Bagasse Pulp

5/7/2018 Bagasse Pulp - slidepdf.com

http://slidepdf.com/reader/full/bagasse-pulp 1/10

COVER SHEET

Rainey, Thomas and Brown, Richard and Martinez, Mark and Doherty, Bill (2006) The Use of

CFD to simulate the Behaviour of Bagasse Pulp Suspensions during the Dewatering Process.

In Proceedings 60th Appita Annual Conference and Exhibition, Australia, Victoria, Melbourne.

Accessed from http://eprints.qut.edu.au

Copyright 2006 the authors.



The use of CFD to simulate the

behaviour of bagasse pulp

suspensions during the dewatering

process

By T.J. Rainey1, R. Brown2, M.Martinez3 & W. Doherty4

1 PhD student, Queensland Universityof Technology2 Senior Lecturer, QueenslandUniversity of Technology3 University of British Columbia4 Principal Research Fellow,Queensland University of Technology

Abstract

Although bagasse is used as a fibresource for paper manufacturethroughout the world, no bagasse pulpindustry exists in Australia. Thereasons for this include cost, given theenormous economics of scale achievedby the hardwood pulp mills, theremoteness of cane farms relative tohardwood pulp mills, and processingand environmental issues. Theprocessing issue arises because of therelatively poor drainage properties of bagasse pulp compared to wood pulp.As paper machines are specificallydesigned to process wood pulp, theyare not very effective to dewaterbagasse pulp suspensions at aneffective rate. This reduces factoryproductivity and increases energy

consumption, thereby impactingnegatively on the overall operatingcosts of a pulp mill.

This paper describes an approach thatwould utilise surface chemistry andcomputational fluid dynamics (CFD)to improve our understanding of bagasse pulp dewatering process. Thiswould lead to developing a process toenhance the physicochemical

properties of bagasse fibres and hencethe dewatering process.

Background

The Australian sugar industry is

amongst the most competitive in theworld. However, it has remainedlargely a single commodity producer.Until recently the industry hasstruggled from low world sugar prices,partly due to cheap Brazilian sugarflooding the market. Also, in manycountries, such as the USA and inEurope, the local sugar industries areheavily protected. This has meant thatlocal consumers are paying many times

more than the world market price. In1990-1998, the sugar price averagedover US 10 c/lb, but has averagedaround US 9 c/lb in 2005. In order toreduce the Australian sugar industry’svulnerability to the world sugar price,the industry is looking atdiversification through the use of thewhole sugarcane. Recently the sugarindustry has focussed its attention onburning bagasse for cogeneration of electricity as an additional incomestream. The value of bagasse as asource of fuel for cogeneration,including renewable energy credits, isaround $40-$60 per tonne on a drybasis. Alternatively, bagasse can befurther processed to manufacturepaper, board and composite materials,increasing its opportunity cost to becomparable with hardwood chips.

Sugarcane is a tropical grass that growsin warm climates and requiresabundant rain or irrigation each year.Sugarcane consists of a pith sectionsurrounded by a rind. Sugarcaneconsists of 11-16% (dry basis) sucrose,which is mainly contained in the pithsection. This is mainly dependant onthe specific cane species and theweather conditions. After the cane is

harvested, it is shredded and milled toextract the juice from the cane which



contains the sugar. The juice is thentransported for further processing intoraw sugar. The fibrous matter that isleft behind after crushing the cane toextract sucrose is known as “bagasse”.

Bagasse consists of fibre from both therind portion of the sugarcane and alsothe pith. The individual fibre length inthe rind are relatively short, around 0.8mm, making them suitable forapplications where papers are requiredto be smooth or soft, such as writingpapers and tissues. Individual fibres inthe pith are extremely small (<< 1 mm)and must be substantially removed

prior to pulping to improve thedrainage properties of the pulp.

Figure 1 Sugarcane plant andstalk section [1]

Table 1 gives a typical breakdown of the fibre composition in bagasse

(protein and ash free basis). Table 2gives a breakdown of bagasse in termsof the pith and rind fractions.

Table 1 Composition of wholebagasse (protein and ash free basis)[2].

