Improving the strength properties of TMP

Improving the strength properties of TMP By K.B. Miles and I. Omholt Abstract: In order to develop the quality of TMP while limiting the drop in fibre length, the refining intensity was reduced in the second and third stages during pilot-scale refining trials, by reducing the rotational speed from 1,200 to 900 rpm. The tear index and, in some cases the TEA index improved. Similar bulk and air resistance were reached at higher long-fibre content, indicating that the fibres were well developed. Both spruce and pine were investigated. K.B. MILES Paprican Pointe-Claire, QC I. OMHOLT Paprican Pointe-Claire, QC iomholt@paprican.ca HE specific energy consumption associated with thermomechanical pulp- T ing is high and has long been a matter of concern. However, the main priority is still to maintain the pulp quality. In cases where serious fibre shortening occurs, or in cases where the strength properties periodically drop due to wood quality variations, the problem may become the inability to apply enough energy to obtain the required pulp quality before the target CSF value is reached. This paper describes experimental work done in a TMP pilot plant with the objective of showing how the quality can be improved by increasing the energy application through reducing the intensity in the second and third refining stage. It represents the continuation of earlier work at Paprican, investigating the potential of refining at reduced rotational speed [1,2]. Many results have been published from work done to explore the effect of varying intensity in the primary stage. In a few pilot studies, the intensity in the post primary stages was investigated specifically by increasing it over conventional levels. This generally led to reduced length weighted average fibre length [3], reduced long fibre and shive content [4,5,6] and lower CSF [4,5] at a given energy consumption. Reduced tear index at a given breaking length has been shown and it has also been observed that breaking length and burst could suffer, depending to some extent on the wood specie [4,6]. BACKGROUND Refining intensity is defined as the specific energy delivered per bar impact, and at a given consistency it is proportional to the square of the rotational speed [7]. Fibre development depends upon the material s ability to deform and absorb this energy at impact stresses below the point of fibre cutting. Through a high number of impacts, the fibre wall will become fatigued, the outer fibre layers will be loosened and partly peeled off, forming fines and ribbon-like material. This is illustrated in Fig. 1. Based upon the physical relationship equating work to the product of pressure and volume change, the stress developed during compression of any material is a function of the absorbed energy divided by the deformation. For practical pur- poses, the unstressed initial bulk is indicative of the potential for such deformation, and so the stresses acting upon impact are related to the ratio of refining intensity to bulk. As specific energy accumulates and the fibres become more developed during the refining process, the uncompressed bulk decreases and the potential for deformation during the impact is reduced. If the average refining intensity is kept on the same level in all the refining stages, impact stress will thus rise for each stage. Consequently, the risk of exceeding the fibre wall strength and ultimately reducing the fibre length, will increase. It is important to keep in mind that the absolute ability of the material to absorb energy, to deform and to withstand stress without severe fibre shortening, probably also depends on the temperature and on the inherent wood related fibre properties. The principle of this relationship can be illustrated using the results from a high-speed impact tester [8]. The apparatus delivers an impact to a pulp pad at a speed comparable to that of a refiner. By increasing the air pressure acting on the piston, impact velocity and energy are increased. As seen in Fig. 2, at a given air pressure, or impact energy, the stress upon the pulp pad increases with decreasing bulk value. The stress on the material at a given bulk may be reduced by reducing the average refining intensity according to this relationship. Even if the forces in the refiner may act on each fibre from many different directions simultaneously, our hypothesis is that the stress level would still be reduced. This would potentially lead to a higher long fibre content at equal bulk, as shown in Fig. 3, where handsheet bulk has been taken to represent the bulk of the material in the refiner. From the foregoing it is likely that when higher quality is needed, application of the additional energy could be facilitated by reducing the average refining intensity in the post primary stages of the process. This was the approach explored in the series of pilot plant trials described below. EXPERIMENTAL Unless otherwise specified, the pulps were produced from black spruce on Paprican s TMP pilot plant. The primary refiner is a pressurized Andritz 22-1CP single disc refiner, preceded by an 46 105:5 (2004) T 123 Pulp & Paper Canada

