or only on the long fibre fraction. However, in some applications, due to the progress in fine slotscreening, the use of hot dispersion is questioned.

Hot dispersion in deinking plants has been utilized since 1978. Dispersion homogenizes the stock

[4][5], residual ink particles which have not been detached from the fibers (such as inks used for offsetnewspapers) are so finely dispersed as to no longer appear as undesirable specks, i.e., to say thatsmaller than 40-60 m, not to be seen by the naked eye. Dispersion or kneading has also beenproposed to simultaneously disperse specks and as a high consistency mixer for bleaching the pulp[6]. Several papers proposing a hot dispersion stage between two flotation stages [7][8][9], or betweena flotation stage and a washing stage [11] were presented during the 1989 EUCEPA symposium inLjubljana. Since the end of the 1980s, dispersion has become a basic treatment in multi-loop deinkingprocesses. The high treatment temperatures also decrease the bacterial content of the pulp, leading todecrease in microbes in the final pulp. Dispersion and/or kneading equipment is now included in allmodern deinking facilities.

The main applications of dispersion and kneading are as follow:

Dispersion of hot-melt contaminants, stickies, specks, residual ink. Reduce dirt specks below the visibility limit & distribute them finely or make them floatable

Break down stickies & distribute them finely or make them floatable.

Distribute wax very finely.

Treat fibre thermally to increase bulk.

Detachment of ink prior to deinking, or removal of residual ink prior to post-deinking.

Bleaching: thermal pretreatment, mixing chemicals or use as bleaching reactor.

Microbiological decontamination: elevated temperatures in the presence of hydrogen peroxidedestroy bacteria and fungi.

Changes in fiber properties: depending on device and operating conditions.


4/38

8th


4

2. PRINCIPLES OF DISPERSION AND KNEADINGEQUIPEMENTS

Hot dispersion consists of mechanical treatment performed generally at high temperature; high

consistency and using appropriate techniques to transfer energy to the pulp. It must be however notedhere that dispersion does not remove contaminants associated with the incoming fibres (ink, tonerparticle, stickes..); but what it does is to break down these large size particles due to high shearingforces applied on the fibres & the contaminants, the result is fine dispersion of these non wantedparticles. It can be however possible that due to high level of forces applied on high consistency pulp,the unwanted particle gets detached from the fibres and get removed in the subsequent processes.

Key components of a dispersing station arethickener, the plug crew & the steam-preheater to raise the temperature. Thedispersion system can normally be operatedin both atmospheric as well as pressurisedconditions, the latter offering advantages to

ink & sticky dispersion in some grades. Plugscrew facilitates the pulp flow to thepreheater and helps avoiding pressure dropand steam losses. Fig. 1 provides a look atthe complete dispersing system with itsvarious sub-parts.

The table below shows the variousapplications of dispersion & kneadingoperations are linked for propertyenhancement or contaminant removal forwhite and brown paper grades.

Fig. 1 Dispersion system with its full assembly consistingof a dewatering screw, heating screw, & disperser.

Grades

Task of dispersionNewsprint

SCpaper

LWC

paper

Tissue

TestLiner

Board

(filler)

Board

(topliner)

Dirt specks & stickiesdispersion

Wax dispersion

Coating grits dispersion

Ink/toner detachment

Bleaching agents/ stockmixing

Strength improvement

Bulk Increase

Microbialdecontamination

Tab. I Dispersion requirement for paper grades produced from recycled fibres[10]

Two principal technologies are present for dispersing operations: high-speed dispersion and lowspeed dispersion, also called low speed kneading. Machinery suppliers, with a few rare exceptions,

supply devices using one or the other technology. Consequently, these two technologies have oftenbeen contrasted. A few papers make comparisons of the two technologies [10][12][13][14][15].


5/38

8th


5

2.1. High Speed Dispersion

2.1.1. Principle

High-speed dispersion corresponds to a very short mechanical treatment (less than one second)applied to a small mass of fibrous material with a strong shearing effect due to a narrow disc gap and

high rotational speeds. Speeds range from 10003000 rpm, but are generally around 12001800 rpm.The dispersion effect results mainly from the impact of contaminants or ink against the tackle surface.

High-speed dispersion (disc dispersers) has been developed through the technology transfer frompulper-refiners for mechanical and thermal mechanical pulp (Asplund-Defibrator, Beloit, Krima-Cellwood, Andritz-Sprout-Bauer, Kvaerner-Hymac, etc.) with bar discs, and by the adaptation of highconcentration deflakers with toothed discs (Voith-Sulzer, Krima-Cellwood). A high-speed disperser isillustrated in Fig. 2.

Thoothed fillings

The unit is comprised of stator and a rotor disc. Stockis fed into the centre and due to centrifugal forces, thepulp moves radially and comes into contact with statorand rotor edges. Contacts with rotor accelerate the

pulp travel while contacts with stator slows it down.These impacts induce a velocity differential resultingin shear forces and dispersion. Two types of platesare commonly used:

- pyramidal design (intermeshing toothedpattern). The pulp is forced radially throughthe small chanels created between the teethon opposing plates,

- refiner bar (fine or coarse bars). The pulp isforced through the high shear zone betweenrotor and stator plates

Fig. 2 High-speed disperser (Voith-Sulzer)

Some high-speed dispersers, (i.e., Krima-Cellwood or Voith-Sulzer) can be equipped with different

types of teeth or bars; others are designed to operate at medium consistency (15%). The use ofrefiners operating at low consistency (5%) has also been proposed for the dispersion of specks [16].

Sunds Defibrator [17] has recently proposed a conical high-speed disperser. Compared to aconventional flat disc high-speed disperser at a constant outer radius, the conical design is said toprovide a larger dispersion area. This increases the likelihood of an ink or sticky particle toexperiencing multiple impacts and fragmenting into smaller particles.

Metso paper[18] alsomanufacturers conical dispersers. They mention that the same conical dispersercan be utilized for viscous or mechanical fibre development by using different fillings (toothed barfillings or straight bar fillings)

According to Heimonen [19], conical high-speed dispersion technology in comparison to disc high-

speed disperser allows to have a pulp flow through the plates more uniform, i.e. the effect ofcentrifugal force on the pulp flow is decreased and used for fibre treatment. Compared to the samediameter disc, conical filling gives larger treatment area, which means that the dispersing energy canbe fed to the pulp more gently through a higher number of impacts into impurities. In multi-disc high-speed disperser, the treatment area is increased but the control of plate gap, and hence the control ofdispersion uniformity is much more complex.


6/38

8th


6

Fig. 3 Principle of ConiDisc [20]

Aikawa[20]has developed a completely new type disperser called ConiDisc Dipserser (Fig. 3). Thistypical disperser has inner conical and outer disc combination fillings. The working area of the conicalpart is much more larger than the disc platepattern. This results in longer retention timesin the working zone, hence increasing theimpacts between fibre and bar fillings. Thedisc plate outside the conical part serves as avalve to fill in the conical part with stock.

The adjustment of power consumption inhigh-speed disperser can be achieved by gapadjustment or by dilution at the inlet of thedisperser. Regarding gap adjustment, thetemperature of dispersing is one of theparameter that needs to be taken into accountin terms of dispersing efficiency (Fig. 4).Indeed, high temperature permits thenarrower filling gap for a given motor load. Asa result, the impurities receive strongdispersion power and are dispersed to verysmall pieces [20].

Fig. 4 Power consumption in Conical disperser as a

function of gap and temperature [20]


7/38

8th


7

2.1.2. Theoretical approach

The easiest way to characterize mechanical treatments is to consider the specific energy consumption(energy per unit of dry material expressed in kWh/T). This way has the advantage to give directlyeconomic consideration but it cannot be used directly to characterize the phenomena involved duringmechanical treatment. Several approaches coming from refining theory has been proposed to

describe the phenomena involved during high-speed dispersing treatment.