Cellulose 55%Pentosans 25%Lignin 15%

Acetic acid 6%

Table 2 Composition of the pithand rind components of bagasse [2]

Pith RindAsh 1.68 1.64Fat and wax 0.41 0.98Protein 1.94 2.19Cellulose 49.0 51.09Pentosans 32.04 26.93Lignin 14.93 17.17

Overcoming the pith problem

As only the fibre from the rind portionof the sugarcane is useful forpapermaking, the pith is alwaysremoved prior to pulping to improvethe drainage of the bagasse pulp, toimprove the paper quality and to

reduce chemical consumption.Depithing is normally undertaken intwo stages, (i) moist depithing at 50%moisture undertaken in hammer millsand screened to remove the pith; and(ii) wet depithing at 90% moisturewhich is processed in centrifugalseparators. Normally about 30-40% of the whole bagasse is removed as pithbefore being dewatered and returned tothe sugar mill boilers.



Despite the depithing operations priorto pulping, there is inevitably asignificant quantity of very short fibreswhen bagasse pulps are processed atthe former. Since bagasse pulps are

suitable for the manufacture of lowquality writing papers and tissues dueto their short fibre length, they must besuspended in very dilute solutions andrapidly dewatered. The presence of any very small fibres from the pith canblock holes between the other fibres,reducing drainage. This can result in:(i) a reduction in processing speed of the paper machine; (ii) have an impacton the quality of the sheet produced;

and (iii) create problems indownstream processing.

Although significant work has beendone to improve the drainageproperties of bagasse pulps, the work has mainly focussed on two areas:improving the depithing operation atthe pulp mill; and the development of chemical additives by commercialchemical companies for processingbagasse pulps at the paper machine. Alarge volume of work has been done indeveloping and comparing chemicaladditives for various wood gradesproduced.

More recently the development of microparticle systems have assisted thedrainage properties of bagasse pulps.Duffy [3], from Nalco Chemicals,

presents a succinct description of theaction of silica microparticle systems.The high molecular weight cationicpolymers attach themselves to thenegative surface of the fibres. Thenegative silica microparticle thenattaches separate polymer chains,which are themselves attached todifferent fibres (see figure 10). Theimportant difference betweenconventional polymer flocculation and

microparticle systems are that underconditions of high shear, such as those

in a paper machine headbox, the bondsformed with polymers are destroyed,but microparticle systems have theability to reflocculate the fibres afterbeing subjected to high shear.

However, the use of chemical additivesis only part of the solution.

Figure 1 Mechanism of silica

microparticles [3]

Understanding the fundamentaldewatering behaviour of bagasse pulpsis important in making furtherimprovements for improving thedrainage properties of bagasse pulps.Modelling the behaviour of bagassepulps can, for example, be used as thefirst step in developing specialisedequipment, such as formingarrangement and headbox design, thatcan be used for the manufacture of bagasse paper products.

Computational Fluid Dynamics

model development

Computational Fluid Dynamics (CFD)

modelling is useful for investigation of complex fluid dynamics problems andpermits design optimisation to beachieved in a wide range of engineering applications. In this case,the CFD model could be used fordeveloping optimised equipment (suchas forming fabrics, headbox design andgeneral former configuration) that willimprove the dewatering of bagassepulps.



CFD models are a solution scheme forthe Navier-Stokes equations whichdescribe the motion of fluids. Inaddition heat transfer, chemicalreactions and the resultant interactions

with physical geometries can beincluded. There is also scope to modelmultiphase flow. The equations arecoupled and solved simultaneously inan iterative process to arrive at a finalsolution. The Navier Stokes andcontinuity equations describe the fluidmotion, and other physical processequations can be linked until anadequate description of the process isfound. The number and form of the

equations used are different for eachcase, since every flow problem hasdifferent parameters of importance.CFD techniques are developed in themost general way possible so that theycan be applied to almost anyapplication, so long as the governingfluid behaviour can be describedmathematically. Industrial applicationsof CFD vary from the design of aircraftthrough to nuclear powered boilers,heart valves, pumps, turbines, mixers,burners and fluidised beds. CFDmodels allow the user to visualise theoperation of equipment in ways thatcannot be matched by real worldmeasurements [4].

A significant advantage of computer-based simulation over pilot testing isthe reduction in the time and capital

cost required to investigate designchanges. However, the biggestadvantage of computer simulations isthat they allow the user to investigatethe process in ways not otherwisepossible, and do this without the highcost and risk to product quality orsafety that exists with conventionalexperimental methods. This allows thetraditional processes of technologydevelopment, through pilot plant

testing and incremental changes at full

scale, to be fast tracked or evenbypassed completely.