FIG. 1. Example of well-developed TMP fibres produced in the pilot plant (black spruce, low refining intensity, total specific energy 7,088 kwh/odt applied in four stages). FIG. 2. The stress on the pulp pad during high-speed impact testing increases with stepwise increments in the air pressure acting on the piston. The stress level at a given air pressure increases with increasing specific energy consumption and decreasing uncompressed bulk of the pulp. The basis weight of the pad was 300 g/m 2 and the consistency 30%. The pulp was produced from black spruce at Paprican s TMP pilot plant. FIG. 3. As the refining intensity is reduced by reducing the refiner rotational speed from 1,500 to 900 rpm, the long fibre content will be maintained at a higher level as the bulk is reduced by increasing the refining energy. FIG. 4. As the rotational speed of the refiner was reduced, the specific energy consumption at a given freeness increased. The effect became particularly evident at low CSF. inclined preheater fed, in turn, by a plug screw. Chip steaming conditions were either 250 kpa for 100 seconds or 400 kpa for 20 seconds. All pressures are given as gauge pressure. In these trials a relatively conventional level of refining intensity was maintained in the primary refiner by operating it at a rotational speed of 1,800 rpm and discharge consistency of 25-30%. The feed rate was in the range of 2.28-3.74 odt/d. Generally, about 1,000 kwh/odt were applied in this stage. Several additional levels of specific energy were applied in the secondary refiner, comprised of an atmospheric double disc Bauer 400 operated at a target discharge consistency of 25% and feed rate of 3.62 odt/d. In this unit a conventional level of refining intensity was obtained by operating at its normal rotational speed of 1,200 rpm. Low intensity refining was achieved by reducing the speed to 900 rpm. Lower intensity could also be achieved by increasing the consistency, but this involves more steam production at a given specific energy. At the long residence time resulting from the combination of a substantially lower refining intensity and high specific energy, this approach leads to increased interference between steam and pulp flow. Speed reduction is, therefore, a more appropriate choice for lowering the intensity under these circumstances. In addition, higher consistency is not an option for large commercial refiners already running at near maximum practical levels of consistency. In one of the trials described here, a relatively high level of secondary intensity was obtained by operating at 1,500 rpm. When extremely high energy levels were required, this refiner was also employed to do a third stage at the same refining intensity as the second. The plate patterns used were D17C002 and 36104 (NiHard) for the Andritz and the Bauer refiner respectively. The pulp was screened on a Somerville screen (0.15-mm slots) before testing. The tests were done according to PAPTAC standard methods. The length weighted average fibre length was measured on unscreened pulp using the Fibre Quality Analyser. RESULTS Varying Secondary Refining Intensity: Beginning with primary pulp in which the chips had been steamed for 100 seconds at 250 kpa, some over-all assessment of the effects of varying secondary refining intensity were obtained by operating the second stage refiner at speeds of 900, 1,200 and 1,500 rpm. The relationship Pulp & Paper Canada T 124 105:5 (2004) 47

FIG. 6. The long fibre content at a given energy level correlates well with the refining intensity. The error bars represent the 95% confidence interval of the regression lines used for the interpolation. FIG. 5a and 5b. The long fibre content and the average length weighted fibre length were maintained at a higher level when the rotational speed of the refiner was reduced. FIG. 7. The amount of specific energy necessary to reach a given freeness is related to the long fibre content. The trend is supported by data from mill TMP and RMP made in Paprican s pilot plant, all made from black spruce. FIG. 8. Reducing the refiner rotational speed in the second and third stages resulted in higher long fibre content at a given CSF. All pairs of runs were done with different batches of chips. between freeness and specific energy is shown in Figure 4. It is clear that the reduced intensity resulting from lower secondary rotational speed brings the pulp to a given freeness with more specific energy in it as intended. At 1,500 rpm the highest energy target had to be reduced because of plate clearance limitations. If, as intended, the stresses on the material have been reduced by using lower intensity, it should aid in the preservation of fibre length as more energy is applied. The plots of long fibre content against specific energy in Figure 5a and 5b demon- FIG. 9. Reducing the refiner rotational speed in the second and third stages resulted in higher long fibre content at a given specific energy consumption. The first-stage sample for each batch of chips, except for the second batch, is shown as the point at the lowest energy level. 48 105:5 (2004) T 125 Pulp & Paper Canada

