development and characterization of new feedstock …

i

DEVELOPMENT AND CHARACTERIZATION OF NEW FEEDSTOCK FROM

BIOMASS BLENDS

by

Noorfidza Yub Harun

M.Sc.FE, University of New Brunswick, Canada, 2011

B.Eng, National University of Malaysia, 1993

A Dissertation Submitted in Partial Fulfillment

of the Requirements for the Degree of

Doctor of Philosophy

in the Graduate Academic Unit of Mechanical Engineering

Supervisor: Muhammad T. Afzal, PhD, Mechanical Engineering

Examining Board: Rickey Dubey, PhD, Mechanical Engineering

Yonghao Ni, PhD, Chemical Engineering

Andy Simoneau, PhD, Mechanical Engineering, Chair

External Examiner: Prabir Basu, PhD, Mechanical Engineering, Dalhousie University

This dissertation is accepted by the

Dean of Graduate Studies

THE UNIVERSITY OF NEW BRUNSWICK

October, 2017

©Noorfidza Yub Harun, 2018

ii

ABSTRACT

This thesis investigates the biomass blends to develop standard quality fuel pellets. New

knowledge in the synthesis of chemical, physical and mechanical properties of pellets made

from the biomass blends is being reported. Pellets made from agricultural biomass are

inherently known for poor quality and do not meet the pellet quality requirements. Through

blending proposed in this study, the heating value, ash content, density and durability were

demonstrated to meet the standard requirements.

Theoretically, most relations developed for compaction of powdery material significantly

depend on diameter, thickness, porosity, density, pressure, and strength. The relevant

relationships were reported in various equations such as Gurnham Model, Jonnes Model

and Fell and Newton Model. The effect of particle size and blending for compaction of

biomass was the greatest interest in this study. Results obtained for this study are in line

with the following hypotheses:

Blending may change in the initial and final degradation temperature of the biomass

and thus chemically improves its content, gases production, thermal stability and

heat capacity.

Smaller particle size can yield higher strength than the coarser particle size of pellet

made from either individual or blended biomass. Thus, pellets quality met the

standard requirement of the density and durability set by Fuel Pellet Institution

(PFI)

iii

Blending may reduce the energy of compaction while increasing the inter-particle

bonding caused by the change in the chemical composition. Forestry biomass may

contribute binding agents such as lignin to compact the particles of biomass blends.

After grinding agricultural biomass, it is difficult to keep in moisture, thus

compaction is hard to achieve. Blending may improve the adsorption rate of

moisture within the required moisture content of pellets (8 – 12 %) that can provide

strong inter-particle bonding and most importantly can sustain the combustion

stability.

Pellet strength for biomass blends may be best evaluated at significant dependants

such as particles size, blend ratio and moisture content, for which an experiment is

done at constant process parameters (Temperature 80 °C, pressure 159 MPa and

holding time 30 seconds)

The result shows that an increase in carbon, hydrogen, volatile and char content, and a

significant decrease in chloride were observed after blending. Since blending helped to

improve in its chemical composition, similarly the strength, durability, and density of the

blended pellets was found improved compared with individual agricultural biomasses.

Corresponding to the theory, the yield stress was found inversely proportional to the

particle size. The pellet durability index (PDI) of pure agricultural biomass pellets was less

than 90, which is well below the PFI (Pellet Fuels Institute) standard requirement. This

study revealed that using a blend ratio of 50:50 it was possible to achieve PDI of more than

95. At this ratio, pellets made from lower particle size (150 – 300 µm) exhibited higher

density (950 – 1178 kg/m3 for spruce and pine; 668 – 800 kg/m3 for reed canary grass,

iv

timothy hay and switchgrass; 900 – 970 kg/m3 for biomass blends). The yield stress

exhibited differences in values for individual forestry (40 MPa) and agricultural biomasses

(27-48 MPa). However, after blending the values converged closest to that for forestry

biomass. The ground biomasses were also evaluated for their compaction energy

requirement into pellets. The compaction conditions were set as at 159 MPa load and a

temperature of 80 °C. The energy required in compacting ground reed canary grass,

timothy hay and switchgrass was lower (1.61, 1.97, and 1.68 kJ) than spruce (2.36 kJ) and

pine (2.35 kJ). After blending, the values were around 2 kJ with the pellet quality

approaching almost similar to that of pellets made from woody biomass.

Besides the particle size and blending, moisture content was also manipulated to see its

effects on the strength and quality of pellets. The proposed model developed based on pine

and Timothy hay can be used to predict the strength of pellets made from similar species

with moisture content range of 8 – 12%, blend ratio in the range of 25 – 50%, and the

particle size in the range of 300µm and below.

Besides evaluating the pellet quality, this study was further investigating the behavior of

pellets made from biomass blends. The heat released from the blended biomass (6.94 –

9.26 kJ/kg) was higher than the individual agricultural biomass (4.59 – 6.78 kJ/kg) but

lower than individual spruce (10.2kJ/kg) and pine (11.13kJ/kg). The findings indicate that

the reactivity of the individual agricultural biomass material changed due to blending.

As for torrefaction, the chemical functional groups were recognized using TG-FTIR and

TG-MS techniques, which were paired to refine the identification of gases. The FTIR

analysis indicated that biomass blends had similar functional groups but their

decomposition was largely at higher torrefaction temperature of 290 °C compared to

v

switchgrass alone, which was largely at lower torrefaction temperature of 230 °C. The MS

quantified the degradation of combustible gases: CH4, C2H4, CO and O2 from individual

biomass in the range of 20-30% at torrefaction temperature of 200 °C and 230 °C. At the

increased torrefaction temperature of 260 °C and 290 °C the degradation of the gases was

more than 30%. The blended biomass degraded into gases in the range of 28-30 %(v/v) at

torrefaction temperature higher than 290 °C. All gaseous products evolved from the

torrefaction of agricultural and forestry biomasses were almost similar in characteristics,

but varied in their proportions. The composition of product gas generated from torrefied

biomass depends on the types of biomass blended and the temperature. In conclusion,

blending can help increase the energy capacity, improve the pellet quality, increase in pellet

fuel production, and improve the combustion performance. The pellet fuel industry can

successfully implement the biomass blending approach as this research has demonstrated

achievement of positive effects in pellet fuel quality.

vi

PREFACE

This thesis is submitted in partial fulfilment of the requirements for the Degree of

Philosophy Doctor (Ph.D.) at the University of New Brunswick (UNB). The work has been

carried out at the Department of Mechanical Engineering between Jan 2012 and March

2015. The doctoral study was conducted under the supervision of Prof. Dr Muhammad T.

Afzal and study leave was granted by Universiti Teknologi PETRONAS, Malaysia.

The work was initiated to establish new knowledge on how blending of biomass affect the

change in composition, compaction characteristics, compaction energy requirement,

combustion stability and the product gas concentration in order to improve the density,

energy content, durability and strength of pellets made from agricultural biomass. Thus,

agricultural biomass can be exploited in the development of biomass based solid fuels.

vii

ACKNOWLEDGEMENTS

First and foremost, the author would like to thank to the supervisor of this study, Dr.

Muhammad T. Afzal for the valuable guidance and advice. All the constructive comments

and suggestions given has helped the author in improving the study and taught to focus on

a lot of areas of research in completing this project, which is often overlooked. His

contribution is indeed precious and educational. I am particularly indebted to him for his

support morally, and always being encouraging and providing useful critiques in finishing this

PhD study.

My sincere thanks go to the committee members Dr. Rickey Dubay, Dr. Yonghao Ni and

Dr. Andy Simoneau for their valuable time and giving constructive criticisms and

suggestions on my thesis, as well as all the instructors with whom I completed my course

work. I would like to express my sincere thanks to Dr. Arshad Salema postdoctoral fellow

and other colleagues at Bioenergy and Bioproducts lab for their support and

encouragement. I would like to express my appreciation for the support provided by the

staff of the Department of Mechanical Engineering: Paulette Steever, Ann Bye, Debbie

Basque, Allan Robinson, and Vincent Boardman. Their constant encouragement, caring

words, and supports were helpful towards the successful completion of this work.

It is also an honor to thank to Universiti Teknologi PETRONAS (UTP) for the financial

support throughout this study. These resources have provided the aids needed in order to

effectively design the desired process and accomplish the work objective.

The authors appreciate the financial assistance from New Brunswick Department of

Agriculture, Aquaculture and Fisheries; New Brunswick Soil and Crop Improvement

viii

Association; New Brunswick Agricultural Council; Agriculture and Agri-Food Canada;

New Brunswick Innovation Foundation.

Lastly, an honorable mention goes to the author’s families, especially husband and lovely

children for their sacrifices, patience and non-stop support. Also, I would like to thanks to

friends for their understanding and support in completing this project. Without helps of the

particular that mentioned above, the author would face many difficulties while doing this

project.

ix

LIST OF PUBLICATIONS

This thesis is based on the following seven papers referred to as Paper I-VII in the text:

Paper I: Harun, N. Y. and Afzal, M. T. (2015): Chemical and Mechanical

Characteristics of Biomass Pellets made from Agricultural and Woody

Biomass Blends. Transactions of the ASABE, 58(4), 921-930

Paper II: Harun, N. Y. and Afzal, M. T. (2016): Effect of Particle Size on Mechanical

Properties of Pellets Made from Biomass Blends. Procedia Engineering,

148, 93–99

Paper III: Harun, N. Y. and Afzal, M. T. (2016). Process and Energy Analysis of

Pelleting Agricultural and Woody Biomass Blends.

