Some key aspects on screening of chemical pulp to achieve a fine fraction – a literature review

BIOECONOMY

Some key aspects on screening of chemical

pulp to achieve a fine fraction – a literature

review

Elisabeth Björk

Summary

This literature review focus on fibre length-based fractionation with screens to achieve a fine fraction, how fractionation results can be evaluated and on the influence of different screening parameters when using pressure screens. The design of the screen basket, and in particular the aperture size and shape, have a predominant effect on fractionation. Smaller aperture size leads to improved fractionation efficiency. Furthermore, it seems clear that screen plates with holes fractionate better than slotted, even when the slots are narrower than the hole diameter. Moreover, smooth screen plates are more efficient for fractionation, as contoured screen plates increase the fibre passage. For slotted screens the fibre passage ratio is affected by the fluid velocity through the aperture, while the fibre passage ratio for screen plates with small holes is independent of the fluid velocity. Microperforated screens, i.e. screens with holes with a diameter of less than 300 micrometre, are very efficient for fines fractionation in order to enrich the fines in the accept fraction.

Keywords: screening, fractionation, chemical pulp, fines

RISE Research Institutes of Sweden RISE Bioeconomy Report No: 140 Stockholm May 2020

Preface

The author is grateful for the financial support from RISE Research Institutes of Sweden and The Knowledge Foundation.

Content

3 Characterisation of fractionation result ... 9

4 Evaluation of screening process ... 10

5 Fractionation studies using screens ... 15

6 Conclusions ... 20

7 Appendix – nomenclature used in this review ... 21

1

Introduction

Screening is an important operation in pulping and papermaking. The primary purpose of screening is to remove contaminants, but screens are increasingly used to fractionate fibres for targeted processing (Olson 2003). Screens are used in pulp mills, secondary fibre processing plants, paper machine approach systems and broke system applications (McCarty 1988). In screening for cleaning and debris removal, the purpose is to remove unwanted components while the acceptable fibres should pass without losses. In waste paper preparation, screening is one of the most important process stages for obtaining optimum finished stock quality (Borschke, Bergfeld et al. 1998). In mechanical pulping, screens usually are used to separate long fibres and shives and send them to reject refining (Allison and Olson 2000). Also, in chemical pulping, screening is used to remove shives and fibre bundles. Furthermore, in nearly all paper machines there is a pressure screen just before the headbox to protect the wet end from debris and dirt and to deflocculate the furnish. Screening is also used for fibre recovery, e.g. in water streams or black liquor filtration, here microperforated screen baskets are used. Nichols describes a method for black liquor filtration where 200 µm drilled holes are used (Nichols 1986).

When screening is used to separate fibres according to characteristics as length or flexibility it is often called fractionation (Scott and Abubakr 1994; Sloane and Hogg 1995; Meltzer 1998; Olson 2003; Höke and Schabel 2009). During fractionation, a single pulp stream is separated into two, each with a modified fibre morphology.

In the industry, pressure screens and hydrocyclones are used to fractionate pulp (Kumar, Julien Saint Amand et al. 2014). Pressure screens are widely recognized as the most efficient and practical unit for achieving fibre length fractionation (Scott and Abubakr 1994). In fractionation with screen, the separation is dominated by fibre length, but fibre flexibility is also known to be a controlling factor (Sloane, Kibblewhite et al. 2006). Whilst in fractionation with hydrocyclone the separation is dominated by density and specific surface (Weise, Terho et al. 2000).

Screening to fractionate fibres has been used for many years, especially in mechanical pulping and fibre recycling. Driven by the need to optimize the product properties, fractionation has become a tool also for chemical pulps (Sloane 1999; Olson 2003; Asikainen, Fuhrmann et al. 2010; Qazi, Mohamad et al. 2015). Improved fibre fractionation and selective processing of the resulting pulp fractions provide a means of producing higher quality, more uniform pulp (Olson, Roberts et al. 1998; Kumar, Julien Saint Amand et al. 2014).

Fractionation allows for a production of pulps containing a large share of a certain fibre fraction, for example, to produce a pulp fraction with high fines content. The fines are important for product properties. The use of fines for increasing the strength has been reported on many occasions. Lobben (1977) showed that fines, separated from refined kraft pulp by a bow screen, added to both unrefined and refined pulp, increased the strength for hand sheets. Retulainen et al (1993) compared the impact on strength for kraft pulp fines and TMP fines and found that kraft fines were much more effective than TMP fines for increasing Scott Bond. For bulky paper and board grades, pulp with high fines content can be added to improve the strength. In board making, this knowledge has

been used for a long time to increase the z-strength in the middle plies of the board. Highly refined chemical pulp is added to improve the strength (Gavelin 1995). This highly refined pulp, also called glue pulp, has high fines content. In many laboratory scale studies, fines have been separated with Britt dynamic drainage jar (BDDJ), with 76 µm holes, or Bauer McNett, with 200 mesh wire, i.e. 76 µm openings. In an industrial scale the openings in the screen plates are often much larger, as the main purposes of the screening are other than separation of fines.

If a fraction with high fines content should be produced through screening of pulp in an industrial scale, the fines must be efficiently separated from the pulp. At the same time, to get an efficient production, the concentration of the fines fraction must be high. Moreover, the production through the screen and the runnability are crucial. The focus of this review is on length fractionation with screens, how the screening results can be evaluated and the effect of screening parameters. The delimitation is on screening of virgin fibres, particularly in pilot or industrial scale, and especially with small screen openings (less than 300 µm), to collect fine material.