FIG. 10. The tear index was higher at low intensity. The effect became evident in particular at high levels of burst. FIG. 11. In order to form a sheet with the same bulk with higher long fibre content, the fibres have to be flexible and conformable. This was accomplished by the use of higher specific energy. FIG. 12. The same air resistance could be reached at higher long fibre content after low-intensity refining, indicating that the fibres were well developed. FIG. 13. There was no statistically significant difference in light scattering coefficient between the two intensities. strate that this is the case. It can also be shown, as in Fig. 6, that the long fibre content at a given specific energy is highly dependent on the refining intensity used in the second stage. The refining intensity is calculated according to [9]. Another relevant observation to be made from Figs 5a and b, is that the actual difference in fibre length and long fibre content between the different intensities, will depend on which energy level is used for the comparison. As previously discussed, the fatigue work required in order to flexibilize the long fibres and improve their bonding ability, demands specific energy. This is quite evident in Fig. 7, which shows that the specific energy needed to achieve a fixed freeness is well related to the amount of long fibre contained by the pulp at that freeness. Adding the points obtained from this investigation to data from pilot plant RMP and three commercial TMP installations, produces a fairly well defined trend. High Quality, Low Freeness Pulps: The potential benefits of using lower intensity to produce high quality, low freeness thermomechanical pulps were also investigated. In this case, a higher preheating temperature was used. To do this, conventional primary pulp made from chips that had been presteamed for 20 seconds at 400 kpa was put through two additional refining stages. One run was also done at 500 kpa at a primary energy consumption of 569 kwh/odt. The second and third stages were done at a rotational speed of 1,200 rpm for the reference pulp and 900 rpm for the low-intensity pulp. Refining at 900 rpm yielded substantially more long fibre on the basis of either freeness or specific energy, as shown in Figs 8 and 9 respectively show. All pairs of runs were done with different batches of chips, which may explain the slight variation in response at 400 kpa. The change in length weighted average fibre length generally followed the same trend. The general effect for the black spruce runs seems to be that the difference in long fibre content for the two refiner speeds increased as the energy increased, confirming that the refining intensity becomes more critical as lower levels of bulk are approached. Thus, as shown in the plot of Fig. 10, the energy needed to increase burst can be applied with better retention of tear strength at lower intensity. This was particularly evident in the trial done at 500 kpa. Again, in addition to containing more long fibre, it can be seen from Figs 11 and 12 that pulps refined at low intensity can give similar bulk and air resistance as pulps refined at conventional intensity. This indicates that these fibres have been well developed. The shive content was slightly higher after low intensity refining. The differences were in the range of 0.1-0.4 percentage points Somerville shives at a CSF level of 200 ml. Compared at a given CSF there was no statistically significant difference in light scattering coefficient between the two intensities, as shown in Fig. 13. Low intensity refining gave higher TEA index at a given CSF in some cases, but Pulp & Paper Canada T 126 105:5 (2004) 49

FIG. 14. Low intensity refining gave higher TEA index at a given CSF in some cases. FIG. 15a. Low intensity refining gave a higher long fibre content at a given freeness. The first-stage sample is shown as the point at the highest freeness level. FIG. 15b. The jack pine used in this trial had low tracheid length. The R28 fraction (R14 + 14/28) gave a more complete assessment of the differences in long fibre content due to the reduction in refining intensity. FIG. 16. As observed for spruce, pine had a higher long fibre content at a given bulk after low-intensity refining. not always, as illustrated in Fig. 14. Application to Other Species: Initial studies have also been carried out on wood species that traditionally provide TMP of poor quality. In this case, the structure of the fibres themselves or other factors may limit the deformation available to control refining stress at conventional levels of refining intensity. Jack pine and southern pine (loblolly pine) are representative of this class of raw materials. After conventional primary refining with chip presteaming for 20 seconds at 400 kpa they were given second and third stage refining at either 1,200 rpm for reference, or at 900 rpm for low intensity. These presteaming conditions were chosen in order to provide a reference for the trials with black spruce. Lower refining intensity improved the long fibre content at a given specific energy for both jack pine and southern pine and, as shown in Figs 15 and 16, resulted in considerably more long fibre at a given freeness or bulk. Other properties also responded in a generally similar manner to those observed for black spruce. The jack pine used in this trial had relatively short wood tracheid length, which is a possible explanation for the low over-all level of the long fibre content in the TMP. Measured on kraft pulp cooked from the chips used in the refining trials, the length weighted average fibre length for the jack pine was 1.89 mm, compared with 2.94 mm for the southern pine. A level of 2.30-2.45 mm was measured for black spruce using the same method. According to data shown by Gullichsen and Paulapuro (10), jack pine normally has tracheids almost as long as black spruce. DISCUSSION Pulp linting is often a concern regarding the quality of TMP. Previous work has shown that linting can be reduced by increasing the specific