Brecht approach

Brecht [21] developed a theory to characterize refining operation. It assumes that refining treatmentoccurs at the edges of refiner bars with pulp response to refining being identical whatever the deviceconsidered (size, type). He defined the average specific edge load as the effective power input pertotal edge length per second. For barred plate patterns, the total edge length per second (LB) and theaverage specific edge load (SELB) are given by the following equations:

For bar-plates:

LB Total edge length per second (m/s)

w Rotational speed (rpm)

ZRB Number of bars on rotorZSB Number of bars on stator

lB Average length of a bar (m)

SELB Average specific edge load (Ws/m)

PA Average total power (W)

BSBRBB lZZ

w

L=

60

B

NLAB

L

PPSEL

=

where

PNL No load power at given refiner speed (W)

Ruzinsky et al. in 2004 [22][23] applied the specific edge load theory to characterize the magnitude offorce involved during high-speed dispersing operations and the specific energy to measure the extentof dispersion. It was assumed that the treatment (ink detachment, particle comminution, etc.) occursprimarily at the edges of dispersing elements. It was also assumed that the magnitude of force appliedat the edge of high-speed disperser element is the critical factor determiningink detachment.

In the case of pyramidal plates, the total edgewas calculated by summing the lengths of theleading (active) edges of the intermeshingpyramids. To perform calculation, theynumber the rows of pyramids beginning at thecentre of disperser as illustrated in Fig. 5. Thecalculation assumes that the first rotor rowintermeshes between the first and secondrows on the stator. The pyramid dimensionschange with radial position due to the bevel ofthe base plate, and is included in thecalculation. The total edge length per secondand the average specific edge load forpyramidal plates are given by:

Fig. 5 Diagram of rotor and stator plates used by

Ruzinsky et al. [22] Numbers represent the positionof pyramidal rows from the centre of the disc

For pyramidal-plates: LP Total edge length per second (m/s)w Rotational speed (rpm)

iRPN

Number of pyramid in row i on the rotor

iSPN

Number of pyramid in row i on the stator

iPIl

Average length in inner pyramidal edge in row i (m)

iPOl

Average length of outer pyramidal edge interactingin row i

n Total number of rows of pyramid on a plateSELP Average specific edge load (Ws/m)PA Average total power (W)

( ) ([

+= =

+

n

i

i

SP

i

PO

i

SP

i

PI

i

RPP NlNlNw

L1

1

60

P

NLAP

L

PPSEL

=

where

PNL No load power at given refiner speed (W)


8/38

8th


8

CTP high-speed disperser was characterized by this approach. A comparison with data reported byRuzinsky et al. is given in the following table.

Specificenergy

SEL LBDisperser type Plate pattern Source

KWh/T Ws/m m/s

Lab refiner Pyramidal plates Ruzinsky et al. [22] 6-36 0.03-0.27 3 900-5 600Lab refiner Barred plates Ruzinsky et al. [24] 50-280 0.02-0.20 49 000

Industrialdisperser

Toothed plates McCarthy [13] 40-100 9-14 *

-

Industrialdisperser

Toothed platesRuzinsky et al. [22]**

35-63 - -

CTP disperserCTP plates (fromBonpertuis)

Calculated 40-220 0.26-0.925 500-10

200*: estimated value by Ruzinsky et al. [22] based on units usually operating at Cm between 25 and 30% with a temperaturebetween 77 and 90C and rotor speed between 1200 an d 1888 rpm**: dispersers used for SOW/MOW or ONP/OMG

Tab. II Data related to various high-speed disperser type according to Brecht approach

2.1.3. Miles and May theory applied to dispersing

In 2001, Ruzinsky et al. [24] characterized high-speed disperser by applying the theory of Miles andMay [25][26] that allows to distinguish the number of bar impacts and the specific energy per barimpact imposed during the high-speed dispersing stage. This theory calculates the radial velocity ofpulp moving through the refiner allowing the residence time and the number of bar impacts imposed tobe calculated. For their test conditions, the following equations were used:

For bar-plates: E Specific energy consumption (J/kg)n Number of impacts imparted to a fibree Specific energy per impact (J/kg/impact)r Radial coefficient of frictiont Tangential coefficient of frictionN Average number of bars per unit length of arch Number of rotating discs in refinera Constant of frictionr2 Outer radius of dispersing zone (m)

enE =

( )

=

1

2

12

2

2r

r

rrw

cEahN

n

t

rln

( )

=

1

2

1222

r

rcahN

rrw

e

r

t

ln

where

r1 Inner radius of dispersing zone (m)w Rotational speed (rad/s)

As mentioned by Ruzinsky et al. [22] in 2004, the Miles and May theory was developed for bar-platesand cannot be used for pyramidal plates in its present form.

Comparison between refining and dispersing is also reported in the following table in terms of specificenergy consumption, number of impact and specific energy per impact.

E (MJ/kg) n (impacts) e (J/kg.impact)Laboratory refiner (Sprout Waldrom 12") [24] 0.17-1.31 300-3600 250-610

Sprout-Bauer 36-1CP Refiner [26] 4.3 8100-16700 260-530

Typical disperser [13] 0.14-0.36 n.d. n.d.

Tab. III Comparison of operating parameters for refiners and high-speed dispersers [24]


9/38

8th


9

2.1.4. CTP approach

Another approach consists in considering volume energy consumption. This idea has beensuccessfully applied to have overall and composite information for pulping phenomena [27]: theknowledge of volume energy consumption allows determining defibering level whatever the pulperdevice (LC pulper, Helico pulper or drum pulper), the consistency and the pulping time.

Based on this idea, the following equations can be written:

pulpM

EEm = and

totV

EEv = where

Em

E

MpulpEvVtot

Specific energy consumption (kWh/T)Energy consumption (kWh)Mass of o.d. pulp treated (T)Volume energy (kWh/m

3)

Total volume of pulp treated (m3)

By combining the two equations,

EmV

MEv

tot

pulp=

If we suppose that the density of pulp suspension corresponds to water density (=1000 kg/m3=1

T/m3), then

EmCmMpulpEv

Vtot

Specific energy consumption (kWh/T)Mass consistency expressed as a fraction (-)Mass of o.d. pulp treated (T)Volume energy (kWh/m

3)

Total volume of treated pulp (m3)

totwater

tot MV =

1

EmCmEmM

MEm

V

MEv

tot

pulp

tot

pulp===

where

waterWater density (1000 kg/m

3)

Note that the expression of this estimated volume energy is present in the equation given by Miles andMay to describe the number of impact imparted to a fibre. Besides, the other parameters are constantin the present study. In other terms, the approach of Miles and May can be summarized by theestimated volume approach.

The different phenomena that are controlled during the dispersing stage can be viewed as 'solid'fragmentation. The fragmentation phenomena can then be described by two main parameters:

- The forces involved during pulping. The overall forces could be described by energyconsideration even if we are not able to determine the energy applied to solid particles [28]. Byusing energy consideration, it should be possible to take into account both the intensity of themechanical forces and the retention time inside the disperser.

- The 'solid' particle strength. Cohesive forces that linked the 'solid' particles each other coulddescribe it. The cohesive forces could be affected by external parameters such astemperature or physico-chemical parameters.

Application of this approach will be given in p. 20. [29]


10/38

8th


10

2.2. Low Speed Kneading

Low speed kneading imparts a rather prolonged mechanical treatment (some minutes) to a large massof fibrous material with a moderate shearing effect. This action is related to the relatively wide interbarclearance and slow rotation. Rotational speeds are normally 100200 rpm, with a few exceptions athigher speed. The dispersion effect results mainly from fibre-to-fibre friction (rubbing action). In

general, low-speed kneaders are devices in which the pulp is fed in by a screw and held by adischarge door as it is transferred under pressure between rows of fingers on a shaft and others onthe stator wall.

Kneading technology for low speed dispersion consists of equipment with a singleshaft (Erwepa,Voith-Sulzer, Lamort-Fiberprep, Maule), or two shafts (Shinhama, Modomekan-Ahlstrom-Kamyr),specially developed for hot dispersion or resulting from technology transfer from stock mixers (MicarBlack-Clawson). A typical single shaft kneader is illustrated in Fig. 6.Various designs are used for double shaft kneaders.

B

A

E

C

F

D

A) Inlet of raw stockB) Shaft with feed screwand rotating kneading cogsC) Body with stationary cogs

D) Steam inletE) Air-operating trapF) Outlet of processed stock

Fig. 6 Single shaft low-speed kneader (Kdant-

Lamort)

Fig. 7 Double shaft low-speed kneader (Shinhama)

The Shinhamalow-speed kneader, illustrated in Fig. 7, has two counter-rotating shafts, one turning at95 rpm, and the other rotating slightly faster at 110 rpm [30]. The device previously called"Frotapulper" (Modemekan-Kamyr), and currently referred to as "MDR Kneader," (Alhlstrom-Kamyr)looks like a low-speed kneader, but the two screws counter-rotate synchronously at a speed of 900-1800 rpm (in the range of the of high-speed dispersers). The device called "Micar" ( Black-Clawson)runs at intermediate speed (400-500 rpm) [12].