(Manuscript is ready to be submitted)

Paper IV: Harun, N. Y. and Afzal, M. T. (2015): Combustion Behavior and Thermal

Analysis of Agricultural and Woody Biomass Blends. Advances in

Environmental Biology, 9 (15), 34-40

Paper V: Noorfidza Yub Harun and Muhammad T. Afzal, “Torrefaction of

Agriculture and Forestry Biomass Using TGA-FTIR-MS”, Dincer et al.

(eds.), Progress in Exergy, Energy, and the Environment, DOI 10.1007/978-

3-319-04681-5_77, # Springer International Publishing Switzerland 2014

Paper VI: Harun, N. Y. and Afzal, M. T. (2013): Characteristics of Product Gas from

Torrefied Biomass Blends. International Journal of Chemical and

Environmental Engineering, 4(4), 266-272

All published papers (Paper I, II, IV, V and VI) were also presented in peer reviewed

conferences as the following:

Technical Conferences:

1. Harun, N. Y. and Afzal, M. T. (2014). Chemical and Mechanical Properties of

Pellets Made from Agricultural and Woody Biomass Blends. 2014 ASABE and

CSBE/SCGAB Annual International Meeting, Montreal, Canada. July 13-17, 2014.

Paper Number: 1908650

2. Harun, N. Y. and Afzal, M. T. (2016). Effect of Particle Size on Mechanical

Properties of Pellets Made from Biomass Blends. 4th International Conference on

Process Engineering and Advanced Materials (ICPEAM 2016), Kuala Lumpur,

Malaysia, 15-17 August 2016.

3. Harun, N. Y. and Afzal, M. T. (2015). Combustion Behavior and Thermal Analysis

of Agricultural and Woody Biomass Blends. International Postgraduate Conference

Agriculture, Biology and Environment (IPCABE 2015), Jakarta, Indonesia, 31 July

– 1 August 2015.

4. Harun, N. Y. and Afzal, M. T. (2013). Characteristics of Product Gas from

https://scholar.google.com/citations?view_op=view_citation&hl=en&user=98vcocgAAAAJ&citation_for_view=98vcocgAAAAJ:dshw04ExmUIC




x

Torrefied Biomass Blends. 2nd International Renewable Energy and Environment

Conference (IREEC-2013), Kuala Lumpur, Malaysia, 4-6 July, 2013.

5. Harun, N. Y. and Afzal, M. T. (2013). Torrefaction of Agriculture and Forestry

Biomass Using TGA-FTIR-MS. Proceeding of the Sixth International Exergy,

Energy and Environment Symposium (IEEES-6), Italy, 1-4 July 2013, p317-325.

xi

TABLE OF CONTENTS

ABSTRACT ........................................................................................................................ ii

PREFACE ......................................................................................................................... vi

ACKNOWLEDGEMENTS .............................................................................................. vii

LIST OF PUBLICATIONS ............................................................................................... ix

TABLE OF CONTENTS ................................................................................................... xi

LIST OF TABLES ......................................................................................................... xviii

LIST OF FIGURES ......................................................................................................... xix

LIST OF SYMBOLS, NOMENCLATURE OR ABBREVIATIONS ............................. xx

Chapter 1: INTRODUCTION ........................................................................................... 1

1.1 Motivation ............................................................................................................ 1

1.2 Objectives of the study ......................................................................................... 2

1.3 Organization of thesis........................................................................................... 3

Chapter 2 BACKGROUND ................................................................................................ 5

2.1 Chemical analysis of biomass blends ................................................................... 5

2.2 Compaction of biomass blends ............................................................................ 8

2.2.1 Density ........................................................................................................... 8

2.2.2 Durability ....................................................................................................... 9

2.2.2.1 Pelleting parameters .................................................................................... 10

i. Pressure ...................................................................................................... 10

ii. Temperature ................................................................................................. 11

xii

iii. Holding time ................................................................................................ 12

2.2.2.2 Material characteristics ................................................................................ 13

i. Moisture content .......................................................................................... 14

ii. Particle size ................................................................................................. 14

iii. Porosity ........................................................................................................ 15

iv. Friability characteristics ............................................................................. 16

a. Solid bridge form ......................................................................................... 16

b. Springback effects ....................................................................................... 17

2.3 Theoretical point of view ................................................................................... 19

2.3.1 Biomass blends pellets ....................................................................................... 19

2.3.2 Blending effect on the compaction capability ............................................. 20

2.3.3 Basic concepts related to powder compaction models ................................ 21

Gurnham Model ........................................................................................................ 21

Jones Model .............................................................................................................. 22

Fell and Newton Model ............................................................................................ 23

2.3.4 Blending effects on thermal behavior .......................................................... 24

Chapter 3 MATERIALS AND METHODOLOGY ......................................................... 26

3.1 Preparation of materials ..................................................................................... 26

3.2 Chemical and physical analysis.......................................................................... 27

3.3 Pelleting process ................................................................................................. 29

3.4 Physical properties of pellets.............................................................................. 30

3.4.1 Pellet density ................................................................................................ 30

xiii

3.4.2 Compression test (crushing) ........................................................................ 30

3.5 Summary of the work ......................................................................................... 32

Chapter 4 SUMMARY OF APPENDED PAPERS.......................................................... 34

Chapter 5 CONCLUSION AND RECOMMENDATIONS ............................................. 37

4.1 Overall conclusions ............................................................................................ 37

4.3 Recommendations for future research................................................................ 38

4.4 Contributions to pellet fuel development field ................................................... 38

Chapter 6 REFERENCES ................................................................................................. 40

Chapter 7 Collection of Papers ......................................................................................... 45

Paper I ........................................................................................................................ 46

Chemical and Mechanical Properties of Pellets Made from Agricultural and Woody

Biomass Blends .............................................................................................. 46

Abstract ......................................................................................................................... 46

Introduction ................................................................................................................... 47

Methodology ................................................................................................................. 51

Sample Preparation ................................................................................................... 51

Pelletization............................................................................................................... 52

Proximate and Ultimate Analysis ............................................................................. 54

Mineral Analysis ....................................................................................................... 54

Mechanical Property ................................................................................................. 55

Results and Discussion ................................................................................................. 55

Proximate and TG Analysis ...................................................................................... 55

xiv

Ultimate Analysis...................................................................................................... 59

Ash Content .............................................................................................................. 61

Minerals Content ....................................................................................................... 63

Blended Biomass Pellet Strength .............................................................................. 64

Porosity ..................................................................................................................... 66

Durability .................................................................................................................. 67

Conclusion .................................................................................................................... 69

Acknowledgement ........................................................................................................ 70

References ..................................................................................................................... 70

Paper II ........................................................................................................................ 78

Effect of particle size on mechanical properties of pellets made from biomass blends ... 78

Abstract ......................................................................................................................... 78

Introduction ................................................................................................................... 79

Methodology ................................................................................................................. 80

Sample Preparation ................................................................................................... 80

Proximate and Ultimate Analysis ............................................................................. 81

Pelletization Process ................................................................................................. 81

Density of Pellets ...................................................................................................... 83

Mechanical Strength ................................................................................................. 83

Results and Discussion ................................................................................................. 83

Elemental Analysis ................................................................................................... 83

Heating Value and Ash Analysis .............................................................................. 85

xv

Physical and Mechanical Characteristics ................................................................. 86

Conclusion .................................................................................................................... 92

Acknowledgements ....................................................................................................... 93

References ..................................................................................................................... 93

Paper III ........................................................................................................................ 98

Process and Energy Analysis of Pelleting Agricultural and Woody Biomass Blends ..... 98

Abstract ......................................................................................................................... 98

Introduction ................................................................................................................... 99

Materials and Methods ................................................................................................ 100

Raw Materials ......................................................................................................... 100

Chemical and Compositional Characterization ...................................................... 100

Process of Compaction for Biomass Pellet ............................................................. 101

Surface Fracture Analysis ....................................................................................... 102

Results and Discussion ............................................................................................... 102

Chemical Analysis .................................................................................................. 102

Work of Compaction............................................................................................... 105

Conclusion .................................................................................................................. 111

Acknowledgements ..................................................................................................... 112

References ................................................................................................................... 112

Paper IV ...................................................................................................................... 115

Combustion behavior and thermal analysis of agricultural and woody biomass blends 115

Abstract ....................................................................................................................... 115

Introduction ................................................................................................................. 116

Methodology ............................................................................................................... 118

xvi

Sample Preparation ................................................................................................. 118

Proximate and Ultimate Analysis ........................................................................... 118

Mineral Analysis ..................................................................................................... 119

Combustion Analysis .............................................................................................. 119


Chemical and Mineral Content of Individual and Blended Biomass ..................... 120

Combustion Behavior ............................................................................................. 122

Heat Release............................................................................................................ 127

Conclusions ................................................................................................................. 128

Acknowledgements. .................................................................................................... 129

References ................................................................................................................... 129

Paper V ...................................................................................................................... 133

Torrefaction of Agriculture and Forestry Biomass using TGA-FTIR-MS ..................... 133