energy consumption [11]. At a given specific energy, the refiner rotational speed and the refining consistency did not seem to affect the risk of linting measured as the linting propensity index (PLPI) [2,6,12]. Another important concern for high value paper grades is fibre rising caused by moistening of the paper surface during coating and offset printing. Fibre rising is related to reduced gloss and increased surface roughness. It may increase with increased long fibre content, and seems to involve particularly thick walled fibres. However, increased fibre bonding ability through refining will counteract this effect [13,14]. In an earlier study, both increasing the intensity in the first stage and increasing the specific energy consumption produced high long fibre quality, which reduced the changes in gloss and roughness on application of water. In the case of increased intensity, the fraction of long fibres was however somewhat reduced [15]. The advantage of lower post primary intensity is its ability to extend greatly the energy input when needed. CONCLUSION Lower rotational speed was used to reduce the refining intensity in the post primary stages of TMP. This resulted in: More long fibres compared at either a given freeness, at a given specific energy, or at a given bulk; and 50 105:5 (2004) T 127 Pulp & Paper Canada

Better tear index at a given burst and in some cases improved TEA at a given CSF. It was also observed that the specific energy required to reach a given freeness, for black spruce, is related to the amount of long fibre contained by the pulp at that freeness. This approach might be useful when more specific energy consumption is needed, for example when refining low quality wood, when attempting to improve the strength properties at a given CSF or when aiming for low CSF levels while preserving a high fibre length. The implications could, for example, be increased use of less desirable wood species in newsprint or reduced content of kraft pulp in higher quality paper. ACKNOWLEDGEMENTS The contribution of Michael Stacey and Derek Dranfield in carrying out the trials, organizing the results and preparing diagrams is gratefully acknowledged. The authors also wish to thank Reza Amiri for useful discussions. LITERATURE 1. US Patent no. 6,336,602 B1 (2002). 2. MILES, K.B., MAY, W.D., KARNIS, A., Refining intensity, energy consumption, and pulp quality in two-stage chip refining, Tappi J. 74 (3): 221-230 (1991). 3. KURE, K.-A., DAHLQUIST, G., HELLE, T., Morphological characteristics of TMP fibres as affected by the rotational speed of the refiner, Nord. Pulp Pap. Res. J. 14 (2): 105-110 (1999). 4. MILES, K.B., KARNIS, A., The response of mechanical and chemical pulps to refining, Tappi J., 74 (1): 157-164 (1991). 5. CHAPMAN, D.L.T., ALLAN, R.S., Some basic considerations in groundwood rejects refining, Pulp Paper Can. 71 (8): 67-72 (1970). 6. SINKEY, J., Mechanical pulp properties and distributions as affected by refining conditions, International symposium on: Fundamental concepts of refining, Institute of paper chemistry, Appleton, WI, Sept. 16-18 (1980). 7. MILES, K.B., MAY, W.D., The flow of pulp in chip refiners, J. Pulp Pap. Sci. 16 (2): J63-J72 (1990). 8. AMIRI, R., HOFMANN, R., Dynamic compressibility of papermaking pulps, Paperi ja Puu, 85 (2): 100-106 (2003). 9. MILES, K.B., A simplified method for calculating the residence time and the refining intensity in a chip refiner, Paperi ja Puu, 73 (9): 852-857 (1991). 10. GULLICHSEN, J., PAULAPURO, H., ed. Papermaking Science and Technology. Book 6A. Chemical Pulping. Helsinki: Fapet Oy (1999). 11. WOOD, J.R., KARNIS, A., Towards a lint-free newsprint sheet, Paperi ja Puu, 59 (10): 660-674 (1977). 12. WOOD, J.R., KARNIS, A., Linting propensity of mechanical pulps, Pulp Paper Can. 93 (7): T191-T198 (1992). 13. HALLAMAA, T., HEIKKURINEN, A., Effect of fibre properties on sheet surface roughening, Proc., International Mechanical Pulping Conference, SPCI, Stockholm, Sweden, 361-363 (1997). 14. HOC, M., Fibre rising in papers containing mechanical pulp, Tappi J. 72 (4): 165-169 (1989). 15. AMIRI, R., STEPIEN,G., WOOD, J.R., TMP process conditions to produce pulps for mechanical printing papers - Effects on surface properties, Proc. 1996 Pulping Conference, Atlanta: TAPPI Press, Book 1: 265-278 (1996). Résumé: Afin d améliorer la qualité de la PTM tout en limitant la réduction de la longueur de la fibre, nous avons réduit l intensité du raffinage aux deuxième et troisième stades lors d essais pilotes sur le raffinage, en diminuant la vitesse de rotation de 1200 à 900 trs/min. L indice de déchirement et, dans certains cas, l indice TEA se sont aussi améliorés. Une résistance au passage de l air et un bouffant similaires ont été atteints à une teneur plus élevée en fibres longues, ce qui indique que les fibres étaient bien développées. Les essais ont porté sur l épinette et le pin. Reference: MILES, K.B., OMHOLT, I. Improving the strength properties of TMP. Pulp & Paper Canada 105(5): T123-128 (May, 2004). Paper presented at the 2003 Intl. Mechanical Pulping Conference in Québec, QC, on June 2 to 5, 2003. Not to be reproduced without permission of PAP- TAC. Manuscript received on September 26, 2003. Revised manuscript approved for publication by the Review Panel on November 21, 2003. Keywords: THERMOMECHANICAL PULPING, MECHANICAL PROPERTIES, BEATING DEGREE, ROTATION, VELOCITY. Pulp & Paper Canada T 128 105:5 (2004) 51