The discharge door controls the volume of pulp in the device. Stock moves through the device at arather low speed. Differences in stock velocity are created inside the machine. The stock near therotor is moving at a higher velocity than the stock near the stator wall. In the double shaft kneader, therotors turn in opposite directions causing a shearing effect when the stock changes its direction ofmovement). These differences in velocity induce the fiber-to-fiber friction and the dispersion effect.

A gentle kneader including dewatering equipment has been proposed [31] and is illustrated in Fig. 8. Itconsits of 3 cylinders. At the first half of each cylinder, a screw boots the stock forward and at thelatter half, the stock proceeds through kneading blades. When the stock proceeds from one cylinder toanother, the rotation of shaft alters vice versa. In the third cylinder chemicals can be introduced. The 3cylinders allow increasing the kneading time. Each cylinder is equipped with a single motor permittingan independent control of each of them so that it is possible to adjust the unit to a variation of stockand/or target quality of the pulp to be produced. In this kind of device, no steam introduction is

required as the temperature can reach spontaneous 50 to 90C due to the friction involved in thekneading parts. The energy consumption is between 40 and 80 kWh/T. The main difference with theother low-speed kneader is the presence of dewatering zone in the first and second cylinder. This


11/38

8th


11

dewatering area permits a free water in the stock to be drained away that is claimed to remove bywashing the small particles generated by the kneading mechanical action. However, it must bebrought to the fore that the inlet consistency is 20% and the outlet about 30% inducing a "low washingeffect". The principle of this device is quite similar to the Bivis from Clextral.

Fig. 8 Gentle Kneading process from Taizen Co, Japan

3. OPERATING CONDITIONS

The dispersion stage is implemented after a thickening stage. Various devices including the doublewire thickener and screw press are used to concentrate the stock.

A heating screw can be implemented between the thickening stage and the dispersion unit. It isrequired in high-speed dispersion when steam is introduced to increase the pulp temperatures to theappropriate level. Voith Paper also proposed a new type of high-speed disperser where the pulp isdirectly heated in it [32]. In low speed kneading, steam can be introduced at the low-speed kneaderitself, and a heating screw is not generally required.

The dispersion consistency is generally 25-30%, with some devices operating at up to 35-40%.Depending upon the pulp characteristics, some devices are designed to operate at mediumconsistency.

The running temperature depends on the requirements of the dispersion and varies within a largerange. Some low-speed kneaders are operated without any steam input for deinking applications. Inthis application, the increase of temperature (up to 40-60 C) is related to mechanical energy

dissipation. Most applications use steam to raise the dispersion temperature to approximately 90C.Some low-speed kneaders and high-speed dispersers are designed as pressurized units, which can


12/38

8th


12

operate at temperatures up to 150C. These conditi ons were required in board mills for asphaltdispersion.

For both types, the energy consumption is in the range of 35100 kWh/t, with typical industrial valuesin the range 60-80 kWh/t. Various techniques can be used to adjust the energy consumptiondepending on the device used. These approaches include adjusting the pressure on the discharge

door, gap between rotor and stator, inlet consistency, etc.

In the deinking plant, dispersion can be combined with bleaching. Bleaching chemicals can beintroduced at the heating screw or at the inlet of the dispersion unit, which can be used as mixer orbleaching reactor. Depending on the operating temperature and the device used, retention time mayor not be required after the dispersion unit.

Generally the pulp discharged from a low-speed kneader does not require dilution and bleaching isconducted at the kneading consistency. The pulp can be discharged at operating consistency in somehigh-speed dispersers but generally a dilution is required for discharge. In this case storage forbleaching extension is not possible.

4. DISPERSION

Various types of contaminants can be dispersed by using high-speed dispersers or low-speedkneaders. These include bitumen, waxes, stickies, specks, residual ink, and hot melt contaminantssuch as bookbinding or container sealing glues. Dispersion of wetstrength paper is also reported[71].

It must be brought to the fore that dispersing treatment induces fragmentation of the particles so thatthey cannot be seen. However, even if the pulp seems to be cleaner, these contaminants are stillpresent in the pulp but they are not visible. If they are not removed during the next step of the process,some problems could appear in some cases during papermaking and during the final use of theproduct.

4.1. Asphalt

Dispersion of asphalt was the first use of dispersion systems. Asphalt dispersion requires hightemperature, therefore a pressurized high-speed disperser should be used in order to perform thedispersion at approximately 150C. These processes have been developed for a long time in NorthAmerica [1][2].

Precautions must be taken in order to avoid excessive losses in mechanical properties due to the hightemperature. For recovered paper mixtures containing 2% of bitumen-containing paper, it has beensuggested, that the pulp is preheated up to 85C a nd processed through a frotapulper at 120-130kWh/t [33].

Some drawbacks regarding problems of picking of asphalt on the wire and dryers and the relative costof the treatment have been reported and alternative treatments including fine slotted screening andreverse cleaning have been proposed [34].

4.2. Hot Melt Contaminants

Hot dispersion has been used for control of hot melt contaminants. The temperature rise decreasesthe viscosity of hot melt products and their internal cohesion, therefore the dispersion under shearforce becomes easier. The required temperature for dispersion depends of the type of material.Bitumen requires a very high temperature (pressurized devices should be used), hot melt adhesives


13/38

8th


13

utilized for bookbinding or package sealing have lower softening temperature and can be dispersed bynon-pressurized devices

1. Wax has an even lower softening temperature and is easily dispersed.

4.2.1. Waxed papers and boards

The recyclability of waxed papers is being questioned. The inclusion of waxed papers and boards in

the recycled pulp leads to the production of paper containing a considerable number of translucentspots.

A group of industrialists including papermakers, suppliers of wax, and converters has asked theCentre Technique du Papier (CTP) to develop or adapt technologies for recycling the various gradesof papers and boards treated with conventional waxes. The main objective of the project was todevelop a process to recycle waxed papers without downgrading. All grades were considered,including white waxed papers and brown waxed board containers. The main conclusions werepublished in a previous paper [36].

BURST INDEX

Chest Screw press Kneader Post refining

1.5

2

2.5

3

3.5

4

%waxed OCC

kPa.m/g

0 2 12 100

Kneading at 95 C, 100 kWh/t after coarse screening andthickening

In recycling plants using a hot dispersionstage, a spot-free pulp can be producedwhen recycling waxed OCC, however,hot dispersion has a negative effect on

physical properties. An additional lightpost refining treatment can restore theinitial level of strength properties, but ifthe same energy is applied to a pulpcontaining no waxed OCC, better resultscan be achieved (Fig. 9).

Hot dispersion of waxed OCC alsopromotes strong hydrophobicity in thepulp fibers, so that a dramatic increase inthe drop test is observed. Post refiningdoes not alter this aspect.

White waxed papers (which are generallywet strength papers) can be recycled in aconventional flotation deinking plantprovided the pulping conditions are

Fig. 9 Burst index after various treatments of pulps

produced by repulping mixtures containing variousamounts of waxed OCC.

modified. A recycled pulp with characteristics similar to that of wood-free deinked pulp can beproduced.

4.2.2. Hot melt glues for book bindings and container sealing

Dispersion consistency (10-14 %). Powerconsumption 54-71 kWh/T

Dispersion temperature(inlet/outlet)

Hot melt speckreduction

25/32 C 21.6 %

25/68 C 43 %

55/90 C 92 %

The small hot melt glue particles which have beendisintegrated and not removed by cleaning and screeningcan be dispersed in non-pressurized low-speed kneaders

or high-speed dispersers working at temperature close to100C. The visual aspect of the paper is improved, andthe translucent specks disappear after dispersion.

Results are reported in Tab. IV regarding the dispersiontemperature required for hot melt and wax introduced in astock from clean corrugated clippings in the mediumconsistency in the Black Clawson low-speed kneader [35].

Tab. IV Effect of Temperature on Hot-

Melt Dispersion [35]

1 Notice that new generation hot melt glues proposed and starting to be used for bookbinding have higher softeningtemperature. They are designed to be resistant to the shearing forces in the pulper and to be completely removed by coarsescreening, they will be probably more difficult to disperse, but due to the high screening removal ability this characteristic is nomore required.