Abstract ....................................................................................................................... 133

Introduction ................................................................................................................. 134

Materials and Method ................................................................................................. 136

TG-FTIR Experiment.............................................................................................. 136

TG-MS Experiment ................................................................................................ 137


Proximate Analysis ................................................................................................. 137

TG-DTG Analysis ................................................................................................... 138

FTIR Analysis ......................................................................................................... 141

xvii

Torrefaction 200 °C ................................................................................................ 141

Torrefaction 230 °C ................................................................................................ 143

Torrefaction 260 °C ................................................................................................ 144

Torrefaction 290 °C ................................................................................................ 145

Gas quantification ................................................................................................... 146

Conclusion .................................................................................................................. 149

Acknowledgment ........................................................................................................ 150

References ................................................................................................................... 150

Paper VI ...................................................................................................................... 153

Characteristics of Product Gas from Torrefied Biomass Blends .................................... 153

Abstract ....................................................................................................................... 153

Introduction ................................................................................................................. 154

Materials and Method ................................................................................................. 156

TG-FTIR-MS Experiment ...................................................................................... 156


Proximate Analysis ................................................................................................. 158

TG-DTG Analysis ................................................................................................... 159

FTIR Analysis ......................................................................................................... 162

Analysis of Product Gas by MS .............................................................................. 165

Conclusion .................................................................................................................. 168

Acknowledgment ........................................................................................................ 169

References ................................................................................................................... 169

Curriculum Vitae

xviii

LIST OF TABLES

Table 2 1 Densification characteristics ............................................................................... 9

Table 3 1 Biomass and blending ratios ............................................................................. 27

Table 3 2 Standard procedure for characterization of biomass blends ............................. 28

xix

LIST OF FIGURES

Figure 2 1: Biomass chemical structure [4, 5, 6 and 7] ...................................................... 6

Figure 2 2 Mechanical strength of pellets made from hardwood, softwood and corn stover

under various applied pressure.......................................................................................... 11

Figure 2 3 Durability of pellets made from Alfalfa, corn stover and switchgrass at

different temperature [17, 29 and 30] ............................................................................... 12

Figure 2 4 Compression behavior of ground biomass for making pellet .......................... 13

Figure 2 5 SEM image fracture surface of spruce pellet at magnification of 200 µm ...... 17

Figure 2 6 Springback effects during compaction of ground hay (agricultural) for

pelletization (prelimanary result from this work) ............................................................. 18

Figure 3 1 Stress and strain curves for brittle materials.................................................... 30

Figure 3 2 Summary of work for the individual and blended biomass ............................. 33

xx

LIST OF SYMBOLS, NOMENCLATURE OR ABBREVIATIONS

RCG Reed Canary Grass

H Timothy Hay

SW Switchgrass

S Spruce

P Pine

SEM Scanning Electron Microscopy

TGA Thermogravimetric Analysis

FTIR Fourier Transform Infrared Spectroscopy

MS Mass Spectrometry

PDI Pellet Durability Index

PFI Pellet Fuels Institute

V Volume

DTG Derivative Thermogravimetric

HHV High Heating Value

ε Porosity

ave Average

BET Brunauer, Emmett and Teller

σy Axial yield stress

σi Inherent compressive stress of biomass

d Diameter

MC Moisture Content

xxi

VM Volatile Matter

FC Fixed Carbon

BR Blend Ratio

PS Particle Size

DSC Differential Scanning Calorimeter

DTA Differential Thermal Analysis

[Ai] Concentration of the gas A

Pi Partial Pressure for m/z = i representing of gas A

PT Total Pressure for all species i

1

Chapter 1: INTRODUCTION

1.1 Motivation

The pellet industry is solely based on biomass from forestry and mills residues, and should

therefore look for alternative feedstock to increase the production of pellets. The use of

biomass from agriculture is expected. However, there is a concern for agricultural biomass

pellets due to inherent low quality. It has complex structure and possesses muddle

characteristics; thus production of pellets from agricultural biomass is slowed and the

research is still on-going to improve its properties to be accepted in pellet fuel industry. It

is crucial to increase the production of fuel pellets to support the pellet industry to meet

the present and future demand.

Biomass is commonly compacted into pellets to provide a standard dimension of solid

fuel to facilitate handling, storage and reducing the transportation costs. Among many

other factors, the quality of pellets mainly depends on the type of biomass species and the

pelleting process parameters. Particularly, agricultural biomass has large variation in its

chemical composition and structure; thus, the density, durability, thermal properties, and

energy efficiency of pellets can be diverse from one species to another. Despite the wealth

of studies on the biomass pellets from woody and agricultural biomasses, it seems that

nothing had been published on the blending of biomass. This study would help increase

the knowledge to understand how the ground particles and contents interact during the

compaction of biomass blends for pellets.

PFI quality standard requires the pellet density higher than 600 kg/m3, dimension of 1.5

inch, durability index (PDI) more than 90%, heating value of more than 16 MJ/kg, ash

2

content of less than 3 % and moisture content of 10 %. Thus, much effort has been put

into this study to develop pellets from biomass blends and know the effect of blending on:

1. The degradation rate, proximate, ultimate and mineral contents

2. The combustion characteristics and combustible gas products of pyrolysis

3. The porosity, density, durability, yield stress, strength, and energy of compaction

for fuel pellets.

However, the pellets industry is also seeking more economical and energy efficient pre-

processing of pelleting process such as drying and grinding of the biomass. Therefore, it

is imperative and critical to further enhance the knowledge in understanding the moisture

content and particle size effect on the density, yield stress, and strength of pellets made

from biomass blends.

1.2 Objectives of the study

The major goal of this research was to develop and characterize a new feedstock for fuel

pellets by blending agricultural and woody biomass, investigating the impact on pellet

quality, the energy of compaction, its combustion performance and combustible gas

products attained. The specific objectives of the study were as follows:

1. To characterize the chemical, physical and mechanical properties of individual

agricultural and forestry biomass pellets.

2. To investigate the effect of particle size and blending on chemical, physical and

compaction behaviour, and strength of pellets.

3. To study the combustion characteristics of pellets made from blends.

4. To investigate the effect of blending on product gas during torrefaction and

pyrolysis process.

3

1.3 Organization of thesis

This thesis comprises results from the papers listed on pp 70 (referred to as Paper I-VII in

the text) and is organized in seven chapters.

Chapter 1: Introduction - contains a general introduction to the motivation and

objectives.

Chapter 2: Background – presents the theoretical background to the research. The

variation in the chemical composition of forestry biomasses includes

softwood and hardwood, and agricultural biomasses are briefly described

on a general basis. Emphasis is focused on four parts, the effect of blending

on

a) the change in chemical composition

b) the product gas of torrefaction

c) the thermal characteristics and release of energy

d) the energy needed to pelletize the biomass materials.

For the latter part, this study has concentrated the effect of particle size on

the density and strength of pellets made from biomass blends. Relevant

literature on the pelletization of biomass, the fuel pellet properties, the

pelleting process parameters and the compaction relations is reviewed.

Chapter 3: Materials and methods – provides an overview of the materials and

methods used. Experimental details can be found in the appended papers.

Chapter 4: Summary of the appended papers – contains a summary of the main

findings in Paper I – VI.

4

Chapter 5: Conclusion and future work – summarizes the overall conclusions drawn

from the work in this thesis, both from the fundamental point of view and

with respect to the pellet commercial applicability. Some suggestions for

future work are given.

Chapter 6: References.

Chapter 7: Collection of papers.

5

Chapter 2 BACKGROUND

This chapter contains background information for the research presented in this thesis.

First is a brief description of chemical composition and structure of biomass, particularly

from forest, hardwood and softwood. Several examples from agricultural biomass such as

corn stover, switchgrass and reed canary grass were also selected to converse on the

variations in the chemical structure for comparison. Next is an overview of relevant

literature on the compaction characteristics of ground biomass from both forestry and

agriculture and includes particles size, porosity, density, durability and strength. Finally,

there is a brief description of the theoretical compaction model related to the particle size,

load and porosity as the main factors contributed to the strength of biomass blends pellet.

More background information is also given in Paper I-VI.

2.1 Chemical analysis of biomass blends

Biomass chemical composition and its structure differ significantly with biomass types.

In forestry biomass, two different types are distinguished, hardwood and softwood.

Hardwoods have cellulose contents ranging from 42 to 49%, hemicellulose from 16 to

23%, and lignin contents from 21 to 29% [1], while softwoods contain 40–50% cellulose,

25–35% hemicellulose and 16–33% lignin [2]. High in cellulose contents, softwood is

generally used for pulp and paper industries, structural and industrial materials. Due to

these differences in structure, residues from forest and mills are making it a desirable

feedstock for pellets industries.

Agricultural biomass also has a wide variety of component structures. Corn stover consists

of 30-38% cellulose, 19-25% hemicellulose and 17-21% lignin [3]. Switchgrass contains

6

much lower cellulose (27–37%), much higher hemicellulose (22-29%) and 13–21% lignin

[3]. The higher content of cellulose (47-49 %) and hemicelluloses (28-39 %) with much

lower content of lignin (8-11.5 %) points to reed canary grass (RCG) [4].