At this point, it is appropriate to clarifythe difference in terminology used in

this paper. CFD now refers solely tocommercial CFD packages andnumerical modelling refers to theunderlying mathematical equationswhich characterise the flow behaviour.

CFD is available throughmanufacturers such as CFX (Ansys)and FLUENT as commercial packagesfor both commercial and academicpurposes. These commercial programs

are very powerful as they can easilyadapt to different geometries, thepresence of particles, and changingphysical parameters, such astemperature and pressure. Thecommercial packages have propertylibraries to determine a broad range of physical properties of many commonfluids, such as water. However, thesecommercial licenses can be quiteexpensive. The use of these packagesfor academic purposes is becomingmore common as academic licenses areonly a fraction of the cost of a fullcommercial license.

In commercial CFD packages, such asCFX, the desired geometry is firstmodelled, then the geometry is brokendown into typically thousands of individual elements (depending on the

desired resolution and computingpower available) to create a “mesh”.The mesh is imported into a pre-processing package where the physicalconstraints are applied to the geometry,such as boundary conditions and fluidproperties. The CFD solver can thencrunch the numerical algorithms andsome time later, the results can bevisualised in post-processingoperations.



In the past, commercial CFD packageshave not been particularly user-friendly, even for experienced users.In the case of Ansys CFX themodelling, meshing, pre- and post-

processing operations were distributedwith quite different graphical userinterfaces (GUI’s). Due toconsolidation in the industry over thepast few years, Ansys have been ableto provide one package with a moreuniform GUI. It is now much easier topick up CFX, and further integration of the Ansys software is expected.However, it takes many years to notonly master the intricacies of the GUI

in commercial CFD packages, but anin-depth understanding of the availablephysics models (e.g. turbulencemodels) takes quite some time.

Commercial CFD packages areexcellent at predicting the fluidmechanics behaviour through vessels.CFD packages predict Eulerian flow,where the fluid and particles are treatedas a continuum, and also provideparticle tracking. Commercial CFDpackages have provision for includingthe shape factor, drag coefficients anddensity of the particles, for examplefibres. CFD can be used to investigatemacro scale problems involving fibres,for example MD basis weight variationduring paper formation. However,commercial packages do not take intoaccount issues unique to paper

manufacture such as particle chargeand particle to particle interactionswhere the particles may be fibres oradditives. For most papermakingnumerical models, it is adequate toconsider the hydrodynamic forcesonly, as these are usually adequate tocharacterise flocculation [5]. In thiscase, the problem is drainage of waterthrough a mat of bagasse fibres. It iswell established that polymer and

microparticle systems can improve thedrainage properties of pulps containing

a large quantity of ionic trash [see forexample 3, 6, 7-13]. Thus commercialCFD packages only provide aframework for the modelling. It isnecessary to develop a suitable

numerical model specialised forbagasse pulp.

Having discussed commercial CFDpackages, the alternative method of modelling, numerical modelling, isdiscussed in more detail. Numericalmodelling refers to the underlyingmathematical equations whichcharacterise the flow behaviour.

There exists a significant volume of work on the numerical modelling of various aspects of papermaking; theUniversity of British Columbia andSweden’s Royal Institute of Technology (KTH) including 1-D [14],2-D [15] and 3-D [16] models.Various components of the papermachine have been numericallymodelled, for example, the headbox,[14] and the effect of jet impingementon one fabric [17]. Both types of Twin-wire formers have been thesubject of numerical investigations todetermine the drainage rate. Forexample Holmqvist [18] providesdetail on the numerical analysis of Blade Formers whilst Martinez [19]characterised the dewatering rate of Roll Formers.

Once these numerical models havebeen derived mathematically, theymust be validated by experimentalinvestigation for a single geometricconfiguration before they can beapplied to other geometricconfigurations. The design of theexperimental equipment for measuringthe velocity profile can be moredifficult and expensive than thenumerical modelling itself.



The University of British Columbiadeveloped a three dimensional fibre-tracking model in 1997 [20]. When theexperimental work was undertaken toconfirm the predictions of the model, it

was found that there was a lack of agreement between the experimentalresults and the model. Salcudeanconcluded that the next phase of work was to include the effects of turbulentdispersion [21]. Despite the lack of agreement between the experimentalinvestigation and the modelpredictions, the model was simplifiedand adapted for hydrocyclones, screensand headboxes where it has been more

successful. It is not known whetherthe turbulence effects have since beenincorporated into the complex model.

