UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of screw feeder compression and its effect on wood-chip cellulose accessibility Villalba Chehab, Miguel Esteban 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2019_february_villalba_miguel.pdf [ 3.16MB ]
Metadata
JSON: 24-1.0375847.json
JSON-LD: 24-1.0375847-ld.json
RDF/XML (Pretty): 24-1.0375847-rdf.xml
RDF/JSON: 24-1.0375847-rdf.json
Turtle: 24-1.0375847-turtle.txt
N-Triples: 24-1.0375847-rdf-ntriples.txt
Original Record: 24-1.0375847-source.json
Full Text
24-1.0375847-fulltext.txt
Citation
24-1.0375847.ris

Full Text

CHARACTERIZATION OF SCREW FEEDER COMPRESSION AND ITS EFFECT ON WOOD-CHIP CELLULOSE ACCESSIBILITY  by  Miguel Esteban Villalba Chehab  B.A., The Pennsylvania State University, 2016  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Mechanical Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   December 2018 © Miguel Esteban Villalba Chehab, 2018 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled:  Characterization of Screw Feeder Compression and its Effect on Wood-chip Cellulose Accessibility  submitted by Miguel Esteban Villalba Chehab in partial fulfillment of the requirements for the degree of Master of Applied Science in Mechanical Engineering  Examining Committee: Dr. James Olson, Mechanical Engineering Supervisor  Dr. Mark Martinez, Chemical and Mechanical Engineering Supervisory Committee Member  Dr. Rodger Beatson, Chemical Engineering Supervisory Committee Member  Additional Examiner   Additional Supervisory Committee Members:  Supervisory Committee Member  Supervisory Committee Member iii  Abstract  Wood-chip compression and enzyme impregnation are used as pre-treatment prior to refining to reduce energy consumption and improve pulp quality. This work aims at characterizing the effect of compression ratios on cellulose accessibility to the enzyme impregnation. This objective was achieved by quantifying the amount of glucose produced in the enzyme treatment of compressed wood-chips with a high-performance liquid chromatograph. A laboratory compressor and a controlled uniaxial load set-up were implemented to test different compression ratios and compression rates. The stress and strain data from the wood-chips compression was applied to torque equations to predict the power consumption of a screw feeder.  The trial results showed that cellulose accessibility increased with compression ratio. High compression ratios were required to improve the hydrolysis rate significantly. Compression rate had no apparent effect on the cellulose accessibility. The improved accessibility is due to changes in morphology of the wood-chips as well as removal of the extractives. Microscopy imaging of wood-chip cross-sections showed buckling and fractures of the cell walls. Fractures lead to improved enzyme penetration and the reduction of chips size lead to higher available surface area.  Using the compression data from the trials, screw feeder performance could be determined for different compression ratios. The calculations predicted that compressing at around 5:1 compression ratio achieved a good balance between power consumption, screw feeder capacity and improvement in cellulose accessibility.   iv  Lay Summary  Mechanical compression is used in the pulp and paper industry as a wood-chip pre-treatment via a screw feeder device. Compression opens the wood-chip structure which makes it more amenable for refining while improving its chemical uptake. Enzymes are used in this industry to further reduce energy consumption in the refining stages.   The purpose of this research is to determine the effect of different compression conditions on the enzyme uptake and wood-chip structure. Additionally, this study aims at understanding the relationship of compression and power consumption in order to optimally operate screw feeders.  Enzyme uptake was quantified by measuring the sugars released during enzyme treatment. Experimental results showed that increasing the degree of compression improved the sugar yields systematically. Data from experimental compression was applied to a torque prediction model to determine the optimal screw feeder operation for effective compression at lower energy consumption.  v  Preface This dissertation is original, unpublished and an independent work by the author, Miguel Esteban Villalba Chehab. Although the main ideas behind this research are his own, this project would be impossible without the continuous dialogue that took place between the author and his supervisor, Dr. James Olson. Compression with the lab compressor and operation of the MTS load frame in this thesis were not conducted by the author. The lab compressor operation was carried out by personnel from FPInnovations in Pointe-Claire. The operation of the MTS load frame was carried out by the personnel of the Materials Department of UBC. The machining of the compressor and compression vessel was accomplished by a machinist from the mechanical engineering machine shop at UBC.  Chapters 1 presents an introduction to the research and its objectives. Chapter 2 contains most of the literature review. Chapter 3 describes experimental procedure to quantify cellulose accessibility which was designed for this study by the author. Additionally, this procedure is tested with different substrates to determine the factors that affect enzyme hydrolysis. Chapter 4 involves the results of two compression tests to study the effect of compression ratios and rates on the cellulose accessibility to enzyme treatment and wood-chip morphology. This includes performing enzyme treatments of compressed wood-chips, quantifying sugars released in hydrolysis reactions and the quantification of extractive content. All these tasks were performed by the author with initial mentorship of Dr. Heather Trajano. Chapter 4 also includes microscopy work which was performed by the author. Chapter 5 presents analytical work of screw feeder compression developed by the author. Finally, chapter 6 includes analysis of the research data and conclusions.  vi  Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................ ix List of Figures .................................................................................................................................x List of Symbols ........................................................................................................................... xiii List of Abbreviations ...................................................................................................................xv Acknowledgements .................................................................................................................... xvi Dedication .................................................................................................................................. xvii Chapter 1: Introduction ................................................................................................................1 1.1 Thermomechanical Pulping ............................................................................................ 1 1.2 Research Objectives ........................................................................................................ 4 1.3 Thesis Overview ............................................................................................................. 5 Chapter 2: Background .................................................................................................................7 2.1 Wood Structure ............................................................................................................... 7 2.2 Wood Chemistry ........................................................................................................... 10 2.3 Compressive Pre-treatment ........................................................................................... 11 2.3.1 Effect of Compression on Morphology .................................................................... 12 2.3.2 Effect of Compression on Liquid Uptake ................................................................. 13 2.3.3 Effect of Compression on Refining Energy Reduction ............................................ 14 2.3.4 Effect of Compression on Extractives and Pulp Properties ...................................... 14 2.4 Enzyme Hydrolysis ....................................................................................................... 15 2.4.1 Cellulase Enzyme...................................................................................................... 15 2.4.2 Factors Affecting Enzymatic Hydrolysis .................................................................. 16 2.5 Characterization of Cellulose Accessibility .................................................................. 18 2.6 Study Approach ............................................................................................................ 21 vii  Chapter 3: Characterizing Cellulose Accessibility ...................................................................23 3.1 Materials and Methodology .......................................................................................... 23 3.1.1 Compositional Analysis ............................................................................................ 23 3.1.2 Quantifying Sugars in Liquid Fraction ..................................................................... 25 3.2 Method Validation: Enzyme treatment of substrates .................................................... 27 3.3 Results ........................................................................................................................... 30 3.3.1 Compositional Analysis ............................................................................................ 30 3.3.2 Enzyme Hydrolysis of Wood .................................................................................... 30 3.3.3 Comparison between Substrates ............................................................................... 32 3.4 Conclusions ................................................................................................................... 35 Chapter 4: Evaluation of Wood-Chip Compression.................................................................36 4.1 Experimental Methodology .......................................................................................... 36 4.1.1 Enzyme Treatment .................................................................................................... 37 4.1.2 Extractive Measurements .......................................................................................... 38 4.1.3 Microscopy ............................................................................................................... 39 4.2 Lab Compressor Trial ................................................................................................... 40 4.2.1 Trial Conditions ........................................................................................................ 41 4.2.2 Effect of Compression on Cellulose Accessibility ................................................... 43 4.2.3 Effect of Compression on Extractives ...................................................................... 46 4.3 MTS Compression Trial ............................................................................................... 47 4.3.1 MTS Trial Conditions ............................................................................................... 48 4.3.2 Effect of Compression Ratio and Compression Rate on Load ................................. 49 4.3.3 Effect of Compression Conditions on Cellulose Accessibility ................................. 51 4.3.4 Effect of Compression on Extractive Removal ........................................................ 54 4.3.5 Effect of Compression on Wood-Chip Morphology ................................................ 56 4.4 Conclusion .................................................................................................................... 59 Chapter 5: Screw Feeder Analysis .............................................................................................60 5.1 Screw Feeder Torque Equations ................................................................................... 60 5.1.1 Formulation ............................................................................................................... 60 5.2 Methodology ................................................................................................................. 63 viii  5.2.1 Screw Configuration ................................................................................................. 63 5.2.2 Assumptions .............................................................................................................. 64 5.2.3 Experimental Data .................................................................................................... 65 5.2.4 Applying the Screw Feeder Equations ...................................................................... 67 5.3 Results ........................................................................................................................... 68 5.3.1 Stress and Torque Predictions ................................................................................... 68 5.3.2 Performance of Screw feeder .................................................................................... 69 5.4 Conclusions ................................................................................................................... 71 Chapter 6: Conclusions and Future Work ................................................................................73 6.1 Characterization of Cellulose Accessibility .................................................................. 73 6.2 The Effect of Compression of Wood-chips on Cellulose Accessibility ....................... 73 6.3 Screw Feeder Characterization ..................................................................................... 74 6.4 Suggestions for Future Work ........................................................................................ 75 Bibliography .................................................................................................................................76 Appendices ....................................................................................................................................81 Appendix A : MTS Compression Design Drawings ................................................................ 81 A.1 Design and Analysis ................................................................................................. 81 A.2 Compression Chamber Drawing ............................................................................... 83 Appendix B : MTS Compression Test ...................................................................................... 84 B.1 Test 1 – Constant Compression time of 5 seconds ................................................... 84 B.2 Test 2 – Variable compression time .......................................................................... 85  ix  List of Tables   Table 2-1: Common methods used for the prediction of cellulose accessibility. ......................... 20 Table 3-1: Enzyme Treatment Conditions .................................................................................... 28 Table 3-2: Chemical compositions of pulps [51] and SPF wood chips as determined in this work........................................................................................................................................................ 30 Table 3-3: Initial Hydrolysis Rates of each substrates at high and low dosages. ......................... 34 Table 4-1. Lab Compressor Trial Conditions ............................................................................... 42 Table 4-2. Enzyme Treatment conditions for the lab compressor trial. ........................................ 42 Table 4-3. MTS compression conditions ...................................................................................... 49 Table 5-1. Screw Feeder and Material parameters. Asterisk represents values taken from a screw feeder............................................................................................................................................. 65 Table 5-2. High degree polynomial equations that describe load as function of strain. ............... 66  x  List of Figures   Figure 1-1: Refiner schematic. Chips are fed into refiners where they are defibrated. A combination of compressive and shear forces breaks down wood structure as the plate bars pass over the wood fibres. Adapted from Tienvieri et al. 1999 [4]. ....................................................... 1 Figure 1-2: Simplified TMP process with chemical impregnation also called Chemi-Thermomechanical Pulping (CTMP). This project focuses on the pre-treatment steps of mechanical compression and chemical impregnation of wood-chips prior to high consistency and low consistency refining. ................................................................................................................ 2 Figure 2-1: Wood Structure. Structure of hardwood (left) and softwood (right). Hardwood features different types of cells including vessels, and ray parenchyma cells. Softwood features mostly tracheids and parenchyma cells. Images extracted from Selecting Timber [15]. ................ 