As illustrated in Figure 2 1, softwood contains more lignin, but less hemicellulose as

compared to agricultural biomass. Due to this wide variability in composition, it is

essential that the effects of blending the agricultural biomass with the forestry biomass on

the chemical (proximate, ultimate and mineral) properties and thermal degradation rate

and region in biomass blends be better understood.

Figure 2 1: Biomass chemical structure [1, 2, 3 and 4]

19.5 %24 % 22 %

25.5 %

33.5 %

Hemicellulose (wt.%)

45.5 % 45 %

34 % 32 %

48 %Cellulose (wt.%)

25 % 24.5 %19 % 17 %

9.75 %

Lignin (wt. %)

7

Agricultural biomasses have recently been paid high attention to be converted into fuel in

different states: liquid, gas and solid. Pyrolysis process using thermogravimetric analyzer

(TGA) is investigated on the purpose to obtain the individual of agricultural and forestry

biomasses; however, the effect of blending with ratio 50:50 (wt% agricultural:wt%

forestry) on the thermal characteristics, rate of reaction zone and temperature were

highlighted as to see its changes.

From a fundamental point of view, the glass transition reflects a change in the molecular

mobility and associated to the portion of free volume. However, this portion depends on

the absolute value of glass transition temperature of biomass, Tg. As the Tg absolute value

increases, more free volume becomes accessible to the thermal motion in the glassy zone.

Biomass fuel undergoes an increase in its heat capacity when it undergoes the glass

transition. Data presented by TGA for individual agricultural and forestry biomass and

their blends are judged for the fundamental mechanism of the glass transition changed in

the blends. The change in the chemical composition and the thermal characteristics of

biomass after blending are expected due to this glass transition temperature, Tg changed.

Biomass with lower Tg will be shifted after blending with biomass with larger Tg. Though

ground finely, biomass blends is an immiscible system. The blends can be so well

homogenized that the phase is not affected by other parameters. A single and uniform Tg

for biomass blends is expected to be valid for Tg not larger than 400 °C as they were when

individually measured. In many cases, Tg is able to identify whether the blends are well

homogenized so that it will not be affected by the processing conditions. The

thermogravimetric (TG) and differential scanning calorimetric (DSC) analyses can be a

useful and practical test for pellet industry to evaluate pellet from biomass blends

8

compatible to standard quality. From these characteristics provided by TG and DSC, the

chemical composition and structure such as proximate, ultimate, lignin and heating value

changed after blending of biomass, which verified the theory.

2.2 Compaction of biomass blends

Pelletization is the suitable technology to be applied for solid-fuel production from

agricultural biomass because it can be adjusted well to feed in the thermal and energy

generation system. The bulk volume of the material can be reduced by mechanical

compaction for easy handling, transportation and storage [5]. High density pellets are an

environmentally friendly solid bio-fuel, very comfortable in use for both households and

industrial plants. In this study, the pelletization of selected agricultural biomass blends

with forestry biomass was investigated on the density, durability, strength and yield stress

under different particle size, moisture content and blend ratio in order to obtain pellets

with the satisfied fuel properties. The Pellet Fuel Institute (PFI) standard is followed and

detailed in Paper I.

2.2.1 Density

Biomass pellet is a favored form of solid fuel because it provides the uniformity of

biomass material structures and can convert into fuel and energy easily [6]. The density

of raw biomass is typically low, thus it withdraws reasonably low energy density. The

bulk density of agricultural biomass is about 40‐150 kg/m3 [7, 8] and woody biomass is

about 150‐400 kg/m3 [9]. The density of woody biomass can be increased to 700 kg/m3

after pelleting [10]. Table 2.1 shows that the densified agricultural biomass has the energy

9

density almost close to the required range for high heating value of biofuels, like densified

forestry biomass.

Biomass pellets allow for a standardized composition, size, shape and quality that ease

global trade and utilization. These characteristics also make the pellets easy to store and

transport in trucks, ships or mobile storage facilities. Additionally, biomass pellets have

a large surface to volume ratio and low moisture content (less than 15 %), which makes

it relatively resistant to decay processes.

Table 2.1: Densification characteristics [6, 7, 8, 9]

Bulk density

(kg/m3)

Pellet density (kg/m3) Energy density

(GJ/m3)

Agricultural 40-150 500-900 15-17

Forestry 150-400 700-1100 18-21

With regard to the effect of blending on the density of the fuel pellets mentioned above,

it is crucial to fill the demand for high energy density pellets from agricultural biomass.

Furthermore, it is important to note that ground biomass of different particle size changes

the density of pellet. This thesis addresses improvement in the density of pellet made from

agricultural biomass in view of blended biomass technique and particle size effects. These

issues have received limited attention in the literature.

2.2.2 Durability

Other than bulk density, pellet quality is usually measured by means of pellet durability.

However, the use of different raw biomasses may have opposite effects on the final

densified products. Mechanical durability is a parameter that is defined as the ability of

10

densified biofuels to remain intact when handled. Durability refers to the amount of fines

that are recovered from pellets after subjected to mechanical agitation. In this study, it is

important to investigate the effect of blending biomass on pellet durability. Many studies

have been focused on the study of the pelleting process parameters affecting the pellet

durability.

2.2.2.1 Pelleting parameters

The challenge of biomass pellets is to keep them bound and compact in the range where

high quality pellets are produced with high durability. Research done regarding the effects

of the pelleting process and pellet quality has typically focused on factors such as pressure,

temperature, die size, press channel length and compression ratio [11, 12a, 12b, 13, 14

and 15]. However, the load applied to press the biomass, the heat supply for formation of

inter-particle bonding to keep compacted and the holding time are often found as the main

influential factors on quality of pellets.

i. Pressure

The compression characteristics of ground biomass vary under various applied pressures.

Many researchers have reported that the optimum values of pressure to form highest

strength pellets is around 100 – 150 MPa [16, 17a, 17b and 17c]. Figure 2.2 shows resulted

compressive strength of hardwood, softwood and corn stover in the increment of applied

pressure. From these studies, agricultural biomass pellets unfortunately showed lower

strength than softwood and hardwood pellets.

11

Figure 2 2 Mechanical strength of pellets made from hardwood, softwood and corn

stover under various applied pressure

ii. Temperature

Kaliyan and Morey [17c] noted that it is essential to provide heat while pelleting to

activate inherent binders from the biomass. The activation of these inherent binders

promotes the formation of solid bridges, a primary mechanism of particle agglomeration.

Furthermore, additional heat facilitates the plastic deformation of particles, thereby

increasing the inter-particle contact area and decreasing the inter-particle distance. A

temperature range of 60 to 130 °C is reported in most studies. However, the optimum

temperature to reach the highest durability of pellets is within the glass transition

temperature, 75 – 100 °C [17c]. Kaliyan and Morey [17c] also reported highest durability

at 80 – 85 °C, which is a temperature range for gelatinization of starch.

0

10

20

30

40

50

0 50 100 150 200C

om

pre

ssiv

e st

rength

(MP

a)Forming pressure (MPa)

hardwood

softwood

12

Figure 2 3 Durability of pellets made from Alfalfa, corn stover and switchgrass at

different temperature [17a and 18]

iii. Holding time

The cooling period is necessary to remove the soft pellet from the heated die. Inadequate

cooling and drying of pellet after the moisture and temperature arise from the frictional

heating of the die during pelleting can lead to poor quality pellets. As illustrated in Figure

2.4, the hold time after compression of ground biomass is less than a minute to keep the

pellet compacted when relieved. However, some researchers have not considered the

holding time as an influential factor to quality pellets [5, 19 and 20]. Mani et al. [20]

reported that a holding time of 30 seconds or less is enough to halt the springback effect

for pellets made from agricultural biomass.

0

20

40

60

80

100

0 20 40 60 80 100 120 140 160

Du

rab

ilit

y (

%)

Temperature (°C)

alfalfa

corn stover

switchgrass

13

Figure 2 4 Compression behavior of ground biomass for making pellet

There are many publications in the literature concerning factors influencing pellet

durability. High durability was found for pellets formed at pressure 159.6 MPa,

temperature of 80 °C and holding time 30 s [21a, 21b]. However, few articles have been

published regarding the durability of pellets made from biomass blends. In this study, the

pelleting parameters are designated uniform as suggested by Dharmodaran and Afzal

[21b] for production of pellets made from individual agricultural and forestry biomass

and their blends.

2.2.2.2 Material characteristics

Not only process parameters, but more recently the characteristics of the raw biomass to

be pelletized have also been studied considerably. Low bulk density agricultural biomass

is difficult to be densified into pellet due to the dry basis residue which denotes low lignin

content resulting in low durability characteristics. The lignin content of woody biomass

Holding time

14

is normally 25 %, and much less variable among woody species. However, agricultural

biomasses are much more variable in lignin content. Also for the densification process of

biomasses, the effects of moisture content, particle size and porosity on inter-particles

bonding focusing on pellet strength and energy of compaction of biomass blends are

scarce.

i. Moisture content

Dry or wet biomass particles have been reported significantly affecting the pellet density

and durability. Low moisture content in agricultural biomass may cause pellet to be brittle,

emitting fine particles which promotes high ash content resulting in the equipment failure.