2

Screening equipment

The concept of screening has followed the development of paper machines, since the first Fourdrinier paper machine in 1799. Before that, the impurities were picked out by hand. There are two basic elements in a screen, some sort of separation barrier and some sort of device to keep it clean (Gallager 1999). The pulp and debris pass through the apertures with differing probabilities (Olson, Roberts et al. 1998). If the particles to be separated are larger than the screen aperture in all three dimensions the particles will be 100% rejected. This type of screening is called barrier screening. The remaining material will be selectively rejected based on length, stiffness and thickness. This screening effect is called probability screening or statistical screening. In probability screening the efficiency of the screen depends on the mass reject rate (Gallager 1999; Höke and Schabel 2009).

From the beginning, the screens were open atmospheric systems. The original screening system was the vibratory flat screen, where the stock flow passed through and the rejects accumulated on top. The best known of all predecessors to the modern pressure screen is the Cowan screen (Gallager 1999; Olson 2003). This screen was fed from a gravity headbox to ensure a constant feed pressure and a flat bladed rotor cleaned the screen plate, the accept chamber was atmospheric. Another screen type is the bow screen. It has few moving parts and low maintenance cost. The screen plate has a bowed surface with narrow horizontal slots, the curved flow generates a dewatering pressure and at the same time keeps the surface clean. There is a wide selection of screen plates, a normal slot width is 150 microns. The pulp suspension is fed on the screen plate, the fibrous material aligns with the flow direction and is retained on the screen surface, while water and fine particles passes through the slots. Bow screens are used for fibre recovery, cleaning, thickening fibre/filler separation and fractionation.

Pressure screens are today the dominant type of screen. Pressure screens present several advantages over older atmospheric centrifugal or vacuum screens (Gallager 1999). Atmospheric screens are a major source for entrained air in the process, and they are much more sensitive to consistency for best operation. Pressure screens are easier to

control, and they are much smaller than atmospheric screens of the same capacity. As atmospheric screens are not under pressure, they must be placed high, while pressure screens can be placed anywhere.

A pressure screen has a cylindrical screen plate and a rotor. During screening, the pulp suspension enters the feed zone of the screen and is accelerated to a high tangential velocity by the rotor. The part of the pulp that passes through the screen plate screen goes out through the accept port, while the remaining pulp continues down on the feed side of the screen and out through the reject port. There are four different types of pressure screens, a flow path from inside to outside with the rotor on the inside is the most common configuration, labelled Outflow in Figure 1.

Figure 1 Different working principles of pressure screens. (Olson 2003)

The equipment design and the operation parameters have a great effect on the screening result. The screen plates have holes or slots with different apertures and the screen surface can be smooth or be contoured with different designs. There are rotors with different design, and the screens can be operated at different rotor speed, pulp consistency and reject rates. The screen plate design and configuration have a larger influence on operation than any other controllable parameter (Gallager 1999).

The first pressure screens had holed screen plates. Holed screen plates are still widely used as they are easier to manufacture and therefore less costly. The holes are normally taper-drilled (cylindrical at the feed side and conical on the accept side) to reduce the pressure drop through the plate. To avoid stapling, the distance between holes in the direction of rotation should be approximately the size of the fibre length. Thus, a trash screen can have an open area of 40-50% while a screen with much smaller holes have 10% or less. Holed screen plates are better at fractionation of fibres by length than slotted screen plates. Hence, they are usually only used for protection or as fractionation screens. To increase the capacity, holed screen plates with a contoured surface can be used, but this also reduce the screens efficiency. Contoured baskets are more able to allow long fibres to pass and are therefore less efficient at fractionation (Retulainen, Nieminen et al. 1993).

Slots generally have their long dimension perpendicular to the tangential direction of the rotor rotation. This increases the probability for rejection of flat and hair-like debris. The slot width is usually much shorter than the hole diameter for holed screens. The slots are narrow, 0.06 - 0.76 mm. Slotted screen plates are better than holes at removing short fat debris by barrier screening.

The main purpose of the rotor is to produce negative pressure pulses to backflush pulp accumulations from the screen aperture. The higher the pressure pulse magnitude the higher the screen capacity. The pressure pulse magnitude is the main difference between

different rotor designs. The pulse magnitude can also be increased by increasing the rotor speed. As the rotation increases the pulse frequency also increases, which can affect the screening result. If the time between backflushing pulses is less than it takes for at least half the length of the particle to pass through the opening it is probably rejected.

The rotor also aligns the long dimension of the fibres with its direction of rotation which tends to align debris and fibres, this is good for debris removal but also discourage especially longer and stiffer fibres from passing the screen and encourages mat formation. Contoured screen plate surfaces in combination with the rotor introduce microturbulence near the cylinder surface which disrupts the fibre mat and lets more fibres pass the screen more easily. Contoured screen plates have much greater capacity and can handle higher feed concentration, Figure 2 and Figure 3. A contoured surface cylinder can be operated at lower pulp reject rate compared to a smooth surface screen cylinder with one size larger openings. The fluidizing effect and its effect on screen capacity, concentration range, reject rate, debris removal etc. is dependent on surface contour and rotor design.

Figure 2 Feed consistency vs accept capacity at constant pressure drop for screen cylinders with 2.0 mm holes and different surfaces. Old corrugated container furnish (McCarty 1988).

Figure 3 Pressure drop vs accept capacity at 3.5% feed consistency for screen cylinders with 2.0 mm holes and different surfaces. Old corrugated container furnish (McCarty 1988).

In general screen baskets with holes are more suitable for fractionation of long flexible fibres, while screen baskets with slots are better for separation of thicker, stiffer and even shorter fibres. For good fractionation results, the rotors circumferential speed and the mean basket hole/slot velocity should generally be lower than those for screening applications. Low consistency fractionation about 1% has advantages, narrower slot widths may be used, and thus screening and fractionation efficiency can be significantly improved. The disadvantage is that larger machines are needed (Borschke, Bergfeld et al. 1998).