8 Figure 2-2: Diagram of the layer and fibril arrangement in the softwood tracheid. ML stands for Middle Lamella. P stands for Primary Wall. S1, S2, and S3 are the three different layers of the Secondary Wall. L stands for Lumen, which is the central channel of the fibre. Each layer is composed of macro-fibrils that contain several microfibrils. These are made of cellulose chains which are surrounded by hemicellulose and lignin. Image extracted and adapted from Smook, 1989 [2] and Sorieul et al. 2016 [16]. ............................................................................................. 9 Figure 2-3: Typical cellulase enzyme system. Cellulases include Endo-β-glucanases, Exo-β-glucanase and β-glucosidases. The first one degrades the amorphous regions of the cellulose, while the second one degrades the crystalline regions. The latter degrades the resulting cellobiose from the action of the Endo-β-glucanases, Exo-β-glucanase into glucose. Image extracted from Singh et al. 2017 [30]. ................................................................................................................... 16 Figure 2-4: Factors that affect the rate and extent of enzymatic hydrolysis. Adapted from Chandra et al. [14]......................................................................................................................... 18 Figure 3-1: Compositional analysis flowchart. The klason lignin procedure is summarized. ...... 24 Figure 3-2: HPLC vial preparation of calibration standards and unknown sample (Left). The vials are analyzed in the HPLC with autosampler and Dionex column (Right). .................................. 26 Figure 3-3: Sample of sugar peaks produced by the HPLC. This sample plot displays the charge peaks over time. The retention times are associated with the species. The calibration curves are then used to assign the peaks to the retention times. The area under the graphs is used to calculate the concentration of each species in [g/L]. .................................................................... 26 Figure 3-4: Sample calibration curve for glucose. ........................................................................ 27 Figure 3-5: Kraft pulp, BCTMP, SPF Sawdust and SPF wood chips (from left to right in order) were treated with cellulase enzyme. The wood-chip samples were screened to a uniform size of ~ 0.8 cm x 0.8 cm x 2.5 mm............................................................................................................. 29 Figure 3-6: Incubator shaker used for the enzyme treatment of substrates. The initial volume of the liquor was ~ 150 ml for all samples. ....................................................................................... 29 Figure 3-7: Dosage comparison. Glucose concentration of sawdust at high and low dosage of 10 mg enzyme/ g odw and 100 mg enzyme / g odw, respectively (Top). Glucose concentration of xi  wood chips at high and low dosages (Bottom). The difference between the two substrates is evident. .......................................................................................................................................... 31 Figure 3-8: Normalized sugar released over time. The first plot (top) shows a low dosage of 10 mg enzyme/ g odw, while second plot (bottom) shows the high dosage of 100 mg enzyme / g odw. Data was normalized for a more accurate comparison between pulps and wood. ............... 33 Figure 4-1: Experimental layout for this work. Step 1: Compositional analysis of wood. Step 2: Mechanical treatment of wood. Step 3: Measurement of extractives and microscopy. Step 4: Enzyme impregnation and hydrolysis of wood. Step 5: Filtrate liquid fraction and perform analysis of sugars released. ........................................................................................................... 36 Figure 4-2: Accelerated Solvent Extractor (ASE). The milled wood chips are placed in the extraction cell at the top. Ethanol and water are injected into the cell removing the extractives. A filter prevents the loss of solids. The extractive liquor is collected in the bottles are the bottom. 38 Figure 4-3: Preparation of sections and microscopy. First, thin sections are cut with microtome. The sections are dyed and washed. Using a Nikon light microscope, the sections are observed. A camera mounted to the microscope takes pictures of the images, which are saved to a computer software for image processing. ..................................................................................................... 40 Figure 4-4: Lab compressor, which is also called chip juicer, is a hydraulic piston used to compress wood-chips. The trial was conducted by FPInnovations in Pointe-Claire. ................... 41 Figure 4-5: Normalized glucose released over time with 0.1 g of enzyme/ g of odw. Wood-chips compressed at different compression ratios were treated with cellulase enzyme for 2 hours. ..... 44 Figure 4-6: Normalized glucose released over time for a high dosage with 1 g of enzyme/ g of odw. Wood-................................................................................................................................... 44 Figure 4-7: Hydrolysis rate as function of the estimated power consumed by the lab compressor. The dashed line represents the rate of hydrolysis of control chips attained by increasing the enzyme dosage by a factor of 10. Compressing at 4:1 achieved greater cellulose accessibility than increasing the dosage. ........................................................................................................... 45 Figure 4-8: Extractive water and ethanol soluble extractive content of the wood chips compressed at increasing compression ratios. .............................................................................. 46 Figure 4-9: Average carbohydrate composition of water extract. ................................................ 47 Figure 4-10: MTS set-up. Left: MTS 810 Load Frame. Right: Steel Compressor, compression chamber and aluminum cup at the bottom. The bottom section is heated. ................................... 48 Figure 4-11: Maximum load measured [kN] at each compression ratio. The compression time was 5 seconds for all cases. ........................................................................................................... 50 Figure 4-12: Maximum load measured at increasing compression times..................................... 51 Figure 4-13: Normalized glucose released over time for a high dosage (1 g of enzyme/ g of odw). Wood-chips were compressed at different compression ratios and at constant compression time of 5 sec. ......................................................................................................................................... 52 Figure 4-14: Normalized glucose released over time for a high dosage (1 g of enzyme/ g of odw). Wood-chips were compressed at different compression times at a constant compression ratio of 5:1. ................................................................................................................................................ 52 Figure 4-15: Normalized enzyme hydrolysis rate as function of compression ratio applied to the wood-chips at increasing compression ratios. .............................................................................. 53 Figure 4-16: Extractive water recovered from the compression chamber at increasing compression ratios for a compression time of 5 sec. .................................................................... 55 xii  Figure 4-17: Extractive water squeezed out of the compression chamber at increasing compression time. ......................................................................................................................... 55 Figure 4-18: Extractive water composition for all compression tests. The composition includes the total amount of carbohydrates as well as the soluble lignin content. Here, LC is 3:1 compression, MC is 5:1 compression ratio and HC is 6:1 compression ratio for 5 sec of compression time. LR is 3 sec of compression and HR is 10 sec of compression time at 5:1 compression. ................................................................................................................................. 56 Figure 4-19: Microscopy samples from (a) control chips, (b) wood-chips compressed at 3:1 compression ratio, (c) wood-chips compressed at 5:1 compression ratio, and (d) wood-chips compressed at 6:1 compression ratio. ........................................................................................... 58 Figure 5-1: Stresses on material element in tapered section. Image extracted from Dai et al. 2007 [55]. ............................................................................................................................................... 61 Figure 5-2 : Pressure surfaces of a screw element. Image extracted from Yu et al. 1997 [53]. ... 61 Figure 5-3: Screw configuration of an increasing diameter shaft screw feeder. The clearance H decreases from H0 to H1 along the pitch length. ........................................................................... 63 Figure 5-4: Total stress in the screw element. An equally distributed load is assumed in the bulk........................................................................................................................................................ 64 Figure 5-5: Load plot with the polynomial fit of the data............................................................. 66 Figure 5-6: Stress applied to the wood-chips when compressing at different compression ratios. The corresponding torque requirements to feed the wood-chips at different compression ratios is also shown. .................................................................................................................................... 68 Figure 5-7: Performance curves of screw feeder operating at a different compression time. ...... 69 Figure 5-8: Performance curve of screw feeder operating at a RT = 5 seconds. Enzyme impregnation was done at a dosage of 1-gram enzyme per gram of dried wood. Capacity, screw load and cellulose accessibility are considered at each compression ratio. .................................. 71  xiii  List of Symbols  σda: Stress on driving side [Pa] σw: Wall stress [Pa] σxf: Stress on trailing side [Pa] Fda: Force on the driving side [N] L: Length of screw [m] D: Diameter [m] R: radius [m] Ro: Radius at through surface [m] Rc: Radius at core shaft [m] Rb: Radius of barrel/housing [m] T: Torque [Nm] Td: Torque due to drive side [Nm] Tc: Torque due to shaft [Nm] Tf: Torque due to trailing flight side [Nm] Ttip: Torque due to tip of flight [Nm] αr: Flight helical angle [rad] αo: Flight angle at outside radius [rad] αc: Flight angle at core shaft [rad] µw: Wall friction coefficient [] ϕf: Wall friction angle of solids on flight surface [rad] xiv  δ: Effective internal friction angle [rad] λs: Ratio of the normal stress to axial stress [] γ: Screw flight thickness [m] Ho: Initial height from shaft to barrel/housing [m] Hf: Final height from shaft to barrel/housing [m] z: Horizontal distance along screw [m] ε: Strain [m/m] RT: Retention or compression time [sec] N: Rotational speed of screw [rpm] Vt:  Theoretical volume of screw feeder [m3] Qt: Theoretical capacity  [m3/sec] Pscrew: Screw load or power [W] ḡ: Normalized glucose grams [g/gi] xv  List of Abbreviations  AIL: Acid Insoluble Lignin ASE: Accelerated Solvent Extractor ASL: Acid Soluble Lignin BCTMP: Bleached Chemi-thermomechanical Pulp CR:  Compression Ratio CTMP Chemi-thermomechanical Pulping HC: High Consistency  HPLC: High Performance Liquid Chromatograph LC: Low Consistency MTS: Material Testing System ODW: Oven Dried Weight  SEM: Scanning Electron Microscope SPF: Spruce, Pine and Fir TMP: Thermomechanical Pulping    xvi  Acknowledgements  First, I want to express my gratitude to Dr. James Olson for his supervision. His technical expertise and guidance have been very helpful for the completion of this study. Additionally, I want to thank Dr. Heather Trajano and Jingqian Chen for the very fruitful discussions and for granting me access to work in the CERC laboratory.  I want to thank the researchers and staff of the Energy Reduction in Mechanical Pulping (ERMP) program in the Pulp and Paper Centre at UBC for the very valuable support. I want to acknowledge this program and UBC for providing me the platform to perform my work.  I place on record my sincere thanks to Dr. Warren Poole and Kim Wonsang for their help with setting up the MTS compression tests. I want to also thank Dr. Kim and Rebecca Leung for additional laboratory assistance. I want to thank Dr. Rodger Beatson from BCIT for his assistance in the preparation of samples.  To my committee, I want to extend my appreciation for your suggestions and assistance. I now take this opportunity to acknowledge the constant and unconditional love and support of my family in Ecuador. Thank you for being the backbone and the source of strength in my life.  I want to thank Adriana Cabrera for her encouragement. You give me the inspiration to always move forward.   To all my friends, thank you for the discussions, for the coffee breaks, for the laughs. Each of you has special place in my mind. I will be forever grateful.  Finally, I want to thank any one who directly or indirectly helped me in the completion of this study.  xvii  Dedication   To my parents,  Marco and Jenny. To my siblings, Marco and Maria 1  Chapter 1: Introduction  1.1 Thermomechanical Pulping  In the thermomechanical pulping (TMP) process, wood is reduced to its constituent fibres by mechanical means to make the pulp that is used in paper and other bioproducts today [1]. TMP process is widely used by pulp and paper mills in British Columbia, Canada. It constitutes 10% of the total electrical energy consumption in the province. The high pulp yield (85-95%) gives TMP advantage over other pulping processes (e.g. chemical pulping) [2]. TMP is a highly electrical energy consuming process. It is estimated that the TMP process consumes between 2000 to 3000 kWh/t for newsprint production [3]. Defibration of wood during refining is the major contributor to this high energy consumption. Refiners are machines consisting of two plates of roughened surface that are separated by a gap (see Figure 1-1). Wood is refined through multiple stages from high consistency (HC) to low consistency (LC) until high quality pulp is produced.    Figure 1-1: Refiner schematic. Chips are fed into refiners where they are defibrated. A combination of compressive and shear forces breaks down wood structure as the plate bars pass over the wood fibres. Adapted from Tienvieri et al. 1999 [4]. 2  The single disc refiner configuration has a stationary plate (stator) and a rotating plate (rotor). The double disc configuration has two rotors that rotate in opposite directions. The grinding action of plates imparts a cyclic loading to the wood fibres. This process tends to be highly energy consuming as a lot of energy is lost in the form of heat. To reduce the loss of energy, initial permanent plastic deformations and cracking via a high amplitude stress can be induced prior to refining to reduce the loading cycles in refining [5]. Considering this, a combination of compression and chemical impregnation is typically employed on wood-chips prior to refining (see Figure 1-2) as a pre-treatment.   Figure 1-2: Simplified TMP process with chemical impregnation also called Chemi-Thermomechanical Pulping (CTMP). This project focuses on the pre-treatment steps of mechanical compression and chemical impregnation of wood-chips prior to high consistency and low consistency refining.   The compression is achieved with a screw feeder, which compresses wood-chips as they are fed through a narrowing channel. Wood-chips are first steamed to soften the wood matrix and to preserve the fibres before compression. The compression opens-up the wood-chip, creating cracks and partially separating the fibres, rendering it more amenable for refining. The increasing diameter screw shaft pushes the extractive water out while effectively defibrating the wood-chips. This process leads to a reduction of energy requirements of about 20% in the subsequent refining 3  stages [1], [6]. Additionally, wood-chip compression improves water uptake of wood-chips which translates to a better chemical impregnation [6], [7]. Impregnation of chemicals is carried out in impregnators at the discharge of the screw. The compression removes the air bubbles from the chips. When the chips are discharged into the impregnator, they expand absorbing the impregnation liquor due to the pressure gradient [7]. The impregnators are either inclined or vertical and consist of a conveying screw.  Sodium sulfite is typically used in the impregnation of wood-chips. Sulfite pre-treatments improve certain properties of the pulp and can help reduce refining energy by increasing fibre delamination [8]. Sulfite treatments significantly reduce the lignin softening temperature resulting in improved pulp strength. However, the use of chemicals such as sulfite results in the formation of inorganic contaminants in the effluent that are difficult to treat [7]. Considering these environmental concerns, enzymes have found use in industry. Enzymes are less toxic to the environment compared to traditional chemicals because they are readily biodegradable [9].   