Generally, as the moisture content of biomass increased, pellet density decreased [20,

21a]. This was found to be a crucial process variable in this study. Dharmodaran and Afzal

[21a and 21b] studied that the moisture content, which ranges between 8 % and 12 %, was

found as the optimum for both forestry and agricultural biomasses that achieved

maximum durability and compression stability of pellets. This range of moisture content

is applied in this study.

ii. Particle size

An important consideration which is often ignored during biomass pellet processing is the

size of the biomass particles. It is not certain whether an increase or a decrease in particle

size will influence pellet mechanical characteristics such as strength and durability.

Several studies have been conducted regarding the influence of particle sizes on pellet

density [20, 22, 23, and 24]. The researchers behind these studies concluded that the pellet

density increases with decreasing particle size for pellets made from beech, corn stover

15

and switchgrass. However, the reverse was found in the case of wheat straw by Mani et

al., [20] and barley straw by Serrano et al. [25]. These pellets were made from ground

biomass with a particle size ranging between 0.1 millimeter and 2 millimeters. The

particle bond strength and binding tendency are important parameters that determine the

mechanical properties of pellets such as tensile strength and the durability. Hence this

study therefore aimed at investigating the effect of particle size on some mechanical

characteristics of pellets made from forestry and agricultural biomass and their blends.

The influence of biomass particle size on biomass blends pellet, break force, yield stress,

porosity and the friability characteristics of the pellets were the eventual goal to be

investigated, as these properties are important and crucial to achieve standard quality

pellets.

iii. Porosity

Porosity has a strong effect on forming biomass pellet since it has strong relationship with

particle size and moisture content. Manickam et al. [26] have studied coconut husks, and

determined that porosity is proportional to particle size, but inversely proportional with

moisture content. This is in agreement with similar results reported by Zhang et al., 2012,

[27] in their study on corn cobs, corn leaves and corn stalks. Agricultural biomass dries

faster than hardwood and softwood. This is because agricultural biomass has larger

porosity than softwood and hardwood, as reported by Barbieri, 2014 [28]. It can be

concluded that agricultural biomass has weaker characteristics than that of woody

biomass, which leads to more disadvantages and poor quality pellets. Blending of both

16

biomasses can overcome the disadvantages, and thus can increase the utilization of

agricultural biomass and the production of pellets.

iv. Friability characteristics

a. Solid bridge form

The elasticity and plasticity behavior of biomass varies that result in different hardness of

the compacted biomass. It is important to understand the fundamental mechanism of the

biomass compaction process, which is required to enhance the quality of biomass pellets

[29]. During the initial stages of compaction, particles rearrange themselves under low

pressure to form close packing. Then, the air located in the interstices of the particles is

removed. During this phase, the particles retain most of their properties; however energy

is dissipated due to inter-particle friction. As pressure increases, the particles are forced

against each other and they undergo elastic and plastic deformation. Plastic deformation

contributes to pellet quality by bringing the particles in closer contact. As the particles

approach each other, short range bonding forces become effective. By increasing yield

stress, the brittle particles can fracture leading to the mechanical interlocking of particles.

But the mechanical interlocking is just a minor contributor to overall strength.

At the highest pressure sufficient to maximize the strength, the volume of compacted

material reduces and approaches the true pellet density. If the pressure is high enough to

generate heat or additional heat is supplied during the pelleting process, then some of the

components of the particles will locally melt and form solid bridges once the material

cools down. The indication of solid bridges at the fracture surface by SEM image at

17

magnification 200 micron is shown in Figure 2.5. Solid bridges largely determine the

strength of the final products [30].

Figure 2 5 SEM image fracture surface of spruce pellet at magnification of 200 µm

b. Springback effects

Process parameters alone are insufficient to cause the biomass achieving the desired

compaction with its length upheld. Since biomass varies in its characteristics, this study

has shown the compression behavior of spruce (forestry biomass), timothy hay

(agricultural biomass) and their blends. Figure 2.6 shows that the elasticity rate of hay is

higher and faster than that of spruce. Therefore, the load requirement for compacting hay

is higher than spruce as illustrated by the compression zone.

18

Figure 2 6 Springback effects during compaction of ground hay (agricultural) for

pelletization (prelimanary result from this work)

One of the most significant characteristics during compaction of hay is the springback

effect, as demonstrated along the duration of compression. However, this behavior was

not presented in spruce, and so was in their blends. The length of pellets made from

spruce and blends biomass was maintained as compacted after it releases. Surprisingly,

pellets made from timothy hay expanded by 1-2 mm as released from pelletizer. This

effect must be due to the springback characteristics. Poor quality of agricultural biomass

pellet might be due to this property. In this regard, very little information is available on

characterization and mechanical behavior of agricultural based pellets. Furthermore,

literature on characterization of agricultural blends with forestry biomass for biofuel

pellets is scarce.

Co

mp

ress

ion

lo

ad (

kN

)

Time (s)

0

1

2

3

4

0 10 20 30 40 50 60 70

Com

pre

ssio

n l

oad

(kN

)

Time (s)

Spruce

Hay

H(75%)+S(25%)

19

2.3 Theoretical point of view

Theories regarding compaction models for powder technology have been extensively

investigated. There seem to be few publications in the literature regarding compaction

models for biomass; however none was found for biomass blends. It is essential that an

association of multiple factors that can describe the quality of biomass pellets must be

addressed. Two major problems need to be addressed in order to find the best

combination of parameters to produce pellets with high quality. Poor mechanical strength

of biomass pellets easily crash into fines due to impact when loading. Another problem is

low bulk density like agricultural biomass, which is not suitable for long term storage as

feedstock. Durable pellets require better binding to hold the inter-particles together. In

most circumstances, deviation in chemical composition of different types of biomass

species has an effect on pellet strength and that pellet strength should be due to inter-

particle bonding. The pellet properties can be improved by changing in the chemical

content of biomass. This study is intended to improve the quality of pellets made from

agricultural biomass by blending it with forestry biomass, and consequently, to establish

a background for the assessment of results obtained in this study.

2.3.1 Biomass blends pellets

There is some evidence that mixing biomass from hardwood species with bark improves

the durability of densified biomass, but literature is limited to support this observation.

Mixed biomass pellets have higher thermal efficiency as compared to pellets made from

one type of biomass [31]. Residues from hardwood species such as red maple, southern

red oak, sweetgum, tupelo, white oak, and yellow poplar did not produce quality pellets

20

because of low durability and problems while processing. In order to improve the

properties, blending with other residues such as softwood biomass can be a potential

option. Combination of biomass in making pellets is potentially more effective for

processing fuel feedstock, and as a new renewable source of energy. Thus blended pellets

open a window to a new pellets composition and a new technological line of feedstock.

The chemical component that is responsible for forming solid bridges and the springback

effect during the pelleting process is lignin. Lignin acts as the “glue” that binds the pellet

together, but, from Lehtikangas, 2001 [32] study, the relationship between lignin content

and pellet durability is vague. A more plausible explanation is that lignin content might

cause stronger pellet as stated by many other researchers [17c, 33, 34, 35, 36, and 37].

The biomass species investigated by these researchers for pellet production include corn

stover, switchgrass, straw, alfalfa, wheat, and barley.

Thus this thesis investigated the pellet quality in terms of the mechanical strength,

durability and density of pellets made from biomass blends that counts the chemical

composition effects as well as physical and process parameters. Also, this study includes

pelletization experimental data which would be able to assist in understanding the

compaction characteristics and the work requirement of biomass blends in order to

produce better quality pellets.

2.3.2 Blending effect on the compaction capability

Particle size influences the density of pellet, and thus the compaction and sintering

capability. The pressure [12b] and the die wall friction [38] of the compaction also can

affect the density of compacted powder. Via physical treatment two steps to achieve

21

compaction capability, firstly small particle size can have large surface energy that attract

and bind each other better than larger particles. Second, adding fine particles with coarser

particles could have reduced porosity, thus showing better compaction capabilities but

this conclusion may not always be accurate since biomass species are quite different.

However, studies investigating the biomass blends effect on the compaction capability is

scarce. Some biomass can have much better compaction capability than other biomass

even if their physical treatment and material compounds are nearly identical. Therefore,

the main focus of this thesis was to study the compaction of biomass blends in relation to

the physical and mechanical characteristics. In this thesis the compaction capability was

quantified by density, energy of compaction and crushing strength. Detail of study was

given in the appended results in Paper I, II and III.

2.3.3 Basic concepts related to powder compaction models

A brief introduction to relevant basic concepts used in powder compaction is given below.

The overview is mainly based on theory presented by Gurnhum and Mason (1946), Jones

(1960) and Fell and Newton (1970).

Gurnham Model

Gurnham and Mason proposed that any increase in load (P) results in a proportional

increase in the apparent density of the mass [39].

𝑑𝑃

𝑃= 𝐴. 𝑑𝐷 (1)

Where P is pressure (load), D is density of compacted powder and A is constant. By

integration Eq. (1) gives

22

𝐷 = 𝑎. 𝑙𝑛(𝑃) + 𝑏 (2)

Where a and b are constants.

Linear relationship between density and ln(P) was obtained for compaction of powder

materials. The volume reduction in powder compaction is also commonly described by

porosity.

𝜀 = 1 −𝐷

𝐷𝑡𝑟𝑢𝑒 (3)

Where ε is porosity and 𝐷𝑡𝑟𝑢𝑒 is true density.

Replacing density with porosity in Eq. (2) yields:

𝜀 = −𝑐. ln(𝑃) + 𝑑 (4)

Where c and d are constants.