14/38

8th


14

4.3. Stickies

Data set1

Data set2

Data set3

90C 93.4 90.7 92.1

105C 95.8 96.1 98.0120C 95.4 96.1 97.5

For more efficient dispersion of stickies, high temperaturesare suggested by American equipment suppliers [37].Trials performed with a pressurized high-speed disperser(deflaker type) gave the results shown in Tab. V.

Mannes has operated a high-speed disperser and low-speed kneader in parallel in their pilot plant to determinethe ability of the two machines to disperse stickies inindustrial stocks [38]. Both machines are able to achievereduction in stickies. Stickies area was reduced by 65-90%

Tab. V Percent reduction in stickies

[37]

with a high-speed disperser. Results obtained with the low-speed kneader were lower and varied overa wider range. The recommendation was: Effective and reliable stickies treatment is only possiblewith high-speed dispersers.

The incidence of retention time inside high-speed disperser through changes in feed flowhas been reported by Kanazwa and Fujita

[20]: the lower the productivity (i.e. the higherthe retention time in the high-speeddisperser), the higher the increase in stickiesfragmentation (increase in very small stickiesnumber as reported in Fig. 10). In the samestudy, they reported the effect of dispersingtemperature: the higher the dispersingtemperature, the narrower filling gap for agiven power load and the better the dispersingeffect on stickies due to stronger powerimparted to the particles.

When comparing the low-speed kneader and

the high-speed disperser in a carton boardmill, Stemmer [39] reported that high-speeddisperser induces higher cuts ofmacrostickies.

Fig. 10 Incidence of retention time through productivity

changes on micro-stickies generation [20]

High consistencies and high temperatures are recommended for effective dispersion. Theeffectiveness of dispersion increases with the circumferential speed. Speeds of 50-60 m/s representthe optimum level.

In the framework of an EU project (FOREST 1991-1993) entitled Use of Surface and RheologicalProperties in Stickies Removal and Control, conducted with Pira (UK), PTS (Germany) and TNO (TheNetherlands), CTP was responsible for studying the changes in the shape of stickies in order toimprove their removal efficiency [40][41]. Various model stickies were used in the study: acrylic and

styrene-butadiene rubber (SBR) labels and acrylic tapes. The trials were performed in the CTP pilotplant facilities, which features a Lamort low-speed kneader.

Kneading induces fragmentation of the stickies in a large particle size range. The fragmentationincreases dramatically when kneading temperature is increased from 60to 90C. Increasing thespecific energy also increases stickies fragmentation, but the effect is lower.


15/38

8th


15

0

5

10

15

20

25

30

RESIDUAL STICKIES AREA (mm/g)

S1+

S2

S1

+CL

S1+

FL

S1 S1

S1+

K S1+K+

S2

S1+K+

CL

S1+K+

FL

FL : Flotation

CL : Cleaning

K : KneadingS1 : Screening 1

S2 : Screening 2

Kneading modifies the shape of the stickiescausing them to become more spherical.The shape modification is particularlysignificant (including small particles) whenkneading is performed at high temperature.This change in shape increases the removal

efficiency by screening, but also increasescleaning and flotation effectiveness due tothe reduction in stickies size (Fig. 11).

The final recommendations of this studywere to implement screening, but alsoflotation and centrifugal cleaning afterkneading in deinking plants including a post-deinking stage in order to enhance stickiesremoval.

Fig. 11 Effect of kneading on the removal of stickies

(from acrylic tapes) by screening, cleaning and

flotation [41]

The first observation dealing with change in shape of contaminants by kneading was reported in 1976

based on experience with a frotapulper: plastic is not defibrated but rolled into spiral fashionedshapes which can be later removed by flat screening... [42]. Similar observations (stickies particlesobserved to roll up into balls rather than be fragmented) have also been reported when using a twin-shaft kneader operating without steam [43][44]. Industrial experiences have also been reported. Theyconfirmed that the shape of the stickies changes from amorphous masses to spherical particles. It wasalso observed that ink can become a part of the sticky balls. The average stickies particle size afterkneading is significantly reduced while their number is dramatically increased. It is considered(contradictory to the CTP observations reported previously) that screens and flotation are not efficientin the removal of these micro stickies [45].We can try to explain these different behaviors of stickies when subjected to the action of a high-speed disperser or low-speed kneader. The dispersion is the consequence of stretching the particlesbeyond a level inducing break-up; it becomes easier due to the increase in temperature, which causessoftening of the stickies particle and reduction of its internal cohesion. The viscoelasticity of the

stickies particle is the key point. The low-speed kneader induces more rubbing action and its ratherlow speed stretches the sticky, allowing the particle to conform and become more spherical. Incontrast, the high-speed disperser produces more impacts and quickly stretches the particle beyondits breaking point leading to break-up of the particle and higher dispersion.

4.4. Residual ink and specks

The use of hot dispersion was proposed in the 1970's in deinking plant to improve the visual aspectsof deinked pulp produced from offset print on uncoated paper (i.e., offset newspaper printed with inkcontaining a high amount of self-setting binder). Hot dispersion was implemented at the end of theprocess in order to reduce the size of large ink particles remaining attached to the fibres whichproduced a mottled appearance [4][5][46].

The drawback of this application was a brightness loss, and therefore, post-deinking stages have beenproposed. They will be discussed in the next section.

Specks are black or colored visible particles. In deinked pulp they appear when ink has not beenbroken up by pulping into particles small enough to be undetectable to the naked eye. Different valuesof this size limit are from 40 to 60 microns and more, depending on the contrast.

The main origins of specks (also called dirts) are [47][48]:

Recovered paper from household collection may contain varnished printed papers (mainlywith UV-cured varnishes). High gloss covers of magazines are the main source of UVvarnishes and specks.

Recovered office paper contains toner printed papers (laser printed papers or photocopies).

Very old offset inks, heat set offset on SC papers, etc


16/38

8th


16

High-quality print on coated papers with inks hardened by drying can also cause specks, which resultfrom binding between the pigment and the coating material. Recovered papers, which produce thesekinds of specks, are also glossy magazines. Papers printed by unconventional processes such as UV-cured inks [49] or some grades of modern digital print also induce speck formation.

A large recycling study of papers focusing on speckled deinked pulp and UV-varnished papers

particularly, has been performed in the Centre Technique du Papier [50].

40 60 80 100 120 140 160 180

99

90

0

99.9 Kneading efficiency (%)

Energy consumption (kWh/t)

UV varnish

Resistant coating

Pulp from unprinted wastepaper contaminated with 4% of a pulp produced by pulping UV varnished printedpapers or 10 % of a pulp produced from water resistantcoated papers (offset printing). Influence of energyconsumption. Pilot plant trials (pulp consistency 32 %,temperature 90C). Efficiency is calculated as spec karea reduction (specks larger than 100 m).

First, the specks from various origins have beencompared regarding their difficulty to bedispersed by kneading. Trials were performedat the CTP deinking pilot plant with a low speedkneader, which is a hot dispersion unit (TL0from Lamort, 20 kg/hr). Results from resistantcoated papers and UV varnished printed papersare reported in Fig. 12 [50].

Low energy is sufficient for dispersion of specksfrom resistant coated papers. With treatment at

current industrial energy consumption (60kWh/t) the pulp looks clean to the naked eye.Higher energy consumption is necessary fordispersal of specks from UV varnished printedpapers. Good dispersion can be obtained athigh temperature and with high energyconsumption in a slow speed kneader.

Fig. 12 Hot dispersion of specks in a low speed

kneader [50].

The work has been focused on UV-varnished papers; kneading and dispersing have been investigatedas part of this study. Trials have been performed in pilot plants of machinery suppliers in order tooptimize the dispersion of visible specks. Pulp for trials was produced at CTP by blending a pulp fromunprinted papers with 2% contaminated pulp. Contaminated pulp contained UV-varnished printed

papers pulped for 20 min at 14% consistency and 50 C with deinking chemicals. Fig. 13 to Fig. 15show results from treatment of clean pulp mixed with 2% UV-varnish contaminated pulp (Unprintedwastepaper including 2 % of UV varnished printed papers. Particles larger than 100 m were considered.

Efficiency is calculated as speck area reduction).

Both low-speed kneaders and high-speed dispersers are efficient in reducing the size of large specksas shown Fig. 13 [51].