Enzymes used in the pulp and paper industry tend to selectively degrade specific components of the fibres to facilitate their separation. Studies have proved that certain enzyme blends positively influence reduction in refining energy and also improve pulp properties [9]–[11]. Close contact between the cellulose fibres and the cellulase enzyme is needed to start the hydrolysis reaction. Nevertheless, fibre accessibility to the enzyme is very limited due to wood’s recalcitrant nature. Therefore, it is of great importance to improve the enzyme impregnation into the wood-chips. Harsh mechanical treatment is needed to alter the wood morphology in favor of improving enzyme transport to the cellulose fibres.   4  1.2 Research Objectives Compression is beneficial for the TMP process in terms of energy savings in refining, chemical impregnation and fibre properties. Indeed, several studies have documented the effect of compression on wood-chip morphology, and water uptake to justify the benefits of having a mechanical pre-treatment (see Section 2.3) [6], [12], [13]. These studies report increase in wood porosity as well as cell wall delamination. Higher pore size leads to improved liquid transport and cell wall delamination facilitates further defibration in refining. The literature is rich in studies that show the importance of compressive pre-treatment. However, the effect screw feeder compression on cellulose accessibility to enzymes is not fully explored. Additionally, there is no study that combines the characterization of compression, enzyme penetration of wood-chips and power consumption of a modular screw feeder.  This study focuses on understanding the effect of different compression conditions on cellulose accessibility. Cellulose accessibility is directly correlated to the exposed surface area of the fibres [14]. Changes in cellulose accessibility should thus give an indication of changes in morphology. Furthermore, this study aims at understanding the operation of a screw feeder. Screw feeder operation for wood-chip compression has not been analyzed in the literature. The long-term impact of this study will be to optimize screw feeder operation to make better use of enzymes. In turn, this is expected to reduce energy consumption in refining. Reducing energy consumption while maintaining the quality of pulp properties is of paramount importance for the pulp and paper industry.    5  To achieve this, the following objectives are proposed: • Quantify the cellulose accessibility of the compressed wood-chips: o Evaluate the effect of compression ratio and rate on the initial hydrolysis rate. o Evaluate the effect of compression ratio and rate on extractive removal. o Evaluate the effect of compression ratio on wood morphology.  • Characterize compression by a screw feeder:  o Develop a model for the relationship between screw feeder operation, geometry and compression stress. o Develop a relationship between compression and power consumption. By fulfilling these objectives, this study will provide useful knowledge for more efficient operation of screw feeder and enzymatic impregnation. Improving the uptake of enzymes into the wood-chips and effectively defibrating the wood-chips is ultimately very important in the reduction of refining energy.  1.3 Thesis Overview This thesis is broken down into 6 main chapters. This first present chapter serves as a brief introduction to the topic at hand. Research objectives and motivation are laid out as well. Chapter 2 presents a detailed literature review on studies related to wood structure and composition, effect of compressive pre-treatment, and enzyme hydrolysis. An overview of typical methods used for the characterization of cellulose accessibility and morphology is presented. An approach to the research objectives is proposed. Chapter 3 outlines the experimental work which involves detailed explanation of the method to measure changes in cellulose accessibility, as well as, initial experiments to test the method. Chapter 4 covers the experimental work on the compression of 6  wood-chips. Here, the result of different degrees of compression on cellulose accessibility, extractive content and morphology is presented. For each individual experiment, detailed and specific methodology is described to help the reader understand how the data was gathered. Chapter 5 outlines a short screw feeder analysis which involves the use of experimental data from chapter 4 and torque prediction equations to analyze the screw feeder performance at different compression ratios. Finally, a concluding chapter brings together all the takeaways of each chapter and brings closure to the research questions.  7  Chapter 2: Background  In this chapter, technical background is provided to facilitate the understanding of the importance of the work presented in this thesis. First, a brief background on wood structure and composition is outlined. Then, an overview of mechanical pre-treatment is provided, including its effects on wood-chip morphology, liquid uptake, energy reduction and extractive content. Moreover, a short introduction to enzymes and the factors that affect their performance are presented. Typical methods to quantify cellulose accessibility are summarized. Finally, the study objectives are reviewed, and the approach is proposed.   2.1  Wood Structure Wood is the principal raw material for pulp and paper production. It is a complex anisotropic and porous material composed of different types of cells. Wood can be classified into two major categories: softwoods and hardwoods. Figure 2-1 depicts the structures of both hardwoods and softwoods.  Softwoods consist of tracheids or fibres, which are long tapered fully developed cells. Ray parenchyma cells, which aid in fibre-to-fibre fluid transport, are also present. On the other hand, hardwoods are composed of different types of cells including long but narrow fibres and short but wide vessels cells, and ray parenchyma cells. The structure of wood and the arrangement of cells depend on the type and species of wood. Another factor that affects the structure is the growth pattern of the wood (i.e. seasonal growth). Early wood is formed during spring season. A greater amount of wider tracheids are produced, which are less dense. Narrower cell walls in early wood allow for greater water transport, thus resulting in increased growth. Then, in summer season, late 8  wood is formed. In contrast to early wood, late wood tissue has narrower lumen with thicker walls. This results in less water transport between tracheids. This work will focus on softwood wood-chips as the raw materials for the experimental work.      Figure 2-1: Wood Structure. Structure of hardwood (left) and softwood (right). Hardwood features different types of cells including vessels, and ray parenchyma cells. Softwood features mostly tracheids and parenchyma cells. Images extracted from Selecting Timber [15].   Tracheids are formed from a collection of several layers of microfibrils [2]. These microfibrils are essentially bundles of cellulose molecules. There are other layers that have different functions in the fibres. A lignin-rich middle lamella is a layer that surrounds the tracheid cells. It serves as a structural adhesive that holds the fibres together. Then a primary wall and secondary walls made of fibrils add structural bulk to the wood. Finally, the lumen is a central void or cavity of the 9  tracheid cell whose function is to hold and transport water. Figure 2-2 shows the schematic of the layers present in a tracheid cell in softwood, as well as, the structure of the fibrils.  Figure 2-2: Diagram of the layer and fibril arrangement in the softwood tracheid. ML stands for Middle Lamella. P stands for Primary Wall. S1, S2, and S3 are the three different layers of the Secondary Wall. L stands for Lumen, which is the central channel of the fibre. Each layer is composed of macro-fibrils that contain several microfibrils. These are made of cellulose chains which are surrounded by hemicellulose and lignin. Image extracted and adapted from Smook, 1989 [2] and Sorieul et al. 2016 [16].  10  2.2 Wood Chemistry The main components of wood fibre are cellulose, hemicellulose, lignin and wood extractives. Figure 2-2 displays the wood components arranged in the fibers. Cellulose is a polysaccharide composed of linear glucose units linked together by β-glucosidic bonds [2]. It is the most important component of the wood fibre and it is also the most abundant renewable polymer in the world. Cellulose bundles can have either crystalline or amorphous regions. A strong fibre to fibre interaction is present due to hydrogen bonds which also gives the fibres strength and insolubility in most solvents. Additionally, the surrounding lignin and hemicellulose makes cellulose highly inaccessible and difficult to isolate [14].   Hemicelluloses are amorphous and branched polymers composed of different sugar units such as xylose, mannose, galactose, arabinose and glucose. Its branched nature makes hemicellulose the most accessible and reactive component of wood. Both cellulose and hemicellulose contribute to the structure of the wood cells. The hemicellulose component also coats the cellulose fibres and thus reduces the cellulose accessibility (see Figure 2-2) [14].   Lignin is a highly complex amorphous macromolecule composed of phenylpropane units. It serves as a glue and structural support that holds the cellulose and hemicellulose together. In contrasts to its other components, lignin is hydrophobic giving wood a moisture resistance. In addition to its moisture resistance properties, lignin also provides natural resistance to enzymes and other chemicals [14]. Lignin in its natural state adds stiffness to the wood structure. However, at high temperature and moisture content, lignin undergoes a glass transition where the modulus 11  of elasticity of wood suddenly decreases. This makes wood soft and easier to deform. Both lignin and hemicellulose give wood its viscoelastic and strength properties [17].  Wood extractives consist of a diverse group of compounds which are not part of the wood structure. These are divided into water-soluble and insoluble. Extractives that dissolve in the process water include hemicellulose carbohydrates, and low mass lignin fragments. The rest of the extractives are lipophilic and include terpenoids, fatty acids, sterols, and phenolics, to name a few [18]. The main function of extractives is to protect wood from decay. Extractives protect wood from enzyme producing fungi. Because of this feature, extractives are believed to enzyme impregnation [19]. Furthermore, extractives can also cause coloring of the pulp [20]. More importantly, extractives decrease the friction in the refiners thus interfering with the refining process [21]. Generally, it is desirable to remove extractives from wood and pulp.   2.3 Compressive Pre-treatment  Compressive pre-treatment has been implemented to induce changes in wood structure in such a way that fibres are easier to separate during refining. This results in improved pulp properties, and most importantly, savings in refining energy. This section discusses the effect of compressive pre-treatment on wood-chip morphology, liquid uptake, energy savings in refining as well as extractive removal and pulp property modification. Experimental studies on wood compression and methods to evaluate change in morphology are also outlined.  12  2.3.1 Effect of Compression on Morphology Determining the effect of compression on wood-chip and fibre morphology has been the goal of many researchers studying the TMP and CTMP processes. Some studies have focused in understanding the effect of compression on wood-chip morphology by measuring changes in porosity [13], [22]. Mercury porosimetry and image analysis via light microscopy, scanning electron microscopy (SEM) and fluorescence microscopy have been common methods used for this purpose.   Mercury porosimetry involves measuring the amount of mercury that is adsorbed onto the cellulosic sample over a range of varying pressures. The pressure needed to force the mercury into the pores is correlated to the diameter of the pores. Peng et al. 1996 collected experimental data of pore size and volume distributions of compressed wood-chips [22]. The study showed that screw feeder compression leads to higher cumulative pore volume, which means that a significant structural change in the wood-chip morphology was induced. Pore size and volume distributions data enables a detailed understanding of microstructural changes in the wood-chips. The same study shows that the increase of pore volume is due to formation of macro cracks (larger pores). On the other hand, the increase of pore surface area is due to the formation of microcracks in significantly lower amount. Determining changes in structure can also be achieved by image analysis of wood via microscopy. Scanning electron microscopy (SEM) and other microscopy techniques have been widely used for studies of wood-chips and fibre morphology  [6], [23], [24]. Image analysis techniques can be used to quantitatively determine pore size distributions with the aid of computer algorithms [13]. Image analysis tests from screw press treated wood-chips show that increasing compression leads to shifts in mean pore size which translates to pore collapse and 13  cell wall delamination. Fluorescence microscopy has also been used for studies of changes in wood-chip structure after Impressafiner (i.e. high-compression screw feeder) compression [6]. Wood-chip treated with Impressafiner compression resulted in cracking and splitting of the wood-chip structure, mainly at the interphase between the primary and inner secondary wall layers (S1/S2) and between middle lamella and S1 interphase.  2.3.2 Effect of Compression on Liquid Uptake Compressive pre-treatment is believed to aid in the impregnation or liquid uptake ability of wood-chips in the impregnation stage. Compression first forces water and excess air bubbles out of the wood-chips thus achieving a homogenous moisture content throughout the extruded, compressed bulk. The compressed plug is fed and released into impregnation liquor. Here, the chips expand and then absorb the impregnation liquor [7]. The mechanical treatment opens the structure of the chips creating fractures which act as penetration paths for chemical reagents. Indeed, studies agree that mechanical compressive pre-treatment improves water uptake of wood-chips, especially in the early wood tracheid [6], [13]. De Choudens et al. 1985 also determined that the water absorption rate and destruction of softwoods does not gradually increase with compression. It seems that at a certain level of compression, fractures spread quickly which rapidly increases the water uptake [12]. At a low degree of compression, morphology modifications are slight because thicker latewood tracheids are spared. Wood-chips subjected to a strong Impressafiner pre-treatment [6] have also showed improvements in water uptake. It is believed that the increase in surface area of the delaminated wood-chips allows for an easier and uniform impregnation of liquor into the chips. It can be speculated that this would correlate to improvement in the penetration of chemicals such as enzymes. 14  2.3.3 Effect of Compression on Refining Energy Reduction Refining is a very energy consuming process that takes 80% of the total energy consumption of a TMP mill [25]. The rest is used for chip handling, pumps, screens, blowers, and other components or lost as heat. It is then of great interest to focus efforts in reducing consumption of refining energy. As mentioned before, mechanical pre-compression of wood-chips leads to changes in morphology, liquid uptake and improved pulp properties. All of this contributes to reduction in subsequent refining energy. To be precise, pre-compression of wood-chips induces a partial defibration which makes the chips more amenable for refining. Lignin in the wood-chips acts as a visco-elastic material that absorbs energy. Pre-treatment of steaming and compression detaches portions of the lignin-rich middle lamella which reduces compressive cycles needed to liberate the fibres during refining and thus reduces energy consumption. Reduction of refining energy of up to 20% have been reported, depending on the pre-treatment method [1], [23].  2.3.4 Effect of Compression on Extractives and Pulp Properties Improved pulp properties with lower energy consumption is the end goal of many studies. During compression, the chips are homogenized in terms of size and moisture content. Consequently, the fibres are developed uniformly during refining. This leads to better fibre properties. Compressive pre-treatment with a Prex-impregnator (3:1 compression ratio) has shown to produce pulps of higher tensile strength and density at the same energy consumption compared to non-treated pulps [26]. Impressafiner pre-treatment (5:1 compression ratio) has produced pulp of superior mechanical and optical properties [6]. Overall, compressive pre-treatment leads to stronger pulps at the same specific energy consumption. Another advantage of including a compressive pre-treatment is to reduce the undesirable extractives [6], [27]. Extractives are known 15  to affect the color of the pulp (i.e. color specks and loss of brightness) as well as pulp yield. Additionally, any chemicals used for treatment are directly affected by the extractives [28]. Impressafiner and other high compressive pre-treatments have proved to reduce extractives in the final pulp by as much as 40% in pilot scale trials and by 15% in mill scale trials [18]. Moreover, Nelsson et al. 2012 determined that the extractive content in wood decreases systematically with increasing compression ratio of an Impressafiner screw [6].  