3

Characterisation of fractionation

result

Pulps are very heterogeneous as the fibres even within individual trees have a quite large variability in width, length and wall thickness. During the pulping process the size distribution is actually increasing so that the pulp will consist of fibres of different size and fines, i.e. smaller particles, often fibre fragments and fibrillar material (Ferritsius, Ferritsius et al. 2018). When screening pulp, a single pulp stream is separated into two streams with different size distributions. To be able to evaluate the fractionation result the pulp streams must be characterised.

The Bauer-McNett classifier and the Clark classifier are probably the oldest standardised methods for analysing fibre distributions, they are both based on a set of serial coupled screens with decreasing sieve openings (Tappi T233 cm-95). The fibres caught in respective fractions are weighted, they thus give a weight-weighted fibre distribution. The screening is primarily done according to fibre length, the effects of fibre width are considered to be minor but long flexible fibres can pass the screens (Ullman, Billing et al. 1968).

Measuring fibre dimensions can be made manually using a microscope, but it has the disadvantage that it is labour-intensive and that only a few particles are imaged at the same time. Thus, it is difficult to obtain quantitative statistics. Automatic image-based fibre analysers have existed since the 1990’s. These fibre analysers combine measurements on flowing pulp with image analysis to obtain quantitative information on a large number of particles. The results are presented as averages and fibre distributions, e.g. fibre length distribution. Different weighting of the data can be used, arithmetical or length weighting are commonly used. The mean fibre length is usually defined as the length weighted mean length of particles longer than 0.2 mm (ISO-16065-2).

Fines are usually defined as the cellulosic pulp fraction that passes through a 76 µm aperture (200 mesh screen) in a solid-liquid separation process (Tappi T261 cm-00; SCAN-M 6:69, ISO 10376:2011). For this a Britt dynamic drainage jar (BDDJ) is often used (SCAN CM 66). In image-based fibre analysers fines are usually defined as all objects with a particle length less than 0.2 mm (ISO 16065-2). Chemical pulp fines are often classified as primary fines and secondary fines. The primary fines are present in the pulp before any refining, while the secondary fines are produced during refining. Secondary fines enhance the strength properties more than primary fines. Mechanical pulp fines can be classified as flaky fines and fibrillar fines. The flaky fines contribute

significantly to light scattering and opacity while the fibrillar fines contribute to strength (Ferluc, Lanouette et al. 2010; Hyll 2015).

4

Evaluation of screening process

Several fundamental studies have been carried out on the factors influencing the screening results. Comprehensive studies of fibre screening have been performed at University of British Columbia starting with the work of Gooding and Kerekes (Gooding 1986; Gooding and Kerekes 1989; Gooding and Kerekes 1989).

The fibres are much narrower than the holes or slots in a screen plate, thus theoretical all fibres could pass the screen, still few longer fibres pass the screen. There are many theories about the mechanisms governing fibre passage in probability screening (Olson 2003). The mat theory states that the pulp form a mat on the screen plate which blocks larger fibres. This is most likely not the case, since high speed video shows there is no mat. Another theory is the turning effect, which refers to the inability of long fibres to turn sufficiently at an aperture to pass through. A third theory is the alignment theory, which implies that long fibres near the screen plat are constrained to be aligned with the flow, this enhances the turning effect. A fourth theory is the wall effect, which assumes a length depending concentration gradient at the screen plate surface, with short fibres closest to the plate. This results in that long fibres do not pass through the aperture as easily.

There are some commonly used basic equations for characterising screening performance. For the nomenclature used in this review see appendix.

Volume reject rate

𝑅𝑣 = 𝑣𝑅

𝑣𝐹 [1]

Mass reject rate

𝑅𝑚 = 𝑚_𝑅 𝑚_𝐹=

𝑐_𝑅 𝑉_𝑅

𝑐_𝐹 𝑉_𝐹 [2]

Reject thickening factor

𝑅𝑇𝐹 =𝑅𝑚

𝑅𝑣 =

𝑐𝑅

𝑐𝐹 [3]

where vR is the volume flow in the reject, vF is the volume flow in the feed, mR is the mass

flow in the reject, mF is the mass flow in the feed, cR is the mass concentration in the

reject, and cF is the mass concentration in the feed.

For probability screening the screen performance can be described by an efficiency – reject rate curve (E-R) curve. In cleaning applications, the debris removal is essential. The efficiency of debris removal Er can be described as the ratio between the amount of

debris removed and the amount of debris in the feed. 𝐸𝑟 =

𝐷_𝑅

where DR is amount of debris in the reject, and DF is the amount of debris in the feed.

The same approach can be used for long fibre removal Lr.

𝐿𝑟 = 𝐿_𝑅

𝐿_𝐹 [5]

where LR is the amount of long fibres in the reject, and LF is the amount of long fibres in

the feed. Both Er and Lr are strongly dependent on Rm.

The separation efficiency as function of reject rate is shown in Figure 4. The diagonal line represents flow separation (i.e. no difference in composition of the reject and accept stream). The other extreme is the ideal separation, where all particles of a certain particle type x, e.g. contaminants or long fibre stays in the reject stream. When the reject rate equals the concentration of x, all x particles will go to the reject and all the other particles go to the accept. But at lower reject rates x particles must go to the accept due to the material balance. And when the reject rate is higher than the concentration of x, also other particles will end up in the reject. Thus, at higher reject rates the purity of the reject will decrease until all the feed ends up in the reject at 100% reject rate Figure 5.

Figure 4 Separation efficiency versus

reject rate. Ideal separation with 20% particles of the type that should be removed. Based on a combination of (Höke and Schabel 2009) and (Bliss 1992).