40 60 80 100 120 140 160 180

99

90

0

99.9 Dispersion efficiency (%)


Dispersing

Kneading

Influence of energy consumption, pulpconsistency and temperature on results in trialsperformed with an industrial low-speed kneaderare shown in Fig. 14. Increasing pulpconsistency and temperature improves thedispersion efficiency of specks from UVvarnishes. Fig. 15 shows the results of trialsperformed with a high-speed disperser fordispersion of specks from UV varnishes [52].Dispersion is more efficient at high temperatureand, according to the machinery supplier, theimplementation of a post deinking stage isrecommended for removal of speck particles,which have been broken in smaller, floatableparticles.

CTP pilot plant trials on industrial devices. Specks fromUV varnishes, only specks larger than 300m are

considered.

Fig. 13 Dispersion efficiency versus energy

consumption [51]


17/38

8th


17

40 60 80 100 120 140 160 180

99

90

0

99.9 Kneading efficiency (%)


90C, 37 %

60C, 37 %

60 and 90C, 29 %

40 60 80 100 120 140 160 180

99

90

0

99.9 Dispersion efficiency (%)


(95C)

(65C)

Influence of energy consumption, pulpconsistency and temperature

Influence of energy consumption andtemperature. Pulp consistency 30%.

Fig. 14 Hot dispersion efficiency of specks from UV

varnishes in a low-speed kneader [52]Fig. 15 Hot dispersion of specks from UV varnishes

in a high-speed disperser [52]

In Fig. 14 and Fig. 15 only particles larger than 100 m were considered. Efficiency is calculated asspeck area reduction.

Voith-Sulzer is supplying both high-speed dispersers and low speed kneaders. They have comparedthe dispersion of specks for various raw materials [53]. Both devices have advantages and theirefficiencies increase with specific energy consumption. The efficiencies of the two machines alsoincrease slightly as temperature increases. With a high-speed disperser, better dispersion of specks(in a deinked pulp after a single flotation stage) from mixtures ONP/OMG is achieved, while the lowspeed kneader presents advantages with laser inks and UV varnishes.

Dispersion of specks by using a refiner operating at low consistency has been proposed. With energyconsumption of 70 kWh/t, 50% of the dirts larger than 50 m are dispersed. Larger particles (>200 m)are more efficiently dispersed (70%), and dispersion efficiency increases with increasing energy input[16].

5. INK FRAGMENTATION AND DETACHMENT

0

1 000

2 000

3 000

4 000

5 000

6 000

7 000

Number of ink particles / mg pulp

before kneading

after kneading

Average size of ink particles (m)

1.0 1.4 2.0 2.8 4.0 5.6 8.0 11 16 23 32

The dispersion of ink particles to improve the visualaspects of deinked pulp brings about a loss ofbrightness. Speck size is reduced by hotdispersion, but smaller ink particles (not visible bythe naked eye) are also broken up. The increase inthe number of small ink particles is responsible forthe brightness loss. Data regarding reduction of inkparticle size by kneading are reported in Fig. 16.

This brightness loss can be very important ifkneading is performed (without chemical) on a

poorly deinked pulp [11]. Brightness losses ofdeinked pulp containing various amounts ofresidual ink are reported in Fig. 17. It can be seenthat only a part of the brightness loss can berecovered by a hyperwashing stage, indicating that

Fig. 16 Reduction of ink particle size by kneading

(industrial plant, 70 kWh/t, 35 %, woodfree

deinked pulp) [54]

a portion of the dispersed ink has been irreversibly redeposited onto the fibers.


18/38

8th


18

ENTIRE PULP

52

56

60

64

68

72

76

80

Increased amount of residual ink

Brightness (%)

Before kneading After kneading

- 1.2

- 3.5

- 4.7

- 7.8

- 4.5

HYPERWASHED PULP

78

80

82

84

86

88

Increased amount of residual ink

Brightness (%)

Before kneading After kneading

- 4.2

- 3.8

- 0.6

- 0.4

- 1.7

Furnish: woodfree printed paper, conventional deinking (pulping 14 %, with soda, sodium silicate, hydrogenperoxide and soap, flotation). Same pulp deinked by using 0, 1, 2, 3 and 4 aeration stages.

Fig. 17 Comparison of brightness loss of poorly deinked pulp and well deinked pulp [11]

5.1. Hot dispersion between 2 deinking stages

The efficient detachment of residual ink by high-speed dispersers and low speed kneaders has led tothe development of multi-loop deinking processes. Although some mills [55][56] are operating post-flotation after a dispersion stage without chemical and produce a DIP for newsprint which meets thebrightness requirement, the advantages of combining hot dispersion and peroxide bleaching havebeen widely described and higher brightness is attained.

5.1.1. Low speed kneader between 2 deinking stages

CTP has cooperated with a deinking mill to develop a low investment cost, two loop deinking systemthat is still in operation [11][54]. The basic concept is that a peroxide bleaching stage restores thebrightness loss caused by kneading. Post-deinking is performed by washing (and flotation of the washwater which is reused for the forward dilution of the washed pulp). The alkaline conditions afterbleaching are suitable for bleaching.

70

72

74

76

78

80

82

84

86

deinked pulp kneaded pulp bleached pulp

(no kneading)

kneaded and

bleached pulp

Brightness (%) Before washing After washing

Conventional alkaline deinking of woodfree wastepaper,washing by inclined screw press, pH 10, inlet consistency3.5%.

The brightness gain obtained bykneading and peroxide bleaching ishigher than the sum of thebrightness loss during kneading andbrightness gained by washing thebleached (and unkneaded pulp).These results are reported in Fig. 18[57].

Additional trials have comparedvarious operating conditions andconcluded that the best approach isto introduce hydrogen peroxide at

the inlet of the low-speed kneader[58][59].

A lot of results demonstrating thebenefit of using hot dispersion incombination with peroxide bleachingbefore post flotation have beenpublished during the past ten years.

Fig. 18 Brightness changes by kneading, bleaching and washingpost deinking [57]

It is not possible to cite all of them, but the following examples illustrate the benefits of thiscombination.


19/38

8th


19

0 10 20 30 40 50 60 70 80 90 100

specific energy, kWh/t

60 -

55 -

50 -45 -

40 -

65 -brightness, ISO

outlet dispersionwith chemicals

outlet post flotation with chemicals

outlet dispersionno chemicals

Results of a joint study examining optimumkneading conditions for peroxide bleachingand post flotation have been published byLamort and Interox [60][61]. Fig. 19 illustratesthat kneading in the presence of bleachingchemicals is an efficient pretreatment step for

flotation post-deinking.The use of the Black Clawson high-speeddisperser (single shaft device which works atmedium rotating speed of 400-500 rpm)between a washing deinking stage and aflotation stage has been proposed [62] whendeinking a mixture of impact CPO, whiteledger and office waste. The dispersioninduces a reduction of ink particle size,thereby improving flotation removal forparticles in the range of 80-300 m.

Fig. 19 Energy optimization with 1 % H2O2 in the high-

speed disperser (70C)[61]

Chemicals : 0.7 % NaOH, 3 % silicate.Brightness at the high-speed disperser inlet : 52.2 % ISO

However, the removal of smaller particles (4-40 m) decreases, so it is suggested that these particlesshould be removed by a washing stage prior to dispersion.

5.1.2. High-speed disperser between 2 deinking stages

Implementation of dispersing stage between two deinking loops

Most of modern deinking mills use a dispersion stage between two flotation stages. As indicated in theintroduction, during the 1989 EUCEPA Symposium in Ljubljana several papers by machinery supplierswere presented proposing a hot dispersion stage between two flotation stages [7][8]. An example of amill running with these conditions was also presented [9]. Dispersion or/and kneading equipment isnow included in all modern deinking plants.

Since that symposium, machinery suppliers have published several papers recommending the use ofa high-speed disperser between two flotation stages, and have proposed the introduction of peroxide

bleaching chemicals into the heating screw before the high-speed disperser (even if few mills usechemicals in the high-speed disperser), particularly for production of higher quality DIP than for use fornewsprint, such as that used for the production of SC papers. A few examples are given.

To improve the quality of a DIP produced from 60% ONP and 40% OMG, dispersion and post flotationhas been investigated. The dispersion with hydrogen peroxide induces a brightness gain of 4.2 points,while dispersion without peroxide results in a brightness loss of 2.7 points. The post-flotationperformed after dispersion/bleaching achieves a brightness increase of 5.8 as compared to only 3.1points for undispersed pulp. The total brightness gain (inlet dispersion/ outlet post flotation) was 10.3[63].