These numerical models are excellentfor predicting flow behaviour resultingfrom one particular set of flowconditions on a micro scale. Thesemodels provide superior predictionsfor flow behaviour to the genericmodels available in commercial CFDpackages. However, these numericalmodels generally do not have theflexibility of a commercial CFDpackage, particularly in terms of geometry.

The ultimate objective is the marriageof a commercial CFD package (whichhas geometric flexibility) with anumerical model specific for bagasse

(that handles the numerical modellingof bagasse extremely well). Thenumerical models can be programmedin Fortran, for example, which acts as asubroutine for the commercial CFDpackage.

It is proposed to develop a numericalmodel of bagasse pulp suspensions bysimulating the forming process in thelaboratory using pulsed Doppler

anemometry equipment at theUniversity of British Columbia. A

CFD model will assist to improve theunderstanding of bagasse pulpbehaviour and possibly permit designof equipment that are used to processbagasse pulps, such as forming fabrics

and headboxes. Numerical and CFDmodels have been developed at theUniversity of British Columbia, forexample, for various wood pulpsuspensions and forming equipmentsuch as Fourdrinier and Twin WireFormers.

Simulating the forming process in

the laboratory

Firstly, it is necessary to adapt anexisting numerical model to develop asimplified model, excluding chargeand inertial terms, to characterise thebehaviour of bagasse pulps withoutcommercial additives. In order tovalidate any numerical model forsimulating the behaviour of bagassepulps, a reasonably accurate andaffordable method of simulating theindustrial forming process must beachieved in the laboratory.

Significant research into the drainageproperties of all kinds of pulps haspreviously been conducted in thelaboratory. The value of freeness(CSF) in predicting behaviour on thepaper machine has long beencontroversial. Research organisations

wanted to develop laboratoryequipment that better represents thephysics occurring at the papermachine. For example the DynamicDrainage Jar (Unbehend, 1977) and theWater Release Analyser (Britt andUnbehend, 1980) were developed tomore accurately simulate the drainageof pulps. More recently, work by Xuand Parker [22] and Helmer [23, 24]have focussed on designing more

sophisticated laboratory formers that



closely resembles the industrialforming process.

Once the critical fluid mechanicsproperties of bagasse pulp suspensionswhich affect its drainage properties areidentified, suitable experimentalequipment will be constructed to

numerically model and optimise thedrainage behaviour of bagasse pulpsuspensions. The equipment mostlikely to be used in this project isoutlined by Kataja and Hirsila [25] and

shown in figure 2.

Figure 2 Ultrasound anemometry for measuring filtration of fibre

suspension [25]

The unit consists of a sealed tank witha riser tube. Inside the riser tube is awire and support grid. The suspensionof fibre is allowed to drain through thewire. The fibre is retained on the wireand forms a fibre mat. The water levelinside the riser is measured with an

ultrasonic surface detector. Thevertical velocity of the fibres ismeasured through the wire by fourpulsed ultrasound Doppleranemometers. The signal sent from thesurface detector and the anemometersare processed and the data is captured.The probes can measure the verticalvelocity of the fibres up to 70 mmabove the wire. The water level in theriser tube is adjusted by valves V1, V2

and V3. Assuming valves V1 and V3have adequate accuracy and response

time, the programmable logic could bemodified to allow a time-varyingpressure pulse, simulating the effect of foils in Fourdrinier forming.



Conclusion

Increasing the value of bagasse isimportant for the long term

profitability of the Australian sugarindustry. Making it attractive as afeedstock to the paper industryincreases its opportunity cost. Tomake bagasse pulps more attractive, anumerical model for bagasse pulps isdesired in order to make furthersignificant advances in improving thedrainage performance of bagasse pulpsover and above that achievable withpolymer and microparticle systems.

Once the numerical model isdeveloped for simulating the criticalflow characteristics involved indewatering bagasse pulp, it must beverified experimentally. Anexperimental rig is proposed forevaluation of the numerical model.After the model is validated, it can thenbe incorporated into a commercialCFD package for further investigative

work, for example, to producespecialised equipment for thedewatering of bagasse pulps.

Acknowledgements

I would like to thank the SugarResearch Development Corporationand Queensland University of Technology (Sugar Research and

Innovation) for their financialassistance.

References

1. Miller, J.D. and R.S. Lentini,Sugarcane Botany: A Brief

View. 1993, University of Florida.