2.4 Enzyme Hydrolysis Enzymes are protein structures that act as catalysts of specific chemical reactions. These proteins are naturally found in wood decaying fungi and bacteria, which can effectively break down the complex wood structure into soluble sugars (i.e. glucose, mannose, arabinose, xylose). This work will be focused on cellulase enzymes which primarily hydrolyze cellulose chains into glucose monomers.   2.4.1 Cellulase Enzyme Cellulases are the most common enzymes employed in the hydrolysis of lignocellulosic material [14]. This enzyme targets the cellulose component of the fibres with the purpose of breaking it down to its monomer components. To achieve optimum hydrolysis, the components of the enzyme system must act in a synergistic fashion. Figure 2-3 displays the components of a cellulase enzyme system degrading cellulose.  Cellulase system includes endo-β-glucanases, exo-β-glucanase or cellobiohydrolase and β-glucosidases [29]. The endo-β-glucanases and exo-β-glucanase target the amorphous and 16  crystalline regions, respectively. The action of the first two components yields glucose monomers as well as a two-glucose chain called cellobiose. The water soluble cellobiose is readily hydrolyzed by β-glucosidase to form glucose.   Figure 2-3: Typical cellulase enzyme system. Cellulases include Endo-β-glucanases, Exo-β-glucanase and β-glucosidases. The first one degrades the amorphous regions of the cellulose, while the second one degrades the crystalline regions. The latter degrades the resulting cellobiose from the action of the Endo-β-glucanases, Exo-β-glucanase into glucose. Image extracted from Singh et al. 2017 [30].  2.4.2 Factors Affecting Enzymatic Hydrolysis  Effective enzymatic hydrolysis of woody biomass is difficult to accomplish because it depends on many factors some of which are related to the enzyme and some which depend on the substrate.  Figure 2-4 is provided as a summary of the factors that contribute to or inhibit the rate and the extent of enzyme hydrolysis.  Substrate properties play a big role in the extent of enzyme hydrolysis because woody biomass is physically and chemically constructed to resist degradation. Chemical composition of the biomass influences the extent of enzyme hydrolysis [14]. Lignin acts as both a barrier and an unproductive binding agent, thus preventing the access of cellulases to 17  cellulose fibers [31], [32]. Even if some of the lignin is removed, the inhibition of the enzyme may not be completely reduced as the concentration of lignin is just as important as its distribution in the biomass [33]. Hemicellulose also affects cellulase hydrolysis, as previous studies show that its removal results in improved enzyme performance [14], [32]. Hemicellulases coat the cellulose fibres thus reducing the fibre accessibility to the enzymes. Lastly, the extractive composition also inhibits the performance of cellulase enzymes. Several studies agree that extractives restrict the access of the enzyme and may also be responsible of its deactivation [19], [34], [35]. Physical complexity of the biomass also adds to its high recalcitrance. Physical attributes that affect enzyme hydrolysis include: crystallinity, degree of polymerization and cellulose accessibility [14]. Throughout this work, the term cellulose accessibility means the ease of the cellulose to be reached by the enzyme. Accessibility is affected by the external and internal surface area of the substrate. The external surface area mainly corresponds to the size of the biomass particles. Smaller particles have higher external surface area and are expected to have better accessibility. Sangseethong et al. [36] studied the effect of  microcrystalline cellulose particle size on enzymatic saccharification and demonstrated that hydrolysis of smaller particles exhibited higher hydrolysis rates compared to the larger counterparts. This means that the size of a substrate contributes to its accessibility to the enzyme. The internal surface area corresponds to the size of pores or cell walls. Substrates with large enough pores can allow access to both small and large components of the enzyme allowing for improved enzyme hydrolysis [37]. Access of all enzyme components is important for the synergistic effect of each component [29]. Pore size and pore distributions seems to play an important role in enzyme accessibility. Mooney et al. 1998 believe that the presence of small pores within the substrate contribute to its surface area. Pore size 18  distribution has been commonly used as a method of predicting the ease of hydrolysis and cellulose accessibility [33], [38]. The rate limiting pore size for the cellulase enzymes is 5.1 nm approximately (i.e. nominal diameter of cellulases) [38].   Figure 2-4: Factors that affect the rate and extent of enzymatic hydrolysis. Adapted from Chandra et al. [14].  2.5 Characterization of Cellulose Accessibility  Cellulose accessibility is affected by the characteristics of the internal and external surface area of the lignocellulosic material. Contact between the cellulose fibres and the cellulase enzyme is necessary for the hydrolysis to take place. Common methods to characterize or predict cellulose accessibility usually involve indirect measurements of pore volume and surface area [39]. These methods are summarized in Table 2-1.  Mercury porosimetry has been used to determine changes in wood structure: pore size distributions and total pore volume, specific surface area and permeability (also see section 2.3.1). 19  While this method provides a lot of information about the wood structure, it requires drying which can result in structural damage of the sample. Additionally, it only measures the entrance size of the pores and disregards the inner size of the pore.  Nitrogen adsorption has also been used to determine surface area of substrates [39]. In a similar way to mercury porosimetry, the amount of nitrogen dioxide adsorbed into the substrate is determined by the pressure change in the gas after it is introduced to the sample. For this method, drying of the sample also imposes problems in the prediction of cellulose accessibility. Moreover, this method tends to overestimate results as the nitrogen molecules are smaller than enzymes. Solute exclusion is another method used to predict cellulose accessibility [40]. Samples are incubated with probe solutions of specific molecule size and concentration. The method involves correlating the change in concentration of solutions with different probe size to the porosity of the sample. This method gives quantitative results but is quite laborious. Consequently, achieving repeatable results is challenging.   Simons’ Staining is perhaps one of the most effective methods for the prediction of cellulose accessibility of fibres [41]. This method uses two dyes with different color, size and affinity to cellulose. The blue dye penetrates small pores and the orange dye penetrates larger pores. The ratio of the blue and orange dye adsorption provides useful information to calculate the relative porosity and accessible surface area using a Langmuir adsorption equation. Larger orange to blue dye ratio means higher cellulose accessibility. While this method is simple and sensitive, it categorizes the structure by the relative concentrations of two dyes only. Protein adsorption has been proved to be an accurate, and substrate specific method to characterize cellulose accessibility [42]. In this method, a cellulose binding module and a fluorescent protein can adsorb onto the cellulose fibres. The adsorption of the proteins to the 20  substrate can be determined with a spectrophotometer and the Langmuir equations. This method is particularly useful as the proteins mimic the enzyme in terms of size and affinity. While this method directly applies to a hydrolysis application, the adsorption results can be affected by the unproductive binding properties of lignin.   Table 2-1: Common methods used for the prediction of cellulose accessibility. Method Description  Typical Application  Ref. Mercury Porosimetry Measuring the amount of mercury that is adsorbed onto the cellulosic sample over a range of varying pressures. Dried samples, such as wood or pulp fibres. [22], [43] Simons’ Staining  Use of two different dyes. The blue dye penetrates small pores that are inaccessible to larger orange dye. Orange dye adsorbs to substrate surface. Porosity is determined as the ratio of the orange to blue dye adsorbed. The total orange dye adsorbed predicts the surface area accessible to the enzymes as orange dye is of similar size to cellulases. Wet biomass, mainly pulp fibres, or chemically treated wood. [41] Nitrogen Adsorption Quantification of the amount of nitrogen dioxide molecules that interacts with the test samples over a range of varying pressures. Gases are adsorbed into sample by using a surface area analyzer. Dried samples, such as wood or pulp. [39] Solute Exclusion This method works by incubating the test sample with a dextran/polyethylene solution and then measuring changes in the concentration of the solution which might occur after the solute or water molecules go into the pores of the samples. Wet biomass, mainly pulp fibres. [40] Protein Absorption Concentration of proteins can be calculated with spectrophotometry and adsorption equations. The adsorption of substrate specific proteins is used to determine cellulose accessibility. Wet biomass, mainly pulp fibres. [42]  21  2.6 Study Approach Literature review on the topic of compressive pre-treatment of wood-chips in TMP exposes the effects of compression on the fibre morphology, reduction in energy consumption and pulp properties [1], [6], [23]. However, the effects of different degrees of compression on cellulose accessibility to enzymes is not explored. Enzyme hydrolysis is directly correlated to the available surface area of the fibres. Therefore, it is expected that the degree of hydrolysis is correlated to the cellulose accessibility. The first objective of this study is to quantify the level of cellulose accessibility (see section 1.2). Several methods have been developed to characterize cellulose accessibility (see section 2.5). These methods were assessed based on applicability to the substrate in this work (i.e. wood-chips), and equipment availability. For the determination of cellulose accessibility of wood-chips, most of the direct methods could not be applied. Therefore, a different method was developed for the characterization of cellulose accessibility. For this work, a cellulase enzyme that selectively targets cellulose is used to treat compressed wood-chips. Direct measurements of the glucose released are used to track changes in cellulose accessibility. Instead of measuring surface area or pore size to estimate fibre accessibility to the enzyme, the enzyme itself is used to directly determine the accessibility. Wiman et al. 2012 determined that the initial rate of enzymatic hydrolysis is well correlated to the enzyme adsorption on pretreated spruce [44]. Therefore, the initial hydrolysis rate of wood can be used as a measurement of cellulose accessibility. This measurement can also give information about changes in the substrate surface area or morphology. Higher hydrolysis rate is correlated to higher available surface area [14]. Other morphology characterization methods will be used to validate these results.    22  For the first objective, the study approach will be to: • Perform sugar compositional analysis to qualitatively evaluate cellulose accessibility. • Use light microscopy and extractive measurements to corroborate results.   Compression of wood-chips have been explored with screw feeder compression in mill trials [6], [23]. However, the operating conditions that can be tested in mills is limited. On the other hand, controlled lab experiments of single chip compressions are readily available as well [5], [13], but such experiments ignore chip-to-chip interaction during compression. Considering this, controlled compression experiments of a bed of wood-chips are proposed to address the second objective of characterizing screw feeder compression (see section 1.2).   For the second objective, this study will: • Compress a bed of wood-chips at different compression ratios and rates. • Use compression data and screw feeder force analysis to estimate screw feeder performance.  23  Chapter 3: Characterizing Cellulose Accessibility  Several methods have been used to measure structural changes in wood and cellulose accessibility (see section 2.3.1 and 2.5, respectively). In this study, an enzyme mixture was used as a marker to track changes in cellulose accessibility. The enzyme mixture was tailored to target cellulose and break it down to glucose. Therefore, changes in hydrolysis would correspond to changes in accessibility of the cellulose to the enzyme. Additionally, changes in enzyme hydrolysis are directly affected by surface area [14]. Measuring the initial hydrolysis of the enzyme would then correlate to both accessibility and morphology. The method of quantifying sugars (i.e. glucose) released into liquid fraction is explained and evaluated in this chapter. The limitation of morphology and composition of different substrates on the mass transfer of enzyme is explored.   3.1 Materials and Methodology 3.1.1 Compositional Analysis Compositional analysis was performed on the spruce, pine, fir (SPF) wood-chips prior to enzyme treatment, following a typical Klason lignin procedure [45]. This procedure typically involves a two-stage acid hydrolysis to break down the biomass into its constitutional carbohydrates and lignin. The wood-chips were obtained from Catalyst, Powell River Division. The substrate was first milled to make it as homogeneous as possible. The extractives were removed with an accelerated solvent extractor (ASE) because these interfered with the compositional analysis [46]. The extractive-free biomass was treated with 72% w/w sulfuric acid for 1 hour. Shortly after, the acid concentration was brought down to a 4% w/w by adding 24  deionized water. The sample was post-hydrolyzed with an autoclave at high temperature and pressure. The sample was then vacuum filtered to separate the solids and liquid fractions.  The solid residue was rinsed with water and then dried in an oven for a day and later burned in a furnace to determine the acid insoluble lignin (AIL) and ash contents, respectively. The liquid fraction could be analyzed for acid soluble lignin (ASL) and carbohydrates. The ASL was quantified using a UV spectrophotometer [45] and the carbohydrates were quantified using a High-performance Liquid Chromatograph (HPLC). Determination of sugars in liquid fraction is explained in detail in the section below. The overall compositional analysis process is summarized in Figure 3-1.   Figure 3-1: Compositional analysis flowchart. The klason lignin procedure is summarized.  25  3.1.2 Quantifying Sugars in Liquid Fraction Sugar calibration standard solution was prepared with different concentrations of Arabinose, Galactose, Glucose, Xylose and Mannose. The amount of each sugar was selected depending on the expected concentration of sugars of the sample. Seven calibration standards were prepared by diluting the sugar standard solution from high concentration to low concentration. Fucose was used as internal standard and a fixed amount was added to each sample and calibration standard alike (see Figure 3-2) [47]. The calibration standards were filtered with a 0.22 µm membrane filter and delivered to a HPLC vial via a syringe. Similarly, the samples were prepared with a fucose internal standard and were diluted as needed, if the expected concentration of sugar was very high and fell outside the linear calibration range. The sugar concentration was determined with HPLC (see Figure 3-2). The sample passed through the HPLC column and its different sugar species were retained in the column at different times. The result of this was a display of sugar peaks at different retention times (see Figure 3-3). Fucose was the first peak in the graph, followed by arabinose, galactose, glucose, xylose and then mannose in that order. The order of the species was always the same.  For each calibration standard, the ratio of sugar to fucose was calculated. The ratio of sugar to fucose was plotted over the actual concentration of the sugar of interest to make a linear calibration curve (see Figure 3-4). From this curve, a best fit equation was derived and was used to calculate the actual sugar concentration of the samples.  This linear equation was in the form y = Ax + B, where y was the fucose to glucose ratio, and x was the actual sample concentration in [g/L]. The parameters A and B were obtained from the calibration equation. Each point in the curve represented a calibration standard. The calibration samples could be removed from the line 26  to achieve a R2 of ~ 0.999. For each sample of known glucose to fucose ratio y, the actual concentration of the sample x could be calculated using the calibration equation.     Figure 3-2: HPLC vial preparation of calibration standards and unknown sample (Left). The vials are analyzed in the HPLC with autosampler and Dionex column (Right).   Figure 3-3: Sample of sugar peaks produced by the HPLC. This sample plot displays the charge peaks over time. The retention times are associated with the species. The calibration curves are then used to assign the peaks to the retention times. The area under the graphs is used to calculate the concentration of each species in [g/L]. 