When deinking ledger containing various hard chemically nondispersible inks, the positions beforeflotation and between two flotation stages have been examined. Based on measurements of ink

particle size, it is recommended to have a first flotation stage to remove the fine ink particles eventhough a large portion of the visible specks are not removed. These visible specks are reduced tomicroscopic size by dispersion and are then removed by post flotation [65].

According to a study performed by Selder et al. [64], North American newspapers require the leastamount of stock preparation process equipment for recycling since it is relatively easy to remove inkfrom the fibres (mainly due to differences in ink composition). For the American newspapers they havetested, it requires no dispersing treatment for better deinkability, which is why the dispersing treatmentin some US deinking lines has been shut down. Besides, it appears that storage time has little effecton high rub-off US and Canadian papers, but significantly affects the rub-off characteristics ofEuropean newspapers and strongly affects the Asian newspapers (that are characterized by highcontent of alkyl resins and vegetable oil, resulting on oxidative bonding reactions between binder, inkand substrate). Besides, Asian countries import a lot of their raw material from Europe and NorthAmerica inducing more difficulty to detach inks (this furnish must travel along the globe resulting innatural and thermal aging). Besides, the usage of colour ink is higher in local newsprint for the Asia-Pacific region as mentioned by Haynes [66]). The presence of dispersing treatment for European andAsian newspapers is required to have a good deinking line efficiency.


20/38

8th


20

Optimisation of dispersing stage through an overall approach [29]

For wood containing grade [29]:

During this study, 60% ONP / 40% OMG mixture has been 100% artificially aged in an oven at 60Cduring 3 days that corresponds to very harsh conditions responsible for poor ink detachment, highspeck content, high ink fragmentation and poor ink removal. This raw material has been subjected toconventional first deinking loop and then fed to the dispersing stage at different consistencies, withand without the introduction of peroxide bleaching chemicals. The main result is reported in Fig. 20 asa function of estimated volume energy described in p. 9. The zero volume energy applicationcorresponds to the pulp that have been thickened and submitted to post-flotation without dispersingtreatment.

According to all the properties determined, it appears that the most significant effects of high-speeddispersion between two deinking loops are:

- Ink fragmentation phenomenon that is responsible for a decrease in ink removal efficiency andtherefore in final brightness. It is therefore necessary to decrease as much as possible theenergy applied during this stage.

- Ink detachment occurs as soon as high-speed dispersing treatment is applied. However, thereis no significant improvement if the volume energy is higher than 10 kWh/m

3. There is

certainly a possibility for energy saving regarding this parameter.- Speck fragmentation requires more energy, but there is no significant improvement if the

estimated volume energy is higher than 20 kWh/m3.

- As soon as the estimated volume energy is greater than 20 kWh/m3, fibre degradation starts

to be significant.

0

50

100

150

200

250

0 20 40 60 80

Volume energy estimated by Cm.Em

ERICHyW,Speckcon

tamination

0

100

200

300

400

500

600

ERICpost-flotation

Ink content on post-flotaed pulp

Ink detachement

Speck content after post flotation

High-speed dispersing at 22% in neutral condition

High-speed dispersing at 33% in neutral condition

High-speed dispersing at 33% in alkaline condition

Fig. 20 Incidence of high-speed dispersing treatment on ink detachment, ink removal and final speck

contamination after post-flotation [29]


21/38

8th


21

The main phenomena areillustrated in the following tableand allow us to determine thestrategy to be applied. Thetendencies are reported as afunction of estimated volume

energy as it allows describingthe phenomena and thecorresponding specific energyconsumption for twoconsistencies are reported.

In order to control thedispersing stage, a reduction inenergy applied is required in Fig. 21 Summary of the effect induced by disperser

order to reduce ink fragmentationand to reduce the negative impact on flotation efficiency. However, itis necessary to apply a sufficient energy level to induce ink detachment (but a stagnant value isreached at 10 kWh/m

3) as well as reduction in speck contamination (stagnation levels commence at

20 kWh/m3). Note that for energy upper than 20 kWh/m

3, fibre degradation becomes significant.

Dispersing parameters should therefore take into account these antagonists phenomena according to

the inlet of this stage:- If cleanliness is good after the first deinking loop, it is not necessary to put high energy level

during dispersing- If cleanliness is "poor" (generally due to raw material composition variations), the energy to be

applied must be adapted to the target of the mill in order to reduce the drawbacks.- In any case, ERIC and speck measurements at the inlet (and/or at the outlet) of the high-speed

disperser is required for the regulation of dispersing running parameters.

For wood free grade:

0

10

20

30

40

50

60

70

80

90

100

0 20 40 60 80 100Specific Energy consumption (kWh/T)

Residualspecksorstic

kies(%)

10

14

18

22

26

30

Finecontent(%)

or

(tearxtensile)

1/2

Stickies

Specks

Fine

Mechanicalproperties

An example of the results obtained during milltrials (woodfree deinking line for market pulp) isrepresented in Fig. 22, where decreased

specks (after post-flotation) and macrostickiesare represented as a function of specific energyconsumption. The fines content and thecompromise between tear and tensileproperties are reported in the same figure.The dispersing treatment induces:

- Fines generation with consequences onprocess yield (part will be removedduring flotation and/or washing) and onmechanical properties. It was observedthat an increase in mechanical forcesapplied (through a decrease in gapresponsible for an increase in specific

Fig. 22 Incidence of energy during high-speed dispersingbetween two deinking loops (Industrial measurements

after post-flotation) [67]

energy consumption) induced higher fine generation. It can be advanced that this finegeneration is much more pronounced when specific energy is over 75 kWh/T.- Specific energy above 75 kWh/T, does not cause more fragmentation of macro-stickies.- Increase in speck removal during post-flotation because they are fragmented during dispersing,

to the size desired for removal by flotation.- Increased mechanical properties (expressed by [tear*tensile]

1/2) due to the refining effect,

although a plateau is reached when energy consumption is the highest.

Practical consequences of such an analysis are some possible energy savings and reduction insludge amounts. Indeed as the above figures clearly illustrates, energy can be economised in post-refining after the dispersing treatment. It appears that energy of >75 kWh/T is excessive and does notsignificantly improve mechanical or optical properties, or the dispersion of stickies. There are two maindrawbacks in using excessive energy. Firstly there is an additional energy of no clear benefit andsecondly this leads to an increase in fine generation (and therefore increase in process losses).

Optimising dispersing by decreasing the energy consumed can be recommended when this treatmentis applied before the deinking loop, particularly when washing is present (fines removal occurs mainlyduring this stage).

6633Specific energy (kWh/t)Cm = 30%

10050Specific energy (kWh/t)Cm = 20 %

Fibre degradation

Speck fragmentation

Ink detachment

Decrease in ink removal

2010Estimated volume energy (kWh/m3)

6633Specific energy (kWh/t)Cm = 30%

10050Specific energy (kWh/t)Cm = 20 %

Fibre degradation

Speck fragmentation

Ink detachment

Decrease in ink removal

2010Estimated volume energy (kWh/m3)

Phe

nomena


22/38

8th


22

Implementation of fractionation between 2 deinking loops

Recent publications [68][69] lead to the implementation of fractionation between the two deinkingloops and to implement the dispersing treatment only on the long fraction. This process option allowsto increase the capacity of existing deinking line and to reduce the specific energy consumption. Themain drawbacks that have been reported concern mainly the specks and the macrostickies as only a

part of the pulp is submitted to the mechanical treatment. This process configuration has beenimplemented in a mill [68].

5.1.3. Comparison of high-speed disperser and low speed kneader between 2deinking stages

A detailed comparison of a low speed kneader and high-speed disperser, operating at variousconditions (energy, temperature) and in the presence of various chemicals has been performed inCTP. The furnish was a wood-containing mixture comprised of printed papers that were difficult todeink such as old offset newspaper and heatset offset-printed SC papers [59]. Some of the mostinteresting results obtained when running at high energy (90-10 kWh/t) and temperature (70-90C)are reported in Tab. VI and Tab. VII.

Low speed kneading without chemicals produces strong ink fragmentation (lower brightness andhigher ERIC value on the entire pulp, particularly when kneading is performed in the presence of alarge amount of ink, i.e., after a first neutral deinking stage). The presence of peroxide bleachingchemicals improves ink detachment and/or reduces ink redeposition (lower ERIC value onhyperwashed pulp).