Figure 5 Reject purity at ideal separation

with 20% particles of the type that should be removed.

It is desirable to have E-R curves in the form of equations, which can be used for modelling of screening systems. If the equations contain a single independent parameter relating to E and R, this parameter can be used as a screening index for characterisation of screen performance.

Kubat and Steenberg (1955) treated the screening process as a statistical process, where particles have an average probability of passing the screen. They modelled a flat screen. Provided that no mixing exists in the direction of the flow, such a screen can be treated as a series of hypothetical micro-screen units. By extending their work, an E-R equation can be obtained (Gooding 1986; Gooding and Kerekes 1989)

𝐸 = 𝑅𝛽 [6] 0% 100% 0% 20% 40% 60% 80% 100% Se parat io n rat io T Reject rate Rm Ideal separation Real separation No separation Se p ar ati o n e ff ic ie n cy 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% R ej ect cl ea nl ine ss Reject rate Rm R ej ec t p u ri ty

The parameter β may be used as a screening index. Decreasing values of β represents better screening. At β=1 no screening takes place (E = R), while at β=0 E = 1 for all R. Nelson (1981) introduced the screening quotient, Q, as a possible way to assess screening performance.

𝐸𝑟 = 𝑅𝑚

1−𝑄+𝑄𝑅𝑚 [7]

Q describes the proportions of debris directed to the accept and reject streams. If Q = 0, then E=Rm and no screening take place, the screen merely splits the feed stream into two

streams. Higher values of Q denote better screen performance. At Q = 1, E=1 for all Rm,

the theoretical perfect screen. Nelson’s equation is widely used in industry.

Some of the first fundamental studies of pressure screening was done by Gooding and Kerekes (Gooding 1986; Gooding and Kerekes 1989) who proposed mathematical models relating pulp passage through a single aperture to pressure screen performance. In this work, they derived the two screen performance equations [6] and [7] from screening at a single aperture and in doing so defined the assumptions on which the equations are based. They define the term passage ratio Px as:

𝑃𝑥= 𝐶_{𝑠 𝑥}

𝐶_{𝑢 𝑥} [8]

where Cu x is the concentration of particles upstream of an aperture and Cs x is the

concentration of particles through the aperture. The subscript x identifies the particle type. For optimal screening of contaminants c from pulp p the passage ratio for pulp Pp=1

and the passage ratio for contaminants Pc=0.

From the material balance for the components p and c the E-R equation can be derived for two cases; perfect mixing where the particle concentration is the same everywhere along the screen plate and plug flow where there is no axial mixing in the screening zone but perfect mixing perpendicular the screen plate. In the case perfect mixing the E-R equation can be written:

𝐸𝑟 =

𝑅𝑚

𝑅𝑚+ 𝑃𝑐_𝑃𝑝− _𝑃𝑝𝑃𝑐𝑅𝑚

[9]

Combining equation [7] and [9] gives: 𝑄 = 1 −𝑃𝑐

𝑃_𝑝 [10]

Where Q is Nelson’s screening quotient.

In the case of plug flow the E-R equation can be written: 𝐸 = 𝑅(𝑃𝑐/𝑃𝑝) [11]

which combined with equation [6] gives the quotient β: 𝛽 = 𝑃𝑐

𝑃𝑝 [12]

The two equations [9] and [11] represent limiting cases of mixing and define the bounds within which real screens work in pure probability screening Figure 6.

Figure 6 The E-R curves for probability screening at Pc/Pp=0.25.

Using the same approach, Gooding and Kerekes (1992) presented a mathematical model that relates consistency drop and reject thickening to reject ratio and passage ratio. They showed that a model assuming plug flow through the screening zone in a pressure screen fitted well with experimental data from a real pressure screen.

𝑅𝑇𝐹 =𝑐𝑅 𝑐𝐹= 𝑅𝑣 (𝑃𝑝−1) _[13] 𝑅𝑣= 𝑣_𝑅 𝑣_𝐹 [14]

During fractionation a single pulp stream is separated into two streams with different fibre length distributions. An average pulp passage ratio as descried above does not describe these changes in fibre length distributions. Olson et al (Olson 1996; Olson and Wherret 1998; Olson, Roberts et al. 1998; Olson, Allison et al. 1999; Olson 2001) have performed several studies regarding the effect of fibre length on passage through screen apertures. Here, the ability of pulp fibres to pass a screen was quantified, they use the fibre passage ratio to describe the pulp’s change in fibre length distribution caused by screening. Performance equations have been derived to describe the changes in fibre concentration and length distribution caused by the screen. These theoretical models have been verified by experiments.

The reject thickening function t(l) and fibre removal efficiency function e(l) are both theoretically related to volumetric reject ratio Rv by a fibre passage ratio function P(l)

which characterizes the length dependence on fibre passage through a single screen aperture.

𝑡(𝑙) = 𝑅𝑣𝑃(𝑙)−1 [15]

𝑒(𝑙) = 𝑅_𝑣𝑃(𝑙) [16]

Fibre passage ratio is defined as the concentration of fibres of length l passing through an aperture cs(l), divided by the concentration of fibres upstream of an aperture cu(l).