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 35.0 37.5 40.0 42.5 45.0 47.5 50.0 52.5 55.0 57.5 60.0-10020040060090006212018 ENDO CELLULASE HBKP #8 STD0.75 IntAmp_1nCmin1 - 1.8842 - Fucose - 5.9003 - Arabinose - 12.9424 - Galactose - 16.3425 - Glucose - 19.2676 - Xylose - 23.3677 - Mannose - 25.7008 - 32.6429 - 33.6841 - Fucose 2 - Arabin... 3 - Galactose4 - Glucose5 - Xylose6 - Mannose27   Figure 3-4: Sample calibration curve for glucose.   3.2 Method Validation: Enzyme treatment of substrates For the first experiment, two ground pulps (Kraft pulp and Bleached Chemi-Thermo-Mechanical pulp) and ground SPF wood were treated with enzyme. The Northern Bleached Softwood Kraft (NBSK) pulp was provided by Canfor Mill. The Bleached Chemi-Thermo-Mechanical pulp (BCTMP) and wood-chips were provided by Catalyst, Powell River. The pulps and wood were milled with a Wiley mill to achieve consistent morphology across all samples (see Figure 3-5). The milling of wood-chips in this case represents an ‘extreme’ case of mechanical treatment to compare against the ‘raw’ wood-chips. The ideal conditions for this type of cellulase enzyme are very well documented by researchers [31], [48]. Most cellulase enzymes work best in temperatures of 50 ℃ and slightly acidic conditions. The cellulase enzyme (cellulase EL2013/000649) was supplied by ABenzymes y = 8.2211x + 0.0073R² = 0.999500.050.10.150.20.250.30.350.40 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045Ratio Glu/FuActual Concentration (mg/ml)28  and its optimal operating conditions are those of typical cellulase enzymes. These experimental conditions are summarized in Table 3-1.   Table 3-1: Enzyme Treatment Conditions Biomass BCTMP, Kraft pulp (ground), wood  Solid content 4 grams  Consistency 2.5 % Enzyme ABenzyme cellulase (Mostly exoglucanase, and some endoglucanse and betaglucosidase)  Dosages 10 mg enzyme / g odw and 100 mg enzyme / g odw Buffer Sodium acetate  Ph ~ 4.80 Temperature 50 ℃ Sampling Times 30 min, 2 hours, 4 hours, 5 hours Starting Volume ~150 ml  The samples were mixed with buffer and enzyme solution to bring the consistency to 2.5%. The experiment was run in an incubator shaker stirring the samples at 120 RPM (Figure 3-6). “Low” (10 mg/g odw) and “High” (100 mg/ g odw) dosages were tested for all samples. Additionally, controls with no enzyme were evaluated as well. Each sample was run in duplicate.  As much as 15 ml of sample was removed for every sampling point (i.e. 30 min, 2 hours, 4 hours, and 5 hours) for analysis with a lab micropipette. The initial liquor volume (~ 150 ml) was significantly larger than the sampling volume. This reduced the impact of sampling on the overall enzyme kinetics. Quickly after sampling, the samples were heated in a water bath at around 90 ℃ to denature the enzyme. The sample was vacuum filtered to ensure there are no solid particles in the test tube. The test tubes could then be stored in the fridge at 4 ℃ for subsequent analysis.   29    Figure 3-5: Kraft pulp, BCTMP, SPF Sawdust and SPF wood chips (from left to right in order) were treated with cellulase enzyme. The wood-chip samples were screened to a uniform size of ~ 0.8 cm x 0.8 cm x 2.5 mm.    Figure 3-6: Incubator shaker used for the enzyme treatment of substrates. The initial volume of the liquor was ~ 150 ml for all samples.   30  3.3 Results 3.3.1 Compositional Analysis Compositional analysis was performed for wood-chips only as it is biomass used for the rest of the work. The composition of the SPF wood-chips was determined using the method of summarized in Figure 3-1. The composition of SPF wood-chips in terms of carbohydrates and lignin is presented in Table 3-2. The composition of wood-chips agrees with typical values found in softwoods [49], [50]. The composition of the pulps was assumed to be that of same furnish determined by Lahtinen et al. 2014 [51].   Table 3-2: Chemical compositions of pulps [51] and SPF wood chips as determined in this work.  Components  BSKP (%) CTMP (%) SPF Wood-chips (%) Glucose 87 52  40.98 Xylan 8.1 21  6.20  Galactan 0.29 0.55  2.26  Arabinan 0.78 0.29 1.74  Mannan 6.8 2.3 14.60  Total Lignin <1 20.6 33.75    3.3.2 Enzyme Hydrolysis of Wood  The glucose concentration of the liquid extracted from the flask was analyzed over time. Figure 3-7 compares the effect of dosages of enzyme on wood-chips and sawdust. Higher dosage had a greater effect on sawdust compared to wood-chips. At low dosage, 4 times more glucose was released from sawdust than from wood-chips. At high dosage, 10 times more glucose was released from sawdust than from wood-chips. This can be attributed to the difference in accessible surface area of each substrate. Greater surface area (~1 mm diameter particles) of the sawdust 31  allows for more effective hydrolysis. Substrate morphology had a significant influence on the effectiveness of enzyme dosage.        Figure 3-7: Dosage comparison. Glucose concentration of sawdust at high and low dosage of 10 mg enzyme/ g odw and 100 mg enzyme / g odw, respectively (Top). Glucose concentration of wood chips at high and low dosages (Bottom). The difference between the two substrates is evident.   00.20.40.60.811.20 1 2 3 4 5 6Glucose (g/L)Hydrolysis Time (h)Sawdust-High dosageSawdust-Low dosage00.020.040.060.080.10.120.140.160.180 1 2 3 4 5 6Glucose (g/L)Hydrolysis Time (h)Wood Chips-HighDosageWood Chips-Low dosage32  3.3.3 Comparison between Substrates The difference in released glucose between all substrates is analyzed in this section. Because the composition of the pulps is very different compared to that of the wood (see Table 3-2), it is necessary to normalize the data. The raw data is normalized by dividing the grams of glucose released into the liquid fraction by the initial glucose content from cellulose of the solid samples. The normalized glucose is represented by [ḡ] in equation 3-1. Figure 3-8 shows the normalized glucose for all the substrates treated with cellulase enzyme. The hydrolysis rate of each substrate is quantified from the slope of each curve. The hydrolysis rate [ḡ/h] for each substrate is shown in Table 3-3.    𝑔𝑙𝑢𝑐𝑜𝑠𝑒 [?̅?] =𝑔𝑙𝑢𝑐𝑜𝑠𝑒 𝑔𝑟𝑎𝑚𝑠 𝑟𝑒𝑙𝑒𝑎𝑠𝑒𝑑 𝑖𝑛 𝑡ℎ𝑒 𝑙𝑖𝑞𝑢𝑖𝑑 [𝑔] 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑔𝑟𝑎𝑚𝑠 𝑜𝑓 𝑔𝑙𝑢𝑐𝑜𝑠𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑜𝑙𝑖𝑑𝑠 [𝑔𝑖] (3-1)   The results show a high enzyme activity for the high dosage Kraft pulp sample. BCTMP samples show a significantly lower amount of released glucose compared to the Kraft pulp sample. BCTMP contains 52% glucose and 20% lignin while the Kraft pulp contains 87% glucose and 1% lignin. It is not surprising that the Kraft pulp show greater enzymatic activity as it has lower lignin content and higher glucose content (see Table 3-2). Although the data is normalized to the initial composition, the distribution of the components in the substrate is just as important as the substrates composition [33]. The results suggest that there are other factors that affect the hydrolysis including the available surface area of the fibres and the enzyme reactivity.   33     Figure 3-8: Normalized sugar released over time. The first plot (top) shows a low dosage of 10 mg enzyme/ g odw, while second plot (bottom) shows the high dosage of 100 mg enzyme / g odw. Data was normalized for a more accurate comparison between pulps and wood.  00.010.020.030.040.050.060.070 1 2 3 4 5 6Glucose (ḡ)Hydrolysis Time (h)Kraft pulpBCTMPSawdustWood Chips00.050.10.150.20.250.30.350 1 2 3 4 5 6Glucose (ḡ)Hydrolysis Time (h)Kraft pulpBCTMPSawdustWood Chips34  The difference in sugar yields is greater for the Kraft pulp at higher dosage in comparison to the other substrates. On the other end of the spectrum, the effects of increasing the dosage for wood-chip treatment are hindered by the mass transfer limitations of wood’s internal and external surface area (i.e. particle size and porosity) [14]. There are two main time scales of the hydrolysis reactions present in these results: the immediate time scale and the short-time scale. In the immediate time scale, a jump in glucose released is noticeable. This time scale corresponds to the hydrolysis of loose soluble particles of the substrate. The short-time scale corresponds to the hydrolysis of the substrate structure as the enzyme penetrates and diffuses through the substrates fibres. The calculated initial hydrolysis rate corresponds to the linear range of the short time scale.  The hydrolysis rate of the Kraft pulp is significantly higher than the rest of the substrates. Increasing the enzyme dosage by ten times leads to an increase of hydrolysis rate of the Kraft pulp by a factor of 5.1. On the other hand, an equivalent increase in enzyme dosage on the BCTMP, sawdust and wood chips only increase the hydrolysis rates by factors of 2.4, 2.8 and 1.3, respectively.   Table 3-3: Initial Hydrolysis Rates of each substrates at high and low dosages.  Rates (ḡ/h) Substrate Low Dosage High Dosage Kraft 0.01269 0.06447 BCTMP 0.00907 0.02138 Sawdust 0.00705 0.01981 Wood Chips 0.00202 0.00257  Wood-chips and sawdust have the same composition but different morphology. The difference between the hydrolysis rate of the substrates is evident. This implies that available internal and external surface area directly affects the enzymatic hydrolysis. This hypothesis is also widely 35  explored by researchers [33], [36], [39]. Composition also affects the hydrolysis rate. The difference between hydrolysis of the pulps and wood is obvious (see Table 3-3). The normalized hydrolysis rate is essentially the change in glucose with respect to the initial glucose from cellulose fibres in the solid sample over time. The initial hydrolysis rate is directly correlated to the enzyme adsorption which is an indicator of cellulose accessibility [44]. This parameter then represents a quantity of accessible cellulose, which also correlates to morphology characteristic of the substrate.  3.4 Conclusions The enzyme used for this work shows evident activity and can be used for the work on compressed wood-chips. These results demonstrate that the extent of enzymatic hydrolysis depends on several conditions. At optimal conditions for the enzyme (i.e. optimal temperature, pH and high loading), glucose yield depends on the composition of the substrate and mass transfer limitation or cellulose accessibility. Kraft pulp produced the greatest amount of glucose over time because of its low lignin content and high surface area. On the other hand, wood-chips showed the lowest sugar concentration over time. The comparison of enzymatic activity of raw wood-chips and sawdust is of special interest. Changes in hydrolysis rates of wood due to different degrees of compression should fall between the raw wood-chip and sawdust enzyme hydrolysis curves, as the mass transfer limitation should be overcome with mechanical treatment. Cellulose accessibility is to be characterized by measuring the normalized glucose hydrolysis rate. The enzyme serves as a marker to measure changes in cellulose accessibility and morphology. Work in this chapter validates the use of this enzyme as well as its utility for evaluating cellulose accessibility. 36  Chapter 4: Evaluation of Wood-Chip Compression  In this section, the results of the compression experiments are reported. Two compression experiments followed by enzyme treatments evaluate the effect of compression ratio and compression rate on cellulose accessibility, extractive content and morphology of wood-chips. It is expected that cellulose accessibility will increase with compression. For each experiment, the specific methodology and procedure for the measurements are explained. The data is presented, and conclusions are drawn.    4.1 Experimental Methodology The general experimental procedure is represented in Figure 4-1. The experimental work started with sample preparation and compositional analysis, followed by steaming, mechanical treatment and enzyme impregnation. In this section, cellulose accessibility, extractive content and microscopy measurements techniques are outlined in detail.   Figure 4-1: Experimental layout for this work. Step 1: Compositional analysis of wood. Step 2: Mechanical treatment of wood. Step 3: Measurement of extractives and microscopy. Step 4: Enzyme impregnation and hydrolysis of wood. Step 5: Filtrate liquid fraction and perform analysis of sugars released. 37  First, the wood-chips were screened by size via a screen shaker. Over-sized wood chips and fines were discarded while keeping medium size chips (~1.5 cm x 1.5 cm x 2.5 mm) for the experimental work. Wood-chips were then manually selected to consistent shapes and thickness.  Compositional analysis of the wood-chips was done according to well established protocols explained in detail in Section 3.1. The other steps of the experimental process are broken down in greater detailed in the sections that follow. In this work, two different mechanical compression trials were performed:  • Lab Compressor (see section  4.2). • Material Testing System (MTS) Compression (see section 4.3).   4.1.1 Enzyme Treatment Enzyme impregnation and hydrolysis in these experiments follow the same procedure as in section 3.1 of this thesis. The wood-chips are first washed with deionized water. The desired dosage of enzyme was mixed with buffer solution and wood-chips in an Erlenmeyer flask to bring the mixture to a 2.5% consistency. Quickly, the chips are vacuum impregnated for 15 seconds. The flasks were placed in an incubator shaker at 50 ℃ and 120 RPM for 2 hours. The enzyme (EL2013/000649) used consist of a cellulase mixture supplied by ABEnzymes. The hydrolysis experiment was run in an incubator shaker stirring the samples at 120 RPM. The sodium acetate buffer used maintained the pH of the samples at an optimum of 4.8. Each sample was run in triplicates. As much as 10 ml of sample was removed for every sampling period (i.e. 15 min, 30 min, 45 min and 2 hours) from the initial ~150 ml liquor. The samples were then filtered with a 0.2 μm syringe driven filter and prepared for HPLC analysis as explained in section 3.1 of this work. 38  4.1.2 Extractive Measurements The extractive content of the wood-chips was determined by conventional extractive determination methods [46]. The Dionex ASE 350 Accelerated Solvent Extraction System (ASE) was used to remove water and ethanol soluble extractives from the wood-chips (Figure 4-2). This is accomplished by injecting hot water and ethanol through the extraction cells of the ASE.    Figure 4-2: Accelerated Solvent Extractor (ASE). The milled wood chips are placed in the extraction cell at the top. Ethanol and water are injected into the cell removing the extractives. A filter prevents the loss of solids. The extractive liquor is collected in the bottles are the bottom.    The amount of extractive in the wood-chips is then determined as the difference of oven dried weight (ODW) of sample before and after the extraction using equations (4-1) and  (4-2).  39   𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 = 𝑂𝐷𝑊𝑠𝑎𝑚𝑝𝑙𝑒 − 𝑂𝐷𝑊𝑠𝑎𝑚𝑝𝑙𝑒 𝑎𝑓𝑡𝑒𝑟 𝐴𝑆𝐸 (4-1)   %𝐸𝑥𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒 =𝑊𝑒𝑖𝑔ℎ𝑡𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑖𝑣𝑒𝑂𝐷𝑊𝑠𝑎𝑚𝑝𝑙𝑒∗ 100  (4-2)   The water extract, which contains sugars in oligomeric form, is treated with 4% w/w sulfuric acid to break down these complex sugars to their measurable monomeric form. The composition of the water extract was determined with the HPLC as explained in section 3.1.   4.1.3 Microscopy Optical microscopy has been used in several works to qualitatively observe the effect of compression on wood chip morphology [6], [12], [13]. For convenience, light microscope and a microscope camera were utilized in this work. Because of equipment availability, a similar protocol to that of McIntosh, 2016 [13] was implemented for the preparation of samples and the subsequent microscopy.  Representative wood samples from the compression experiments were randomly selected. These samples were soaked in deionized water for two days. Then, with a sledge microtome 60 μm sections were carefully cut. These were thicker cross-sections compared to those in other works [12], [13] but were necessary due to the long, wide shape of the wood-chips. Thinner than 60 μm cross-section cuts resulted in damage of the sample. Then, the samples were dyed with Rit dye to improve the contrast in the image features [13]. The samples were mounted to microscope slides with Permount solution and were left to dry for a day. Upon drying, the slides were ready for 40  microscope observation. Figure 4-3 displays the sample preparation and microscopy process. The slides were observed with a Nikon transmission stereo microscope coupled to a 10 Mpx top-mount digital camera to capture pictures. The images were enhanced with computer software to further improve the contrast and quality of the images.     Figure 4-3: Preparation of sections and microscopy. First, thin sections are cut with microtome. The sections are dyed and washed. Using a Nikon light microscope, the sections are observed. A camera mounted to the microscope takes pictures of the images, which are saved to a computer software for image processing.     4.2 Lab Compressor Trial In preparation of this trial, screened wood-chips were shipped to FPInnovations, Pointe-Claire. The lab compressor (see Figure 4-4) is typically used for compression testing at small scales. The lab compressor is capable of compressing wood-chips at different degrees of compression or at different pressures. The objective of this trial is to characterize the effect of 41  compression of wood-chips on their cellulose accessibility in enzyme impregnation. The effect of the compression on the extractive content of the chips is also evaluated.   