Eq. (4) describes a linear relationship between ln (P) and porosity for powder compaction,

Therefore its differential form is

𝑑𝜀 = −𝑐𝑑𝑃

𝑃 (5)

The constant c expresses the effect of a change in load on porosity.

Jones Model

Similarly Jones (1960) expressed the density relationship with pressure for metal powder

compaction. He derived a linear equation which expressed the logarithmic value of

density as a function of the logarithmic of pressure (P) [40].

ln(𝐷) = 𝑎. ln(𝑃) + 𝑏 (6)

Where D is density of compact powder mixture (kg/m3), a and b are constants.

A review of relevant models above is given below.

23

The dependence between load (P), density (D) and porosity (ε) of the materials in all form

of models was quoted. The behavior of ground biomass subjected to compaction is of

great interest in this study. Some other factors that may affect the pelleting of biomass are

packing density and particle size distribution. Compaction of biomass is a complex and

irreversible dynamic process. The packing density is inversely proportional to the particle

density [41], and thus particle size is an important parameter to consider during pelleting.

Depending on the chemical characteristics of biomass species, which is the plastic

deformation characteristics, ground biomass at some stage may not be easy to handle as

it is greatly influenced if the particle is free flowing. Poor flow characteristics often result

in a wide variation in the strength of pellet (S) and also effects on energy of compaction

(E).

Fell and Newton Model

Fell and Newton (1970) expressed the relationship between fracture load (P) and tensile

strength (T) as below [42].

𝑇 =2𝑃

𝜋𝑑𝑡 (7)

Where d is compacted diameter and t is compacted thickness.

On the basis of limited publications regarding compaction energy of ground biomass for

pellet, it is however difficult to draw general conclusions on how different biomass

species quantitatively affect the energy of pelleting biomass. In this thesis,

displacement vs. force profiles of compaction experiments have been used to obtain

information regarding the compaction behavior of ground forestry and agricultural

biomass and their blends. Subsequent events that occur in process of compaction are:

24

a. Flow of particles

b. Particles repacking

c. Fusion and bonding

d. Decompression

e. Ejection

No attempt is made to analyze the profile and event for mechanical energy estimation in

biomass pellet. In this thesis, the influence of particle size and blends on the compaction

energy was thus determined using force–displacement profiles. Solid pellet body is then

tested for break force response to determine the strength. These results are discussed in

Paper I and III, and the conclusions are summarized in Chapter 4.

2.3.4 Blending effects on thermal behavior

Agricultural biomass comes naturally with low bulk density and its presence is rich in

nature. Thus the low bulk density agricultural biomass was classified as a potential residue

that can be transformed into energy based solid fuel. However, problems were

encountered during combustion by promoting high ash content. Agricultural biomass was

difficult to be densified due to low moisture content which denotes low durability

characteristics. Low durability characteristics may cause to be brittle, emit fine particles,

thus resulting in equipment failure. Blending it with forestry biomass may reduce the ash

content as well as increase the durability of pellet fuel since woody is more moist and

high in lignin content that can bind the particles. The process variables such as pressure,

temperature, moisture content and particle size are the main effect for the low durability

of pellets. This fuel property and differences between biomass affect the combustion and

25

pyrolysis process. They are expected to have additional significant impacts. To accelerate

the deployment of agricultural biomass in combustion and pyrolysis, the development of

biomass blends fuel was a potentially effective approach to account for a measure of fuel

flexibility. This fuel flexibility should not only reduce the wide range of variability

expected in agricultural biomass compositions but should also include the possibility of

agricultural and forestry biomass co-utilization. The development of biomass blends

pellet requires accurate characterization of the thermal behavior in different gas

environments. Experimental studies are necessary to understand the mechanisms of the

combustion and pyrolysis processes and to obtain experimental data needed to validate

the product gas of biomass blends. This study was found limited for biomass blends fuel.

Therefore this thesis was to investigate the impact of inert and oxidative gaseous

environments on thermal behavior and the effect of biomass blends on the ignition

temperature, volatility zone temperature and heat release contribution. The detail of

results was given in the appended Paper IV, V and VI.

26

Chapter 3 MATERIALS AND METHODOLOGY

This chapter provides a description of the materials and a brief overview of the various

preparations used in this thesis. More detailed analyses are discussed and summarized in

Paper I-VI. The design of experimental details can be found in the appended papers.

3.1 Preparation of materials

The biomass used in this thesis was from the agricultural (reed canary grass (RCG),

timothy hay (H), and switchgrass (SW)) and forestry biomass (spruce, S and pine, P).

Prior to analysis, the samples were dried in the oven at 105 oC for 24 h, and then ground

using Wiley mill to less than 1 mm particle size. The samples were then be sieved and

separated into 3 groups of particle sizes: 150–300, 300–425 and 425–600 µm. The

acquired range of moisture content of samples was prepared so that they are of 8, 10 and

12 % each. Blending of the biomass was done manually prior to analyses and pelleting

process. Table 3.1 presents the individual biomasses and their blend ratios. The individual

biomasses and their blends were stored in an airtight container for analysis (chemical,

physical and thermal characteristics) and development of blended biomass pellets.

27

Table 3 1 Biomass and blending ratios

Biomass Blending

Spruce (S) 100 %

Pine (P) 100 %

Reed Canary Grass (RCG) 100 %

Timothy hay (H) 100 %

Switchgrass (SW) 100 %

Spruce + RCG (S+RCG) 25 % + 75 % 50 % + 50 % 75 % + 25 %

Spruce + Timothy Hay (S+H) 25 % + 75 % 50 % + 50 % 75 % + 25 %

Spruce + Switchgrass (S+SW) 25 % + 75 % 50 % + 50 % 75 % + 25 %

Pine + RCG (P+RCG) 25 % + 75 % 50 % + 50 % 75 % + 25 %

Pine + Timothy Hay (P+H) 25 % + 75 % 50 % + 50 % 75 % + 25 %

Pine + Switchgrass (P+SW) 25 % + 75 % 50 % + 50 % 75 % + 25 %

3.2 Chemical and physical analysis

A series of tests including proximate, ultimate, mineral, heating value, weight loss

(thermogravimetric), heat loss (differential scanning calorimetric) and torrefaction were

conducted for chemical analysis. Detail of methods for the analysis of individual and

blended biomass was given in the appended papers. Table 3.2 shows each analysis and

equipment used in accordance to the standard procedure, American Standard Testing &

Materials (ASTM) and Environmental Protection Agency (EPA).

In this thesis the effects of blending by comparing the characteristics of individual

biomass with the blended biomass fuel were presented. The effects of blending biomass

28

on heating value, chemical and mineral contents, and thermal degradation of each set of

samples were examined. The changed in accordance to the theory immiscible

characteristics of mixing materials i.e. glass transition temperature was proven that the

chemical component was altered in biomass after blending. Consequently, the

performance of agricultural biomass improved after blending since it followed the

characteristics of forestry biomass.

Table 3 2 Standard procedure for characterization of biomass blends

Properties Equipment Standard procedure

Chemical analysis

Proximate analysis

(moisture content,

volatiles matter & fixed

carbon)

TGA ASTM E1641-04

Decomposition Kinetics

by TGA

Ultimate analysis

(CHNSO)

LECO VTF-900/CHNS-932

Elemental Analyser

ASTM D5291

Ash analysis Muffle furnace ASTM D3174-04

Heating value Parr 6000 bomb calorimeter ASTM

D240

Mineral analysis ICP-OES EPA Method 200.8

Specific heat analysis TGA-DSC TA instrument ASTM E1269 - 11

Physical analysis

Surface area and

porosity

BET ASTM D4641-12

Microstructure and

morphology

SEM-EDX ASTM E1588

29

3.3 Pelleting process

The compaction process for producing pellets used single pelletizer built in-house

attached to Instron compression machine. The pelletizer has the control on load and

holding time. The temperature (80 °C) was controlled with the help of a PID controller.

Pellets were individually produced using a cylindrical die with diameter of 8 mm. 1.2 g

of material was compressed to a preset compression pressure (160 MPa) and then held for

5 seconds. After compaction the pressure was released, the bottom of the cylinder was

removed and the pellet was pressed out of the cylinder. The data of compaction (force and

displacement) was collected for the analysis of compaction behavior and energy of

compaction with respect to the particle size, individual and blended biomass. Details of

methods were given in the appended papers.

The influence of particle size, moisture content and blend ratio on the pellet density,

durability, yield stress and energy of compaction needed was investigated. Subsequently,

three replications of processing pellet from each sample categories were made leading to

a total of 621 samples, which is a multiple of 23 groups of selected biomass (from Table

3) by 3 levels of particle size (150, 300 and 425 µm) by 3 levels of moisture content (8,

10 and 12 %) by 3 levels of blend ratio (25, 50 and 75 %). The pellets were then used for

crushing test. These characteristics served as the basis for the evaluation of the blended

biomass pellets that can produce better quality pellets.

30

3.4 Physical properties of pellets

3.4.1 Pellet density

Immediately after removal from the die, pellet diameter and length were measured using

a caliper with 0.01 mm resolution and pellet mass was recorded to determine the initial

density of newly formed pellets, which was calculated by dividing the weight by volume.

After that, the pellets were sealed in a closed airtight container.