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 E R_m Plug flow Mixed flow

𝑃(𝑙) = 𝐶𝑠(𝑙)

𝐶𝑢 (𝑙) [17]

The passage ratio P(l) was nearly 1.0 for the fines material, it decreased with increasing fibre length to nearly 0.0 for fibre longer than 4 mm, Figure 7. This means that the shortest material passes freely and thus has the same concentration in passing the aperture and upstream the aperture. However, longer fibres are hindered by the aperture and have a lower concentration after passing the screen. For small fibres the reject thickening function t(l) was approximately 1.0, while with increasing fibre length the reject thickening function increased as longer fibre were retained by the screen and became concentrated in the reject stream, Figure 8. Moreover, the reject thickening function increased with lower Rv, as more fluid passes the screen at low Rv. The fibre

removal efficiency also increases with increasing fibre length, Figure 9. For the shortest material, the fibre removal efficiency is approximately equal to the volumetric reject ratio Rv, as the smallest material passes through the screen with the fluid. Longer fibres do not

pass the screen as easily as short fibres, thus the fibre removal efficiency increase with fibre length. The fibre removal efficiency also increases with reject ratio because more material is passed into the reject stream. At Rv=1.0 e(l)=1.0 by definition.

Figure 7 The measured fibre passage ratio function for Rv range tested. The solid line represents the average passage ratio function P(l) (Olson, Roberts et al. 1998).

Figure 8 The reject thickening function t(l), for the Rv range tested. The solid lines are t(l) calculated from P(l) and equation [15] (Olson, Roberts et al. 1998)

Figure 9 The fibre removal efficiency function, e(l), for the Rv range tested. The solid line represents the fibre removal efficiency function calculated from P(l) and equation [16] (Olson, Roberts et al. 1998).

5

Fractionation studies using screens

Several studies have been carried out on the factors influencing the fractionation results and the resulting pulp and sheet properties. A common conclusion is that smaller aperture size gives better fractionation, i.e. larger difference between the fractions. Gooding et al. (2001) fractionated softwood CTMP in pilot scale using two small industrial pressure screens. Screen baskets with holes 1.0, 1.4 and 1.8 mm was used. Both passage ratio, which is directly related to reject thickening, and fractionation result were analysed. Passage ratio increased with rotor speed in some cases and decreased in others, the relationship depended on passing velocity, rotor type and aperture size. The best fractionation occurred with the smallest apertures, lowest passing velocities and lowest value of passage ratio. The screen basket and in particular the aperture size and shape have a predominant effect on fractionation. The effect of rotor variables was shown to be significant, but secondary.

Even smaller holes increase the fractionation and results in high fines content in the accept fraction. Qazi et al (2015) fractionated bleached softwood kraft pulp using a pressure screen, with different smooth hole screen baskets. The hole sizes were 0.5, 0.8 and 1.0 mm. The smallest holes (0.5 mm) gave the most effective fractionation, with longer fibres in the reject fraction and high fines content in the accept fraction.

Figure 10 Fibre passage ratio, P, plotted against fibre length for three aperture diameters at constant Rv of 0.6 and constant aperture velocity of 0.3 m/s (Qazi, Mohamad et al. 2015).

Smaller aperture size gives better fractionation also for slotted screen baskets. Ämmälä (2001) investigated fractionation of TMP with slotted screens in lab scale and in mill scale. The trials showed that fractionation increased with reduced slot width and reduced contouring.

In several studies, fractionation with holed screen baskets are compared with slotted screen baskets. Holes from 0.8 up to 2.1 mm diameter are compared with much narrower slots, 0.1 to 0.3 mm slot width. All studies draw the same conclusion, holed screen plates fractionate better than slotted, even when the slots are narrower than the hole diameter. Moreover, smooth screen plates are more efficient for fractionation, as contoured screen plates increase the fibre passage.

Olson and Wherret (1998) screened TMP with a laboratory screen with slot width between 0.1 mm and 0.25 mm. It was shown that important factors affecting fibre passage ratio for slotted screens are fibre length, aperture size, average fluid velocity through the aperture and the average fluid velocity upstream of the aperture.

In another study, Olson et. al. (1999) screened softwood TMP in a small industrial pressure screen, with smooth holes between 0.8 mm and 2.1 mm in diameter. These trials showed that fibre passage ratio was independent of volumetric reject ratio and independent of aperture velocity for hole screens with small apertures.

The separation efficiency also differs between slots and holes. Olson (2001) compared the fractionation performance of contour-slotted screen plates and smooth-holed screen plates. Softwood CTMP was screened in a small industrial pressure screen. These trials showed that smooth-holed screen plates fractionate more efficiently than slotted screen plates and that it can be seen in the different shape of the fibre passage ratio, P(l), curve, Figure 11.

Figure 11 Fibre passage ratio for a smooth-hole screen plate, a slotted screen plate and an ideal screen plate all with the same passage ratio for fibres 2.0 mm, P2.0 = 0.5, based on experimental studies (Olson 2001).

The fibre passage ratio for smooth-holed screen plates was independent of aperture fluid velocity, while for slotted screen plates the passage ratio was strongly dependent of aperture fluid velocity.

In another study, (Sloane 1999), bleached radiata pine kraft was screened in pilot scale using an Ahlstrom F1 pressure screen. The effect of equipment design and operating variables were examined. Different screen baskets were used; smooth hole screen baskets with aperture size 1.0, 1.4, 1.6 and 2.0 mm, contoured screen with 1.4 mm holes, and contoured slotted screens with 0.3 mm and 0.15 mm slots. The effectiveness of separation was highly dependent on the design of the screen basket and rotor and to a lesser extent on operating variables. Generally, a decrease in aperture dimension improved separation, as did less vigorous rotor action. Contoured screen surface was detrimental to separation.

Panula-Ontto (2002) fractionated unbleached unrefined never-dried softwood kraft pulp in pilot scale with a pressure screen, Valmet Tampella TAP-50. Smooth-hole screen baskets with aperture size 1.0, 1.2 and 1.4 mm were compared to an earlier study with wedge wire screen baskets, 0.06, 0.1 and 0.15 mm. Pressure screening using smooth-hole screen separated chemical pulp very efficiently according to fibre length. The smooth screen basket with 1.0 mm holes separated the pulp better than 0.06 mm wedge wire basket.