Figure 4-4: Lab compressor, which is also called chip juicer, is a hydraulic piston used to compress wood-chips. The trial was conducted by FPInnovations in Pointe-Claire.   4.2.1 Trial Conditions The wood-chips were first pre-steamed with saturated steam at atmospheric conditions for 20 min to soften the wood structure. Then, the chips were compressed with the lab compressor. The trial conditions are summarized in table 4-1. Each sample (~ 200 g) was compressed in duplicates. The time needed to reach each compression ratio (CR) was recorded. The power input of the piston was estimated from the motor design specifications (115 V and 9.5 A) and the time needed to achieve each compression. 42  Table 4-1. Lab Compressor Trial Conditions Time to reach target ratio Trial 1 Trial 2 Power Input Estimate (Kw-h/t) Compression ratio = 2:1 15.30 seconds 15.10 seconds 23.1 Compression ratio = 3:1 21.34 seconds 21.47 seconds 32.5 Compression ratio = 4:1 35.32 seconds 32.78 seconds 51.7  Control wood-chips were selected as well for comparison purposes. The compressed chips were sealed and returned to UBC for enzyme impregnation. The impregnation included a 15 second vacuum impregnation prior to hydrolysis to simulate the compression and uptake effect that takes place between screw feeder compression and impregnation (see Section 2.3.2). The enzyme hydrolysis was performed in the same way as in section 3.1 of this work. The conditions of the hydrolysis of the samples are summarized in table 4-2. Unfortunately, extractive liquor samples were not collected. Therefore, the extractive content is determined directly from the chips as explained in Section 4.1.2 of this work.   Table 4-2. Enzyme Treatment conditions for the lab compressor trial. Biomass SPF Wood chips (compressed, untreated) Solid content ~ 4 grams  Consistency 2.5 % Enzyme ABenzyme cellulase  Dosages 100 mg enzyme / g odw and 1000 mg enzyme / g odw Buffer Sodium acetate  pH ~4.75 Temperature 50 ℃ Sampling 15 min, 30 min, 45 min, 2 hours Starting Volume ~150 ml  43  4.2.2 Effect of Compression on Cellulose Accessibility Figure 4-5 displays the normalized amount of glucose over time at 0.1 g enzyme/g odw. Figure 4-6 shows the results of increasing the dosage by a factor of 10 (1 g/g odw). The HPLC results give concentration of glucose and other sugars in grams per liter [g/L]. The normalized glucose amount [ḡ] was calculated based on the initial composition of the biomass and the reaction liquor. At first sight, a high increase of released glucose for the first 15 min is noticeable in Figure 4-5 and Figure 4-6. Typically, cellulases take about 20 min to show significant hydrolysis after their introduction to the substrate [52]. However, this enzyme mixture adsorbs on to the external surface of the chips immediately after being introduced and starts degrading shortly after. This can be attributed to the high dosages used. Considering this, the linear range from 15 min to 2 hours (i.e. after the sampling started) is considered. It is also very apparent the increase of enzyme activity with increasing compression ratio for both dosages. Compressing the wood chips with the lab compressor to a CR of 4:1, leads to an increase in sugar released by a factor of approximately 1.4. This could be attributed to increasing cellulose accessibility due to changes in wood-chip morphology (i.e. formation of cracks and changes in pore volume).  The linear range of hydrolysis from 15 min to 2 hours is used to calculate the initial hydrolysis rate [ḡ/h]. The initial hydrolysis rate was evaluated in terms of energy input of the lab compressor as displayed in Figure 4-7.  Hydrolysis rate does not increase significantly at 2:1 compression ratio. The sharper increase starts at higher compression ratio of 3:1 at lower dosage and at 4:1 for higher dosage. This observation suggests that 2:1 compression ratio was insufficient to cause changes in the structure of the wood-chips. Further increase in compression may lead to the formation of more pathways (i.e. cracks) for enzyme to attack the cellulose fibres.  44   Figure 4-5: Normalized glucose released over time with 0.1 g of enzyme/ g of odw. Wood-chips compressed at different compression ratios were treated with cellulase enzyme for 2 hours.     Figure 4-6: Normalized glucose released over time for a high dosage with 1 g of enzyme/ g of odw. Wood- chips compressed at different compression ratios were treated with cellulase enzyme for 2 hours.   00.00050.0010.00150.0020.00250.0030.00350 0.5 1 1.5 2 2.5Glucose (ḡ)Hydrolysis Time (h)4:13:12:1control0.0040.0050.0060.0070.0080.0090.010 0.5 1 1.5 2 2.5Glucose (ḡ)Hydrolysis Time (h)4:13:12:1control45  Compression via the lab compressor improved the cellulose accessibility of the wood-chips and could be a better alternative than increasing the enzyme dosage by small amounts. In this case, sugar production rate increased by 34% when the enzyme dosage was increased by a factor of 10 in the control chips. In comparison, sugar production rate increased by 40% when the highest compression of 4:1 was applied at a lower enzyme dosage. This agrees with Grönqvist et al. 2014 claims that the amount of accessible fibre is the more limiting factor for enzyme hydrolysis than the enzyme dosage [40]. Furthermore, this highlights the importance of compressing chips to improve enzyme uptake and reduce chemical usage.   Figure 4-7: Hydrolysis rate as function of the estimated power consumed by the lab compressor. The dashed line represents the rate of hydrolysis of control chips attained by increasing the enzyme dosage by a factor of 10. Compressing at 4:1 achieved greater cellulose accessibility than increasing the dosage.   2:1 3:14:12:13:14:100.00050.0010.00150.0020.00250 10 20 30 40 50 60Hydrolysis Rate (ḡ/h)Power Input (Kw-h/t)1 g enzyme/ g odw0.1 g enzyme/ g odw46  4.2.3 Effect of Compression on Extractives The extractive content of the wood-chips decreases with compression ratio as shown in Figure 4-8. This agrees with the findings by Nelsson et al. 2012 [6] and it is expected as compression squeezes water and extractives out of the wood chips. In this case, a maximum geometrical CR of 4:1 resulted in a removal of ~3.16% ethanol and water-soluble extractives compared to the control wood-chips. Compression of 2:1 and 3:1 resulted in the removal of ~0.98% and ~1.68% extractives, respectively. The carbohydrate composition of the extractives was determined by measuring the monomeric and oligomeric sugars of the liquor extracted from the wood by treating the liquor to 4% w/w acid hydrolysis prior to determination of sugar concentration. Figure 4-9 shows that the average composition of the carbohydrates was mostly galactose and mannose. The glucose content contribution to the results from Section 4.2.2 is minimal.  Figure 4-8: Extractive water and ethanol soluble extractive content of the wood chips compressed at increasing compression ratios.  0.01.02.03.04.05.06.0Control 2:1 3:1 4:1Extractives  in wood (%)47   Figure 4-9: Average carbohydrate composition of water extract.      4.3 MTS Compression Trial The following experiment was designed to test higher compression ratios and different compression rates. A cylindrical compressor and a compression chamber were designed to simulate a pocket element in a screw feeder. The compression chamber has holes on the sides for the extrusion of water. A small aluminum cap catches the liquid. The bottom of the chamber is heated so that the chamber’s temperature is 90 ℃. The compressor and compression chamber were mounted in an MTS 810 Load Frame (See Figure 4-10). A highly accurate load cell measured the uni-axial force applied over time. The force and displacement data were saved for further analysis (see Chapter 5). The design drawings and engineering considerations for the construction of the compressor and compression vessel are outlined in Appendix A.  051015202530354045Arabinose Galactose Glucose Xylose MannosePercent Composition (%)48    Figure 4-10: MTS set-up. Left: MTS 810 Load Frame. Right: Steel Compressor, compression chamber and aluminum cup at the bottom. The bottom section is heated.   4.3.1 MTS Trial Conditions Wood-chip samples of ~ 25 grams were steamed in boiling water for 10 minutes, and then weighed. The wood-chips were horizontally stacked in the compression chamber so that they are compressed perpendicular to the grain. The compression conditions are outlined in Table 4-3. For the first test, wood-chips were compressed at different compression ratios for a constant compression time. The compression ratios ranged from 3:1 to the highest achievable of 6:1. For the second test the compression was done at variable times but at constant compression ratios. The range of compression times was selected based on screw feeder retention times in mill trials by other authors [6], [23]. 49  The extracted liquid and the compressed chips were weighed at the end of each compression and then stored for further analysis. For each compression, the force and displacement over time was recorded by data acquisition and computer software. The data was then transferred to an Excel file for data processing. The weight of extractive water for each compression test was also recorded. Each measurement was done in duplicates to ensure there is no significant difference for each test due to the stacking of the wood-chips.   Table 4-3. MTS compression conditions Test 1: Fixed Compression Time Test 2: Variable Compression Time CR Time (sec) CR Time (sec) 3:1 5 3:1 3, 5, 10 4:1 5 - - 5:1 5 5:1 3, 5, 10 6:1 5 6:1 3, 5, 10   4.3.2 Effect of Compression Ratio and Compression Rate on Load In this section, the effect of compression ratio and compression rate on the maximum load is presented. As expected, load increases with compression ratio as shown in Figure 4-11. It is worth noting the sharper increase in load at smaller compression ratio, starting from 3:1 compression. Further increase in compression ratio resulted in smaller increments in maximum load. Increasing compression from 3:1 to 4:1 resulted in a load increment of 13 kN. Meanwhile, increasing compression ratio from 5:1 to 6:1 yielded a load increment of 11 kN and 7 kN, respectively. At a compression ratio of 6:1, the wood-chips could not be compressed any further, resulting in no further increment in the compressor’s displacement and no consequent increment 50  in load. Maximum load can be an indicator of how much compression ratio affects the structure of wood. Sample load and displacement plots are provided in Appendix B.   Figure 4-11: Maximum load measured [kN] at each compression ratio. The compression time was 5 seconds for all cases.  The effect of compression time or compression rate was also studied. Figure 4-12 shows maximum load measured at changing compression times for given compression ratios (i.e. 3:1, 5:1 and 6:1). Compression time has little to no effect on the maximum load applied to the wood-chips. Load seems to barely increase with compression time for the 3:1 and 6:1 compression ratio. Nevertheless, the maximum load decreases slightly with compression time for the 5:1 compression case. The discrepancy across the cases could be attributed to variation in wood-chip stacking. Overall, maximum load was not significantly affected by compression time.  05101520253035400 1 2 3 4 5 6 7Max Load (kN)Compression Ratio51   Figure 4-12: Maximum load measured at increasing compression times.    4.3.3 Effect of Compression Conditions on Cellulose Accessibility Cellulose accessibility was assessed in the same way as in the lab compressor trial (see section 4.2.1). In this case, a cellulase dosage of 1 g enzyme/g odw on highly compressed wood-chips was used. Figure 4-13 displays the normalized glucose released into the liquor over time at varying compression ratios. Figure 4-14, on the other hand, displays the normalized glucose released over time at varying compression times. Enzyme hydrolysis improved with increasing compression ratio, which agrees with findings in section 4.2.2. The normalized glucose yield increases in a systematic fashion with compression ratio. In other words, compression increased enzyme’s access to the cellulose fibres at every compression ratio. The method of quantifying sugars shows that there are differences in accessibility when compressing at different CRs. 05101520253035400 2 4 6 8 10 12Max Load (kN)Compression Time (sec)3:1 ratio 5:1 ratio 6:1 ratio52   Figure 4-13: Normalized glucose released over time for a high dosage (1 g of enzyme/ g of odw). Wood-chips were compressed at different compression ratios and at constant compression time of 5 sec.   Figure 4-14: Normalized glucose released over time for a high dosage (1 g of enzyme/ g of odw). Wood-chips were compressed at different compression times at a constant compression ratio of 5:1.  0.0040.0050.0060.0070.0080.0090.010.0110.0120.0130.0140 0.5 1 1.5 2 2.5Glucose (ḡ)Hydrolysis Time (h)6:15:13:1Control0.0040.0050.0060.0070.0080.0090.010.0110.0120 0.5 1 1.5 2 2.5Glucose (ḡ)Hydrolysis Time (h)5 sec10 sec3 sec53  Normalized glucose yield was not dependent on the compression time or rate. The highest enzyme hydrolysis was shown for the sample compressed for 5 seconds (Figure 4-14), which had the highest load (see Figure 4-12). This could imply that hydrolysis is dependent on the load rather than compression time. The hydrolysis rate (see Figure 4-15) increases first rapidly with compression ratio, then it slows at very high loads. This indicates that further compression leads to no significant increases in hydrolysis rate, as the wood could not be compressed any further.    Figure 4-15: Normalized enzyme hydrolysis rate as function of compression ratio applied to the wood-chips at increasing compression ratios and at a compression time of 5 seconds.     0.00000.00100.00200.00300.00400.00500.00600 1 2 3 4 5 6 7Hydrolysis Rate  (ḡ/h)Compression Ratio54  4.3.4 Effect of Compression on Extractive Removal The extractive water was weighed in the capture cup for every compression run. Figure 4-16 displays the amount of water extract collected at each compression ratio for a compression time of 5 seconds. The amount of water expelled varies almost linearly with increasing compression ratio. Figure 4-17 displays the relationship of the amount of extractive water with compression time. For the 3:1 and 6:1 compression case, the amount of water expelled increases with compression time. That is not the case for the 5:1 compression, which follows the same trend as the force and compression time plots (See Figure 4-12). This implies that the amount of water expelled is mostly affected by the compression load and to a lesser extent by the compression time or rate. This agrees with Thornton et al. 1978, who noted that the amount of ether soluble extractives did not vary significantly with screw feeder feed rate [27]. It is important to consider that in conventional screw feeders, expulsion of water is aided by gravity because of the horizontal configuration of the housing screen. Therefore, higher extractive removal is expected at slower feed rates. In either case, the amount of extractive expelled is an indication of how much compression was achieved. This water removal could explain the increase in enzymatic hydrolysis for the compressed chips because extractives act as inhibitors of cellulase enzyme by restricting the enzyme’s access to the cellulose fibres [34], [35].  The carbohydrates and lignin composition of water extracts was measured (see Figure 4-18). The extractives contains hemicellulose sugars and lignin (apart from the terpenoids and other fatty acids) [18], which are also known to inhibit cellulase hydrolysis [14]. The concentration of dissolved lignin and hemicellulose sugars in the water extract increases with increasing CR. Therefore, the compressed wood showed improved enzyme performance in part due to the removal of extractives. 55   Figure 4-16: Extractive water recovered from the compression chamber at increasing compression ratios for a compression time of 5 sec.   Figure 4-17: Extractive water squeezed out of the compression chamber at increasing compression time. 0246810123 4 5 6Extractives Water  (g)Compression Ratio024681012143 5 10Extractive Water (g)Compression Time (s)3:1 ratio5:1 ratio6:1 ratio56   Figure 4-18: Extractive water composition for all compression tests. The composition includes the total amount of carbohydrates as well as the soluble lignin content. Here, LC is 3:1 compression, MC is 5:1 compression ratio and HC is 6:1 compression ratio for 5 sec of compression time. LR is 3 sec of compression and HR is 10 sec of compression time at 5:1 compression.   4.3.5 Effect of Compression on Wood-Chip Morphology  Microscopy images illustrate how compression affected the morphology of wood-chips at the microscopic level. Figure 4-19 shows micrographs of wood-chip cross-sections that were compressed at different compression ratios. The control wood-chips are also shown for comparison. The figure also shows pictures of wood-chips from each run. First, Figure 4-19 (a) displays the cross-section of a wood-chip that has not been compressed, described as 'control'. Normal arrangement and shape of cell walls is present. Upon compressing the wood-chips at a 3:1 ratio, buckling of the cell walls is noticeable in some areas, and the formation of small cracks are 00.20.40.60.811.2Arabinose Galactose Glucose Xylose Mannose TotalCarbohydratesLigninConcentration [g/L]LC- 3:1 MC- 5:1 HC- 6:1 LR - 3 sec HR - 10 sec57  observed in Figure 4-19 (b). The buckling or distortion of the cell wall is a sign that irrecoverable damage was done to the cell structure. Further compression leads to the formation of cracks in the buckled wall.  