3.4.2 Compression test (crushing)

The strength of the pellets was determined by radially compressing a pellet using two

parallel plates set up in an Instron compression machine. The plates moved with a velocity of

0.43 mm/s until the pellet collapsed. The length of the pellet was measured before

compression. Stress and strain data, which curve like in Figure 3.1, were recorded during the

test.

Figure 3 1 Stress and strain curves for brittle materials

Brittle materials like woody on compression typically have an initial linear region

followed by a region in which the shortening increases at a higher rate than does the load.

31

Thus, the compression stress – strain diagram has a shape that is similar to the shape of

the tensile diagram. However, brittle materials usually reach much higher ultimate

stresses in compression than in tension. The pellet strength (N/mm) was determined as the

maximum force recorded divided by the length of the pellet (Finell, 2009). Brittle materials

in compression behave elastically up to certain load, and then fail suddenly by splitting or by

cracking in the way as shown in Figure 3.1. The brittle fracture is performed by separation

and is not accompanied by noticeable plastic deformation.

Compact tensile strength, 𝜎 is proportional to the crushing strength, S as the relation given

below:

𝜎 =2𝑆

𝜋𝐷𝑡

Where D is the diameter of pellet and t is the length of pellet. However this relation does

not account the effect of blending, particle size and moisture content. In this thesis, an

approach was introduced that can be utilized to study the material sensitivity on the impact

of blending as given below:

𝐵𝑅(%) =𝑆𝑜 − 𝑆

𝑆𝑜100%

Where BR is blend ratio, So is the crushing strength of forestry biomass without

agricultural biomass and S is the crushing strength of biomass blends. In this thesis the

expected result by calculation was comparable to the experimental result. The larger the

blend ratio, the more sensitive the agricultural biomass is to the forestry biomass addition.

By adding 50 - 75% of forestry biomass into agricultural biomass, most results from

biomass blends pellet achieved higher crushing strength and almost the crushing strength

of forestry biomass pellet.

32

Thus, the aim of the study in this whole thesis is to examine the effect of blending biomass

on the pellet qualities regarding density, durability, strength, energy compaction

requirement, combustion properties, heating value, and torrefaction performance.

3.5 Summary of the work

There were four stages of process involved in this thesis as shown in Figure 3.2. The

selected agricultural, forestry and their blends underwent biomass preparation, pelleting

and evaluation. Detailed discussion and findings of each part of the works were given in

the appended papers.

33

Figure 3 2 Summary of work for the individual and blended biomass

Individual Biomass

Compaction Process

Individual Biomass

Drying Process

Grinding Process

Blended Biomass

Blended Pellets Individual Pellet

Energy of Compaction

Density, Durability

Strength Test via Crushing

MA

TE

RIA

LS

PR

EP

AR

AT

ION

P

EL

LE

TIN

G

EV

AL

UA

TIO

N

Physical, Chemical and Thermal Analysis

Sieving Process

34

Chapter 4 SUMMARY OF APPENDED PAPERS

This thesis provides a number of important insights, especially with regard to material

effect on strength, durability and density of pellets made from wood, agricultural and their

blends.

This research highlights on the chemical (mineral content, proximate and ultimate

analysis, thermogravimetric analysis), physical (porosity, particle size) and mechanical

(durability and strength) properties of agricultural (reed canary grass, timothy, and

switchgrass) and woody (spruce and pine) biomass pellets and their blends in the ratio of

25:50, 50:50, and 75:25 percentages. An increase in carbon, hydrogen, volatile and char

content, and a significant decrease in chloride were observed after blending. The strength

of the blended biomass pellets was inversely proportional to the particle size. The pellet

durability index (PDI) of pure agricultural biomass pellets was less than 90, which is well

below the PFI (Pellet Fuels Institute) standard requirement. This thesis revealed that using

a blend ratio of 50:50 it was possible to achieve PDI more than 95. Similarly, the pellet

strength of blended biomass was higher compared to individual agricultural biomass.

This study intended to narrate with theoretical understanding regarding

compaction of blended biomass particles in producing the yield stress and densified

volume of pellet. For all biomasses, pellets made from lower particle size (150 – 300 µm)

exhibited higher density (950 – 1178 kg/m3 for spruce and pine; 668 – 800 kg/m3 for reed

canary grass, timothy hay and switchgrass; 900 – 970 kg/m3 for blended biomass). The

yield stress exhibited differences in values for individual forestry (40 MPa) and

agricultural biomass (27-48 MPa); however, after blending, the values converged closest

35

to that value for forestry biomass. This investigation concluded that blending agricultural

biomass with woody biomass would not only result in better mechanical properties but

would also help to expand the biomass feedstock resource for pellet fuel industry to meet

the growing demand in future.

The energy required for the compaction of the biomass, in particular the

agricultural and woody biomass blends, was studied. The energy required in compacting

ground reed canary grass, timothy hay and switchgrass was found lower (1.61, 1.97, and

1.68 kJ) than spruce (2.36 kJ) and pine (2.35 kJ) evaluated at 159 MPa load and

temperature about 80 °C. After blending, the values were around 2 kJ with the pellet

quality approaching almost similar to pellet made from woody biomass. Agricultural

biomass could result in better quality of pellet but low in compaction energy requirements

when blending it with woody biomass.

The study investigated combustion behavior and thermal properties of both

individual and blended agricultural (reed canary grass, timothy hay and switchgrass) and

forestry biomass (spruce and pine). The experiment was carried out using

thermogravimetry (TG) analyzer with an air flow rate of 100 ml/min and heating rate of

20 °C/min. Heat required for the biomass reaction was measured using differential

scanning calorimeter (DSC). The TG results showed that the combustion behavior of all

the biomass samples including the blends were almost similar. The heat released from the

blended biomass (6.94 – 9.26 kJ/kg) was higher than the individual agricultural biomass

(4.59 – 6.78 kJ/kg) but lower than individual spruce (10.2kJ/kg) and pine (11.13kJ/kg).

The findings indicate that the reactivity of the individual agricultural biomass material

36

changed due to blending. Blending can help to increase the energy capacity and can

improve the combustion characteristics.

Under the torrefaction condition, the process using thermogravimetric analyzer

(TGA) coupled with Fourier transform infrared (FTIR), and mass spectrometer (MS),

simultaneously. The chemical functional groups present in the gases were identified by

FTIR and the concentration of gases was determined using MS at different torrefaction

temperatures ranging from 200 to 290 oC. TG-FTIR and TG-MS techniques are paired to

refine the identification of gases. The behaviour of the agricultural and forestry biomass

was not the same due to their composition variation. Both switchgrass and timothy have

two peaks of degradation rate compared to only one peak present for forestry biomass.

The mass spectrometric analysis at torrefaction temperature 200 °C and 230 °C quantified

the degradation of combustible gases: CH4, C2H4, CO and O2 around 20-30%. Whilst at

torrefaction temperature 260 °C and 290 °C the degradation of combustible gases were

more than 30%. All gases evolved from the torrefaction of agricultural and forestry

biomasses were almost similar in characteristics, but vary in proportions.

The FTIR analysis indicated that biomass blends had similar functional groups but

their decomposition was largely at higher torrefaction temperature of 290 °C. Major

product gas (CH4, CO, C2H2, O2) detected by MS during torrefaction was in the range of

28-30 % (v/v). The composition of product gas generated from torrefied biomass depends

on types of biomass blended and temperature. The change in the glass transition

temperature, Tg after blending caused the change in the composition of any agricultural

mixed with woody species.

37

Chapter 5 CONCLUSION AND RECOMMENDATIONS

4.1 Overall conclusions

Agricultural biomass is well known to have variation in the characteristics, which do not

fit well with the standard solid fuel requirements based on the mechanical properties,

heating values, ash content, moisture content and the fuel conversion yield. This research

project was developed based on the idea of blending agricultural biomass with woody

biomass, anticipating that it will generate positive effects in improving the pellet fuel

quality, creating a new feedstock for pellet industry, alleviating dependence on forestry

biomass alone, and thus making more use of agricultural based biomass. Due to blending,

the chemical characteristics were found remarkably altered, the ash content was reduced,

and an increase in carbon and hydrogen content was observed. The energy of compaction

needed for pellet from blended biomass was also found lower than the energy of

compaction for pellets made from woody biomass. In addition, the intrinsic yield stress

of blended biomass pellets with particle size lower than 150 µm follows almost close

range to the compressive stress of the parent woody biomass (pine and spruce).

The fuel conversion yield of the blended biomass has been synthesized and analyzed for

its application in the combustion, pyrolysis and torrefaction. Temperature was found to

be the most influencing parameter significantly affecting the kinetic reaction, product gas

compositions, and environmental emissions (ash content). In conclusion the pellet fuel

industry can successfully implement the blending approach as this study has demonstrated

achievement of positive effects in pellet fuel quality.

38

4.3 Recommendations for future research

As the research is continuing and fuel pellet biomass is indefinitely wide, heterogeneous,

and complex, various future works can be performed to extend the existing knowledge.

The following recommendations are made for future study:

More data can be generated by incorporating the other process parameters (such

as temperature, pressure) and tools used (such as die diameter).

Specific composition of the species needs to be taken into account to correlate the

similarity with the reference species.