Sloane et al. (Sloane, Kibblewhite et al. 2006) examined single stage screening of radiata pine chemical pulp. Both slotted and holed screens were used. The smooth screens had different hole diameters (1.0, 1.2, 1.4, 1.6, 2.0, 2,4 mm). The profiled screens had 1.4 mm holes or slots (0.3 and 0.15 mm). The smooth hole screens gave more extreme separations, and the differences were larger with smaller holes.

The purpose with fractionation differs. It can be all from to produce a long fraction with less fines, to the other extreme, to produce a fine fraction with high fines content. The fractions can be treated separately and the combined for a better final product or be used in different layers or products.

Fractionation to remove short fibres and fines before refining to increase porosity of sack paper was studied by Olson et al. (2001). They fractionated unbleached dried softwood kraft pulp in a pilot pressure screen with smooth hole screens. Two strategies were tested; small aperture size 0.8 and 1.0 mm with 80% of the mass in the long fraction, and large holes 1.75 mm with 40% of the mass in the long fraction. Both fractionations gave a long fraction with higher TEA and porosity performance. The big difference was in the short fractions, where the small holes gave a short fraction with high fines content and low porosity, while the larger holes gave a short fraction with properties close to those of the feed.

The fractionation and refining of eucalyptus kraft pulp was studied by El-Sharkawy et. al. (2008). Pilot screening with holes 0.6 and 1.0 mm and slots 0.12 and 0.15 mm was compared. The 0.6 mm holes gave the largest difference between the fractions, Figure 12. The difference in average fibre length between accept and reject remained when the reject rate was increased, but the mean fibre length in both accept and reject decreased with higher reject rate. This decrease in mean fibre length for the reject can be explained with that the reject contains a larger amount of the total material at higher reject rates, and thus approach the composition of the feed. The accept on the other hand will contain more material and be closer to the feed composition at low reject rate.

Figure 12 Average fibre length as a function of mass reject rate. The cross symbol refers to feed pulp, triangles refer to hole screen of 0.6 mm, circles to a hole screen of 0.1 mm, diamonds to a slot screen of 0.12 mm, squares to a slot screen of 0.15 mm, with filled symbols denoting reject and open symbols accept (El-Sharkawy, Koskenhely et al. 2008).

The accept fraction had also lower freeness, and high reject rate gave the lowest freeness. The reject fraction was refined, sheets with a mixture of eucalyptus and softwood pulp showed that the eucalyptus reject fraction gave better tensile index, fracture toughness and bulk compared to original eucalyptus pulp. This makes it possible to reduce the amount of softwood pulp.

In another study El-Sharkawy (2008)et. al. screened softwood kraft with first with hole screen 1.6 mm and then the accept was screened with slots 0.15 mm. Refining studies were performed on the reject fraction. The accept fraction consisted mainly of short thin fibre and fines. The accept fraction was added to unrefined pulp before refining, this led to lower refining energy to reach the targeted freeness and strength and increased Scott Bond.

A fine fraction produced from a screen with very small holes will contain a large proportion of fines. The high fines content contributes to increased strength properties. Sloane and Hogg (1995) investigated the effect of different fractions of pine kraft pulp. The study was performed in lab scale. First the fines were separated from the pulp with a 200 mesh screen (resembling Britt jar), i.e. 0.076 mm aperture. The remaining pulp was fractionated into a long and a short fraction with an 8-mesh screen (3 mm aperture). The three fractions; fines, short fibres and long fibres were recombined in different proportions in the base layer of a two-ply formette dynamic sheet. The combinations with increased fines content gave increased short span compression strength, burst index and tensile index, while increased short fibre content did not improve the properties. In more recent studies, microperforated screens with hole size 0.2 and 0.25 mm have been used in pilot scale. Asikainen et. al. (2010) fractionated birch pulp using both hydrocyclone and screen. For the screening, a smooth hole screen with 0.2 mm holes was used and the mass reject was 94% to remove only the fines material. This resulted in a fine fraction with high Scott bond and high air resistance.

Huber et. al. (2016) investigated the potential of fractionation to maximize the bending stiffness of a stratified sheet. BCTMP was fractionated with a microperforated screen, hole size 0.25 mm, and both mono-layer and multi-layer handsheets were made on a Formette Dynamique handsheet former. By combining a fractionation model with stratification models, the combination of volumetric reject rate and layer structure which gave the highest bending stiffness could be predicted. The optimal combination found for the tested BCTMP was volumetric reject rate 0.2 and a sheet with long fraction in inner layer and mixture of long fraction and short fraction in the outer layers.

Ferluc et. al. (2010) studied fractionation and refining of TMP. The objective was to study fractionation after the first refining stage, followed by refining of each fraction separately, and then recombining the fractions. The goal was to reduce specific refining energy and increase physical properties in TMP production. They compared different fractionation processes based on pressure screens, two with a combination of 1.2 mm holes and 0.1 mm slots, and a third combination with 0.25 mm holes and 0.1 mm slots. The 0.25 mm holes gave an efficient separation by length but did not separate according to coarseness or fibre wall thickness. Furthermore, the small aperture required low reject rates due to high thickening. However, coarser and less flexible fibres were retained by the 0.1 mm slots. The conclusion was that a screen with larger apertures should be used as a posterior stage in this application.