At a higher compression ratio of 5:1, the smaller cracks start to propagate forming even larger and wider cracks across the cell walls as seen in Figure 4-19 (c). At this point, the cracks are so large that they are visible in the wood-chip surface. Finally, further compression of the wood-chip to the highest possible compression ratio results in further crack formation as well as distortion of both early and late wood cell walls as seen in Figure 4-19 (d). This leads to fragmentations of the cell walls and separation in the middle lamella. Moreover, wood-chips at the large scale are broken into smaller chips because of the high load.  This visual study explains the increase in enzyme hydrolysis rate by: the formation of internal cracks in the wood structure and by size reduction of wood-chips. The former results in an increase of internal surface area and the latter, an increase in external surface area [12].   58    Figure 4-19: Microscopy samples from (a) control chips, (b) wood-chips compressed at 3:1 compression ratio, (c) wood-chips compressed at 5:1 compression ratio, and (d) wood-chips compressed at 6:1 compression ratio.59   4.4 Conclusion Compression is a pre-treatment method that has several advantages in the pulp production. In this work, it was observed that compression leads to a removal of extractives and induces changes in wood-chip morphology through cell buckling and crack formation and propagation. These behaviors have all been reported in literature [6], [13], [18]. In both the lab compressor trial (see section 4.2.2) and in the MTS compression trial (section 4.3.3), an increase in enzyme reactivity was observed when increasing compression ratio. This increase in cellulose accessibility to enzymes is explained by the removal of extractives [19], as well as the formation of cracks that serve as pathways for the enzyme [14]. However, it is not clear which of these factors (extractive removal and surface area) contributes more to the increase in accessibility. In both hydrolysis rate curves (see Figure 4-7 and Figure 4-15), it is evident that high compression ratios are necessary in order to impart sufficient damage to the wood-chip morphology. In the case of the lab compressor, compression ratios greater than 3:1 lead to a noticeable increase in hydrolysis rate. At 4:1 compression ratio, the hydrolysis rate was higher than the hydrolysis rate of the control chips treated with 10 times higher dosage. Accessibility increases with increasing MTS compression ratio until CR=5:1. Further increases in compression ratio led to diminishing changes in hydrolysis rate. The smaller increment was mainly due to the decrease in chip size.  Sugar quantification provides a method to characterize changes in cellulose accessibility due to compression.  Increase in glucose yields gives information of the cellulose fibre accessibility to the enzyme. This represents a measurement of morphology.60  Chapter 5: Screw Feeder Analysis  A few screw feeder models have been proposed for the characterization screw feeding of biomass materials [53], [54]. Particularly, Dai et al. 2007 [55] proposed a model for biomass screw feeding with tapered and extended section. This model estimates the torque requirements for different screw sections given the material feed in the hopper. A simplified formulation of the equations from the model is outlined in this section, as well as, the application of such equations to the experimental data from the MTS compression (see Section 4.3.2). Predicted stress and torque curves for given compression ratios is provided. Finally, screw feeder performance curves demonstrate the effect of increasing compression on the power consumption, capacity and changes in wood-chip cellulose accessibility.  5.1 Screw Feeder Torque Equations  5.1.1 Formulation In the model developed by Dai et al. 2007, a screw feeder of constant screw shaft diameter and tapered barrel is considered. Figure 5-1 displays a diagram of the surface stresses acting on an element of a tapered section. There are five boundary surfaces (see Figure 5-2) where the pressure acts: the driving side, trailing side, shear surface, through surface and core shaft surface.  The net force (i.e. summation of all forces in the element) is solved in terms of the force applied by the driving side. The net stress acting on the element is then described in equation 5-1:   𝜎𝑑𝑎 =  𝐹𝑑𝑎𝜋(𝑅𝑜2 − 𝑅𝑐2) (5-1) 61  Where Fda is the force on the driving side, Ro is the radius of the screw flight and Rc is the radius of the screw core shaft.    Figure 5-1: Stresses on material element in tapered section. Image extracted from Dai et al. 2007 [55].   Figure 5-2 : Pressure surfaces of a screw element. Image extracted from Yu et al. 1997 [53].  62  The torque requirements for a screw feeder are dependent on the resisting forces acting on each surface of the screw and flights [56]. The torque generated in an element is then described as summation of torques generated by each surface of the screw and flights:  𝑇𝑖 =  𝑇𝑑 + 𝑇𝑐 + 𝑇𝑓 + 𝑇𝑡𝑖𝑝 (5-2) where Td is the torque due to driving flight, Tc is the torque due to core shaft surface, Tf is the torque due to trailing flight, Ttip is the torque due to the flight tip, all in units of Nm. The torque requirement for the driving side is:   𝑇𝑑 =  2𝜋𝜎𝑑𝑎 ∫ 𝑟2 tan(𝛼𝑟 + ∅𝑓)𝑑𝑟𝑅𝑜𝑅𝑐 (5-3) Here, tan 𝛼𝑟 =𝑃2𝜋𝑟 and tan ∅𝑓 =  𝜇𝑓 , where µf is the wall friction coefficient between bulk solids and screw flight, and αr is flight helical angle at screw radius, r.  The torque requirement for the core surface is:   𝑇𝑐 =  2𝜋𝑅𝑐2𝑝𝜇𝑤𝜎𝑤𝑎 cos 𝛼𝑐 (5-4)  Where µw is the friction coefficient between the shaft and bulk, αc is the flight helical angle at the shaft and p is the pitch length. The torque generated by the trailing side of the screw flight is obtained by integration of equation 5-5:   𝑇𝑓 =  2𝜋λ𝑠𝜎𝑥𝑓 ∫ (tan ∅𝑓 − tan 𝛼𝑟)𝑟2𝑑𝑟𝑅𝑜𝑅𝑐 (5-5) Where λs is the ratio of the normal stress to axial stress for bulk solids sliding on the surface, σxf is the stress of the trailing side. The torque generated by the flight tip is:   𝑇𝑡𝑖𝑝 =  𝛾𝜎𝑤𝑎𝑝𝜇𝑓𝑅𝑜sin 𝛼𝑜 (5-6) 63  Where γ is the screw flight thickness, and tan 𝛼𝑜 =𝑝2𝜋𝑅𝑜 is the screw flight angle at the outside radius. The σwa is the average normal stress perpendicular to trough wall and core shaft.    5.2 Methodology 5.2.1 Screw Configuration  Here, the equations are adapted to the geometries of the TMP screw feeder. In the case of the conventional screw geometry, the screw shaft diameter increases, and the casing diameter stays relatively constant along the screw channel. Figure 5-3 displays the configuration and geometries of the screw used in TMP. The same equations are applied but in this case the radius of the screw shaft changes along the z axis and the stresses are simplified to a single compressive stress in all surfaces.    Figure 5-3: Screw configuration of an increasing diameter shaft screw feeder. The clearance H decreases from H0 to H1 along the pitch length.   64  The screw radius increases as a function of the length of the screw, z. The change in radius can be easily calculated at each point by knowing the initial screw diameter and the final compression ratio:  𝑟(𝑧) = 𝐻𝑜 − (𝐻𝑜 − 𝐻𝑓)𝑧𝐿 (5-7) Here H represents the clearance between the barrel surface and shaft surface.   5.2.2 Assumptions  In this simple analysis, all surface stresses are assumed to equal a total compressive stress that pushes the water out the wood-chips (see Figure 5-4). This stress will increase because of the friction of the bulk being displaced along the narrowing channel of the screw feeder. The mechanical properties of the wood-chips including friction coefficient of wood with steel and angles of internal friction were approximated based on the work by Stasiak et al. 2015 [57]. These properties are assumed to remain constant. For reference, the initial parameters and geometries for the calculations are summarized in table 5-1.   Figure 5-4: Total stress in the screw element. An equally distributed load is assumed in the bulk. 65  The typical compression ratios of modern screw feeders range in the order of 3:1 in plug screw feeders to 5:1 in Impressafiner [1], [23]. Compression ratios of 3.6:1 to 5:1 are analyzed as these are typical of high compression screws. Additionally, high compression ratios are needed to achieve sufficiently high cellulose accessibility.   Table 5-1. Screw Feeder and Material parameters. Asterisk represents values taken from a screw feeder. Parameter Symbol Value Ref. Screw Length L 1.7 m * Radius of Barrel Rb 0.1875 m * Initial Diameter of Screw Dc (0) 0.08 m * Wall Friction Coefficient µw 0.53  [57] Stress Ratio λs 0.704 [53] Angle of Internal Friction δ 0.55 rad [57] Compression Ratios CR 3.6 - 5.5  [1], [6], [23] Flight Thickness γ 10 mm * Pitch Length p 0.34 m  *   5.2.3 Experimental Data Wood-chip compression was achieved via an MTS compressor to simulate the compression in a screw element (see Figure 4-10). The MTS load frame was coupled to a load cell that could measure load over time. The results of the experiments are load curves which displays the vertical load applied to the chips as a function of the vertical strain. The strain is directly related to the compression ratio. The data from these plots is used to experimentally derive equations of the load applied to wood-chips with a best fit polynomial curve (see Figure 5-5). The polynomial equations that characterize the compression of wood-chips are displayed in Table 5-2. To fully characterize 66  the compression of the wood-chips, two polynomial equations were developed for two regions: region of low strain and high strain. This was necessary to accurately fit all the points in the data.  The load applied is then assumed to be the total net stress applied to the element (see Figure 5-4). This load was used in equations (5-1 to 5-6), following the simplification stated in this chapter. Additionally, the hydrolysis rate curve of the experimental work is included into this analysis.   Figure 5-5: Load plot with the polynomial fit of the data.  Table 5-2. High degree polynomial equations that describe load as function of strain. Equation First Region (ε <0.729) Second Region (ε >0.729) 𝑭(𝜺) = 𝑴𝟓𝜺𝟓 + 𝑴𝟒𝜺𝟒 + 𝑴𝟑𝜺𝟑 +𝑴𝟐𝜺𝟐 + 𝑴𝟏𝜺 + 𝑩   M1 = 4.095004706; M2 = -52.29854327; M3 = 266.8233384; M4 = -517.6093807; M5 = 361.8618811; B = 0; M1 = 904193.77; M2 = -1179436.44; M3 = 0; M4 = 1000099.875; M5 = -519942.5452; B = -207755.3847;  05101520253035400 0.2 0.4 0.6 0.8 1Load (kN)Strain (mm/mm)67  5.2.4 Applying the Screw Feeder Equations Using the load and strain equations from the polynomial fit and the compression experiments, the stress and strain function is implemented into the equations (5-1 to 5-6) to determine torque requirements at each point in the screw. Equation 5-8 is used to determine the ideal retention time RT in a screw feeder [58]. Conversely, the rotational speed N of the screw can be computed if the screw length L, pitch and compression time are given:   𝑅𝑇 =𝐿𝑝 ∗ 𝑁 (5-8) The screw load or power can be computed simply by multiplying the torque T and screw speed N as in equation 5-9:  𝑃𝑠𝑐𝑟𝑒𝑤 = 𝑇 ∗ 𝑁 (5-9)  The theoretical capacity or flowrate can also be determined by computing the volume of the screw channel and assuming a constant feed of the wood-chips. The theoretical volume and capacity of the screw feeder are computed as shown in equations 5-10 and 5-11, respectively:  𝑉𝑡 = 𝜋𝑅𝑏2𝐿 − (𝜋3𝐿 ∗ (𝑅𝑐2 + 𝑅𝑐𝑅𝑐𝑓 + 𝑅𝑐𝑓2 )) (5-10)  𝑄𝑡 =𝑉𝑡𝑅𝑇 (5-11)  Where L is the screw length, Rb is the radius of the barrel, Rc is the radius of the core shaft and Rcf is the final radius of the screw shaft. In this case, the capacity of the screw feeder is idealized to be the available volume of the channel (i.e. the difference between the barrel volume and screw shaft volume). The theoretical capacity is assumed to be the volume available fed for a ideal retention time. 68  5.3 Results 5.3.1 Stress and Torque Predictions Figure 5-6 displays the maximum stress and torque achieved at the end of the screw channel for different compression ratios. Both the stress and the torque increase in an exponential manner when increasing the degree of compression. The stress build-up follows similar behavior as reported in a screw press model [59] and in experimental work involving the compression of wood-chips beds [60]. The order of magnitude of stress is similar to that calculated by a tapered screw press model by Zhong et al. 1992 [54], [59].  Figure 5-6: Stress applied to the wood-chips when compressing at different compression ratios. The corresponding torque requirements to feed the wood-chips at different compression ratios is also shown. 69  5.3.2 Performance of Screw feeder The performance of the screw feeder is assessed in terms of screw load and capacity. Screw load is the energy consumed by the screw when compressing the wood-chips. When the compression ratio increases, the load increases because of larger pressure buildup along the screw channel. Higher torque is necessary to push the wood-chips into a smaller channel; therefore, higher power is required. Conversely, increasing the compression ratio results in a reduced capacity because of the smaller volume available for the wood-chip feed. Increasing the compression rate (or decreasing RT) both increases the screw load and the capacity according to Equations 5-8 and 5-11. Figure 5-7 depicts the screw load and capacity as function of the compression ratio for different RTs of 3, 5 and 10 sec.   Figure 5-7: Performance curves of screw feeder operating at a different compression time.  70  Increasing the compression ratio involves a trade-off of power consumption and applied forces on the wood-chips (i.e. screw load) and capacity (i.e. production rate). Considering a desired RT, the balance of screw load and capacity is achieved at the intersection of the two curves. Industrially, however, the screw feeders operate at screw loads between 400 and 900 kW [23]. This discrepancy could be attributed the simplifications and assumptions made for these calculations. The calculations do not account the frictional shear force of the wood-chips as the result of the rotation of the shaft. Additionally, a constant pitch length is assumed; however, pitch usually decreases slightly along the screw channel. This would result in higher compaction pressure.  Moreover, the size of wood-chips used industrially is larger, which would result in even higher torque requirements [56] and consequently in higher power consumption. To compensate for this, stress relaxation zones (areas of increasing barrel diameter) are typically incorporated in the screw housing.  Factoring the cellulose accessibility of wood-chips to enzyme treatment after they are compressed gives a complete profile of the screw feeder performance. The cellulose accessibility was quantified as the rate of enzymatic hydrolysis to glucose. Figure 5-8 brings all the factors of screw feeder and enzyme impregnation together. The hydrolysis rate data is the same from Figure 4-15. Compressing higher than 4:1 ratio leads to substantially higher stress and energy consumption. In turn, the enzyme rate increases steadily until a compression ratio of ~5:1. The increase of enzyme rate diminishes past CR of ~5:1. In terms of production rate, increasing the compression ratio is clearly detrimental. Operating below the 5:1 compression ratio mark means significantly increasing the capacity while reducing energy consumption of the screw feeder at the cost of decreasing the stress and cellulose accessibility.  71   Figure 5-8: Performance curve of screw feeder operating at a RT = 5 seconds. Enzyme impregnation was done at a dosage of 1-gram enzyme per gram of dried wood. Capacity, screw load and cellulose accessibility are considered at each compression ratio.   5.4 Conclusions In this section, a method of combining experimental data from MTS compression and a simplified version of a model by Dai et al. 2007 [55] is presented. The screw load results fall within the same order of magnitude of typical values [23] while the torque is overestimated in these calculations. Although this model is not exact, it reflects realistic changes in performance (capacity and power) as a result of changing operating conditions (compression ratio and compression time). The best operation of screw feeder depends on the needs of the mill.  72  To increase the cellulose accessibility, the mill operator should: • Increase the compression ratio to increase the pressure. To reduce power consumption, the screw feeder operator should: • Decrease compression ratio to reduce the pressure. • Decrease screw speed. To increase capacity, the screw feeder operator should: • Decrease compression ratio to improve the volumetric flow. • Increase screw speed.  