High-speed dispersion is efficient for breaking up specks (particularly when the speck content is high,i.e., after a first neutral deinking stage) and detaching large ink particles from long fibers (highbrightness and low ERIC value of hyperwashed pulp) even if performed without chemicals.

Dispersion with bleachingchemicals (after post-flotation)

Dispersion without chemicals(after post-flotation)Before

dispersionL.S. kneader H.S. disperser L.S. kneader H.S. disperser

57.3 Brightness % (entire pulp) 64.6 60.3 54.1 57.9

55.0Fiber brightness %(hyperwashed pulp)

62.5 60.5 56.2 58.7

306Total ink (ppm) (ERIC onentire pulp)

120 230 359 260

259Attached ink (ERIC onhyperwashed pulp)

94 118 170 149

9477 Black specks (mm/m) 1858 2935 2497 2228

Tab. VI Kneading and dispersion after a first deinking in alkaline conditions (pulping 15 %, 40C,

15 min, 1 % caustic soda, 2,5 % sodium silicate, 1 % hydrogen peroxide, 0.6 % soap)

Dispersion with bleachingchemicals (after post-flotation)

Dispersion without chemicals(after post-flotation)Before

dispersionL.S. kneader H.S. disperser L.S. kneader H.S. disperser

49.2 Brightness % (entire pulp) 59.1 58.5 44.8 49.5

48.9Fiber brightness %(hyperwashed pulp)

59.2 60.0 48.2 52.1

476Total ink (ppm) (ERIC onentire pulp)

272 309 614 416

387Attached ink (ERIC onhyperwashed pulp)

197 172 348 248

18341 Black specks (mm/m) 7018 4976 6882 3895

Tab. VII Kneading and dispersion after a first deinking in neutral conditions (pulping 15 %, 40C, 15

min, 0.5 % surfactant)


23/38

8th


23

Both high-speed dispersion and low speed kneading improve ink detachment and subsequent postflotation efficiency, especially if performed in the presence of bleaching chemicals. When hotdispersion is performed after a first deinking stage, the best results after post flotation are obtainedwhen using a low speed kneader (higher brightness: +4%, lower ERIC value: 120 vs. 230 ppm). Thisdifference is mainly due to a better detachment of residual ink from cellulosic fines and fillers, while asimilar ink detachment from long fibers has been obtained.

A comparison study between a high-speed disperser and a low speed kneader (Shinhama) has alsobeen performed with wood free recovered papers [70], however the comparison was not done withstrictly identical recovered papers and moreover the deinking processes tested were different: for thehigh-speed disperser, the later was performed after a thickening stage before any deinking whereasthe kneading stage was performed after a first washing stage. The comparison becomes therefore notvery significant.

5.2. Dispersion before deinking

Dispersion has also been proposed for use before deinking. For producing DIP for use in tissue fromledger, computer paper and coated grades, it has been suggested a high-speed disperser be usedprior to deinking. This approach can promote the release of ink and coating from the fibers to get

smaller particles more easily removed by flotation or washing. This treatment is not recommended fornewsprint stock in order to avoid a grayish appearance [46].

For deinking of laser printed papers the use of a high-speed disperser has also been proposed [71].The treatment caused a reduction in the number of visible toner ink particles but the results were notgood enough so flotation was necessary. Flotation of non-dispersed pulp was unsuccessful. However,flotation after dispersion yielded good results.

In order to obtain better contaminant removal, a process with a separate stage for removal ofcontaminants and deinking has been proposed. It includes pulping, cleaning and fine screening at lowtemperature without chemicals in order to remove non-ink contraries before they are degraded intosmall particles. A second alkaline stage is used to release and remove ink. A hot dispersion stage athigh temperature with peroxide bleaching chemicals is used for ink detachment and bleaching prior to

flotation deinking [72].

The use of a twin-shaft kneader running without steam after pulping and coarse screening and beforethe first flotation stage is the basic part of the Pacific Rim Deinking Technology" [30][73]. Thistechnology is recommended mainly for deinking mixed office papers containing computer printout, andcopy papers with laser and electrostatic inks, but it also allows the use of ONP as part of the furnish[30]. In old mills the low-speed kneader can be implemented before a soaking tower designated tocomplete the separation of ink from the fibers [30]. The use of two kneading stages, including onebefore deinking, is also proposed [74].This technology has been suggested for deinking mixed office paper containing laser print or mixturescontaining other difficult to deink print, such as UV-cured and heavy inks and varnishes [43][44]. Thetwin shaft kneader works in cold conditions (i.e., without steam introduction; the power consumptionof 50-60 kWh/t produces a temperature increase from 22 C up to 44-47 C). It has been

recommended for reduction of speck size and to prepare them for removal by flotation.

The kneader discharge temperature should be below the glass transition temperature to prevent thetoner inks from smearing back on to the fibers. In an effort to replace an agglomeration technology, aprocess based on kneading and flotation has been implemented in a mill recycling office waste. Thefinal pulp has lower final dirt counts. The agglomeration technology is considered as a validtechnology but with more risk than kneader/flotation technology [75].

The following analogy has been forwarded by Ferguson and McBride [43] in order to explain whykneading at low temperature is more efficient: think of a piece of chewing gum stuck on a shirt- it iseasier to remove the gum with an ice cube and pull it intact from the cloth rather than use a steam ironand smear the gum over the cloth.

The effect of dispersion/bleaching with hydrogen peroxide in the high-speed disperser (Frotapulperwhich is a high speed kneader) on ink detachment has also been studied by Dubreuil et al. [76].They showed that with a wood-free furnish (mainly sorted office paper), high-speed disperser


24/38

8th


24

bleaching with H2O2 induces a higher brightness gain during flotation. Lower flotation losses (andlower ash removal) were observed when dispersion was performed with peroxide bleaching chemicalsthan occurred with flotation performed after dispersion without chemical. The presence of sodiumsilicate during flotation may have been responsible for this effect.

A more fundamental study has been carried out by Ruzinsky et al. [24] (the approach from refining

theory is also described in page 8). ONP was pulped in laboratory Helico pulper during 40 min. Thereslushed pulp pass then through a laboratory high-speed disperser for several consistencies (12-30%) and for various gap (0.51 to 1.14 mm). The following points have been observed in theirlaboratory refiner:

- ink detachment increased with increased specific energy application until 2.3 MJ/kg- At higher energy application (when 30% consistency was reached), strong ink fragmentation

was observed as well as ink redeposition on the fibres.- An increase in number of impact imparted to a fibre induces an increase in ink redeposition

whereas an increase in specific energy per impact facilitates the ink detachment.With similar experiences, Ruzinsky et al.[22] by using the Brecht approach (page 7), demonstratesthat ink detachment increased as the intensity of the treatment increased when measured using thespecific edge load. Besides, the effect of rotor speed was also brought to the fore:

- dispersing at 2500 rpm decreased ink content on fibres up to approximately 40 kWh/T

- dispersing at 1750 rpm decreased ERIC at lower rateThe effect of rotor speed was then explained by a decrease in treatment intensity (0.08 W.s/m at 1750rpm compared to 0.12 W.s/m at 2500 rpm). A rise in treatment intensity to 0.12 W.s/m improve the inkdetachment whereas higher treatment intensity (>0.12 W.s/m) did not show additional effect.For both approach, no indication is given regarding the ink removal by flotation.

53

55

57

59

61

63

65

0 10 20 30 40 50

Dispersing consistency, %

Brightness,%

0

1000

2000

3000

4000

5000

6000

7000

Specksarea,mm/m

brightness HW brightness specks area

30%.old offset heat set on SC 30% ONP 40% OMGBrightness measurement on entire pulp and hyperwashedpulp pad specks on handsheet

The effect of dispersing or kneadingjust after pulping+screening hasbeen investigated by CTP duringthe last years [59][77]. Theparameters studied included

- The raw materialcomposition: wood

containing and wood freerecovered papers,

- The temperature, thechemical introduced, theenergy input in the high-speed disperser or the low-speed kneader,

- Unconventional runningconditions for kneading,

- The consistency.

According to the results reported, itappears that the best

brightness/speck compromise giving

Fig. 23 Brightness and cleanliness of the pulp after high speed

dispersion at different inlet consistency and flotation [77]

also the best ink detachment / ink redeposition compromise after dispersion and flotation, is obtainedfor a low consistency dispersing.