Kumar et. al. (2014) studied fractionation of SGW (stone ground wood) and TMP in pilot scale. They used a microperforated pressure screen with 0.25 mm holes, the fractions obtained from the pressure screen was further fractionated with hydrocyclones. For both SGW and TMP the microperforated screen was efficient for separating mainly the fines to the accept and fibres to the reject. The SGW fine fraction was fractionated with a hydrocyclone, this resulted in fibrillar fines in the accept and flaky fines in the reject. Handsheets with long fraction and different amounts of the two fines fractions showed that half the amount of fibrillar fines gave same opacity, light scattering, bulk and air permeance as the flaky fines. The fibrillar fines gave increased breaking length, smoothness and Scott bond, whereas flaky fines had no effect. For the TMP both fractions from the microperforated screen were fractionated with hydrocyclones, which resulted

in 4 fractions with different properties. This can provide more flexibility in a multi-layer sheet structure.

6

Conclusions

The most efficient and practical unit for achieving fibre length fractionation in industrial scale are pressure screens. The separation efficiency is important for the screening result. In the case of length fractionation separation efficiency means to get enough difference in fibre length distribution between the accept and reject fractions.

Several fundamental studies have been carried out on the factors influencing the screening results. Mathematical models relating pulp passage through a single aperture to pressure screen performance have been proposed. However, an average pulp passage ratio does not describe the changes in fibre length distributions. Thus, several studies have been performed regarding the effect of fibre length on passage through a screen. In these studies, the ability of pulp fibres to pass a screen has been quantified, the fibre passage ratio describes the pulp’s change in fibre length distribution caused by screening. The screen plate design and configuration have a larger influence on operation than any other controllable parameter. Smaller aperture size separates the fractions better. Furthermore, holed screen plates fractionate better than slotted, even when the slots are narrower than the hole diameter. Moreover, smooth screen plates are more efficient for fractionation, as contoured screen plates increase the fibre passage, and thus reduces the separation efficiency.

Most studies on holed screens are done on screens with larger aperture size, over 500 µm. In more recent studies, microperforated screens with hole size 0.2 and 0.25 mm have been used in pilot scale. These microperforated screens are very efficient for length fractionation, the fines are enriched in the accept fraction and the longer fibres stay in the reject fraction. The problem is that the fines fraction has a low concentration. To get an efficient production, the concentration of the fines fraction must be high. The production capacity and the runnability are important. To get a high production, a high feed concentration is beneficial, but this will at the same time increase the thickening of the reject fraction. The concentration of the reject fraction must not be too high as this deteriorate the runnability.

The running conditions for optimal length fractionation with microperforated screens to achieve an efficient production of a fines fraction was not found in this literature review. Neither was the limits in flow and concentration for the screens to work properly, without plugging found. Thus, more studies are needed.

7

Appendix – nomenclature used in

this review

β

screening index derived from Kubat &Steenberg

c

mass concentration in feed

c

mass concentration in reject

C

concentration of particles x through the aperture

c

(l)

concentration of fibres of length l passing through an aperture

C

concentration of particles x upstream of an aperture

c

(l)

concentration of fibres of length l upstream of an aperture

D

amount of debris in feed

D

amount of debris in reject

E

removal efficiency

e(l)

removal efficiency function

E

removal efficiency (debris), efficiency of debris removal

l

fibre length

L

amount of long fibres in feed

L

removal efficiency (fibres), reject of long fibres

L

amount of long fibres in reject

m

mass flow in feed

m

mass flow in reject

P(l)

fibre passage ratio for fibre length l

P

passage ratio for contaminants

P

passage ratio for pulp

P

passage ratio for particles x

Q

Nelson’s screening quotient

R

reject rate

R

mass reject rate

RTF

reject thickening factor

R

volume reject rate

t(l)

reject thickening function

v

volume flow in feed

v

volume flow in reject

References

Allison, B. and J. Olson (2000): Optimization of multiple screening stages for fibre length fractionation: two-stage case, Journal of pulp and paper science, 26(3), 113-119.

Asikainen, S., A. Fuhrmann, et al. (2010): Birch pulp fractions for fine paper and board, Nordic Pulp and Paper Research Journal, 25(11), 15-22.

Bliss, T. (1992): Screening, Tappi, p.^pp. 224-247.

Borschke, D., D. Bergfeld, et al. (1998): Screening and fractionation - Traditional technology now with new perspectives, Screening and fractionation - Traditional technology now with new perspectives, brisbane, Australia, pp. 279-286.

El-Sharkawy, K., K. Koskenhely, et al. (2008): The fractionation and refining of eucalyptus kraft pulp, Nordic Pulp and Paper Research Journal, 23(2), 172-180. El-Sharkawy, K., K. Koskenhely, et al. (2008): Tailoring softwood kraft pulp properties

by fractionation and refining, Tappi Journal, 7(11), 15-22.

Ferluc, A., R. Lanouette, et al. (2010): Optimum refining of TMP pulp by fractionation after the first refining stage, Appita, 63(4), 308-314.

Ferritsius, O., R. Ferritsius, et al. (2018): Average fibre length as a measure of the amount of long fibres in mechanical pulps–ranking of pulps may shift, Nordic Pulp & Paper Research Journal, 33(3), 468-481.

Gallager, B. J. (1999): Improving screening and cleaning efficiencies, short course, Tappi press, Atlanta.

Gavelin, G. (1995): Papp och kartong, Skogsindustrins utbildning i Markaryd AB, Markaryd.

Gooding, R. (1986): The passage of fibres through slots in pulp screening. University of British Columbia.

Gooding, R. and R. Kerekes (1989): Derivation of performance equations for solid-solid screens, The Canadian journal of chemical engineering, 67.

Gooding, R. and R. Kerekes (1989): The motion of fibres near a screen slot, Journal of pulp and paper science, 15(2), 59-62.