2. Deerr, N., Cane Sugar. A

textbook on the agriculture of

the sugar cane, themanufacture of cane sugar &

the analysis of sugar house

products. 1911, London:Norman Rodger.

3. Duffy, B. A microparticle

retention approach to

papermaking. in TappiPapermakers Conference

Proceedings. 1993.4. Rainey, T., S. Pennisi, and J.

Joyce, Applying Computational

Fluid Dynamics modelling to

optimise environmental

performance at minimum cost. Proceedings of the 2004 Appitaconference, 2004.

5. Parker, J.D., The sheet-forming

process, in A project of theFluid Mechanics Committee.1972, Tappi: Atlanta.

6. Abril, A. and O. Alvarez,Comparative study of retention

agents in the dynamic drainage

jar. Revista sobre losDerivados de la Cana deAzucar, 1984.

7. Abril, A. and O. Alvarez, Influence of pH, degree of

turbulence and dose on the

effect of a retention agent on a

model pulping system. Revistasobre los Derivados de la Canade Azucar, 1984.

8. Abril, A., et al., Use of

retention agents in the

production of papers with a

high content of bagasse pulp. Revista sobre los Derivados de

la Cana de Azucar, 1984.9. Carr, D. and M. Tokarz, Silica

Nanoparticles Build Bridge to

Better Retention, Sheet

Formation. Pulp & Paper,2006. 80(2): p. 34.

10. Carr, D.S., Interaction of

Nanoparticle Chemistry, Starch

Yields Better Groundwood

Grade Runnability. Pulp &Paper, 2004. 78(2): p. 34.

11. Carr, D.S., NanoparticleControl of Wet End Extractives



Yields Better Runnability. Pulp& Paper, 2004. 78(3): p. 44.

12. Carr, D.S., New Anionic Trash

Catchers Boost

Retention/Drainage, Sizing

Efficiency. Pulp & Paper, 2005.79(2): p. 50.

13. Ledda, J., et al., Ciba

TELIOFORM system - a new

multi component

organic/inorganic system. Kami Pa Gikyoshi, 2005.59(8): p. 1181-1187.

14. Olson, J.A., et al., Modeling a

turbulent fibre suspension

flowing in a planar

contraction: The one-dimensional headbox. International Journal of Multiphase Flow, 2004. 30: p.51-66.

15. Zahrai, S., F.H. Bark, and D.M.Martinez, A numberical study

of cake formation in 2-D cross-

flow filtration. Journal of Pulpand Paper Science, 1998. 24(9):p. 281-285.

16. Dong, S., et al., Concentration

of pulp fibers in 3D turbulent

channel flow. InternationalJournal of Multiphase Flow,2003. 29: p. 1-21.

17. Dalpke, B., S.I. Green, and R.J.Kerekes, Modelling of jet

impringement in twin-wire

papermachines: Impingement

on one fabric. Tappi Journal,

2001. 84(2).18. Holmqvist, C., Mechanical

modelling of blade forming and

drainage of flocculated

suspensions, in Department of

Mechanics. 2005, Royal

Institute of Technology:Stockholm.

19. Martinez, D.M., Characterising

the dewatering rate in roll gap

formers. Journal of Pulp and

Paper Science, 1998. 24(1): p.7-13.

20. Wherret, G., I. Gartshore, andM. Salcudean. Numerical

model of fibre motion in shear .in 1997 ASME Symposium on

separated and complex flows.1997. Vancouver, Canada.

21. Salcudean, M., I. Gartshore,and P. Gorog. Computational

Fluid Dynamics (CFD): A tool

for simulating pulp and paper processes. in Tappi

Engineering/Finishing &

Converting Conference &

Trade Fair . 2001. San AntonioMarriott Rivercenter, SanAntonio, TX.

22. Xu, L. and I. Parker, Simulating

the forming process with the

Moving Belt Drainage Former. Appita Journal, 2000. 53:4: p.282-286.

23. Helmer, R.J.N., et al.,Optimising simulated

commercial paper for pulp

quality analysis. AppitaJournal, 2001. 5: p. 289-296.

24. Helmer, R.J.N., Laboratory

simulation of paper forming, inChemical Engineering. 2000,University of Melbourne:

Melbourne.25. Kataja, M. and P. Hirsila.

Application of ultrasound

anemometry for measuring

filtration of fibre suspensions.in 12th Fundamental Research

Symposium. 2001. Oxford.

Bagasse Pulp

Documents

Transcript of Bagasse Pulp