Changing the geometrical compression ratio involves a trade-off between improving cellulose accessibility by increasing the screw load, and the capacity or production rate. If the mill operator wants to increase both the pressure and capacity, the screw speed must be increased but this results in higher power consumption. According to this prediction, approaching the 5:1 compression ratio was optimum for achieving improved enzyme transport without significantly reducing the production rate.   73  Chapter 6: Conclusions and Future Work  This work started with the development of a method to characterize changes in cellulose accessibility in chapter 3. Chapter 4 outlined the experimental work of the compression of wood-chips. The method from chapter 3 was implemented to the experimental work from chapter 4. A simple screw feeder analysis was presented in chapter 5. This chapter lists the main takeaways and conclusions of each chapter. To conclude, suggestions for future work are presented.   6.1 Characterization of Cellulose Accessibility In this work, the sugar quantification is used as a metric to quantify the extent of cellulose hydrolysis which is correlated to the cellulose accessibility of the substrate. Enzymatic hydrolysis is affected by the composition of the biomass, as well as the internal and external surface area. The method of quantifying carbohydrate hydrolysis showed that refined pulp with minimal lignin content and high specific surface area had higher glucose released compared to wood. Moreover, the extreme mechanical treatment of wood-chip milling was compared to raw wood-chips. The difference of hydrolysis between these two extreme cases is attributed to a mass transfer limitation which may be overcome by mechanical treatment.    6.2 The Effect of Compression of Wood-chips on Cellulose Accessibility  The cellulose accessibility was affected by the compression of wood-chips. In both experiments performed in this work, the rate of enzymatic hydrolysis increased with compression ratio. This implies that mass transfer limitation was reduced by changing the morphology of the 74  wood-chips to allow for higher enzyme access. The main takeaways from wood compression trials are: • Cellulose accessibility increases systematically with increasing compression ratio. • Compression rate had no apparent effect on load or on cellulose accessibility.  • High compression ratios of at least 4:1 were needed to improve the enzyme performance significantly. It was possible to surpass the effect of increasing the enzyme dosage by 10 times.  • Compressing the wood-chips lead to reduction of extractives which are detrimental to the enzyme performance. • Microscopy showed that compression lead to morphological changes in the form of cell buckling and fractures formation, and fragmentation which improved the enzyme transport.  6.3 Screw Feeder Characterization The screw feeder operation was characterized with experimental data from wood-chip compression trials. The stress and strain relationship were applied to equations derived from a torque prediction model. The equations were modified for a screw feeder of increasing screw shaft diameter. With this model the stress buildup, torque requirements and power consumption were predicted. The capacity of the screw feeder was estimated to evaluate changes with compression ratio. The main takeaways from this analysis are: • The power predictions are in the same order of magnitude as typical values but are not identical. The simplifications and assumed geometries may be the reason behind the discrepancy.  75  • Increasing the degree of compression increases the stress applied to the wood-chips. This improves cellulose accessibility at the cost of decreasing capacity and increasing energy consumption.  • According to this prediction, the geometrical compression ratio of ~ 5:1 was the optimum area where the power consumption, capacity, cellulose accessibility compromise. Enzyme performance at this point is sufficiently high.  • Compressing higher than 5:1 lead to diminishing returns in enzyme uptake.   6.4 Suggestions for Future Work In this work, the effect of compression on cellulose accessibility of the wood-chips to enzyme treatment in terms of sugar released is successfully evaluated. Enzymes in the pulp and paper industry are primarily used for reduction in refining energy. Improvement in hydrolysis rate indicates that the enzyme gained more access to the cellulose fibres. However, it is not clear if this small improvement in hydrolysis rates due to compression will translate to corresponding reductions in refining energy.  Future work would involve a trial to determine the effect of compression ratios on the enzyme’s performance to further reduce refining energy. In theory, improved enzyme accessibility should induce greater reduction in refining energy. For this trial, compressed wood-chips would be treated with typical enzyme dosage. The chips would then undergo refining. The improvement in cellulose accessibility and enzyme performance would be assessed in terms of the enzyme’s ability to reduce refining energy. The study would link cellulose accessibility to enzyme performance.  76  Bibliography [1] D. Gorski, J. Hill, P. Engstrand, and L. Johansson, “Review: Reduction of energy consumption in TMP refining through mechanical pre-treatment of wood chips,” Nord. Pulp Pap. Res. J., vol. 25, no. 2, pp. 156–161, 2010. [2] A. Smook, G, Handbook for Pulp and Paper Technologists. 1989. [3] D. W. Francis, M. T. Towers, and T. C. Browne, “Energy Cost Reduction in the Pulp and Paper Industry – An Energy Benchmarking Perspective,” Montreal, 2002. [4] A. A. T. Tienvieri, E. Huusari, J. Sundholm, P. Vuorio, J. Kortelainen, H. Nystedt, Thermomechanical Pulping. Helsinki: Fapet Oy, 1999. [5] A. Salmi, L. I. Salminen, M. Lucander, and E. Hæggström, “Significance of fatigue for mechanical defibration,” Cellulose, vol. 19, no. 2, pp. 575–579, 2012. [6] E. Nelsson, S. Christer, L. Hilden, and G. Daniel, “Pressurised compressive chip pre-treatment of Norway spruce with a mill scale Impressafiner,” Nord. Pulp Pap. Res. J., vol. 27, no. 01, pp. 056–062, 2012. [7] C. J. Biermann, Handbook of Pulping and Papermaking, no. July. 1996. [8] E. Nelsson et al., “Low dosage sulfite pretreatment in a modern TMP-line,” Nord. Pulp Pap. Res. J., 2015. [9] K. R. Jegannathan and P. H. Nielsen, “Environmental assessment of enzyme use in industrial production-a literature review,” J. Clean. Prod., vol. 42, pp. 228–240, 2013. [10] P. W. Hart et al., “Selective enzyme impregnation of chips to reduce specific refining energy in alkaline peroxide mechanical pulping,” Holzforschung, vol. 63, no. 4, pp. 418–423, 2009. [11] A. Oksanen, K. Edelmann, J. Kataja-Aho, and A. Suurnäkki, “Enhancing dewatering of thermo-mechanical pulp (TMP) based papermaking through enzymatic treatment,” Holzforschung, vol. 65, no. 6, pp. 787–795, 2011. [12] C. de Choudens, G. Lombardo, G. Michalowicz, and A. Robert, “Destructuration of Wood Chips in a Cylinder Press,” Pap. ja puu, vol. 67, no. 3, pp. 108–116, 1985. [13] N. J. Mcintosh, “Measurement of the Changes to the Liquid Transport Properties of Wood Due to Compression,” The University of British Columbia, 2016. [14] R. P. Chandra, R. Bura, W. E. Mabee, A. Berlin, X. Pan, and J. N. Saddler, “Substrate 77  pretreatment: The key to effective enzymatic hydrolysis of lignocellulosics?,” Adv. Biochem. Eng. Biotechnol., 2007. [15] Government of Western Australia, “Selecting Timber,” 2018. [Online]. Available: tle.westone.wa.gov.au. [16] M. Sorieul, A. Dickson, S. J. Hill, and H. Pearson, “Plant fibre: Molecular structure and biomechanical properties, of a complex living material, influencing its deconstruction towards a biobased composite,” Materials (Basel)., 2016. [17] E. L. Back and L. Salmen, “Glass transitions of wood components hold implications for molding and pulping processes,” Tappi, 1982. [18] M. Tanase, P. Stenius, L. Johansson, J. Hill, and C. Sandberg, “Mass balance of lipophilic extractives around impressafiner in mill and pilot scale,” Nord. Pulp Pap. Res. J., vol. 25, no. 2, pp. 162–169, 2010. [19] T. Belt, F. Mollerup, T. Hänninen, and L. Rautkari, “Inhibitory effects of Scots pine heartwood extractives on enzymatic holocellulose hydrolysis by wood decaying fungi,” Int. Biodeterior. Biodegrad., vol. 132, no. March, pp. 150–156, 2018. [20] D. Fengel and G. Wegener, Wood: chemistry, ultrastructure, reactions. 1989. [21] B. A. Svensson, M. Rundlof, and H. Hoglund, “Sliding friction between wood and steel in a saturated steam environment,” J. Pulp Pap. Sci., 2006. [22] F. Peng and T. Granfeldt, “Changes in the microstructure of spruce wood chips after screw press treatment,” J. Pulp Pap. Sci., vol. 22, no. 4, pp. J140–J145, 1996. [23] M. Sabourin, J. Vaughn, N. Wiseman, J. B. Cort, and P. Galatti, “Mill scale results on TMP pulping of southern pine with pressurized chip pretreatment,” Pulp Pap. Canada, vol. 103, no. 6, pp. 37–42, 2002. [24] S. Azhar, G. Henriksson, H. Theliander, and M. E. Lindström, “Extraction of hemicelluloses from fiberized spruce wood,” Carbohydr. Polym., vol. 117, pp. 169–176, 2015. [25] M. Jackson and N. Wild, “Mechanical pulp mills,” in Energy cost reduction in the pulp and paper industry, T.C. Browne, 1999. [26] L. Johansson, F. F. Peng, R. Simonson, and T. Granfeldt, “Effects of chip pretreatment on CTMP pulp properties,” 1998. 78  [27] D. S. Thornton and B. E. Nunn, “The effect of a plug screw feeder on ether-solubles removal and power reduction during TMP manufacutre,” Tappi J., 1978. [28] J. W. Rowe, Natural products of woody plants I, vol. 1. 1989. [29] H. B. M. Lenting and M. M. C. G. Warmoeskerken, “Mechanism of interaction between cellulase action and applied shear force, an hypothesis,” J. Biotechnol., vol. 89, no. 2–3, pp. 217–226, 2001. [30] L. K. Singh and G. Chaudhary, Advances in Biofeedstocks and Biofuels. 2017. [31] C. Álvarez, F. M. Reyes-Sosa, and B. Díez, “Enzymatic hydrolysis of biomass from wood,” Microb. Biotechnol., vol. 9, no. 2, pp. 149–156, 2016. [32] M. Yoshida et al., “Effects of Cellulose Crystallinity, Hemicellulose, and Lignin on the Enzymatic Hydrolysis of Miscanthus sinensis to Monosaccharides,” Biosci. Biotechnol. Biochem., vol. 72, no. 3, pp. 805–810, 2008. [33] C. A. Mooney, S. D. Mansfield, M. G. Touhy, and J. N. Saddler, “The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods,” Bioresour. Technol., 1998. [34] Z. Li, Y. Yu, J. Sun, D. Li, Y. Huang, and Y. Feng, “Effect of extractives on digestibility of cellulose in corn stover with liquid hot water pretreatment,” BioResources, vol. 11, no. 1, pp. 54–70, 2016. [35] M. T. Holtzapple and A. E. Humphrey, “the Effect of Organosolv Pretreatment on the Enzymatic-Hydrolysis of Poplar,” Biotechnol. Bioeng., vol. 26, no. 7, pp. 670–676, 1984. [36] K. Sangseethong, L. Meunier-Goddik, U. Tantasucharit, E. T. Liaw, and M. H. Penner, “Rationale for particle size effect on rates of enzymatic saccharification of microcrystalline cellulose,” J. Food Biochem., 1998. [37] S. K. Dutta and S. Chakraborty, “Pore-scale dynamics of enzyme adsorption , swelling and reactive dissolution determine sugar yield in hemicellulose hydrolysis for biofuel production,” Nat. Publ. Gr., no. November, pp. 1–13, 2016. [38] X. Meng and A. J. Ragauskas, “Recent advances in understanding the role of cellulose accessibility in enzymatic hydrolysis of lignocellulosic substrates,” Curr. Opin. Biotechnol., vol. 27, pp. 150–158, 2014. [39] R. P. Chandra et al., “Comparison of methods to assess the enzyme accessibility and 79  hydrolysis of pretreated lignocellulosic substrates,” Biotechnol. Lett., vol. 31, no. 8, pp. 1217–1222, 2009. [40] S. Grönqvist et al., “Fibre porosity development of dissolving pulp during mechanical and enzymatic processing,” Cellulose, vol. 21, no. 5, pp. 3667–3676, 2014. [41] R. P. Chandra and J. N. Saddler, “Use of the Simons’ Staining Technique to Assess Cellulose Accessibility in Pretreated Substrates,” Ind. Biotechnol., vol. 8, no. 4, pp. 230–237, 2012. [42] K. Gourlay, V. Arantes, and J. N. Saddler, “Use of substructure-specific carbohydrate binding modules to track changes in cellulose accessibility and surface morphology during the amorphogenesis step of enzymatic hydrolysis,” Biotechnol. Biofuels, vol. 5, pp. 1–14, 2012. [43] M. J. Moura, P. J. Ferreira, and M. M. Figueiredo, “Mercury intrusion porosimetry in pulp and paper technology,” Powder Technol., vol. 160, no. 2, pp. 61–66, 2005. [44] M. Wiman, D. Dienes, M. A. T. Hansen, T. Van Der Meulen, G. Zacchi, and G. Lidén, “Cellulose accessibility determines the rate of enzymatic hydrolysis of steam-pretreated spruce,” Bioresour. Technol., vol. 126, pp. 208–215, 2012. [45] A. Sluiter et al., “NREL/TP-510-42618 analytical procedure - Determination of structural carbohydrates and lignin in Biomass,” 2012. [46] A. A. Sluiter, R. Ruiz, C. Scarlata, J. Sluiter and D. Templeton, “Determination of Extractives in Biomass: Laboratory Analytical Procedure (LAP); Issue Date 7/17/2005 - 42619.pdf,” Tech. Rep. NREL/TP-510-42619, no. January, pp. 1–9, 2008. [47] A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, and D. Templeton, “Determination of Sugars , Byproducts , and Degradation Products in Liquid Fraction Process Samples Laboratory Analytical Procedure ( LAP ) Issue Date : 12 / 08 / 2006 Determination of Sugars , Byproducts , and Degradation Products in Liquid Fraction Proce,” 2008. [48] J. Viell et al., “An efficient process for the saccharification of wood chips by combined ionic liquid pretreatment and enzymatic hydrolysis,” Bioresour. Technol., vol. 146, pp. 144–151, 2013. [49] S. Saka, “Wood as natural raw materials for cellulose acetate production,” Macromol. Symp., vol. 208, pp. 7–28, 2004. 80  [50] E. Sjostrom, Wood chemistry: fundamentals and applications. 1993. [51] P. Lahtinen, S. Liukkonen, J. Pere, A. Sneck, and H. Kangas, “A Comparative study of fibrillated fibers from different mechanical and chemical pulps,” BioResources, vol. 9, no. 2, pp. 2115–2127, 2014. [52] C. Jeong, A. M. Maciel, J. Pawlak, J. A. Heitmann, D. Argyropoulos, and O. Rojas, Following cellulase activity by the quartz crystal microbalance technique, vol. 2. 2005. [53] Y. Yu and P. C. Arnold, “Theoretical modelling of torque requirements for single screw feeders,” Powder Technol., vol. 93, no. 2, pp. 151–162, 1997. [54] Z. Z. B, “Theoretical and Experimental Analysis of the Compaction Process in a Tapered Screw Press,” The University of Newcastel upon Tyne, 1991. [55] J. Dai and J. R. Grace, “Biomass screw feeding with tapered and extended sections,” Powder Technol., vol. 186, no. 1, pp. 56–64, 2008. [56] J. Dai, “Biomass Granular Feeding For Gasification and Combustion,” The University of British Columbia, 2007. [57] M. Stasiak, M. Molenda, M. Bańda, and E. Gondek, “Mechanical properties of sawdust and woodchips,” Fuel, vol. 159, no. October, pp. 900–908, 2015. [58] C. Chamberlin, D. Carter, and A. Jacobson, “Measuring residence time distributions of wood chips in a screw conveyor reactor,” Fuel Process. Technol., vol. 178, no. February, pp. 271–282, 2018. [59] Z. Zhong and J. R. O’Callaghan, “A theoretical analysis of the extrusion pressure in a tapered screw conveyor,” J. Agric. Eng. Res., 1992. [60] M. Alaqqad, C. P. J. Bennington, and D. M. Martinez, “The permeability of wood-chip beds: The effect of compressibility,” Can. J. Chem. Eng., vol. 90, no. 5, pp. 1278–1288, 2012.  81  Appendices  Appendix A  : MTS Compression Design Drawings   A.1  Design and Analysis   Figure A1: Diagram of pressure developed inside a vessel. Pressure is higher in the hoop direction than in the longitudinal direction. More specifically, the hoop stress is twice the longitudinal stress developed inside a pressure vessel. Hoop stress is therefore considered in the calculations of geometry of vessel for safe operation.  For a given length and diameter of vessel, material strength properties and expected stress, the safe thickness of the vessel can be calculated according to Equation A1. Based on preliminary tests on compression of chips the highest load expected is as high as 40 kN. Pressure is calculated by dividing the expected force by the area of the cylinder wall. The hoop stress is assumed to be the highest material strength. With that in mind, the thickness of the pressure vessel can be calculated to ensure a safe operation of the parts. 82      𝜎ℎ𝑜𝑜𝑝 =𝑝𝐷2𝑡 (A-1)  For a D = 0.0508 m, L = 0.04 m, p = 19.6 Mpa, and a material limit of 115 MPa for stainless steel a minimum thickness can be calculated.    𝑡 =𝑝𝐷2 𝜎ℎ𝑜𝑜𝑝 (A-2)    𝑡 =(19.6 𝑀𝑝𝑎)(0.0508 𝑚)2 (115 𝑀𝑝𝑎)= 0.00432 𝑚        83  A.2 Compression Chamber Drawing   Figure A12. Compression vessel design drawing (Top). Compressor design drawing (Bottom). Both parts are made of stainless steel 304. An Aluminum cup was made to hold any water that comes out from the compression test (not shown).   84  Appendix B  : MTS Compression Test  B.1 Test 1 – Constant Compression time of 5 seconds  (a)  (b)  (c)  Figure B11. Load curves for test 1 of the MTS compression trial. Load vs displacement curve and load vs time curve for (a) 3:1 compression ratio, (b) 5:1 compression ratio and (c) 6:1 compression ratio.  00.511.522.533.540 10 20 30 40 50Load (kN)Displacement (mm)00.511.522.533.540 2 4 6Time (s)0510152025300 20 40 60Load (kN)Displacement (mm)0510152025300 2 4 6Time (s)05101520253035400 20 40 60Load (kN)Displacement (mm)05101520253035400 2 4 6Time (s)85  B.2 Test 2 – Variable compression time  (a)  (b)  (c)  Figure B21. Load curves for test 2 of the MTS compression trial. Load vs displacement curve and load vs time curve for compression times of (a) 3 seconds, (b) 5 seconds and (c) 10 seconds.   0510152025300 20 40 60Load (kN)Displacement (mm)0510152025300 1 2 3 4Time (s)051015202530350 20 40 60Load (kN)Displacement (mm)051015202530350 2 4 6Time (s)05101520250 20 40 60Load (kN)Displacement (mm)05101520250 5 10Time (sec)

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0375847/manifest

Comment

Related Items