Gooding, R. and R. Kerekes (1992): Consistency changes caused by pulp screening, Tappi Journal, 75(11), 109-118.

Gooding, R., J. Olson, et al. (2001): Parameters for assessing fibre fractionation and their application to screen rotor effects, Parameters for assessing fibre fractionation and their application to screen rotor effects, Helsinki, Finland, pp. 407-424. Huber, P., S. Kumar, et al. (2016): Preliminary investigation on the potential of

fractionation for stratifies forming: application to maximising bending stiffness, Nordic Pulp and Paper Research Journal, 31(2).

Hyll, K. (2015): Size and shape characterisation of fines and filler - a review, Nordic Pulp and Paper Research Journal, 30(3), 466-487.

Höke, U. and S. Schabel (2009): Recycled fibre and deinking, Paper engineers association, Paperi ja puu, Helsinki, Finland.

Kubat, J. and B. Steenberg (1955): Screeining at low particles concentrations, Svensk Papperstidning, 58(9), 319-324.

Kumar, S., F. Julien Saint Amand, et al. (2014): Micro-hole pressure screening and hydrocyclone fractionation applied to mechanical pulps for fractions generation. IMPC 2014. 29th International Mechanical Pulping Conference 2014, Helsinki, Finland.

Lobben, T. H. (1977): Effekter av finstoff på styrkeegenskaper hos kjemiske massor. (Effect of fines on strength properties of chemical pulp) Norsk Skogsindustri, 77(4), 93-97.

McCarty, C. (1988): Various factors affect pressure screen operation and capacity, Pulp & paper, 62(9), 233-237.

Meltzer, F. (1998): Fractionation: Basics, development, and application, Progress in paper recycling, 7(3), 60-66.

Nelson, G. (1981): Screening quotient - an index of performance for screening evaluation. Papermakers conference 6-8 Apr 1981, Chicago. pp. 165-167.

Nichols, R. L. (1986): A new method of black liquor filtration. 72nd CPPA anuual meeting.

Olson, J. (1996): The effect of fibre length on passage through narrow apertures. University of British Columbia.

Olson, J. (2001): Fibre length fractionation caused by pulp screening, slotted screen plate, Journal of pulp and paper science, 27(8), 255-261.

Olson, J. (2003): A lecture on pressure screening.

Olson, J., B. Allison, et al. (2001): Fibre fractionation for high porosity sack kraft paper, Tappi Journal, 84(6).

Olson, J., B. Allison, et al. (1999): Fibre length fractionation caused by pulp screening, smooth-hole screen plates. 85th Annual meeting, Monteral, Que, Canada 26-29 Jan 1999. pp. B207-B212.

Olson, J. and G. Wherret (1998): A model of fibre fractionation by slotted screen apertures, Journal of pulp and paper science, 24(12), 398-403.

Olson, J. A., N. Roberts, et al. (1998): Fibre length fractionation caused by pulp screening, Journal of pulp and paper science, 24(12), 393-397.

Panula-Ontto, S. (2002): Fractionation of unbleached kraft pulp – pressure screening using smooth-hole screen baskets. In KCL (ed.), KCL reports 2625, STFI report CHEM 56.

Qazi, S. J. S., M. Mohamad, et al. (2015): Multistage fibre fractionation of softwood chemical pulp throug smooth-hole screen cylinders and its effect, Tappi Journal, 14(4).

Retulainen, E., K. Nieminen, et al. (1993): Enhancing strength properties of kraft and CTMP fibre network, Appita, 46(1), 31-38.

Scott, G. and S. Abubakr (1994): Fractionation of secondary fibre - a review, Progress in paper recycling, 3(3), 50-59.

Sloane, M. (1999): Kraft pulp processing - pressure screen fractionation, Kraft pulp processing - pressure screen fractionation, Rotorua, New Zealand, pp. 221-227. Sloane, M. and D. Hogg (1995): Sheet engineering: The potential use of fibre

fractionation to improve chemical pulp and paper quality, Sheet engineering: The potential use of fibre fractionation to improve chemical pulp and paper quality, hobart, Australia, pp. 443-454.

Sloane, M., P. Kibblewhite, et al. (2006): The pressure screen - a tool for the modern alchemist.

Ullman, U., O. Billing, et al. (1968): Fibre classification as a method of characterizing pulp, Pulp and Paper Magazine of Canada, 69(17), 69-83.

Weise, U., J. Terho, et al. (2000): Stock and water systems of the paper machine. In Paulapuro, H. (ed.), papermakin part 1, Stock preparation and wet end Fapet Oy, Helsiniki, Finland.

Ämmälä, A. (2001): Fractionation of thermomechanical pulp in pressure screening. Oulo university, Finland.

Through our international collaboration programmes with academia, industry, and the public sector, we ensure the competitiveness of the Swedish business community on an international level and contribute to a sustainable society. Our 2,200 employees support and promote all manner of innovative processes, and our roughly 100 testbeds and demonstration facilities are instrumental in developing the future-proofing of products, technologies, and services. RISE Research Institutes of Sweden is fully owned by the Swedish state.

I internationell samverkan med akademi, näringsliv och offentlig sektor bidrar vi till ett

konkurrenskraftigt näringsliv och ett hållbart samhälle. RISE 2 200 medarbetare driver och stöder alla typer av innovationsprocesser. Vi erbjuder ett 100-tal test- och demonstrationsmiljöer för framtidssäkra produkter, tekniker och tjänster. RISE Research Institutes of Sweden ägs av svenska staten.

RISE Research Institutes of Sweden Box 5604, 114 86 STOCKHOLM, Sweden Telephone: + 46-8-676 70 00

E-mail: info.innventia@ri.se, Internet: www.ri.se

RISE Bioeconomy Report: 140