Blank with dispersion but floated


25/38

8th


25

0

100

200

300

400

500

0 10 20 30 40 50

Dispersing consistency, %

E.R.I.C.,ppm

0

2000

4000

6000

8000

Specksarea,

mm/m

E.R.I.C. HW E.R.I.C. Specks area

30%.old offset heat set on SC 30% ONP 40% OMGResidual ink content (ERIC) on entire pulp and hyperwashedpulp pad specks on handsheet

In other terms, the "optimal"sequence would be[Pulping+screening+dispersion+flotation] without thickening. Comparedto similar sequence but with athickening to perform dispersing at

high consistency (25-45% range), alarge final brightness decrease canbe observed (it can reach 5 to 10%after flotation step): 64% ifdispersion is performed at 3% butonly 59% and even 54% ifdispersing is performed at 25 and45% respectively. Even if dispersingat 3% is able to decrease the speckcontamination, it represents highercontamination (2 to 3 times higher)than more conventional dispersingconsistency.

Fig. 24 Residual ink and cleanliness of the pulp after high speed

dispersion at different inlet consistency and flotation [77]

For kneading before the first deinking loop, it can only be performed at high consistency (requirementfor such a technology) and induces a large ink fragmentation as well as a large ink redepositionleading to poor efficiency of the deinking step. Better results are obtained if the kneading treatment isperformed with conventional chemistry used for peroxide bleaching (soda, silicate and peroxide).However, such a solution is not sufficient to reach acceptable deinked pulp.

Some unconventional kneading conditions (longer kneading time and/or higher energy input with orwithout peroxide) have been investigated. However, the ink detachment improvement by a kneadingtreatment before any deinking cannot be obtained neither by an increase in kneading time or inkneading energy input. Indeed, the ink amount in the pulp is too large and the compromise betweenink detachment and ink redeposition is not in favour of the ink detachment (even if the forces presentin such a device are able to improve the ink detachment, the ink fragmentation is more pronouncedand is responsible for higher ink redeposition). The ink detachment from fines elements by kneadingwith alkaline bleaching liquor is thus possible only after a first deinking stage [59] (in that case, theremoval of ink particles reduces the potential risk of redeposition caused by a too high ink content anda too high ink fragmentation).

Finally, if a high quality deinked pulp is required keeping inside the fine elements, the first deinkingloop with a previous low consistency dispersing treatment could be followed by a low-speed kneaderwith peroxide in order to detach the ink particles from the fines before their removal in a post flotationloop.

Trials have been also reported regarding the influence of dispersing and kneading conditions for awood free recovered papers that contains 50% laser prints and 50% photocopies [77]. The wood freerecovered papers have been pulped in Helico pulper in neutral condition including 0.15% of Rhoditec1000. The dispersing treatment is then performed just after screening at 30% consistency and 90kWh/T. The influence of alkaline bleaching liquor as well as temperature have been investigated:

- The best "final ink content / speck contamination " compromise is reached for high speeddispersion performed after the peroxide bleaching when dispersing temperature is below thesoftening temperature of the toners,

- Good efficiencies are also obtained with the low-speed kneader in presence of alkalinebleaching liquor, still below the softening temperature of the toners.

- The detrimental effect of high temperature is more pronounced for kneader than for high-speed disperser. At high temperature, the kneading treatment leads to very bad results both interms of brightness and cleanliness of the pulp

The negative impact of high temperature observed during this study confirms results published byFerguson et al. [78].

More recently, full pilot plant simulation of deinking lines for tissue mill (wood free grade) including 2deinking loops combining each of them flotation and washing steps have been presented [79] (moredetails are given in the paper devoted to Pulping and Ink detachment). Two options were investigated:

Blank without dispersion but floated


26/38

8th


26

- [Pulper][screening][High-speed dispersing][1st

deinking loop][Thickening][second deinkingloop] with high speed treatment performed at 3% consistency, 45C and 90 kWh/T

- Pulper][screening][1st

deinking loop][Thickening+steam][High-speed dispersing][seconddeinking loop] with high speed treatment performed at 30% consistency, 70C and 90 kWh/T

The implementation of high-speed dispersing stage at low consistency allowed to improve the inkdetachment and to reduce the speck contamination so that the final pulp was equivalent between the

two options proposed.

For wood-free grade, Ruzinsky et al. [34] still using Brecht approach has determined the efficiency ofvarious pyramidal and barred plate patterns for high-speed dispersing performed directly after pulpingof printed toner. The efficiency was determined in terms of specks ( >113 m) and small particles (byERIC measurements). The following results were brought to the fore:

- Speck contamination decreased exponentially with specific energy consumed, with dispersionmore efficient when carried out at higher consistency.

- Toner particle size reduction was attributed to the magnitude of the forces applied to theparticles and their probability to be affected by the mechanical forces.

- The pyramidal plates induced a mider range of forces, resulting in a greater particle sizedistribution than the barred plates that develop more uniform plate gaps.

The Brecht theory was applied to describe high-speed dispersion by assuming that all treatments were

done by the edges of the bars/pyramids. A Specific Edge Load (SEL) of 0.1 W.s/m was found to givethe best reduction in particle size for barred patterns. However the knowledge of the SEL did notreconcile the action of the different plate patterns and did not explain the behaviour of pyramidalpattern. One more time, no data are given regarding the possible effect of such treatment on inkremoval efficiencies.

Regarding the implementation of mechanical treatment before any deinking sequence, the mostrecent published research works performed in this area [68][80] consist in introducing fractionation andto have optimized separate treatments on each fraction including mechanically actions on the longfraction rich in specks and attached inks. A presentation on this aspect will be done during the presentATC by Lascar and Borrego [80].

5.3. Two hot dispersion stages

The use of low-speed kneaders and high-speed dispersers is widely developed in Japan [74]. Thesedevices are considered as useful for both ONP

Particle diameter, m

Frequency

0.1 1 10 100

One step

Two steps Brightness

(%)

Deinkability

(%)

Ink particles

diameter average

(m)

84.4 9228.1

88.8 9714.5

and OWP (recovered officepaper) deinking. High-speed dispersers arerecommended for strippingink from cellulose fibers andkneaders are used to adjustink particle size. In somecases (deinking of toner-printed papers), twokneading steps seem to be

necessary before flotationto adjust ink particle sizedistribution (Fig. 25).

Fig. 25 The effect of kneading (toner-printed paper) [74]

Modern mills designed for the production of wood-containing DIP for SC or LWC paper or wood-freeDIP for market pulp are more and more complex and can include three (and sometimes four) loopsand flow sheets including two dispersion stages are often proposed.

Regarding multi-loop deinking plants including two dispersion stages, the main questions are Whichdevices? In what order implement them? In which conditions to operate them?

Some Japanese deinking plants include two kneading stages. The use of two twin-shaft kneadersworking in cold conditions is proposed for double-loop flotation deinking [43][81]. These processesare proposed to deink furnish containing a large part of toner printed papers.


27/38

8th


27

For the production of SC paper from European household waste collection (mixture ONP/OMG) it hasbeen proposed a three loop process including two dispersion stages (with bleaching chemicals)located after flotation stages and a final low consistency refining stage [82].

The operations are very sophisticated in mills producing market deinked pulp from office paper. Someof them in North America [83][84][85] and also in Europe [86][87] include two dispersion stages before

the flotation stages.

It has also been suggested to combine the advantages of high-speed disperser and low-speedkneader.

Voith-Sulzer recommends positioning the high-speed disperser in the first dispersion stage due to itsmore efficient sticky reduction and the low-speed kneader in the second stage because it treats thefibers more gently. The low-speed kneader can be operated at low temperature, thereby preservingthe integrity of the fibers [53].

Some North American mills producing market deinked pulp from office paper, operate a kneadingstage at the beginning of the process before flotation, and dispersion after this first flotation stage. Anadditional deinking stage by flotation or washing follows the dispersion stage [83][84][85].

Regarding our results, we recommend the opposite in order to achieve higher brightness and lowerspecks content. The high-speed disperser should be operated early in the process in order to break uplarger specks and increase their removal by the flotation stage. Alternatively, this first loop can beoperated at non-alkaline conditions for the removal of waterbased inks. After a first flotation stage andthe removal of a signif

Documents

Advanced Training Course on Deinking - Dispersion and Kneading