Fixed pressure and temperature during the pelletization process can help to

standardize compaction process of the pellet, but optimum parameters should not

be neglected.

The compaction of ground biomass needs to evaluate for cohesion of particles to

endorse the pellet quality.

Moisture content can have its effects on the energy consumption in biomass

compaction process; thus further investigation is necessary to optimize the energy

consumption at various moisture contents.

Energy contribution by different types of biomass pellets is also different due to

difference in their chemical characteristics; thus further investigation for scale up

pyrolysis and combustion is essential.

4.4 Contributions to pellet fuel development field

Prior research mainly focused on production of pellets made from woody biomass only

or using individual energy crops. This thesis has extended the knowledge on new

39

feedstock development from biomass blends, which was considered to be the first idea in

improving the pellet quality incorporating the agricultural biomass. The research

generated the new data on chemical and physio-mechanical properties of woody and

agricultural biomass blends, intended to make contributions to an understanding of the

differences in pellet fuel quality.

Research also progressed further on compaction behavior of ground biomass relating the

particle size, density and yield stress of pellet made from biomass blends. While this

research provides a number of important insights, especially with regard to material effect

on strength, durability and density of pellets made from wood, agricultural and their

blends, it gives slight attention to material influences upon energy requirement for

compaction and energy contribution by the fuel pellets.

This research demonstrated has demonstrated that pyrolysis and combustion of blended

biomass using TG coupled with FTIR and MS was the first attempt to investigate the gas

composition and chemical functional groups. Thus, this research provides unique

contributions to literature and to the fuel pellet industry by advancing the understanding

of blending biomass and processes associated with fuel pellet quality.

40

Chapter 6 REFERENCES

[1] Sunnigrahi, P and Ragauskas, A. J. (2010). Poplar as a feedstock for biofuels: A

review of compositional characteristics. Biofuels Bioproduct Biorefinery.

4, 209–226.

[2] Mohan, D., Pittman, C. U., and Steele, P.H. (2006). Pyrolysis of wood/biomass for

bio-oil: A critical review. Energy Fuels. 20, 848-889.

[3] US Department of Energy. Biomass Feedstock Composition and Property Database,

Available at

http://www.afdc.energy.gov/biomass/progs/search3.cgi?12701 as of August 1,

2014.

[4] Bridgeman, T.G., Jones, J.M., Shield, J., and Williams, P.T. (2008) Torrefaction of

reed canary grass, wheat straw and willow to enhance solid qualities and

combustion properties. Fuel. 87, 844.

[5] Tabil, L., Adapa, P and Kashaninejad, M. (2011) Biomass Feedstock Pre-

Processing – Part 2: Densification. www.intechopen.com

[6] US Department of Energy. Biomass Feedstock Composition and Property Database.

http://www.afdc.energy.gov/biomass/progs/search3.cgi?12701 as of August 1,

2014

[7] Lam, P. S., S. Sokhansaj, X. Bi, C. Lim, L. J. Naimi, M. Hoque, S. Mani, A. R.

Womac, X. P. Ye, and S. Narayan. (2008) Bulk density of wet and dry wheat straw

and dry particles. Applied Engineering Agricultural. 24: 351-358.

[8] Tulumuru, J. S., Wright, C. T., Hess, J. R. and Kenney, K. L. (2011) A review of

biomass densification systems to develop uniform feedstock commodities for

bioenergy application Biofuel bioprocess biorefinery. 5(6): Pages 683–707.

[9] Dhamodaran, A., and Afzal, M.T. (2012) Compression and Springback Properties

of Hardwood and Softwwod Pellets. BioResources. 7(3), 4362-4376.

41

[10] Arshadi, M., Gref, R., Geladi, P., Dahlqvist, S., and Lestander, T. (2008). The

influence of raw material characteristics on the industrial pelletizing process and

pellet quality. Fuel Processing & Technology., 89, 1442‐1447.

[11] Larsson S. H. Kinematic wall friction properties of reed canary grass powder at high

and low normal stress. Powder Technology. 198(1), 108-113.

[12a] Nielsen, N. P. K., Gardner, D. J., Poulsen, T., and Felby, C. (2009). Importance of

temperature, moisture content, and species for the conversion process of wood

residues into fuel pellets. Wood and Fiber Science. 41, 414‐425.

[12b] Nielsen, N. P. K., Holm, J. K., and Felby, C. (2009). Effect of Fiber Orientation on

Compression and Frictional Properties of Sawdust Particles in Fuel Pellet

Production. Energy & Fuels. 23, 3211‐3216

[13] Holm, J. K., Henriksen, U. B., Hustad, J. E., and Sorensen, L. H. (2006). Toward

an understanding of controlling parameters in softwood and hardwood pellets

production. Energy & Fuels. 20, 2686‐2694.

[14] Li, Y. D.; Liu, H. (2000). High‐pressure densification of wood residues to form an

upgraded fuel. Biomass & Bioenergy. 19, 177‐186.

[15] Holm, J.K., Stelte, W., Posselt, D., Ahrenfeldt, J., and Henriksen, U. B. (2011)

Optimization of a multiparameter model for biomass pelletization to investigate

temperature dependence and to facilitate fast testing of pelletization behavior.

Energy & Fuels. 25, 3706‐3711.

[16] Larsson, S. H., Thyrel, M., Geladi, P., and Lestander, T. A. (2008). High quality

biofuel pellet production from pre‐compacted low density raw materials.

Bioresource Technology. 99, 7176‐7182.

[17a]Kaliyan, N., and R. V. Morey. (2005). Densification characteristics of corn stover

and switchgrass. Transactions of the ASABE. 52(3), 907-920

42

[17b]Kaliyan, N., Morey, R. V., White, M. D., and Doering, A., (2009). Roll press

briquetting and pelleting of corn stover and switchgrass. Transactions of the

ASABE. 52, 543‐555.

[17c]Kaliyan, N., and Morey, R. V. (2009). Strategies to improve durability of

switchgrass briquettes. Transactions of the ASABE. 52, 1943‐1953.

[18] Adapa, P. K., Tabil, L. G., Schoenau, G. J., Crerar, B. and Sokhansanj, S. (2002).

Compression characteristics of fractionated alfalfa grinds. Powder Handling &

Processing. 252‐ 259.

[19] Gonza´lez, M., and Mun˜ oz, G. (2002). Experimental determination of the optimal

parameters of the mechanical densification process of fibrous materials for animal

feed. ASAE Paper no. 02-1072. American Society of Agricultural Engineers. St

Joseph, MI, USA.

[20] Mani, S, Tabil, L. G., and Sokhansaj, S. (2006) Effect of compressive force, particle

size and moisture content on mechanical properties of biomass pellets from grasses.

Biomass Bioenergy, 30 (7): 648-654

[21a] Dhamodaran, A. and Afzal, M. T. (2012). Compression and Springback Properties

of Hardwood and Softwwod Pellets. BioResources. 7(3): 4362-4376

[21b]Dhamodaran, A., and Afzal, M.T. (2013). Modeling and Characterization of Reed

Canary Grass Pellet Formation Phenomenon. International Journal of Renewable

and Sustainable Energy. 2013; 2(2): 63-73

[22] Stelte, W.; Holm, J. K.; Sanadi, A. R.; Barsberg, S.; Ahrenfeldt, J.; Henriksen, U.

B. (2011). A study of bonding and failure mechanisms in fuel pellets from different

biomass resources. Biomass & Bioenergy. 35, 910‐918.

[23] Holm, J. K., Henriksen, U. B., Wand, K., Hustad, J. E., and Posselt, D. (2007)

Experimental verification of novel pellet model using a single pelleter unit. Energy

& Fuels. 21, 2446‐2449.

43

[24] Jensen, P. D., Hartmann, H., Böhm, T., Temmerman, M., Rabier, F., and Morsing,

M. (2006). Moisture content determination in solid biofuels by dielectric and NIR

reflection methods. Biomass Bioenergy. 30, 935–943.

doi:10.1016/j.biombioe.2006.06.005

[25] Serrano, C., E, Monedero, M. Lapuerta, and H. Portero. (2011). Effect of moisture

content, particle size, and pine addition on quality parameters of barley straw

pellets. Fuel Processing Technology. 92(3), 699-706.

[26] Manickam, I. N. and Surresh S. R. (2011). Effect of Moisture Content and Particle

Size on Bulk Density, Porosity, Particle Density and Coefficient of Friction of Coir

Pith. International Journal of Engineering Science and Technology (IJEST).

3(4).

[27] Zhang, Y., Ghaly, A. E. Ghaly and Li, B. (2012). Physical Properties of Corn

Residues. American Journal of Biochemistry and Biotechnology, 8(2), 44-53.

[28] Barbieri, L., Andreola F, Lancellotti, I, Taurino R. (2013). Management of

agricultural biomass wastes: preliminary study on characterization and valorisation

in clay matrix bricks. Waste Management, 33(11), 2307-2315.

[29] Mani, S., Tabil, L. G., and Sokhansanj, S. (2004). Evaluation of Compaction

Equations Applied to Four Biomass Species. Canadian Biosystems Engineering. 46,

3.55–3.61.

[30] Wilson, O. T. (2010). Factors Affecting Wood Pellet Durability. MSc. Thesis. The

Pennsylvania State University.

[31] Palsauskas, M and Mauricas, T. (2013). Use of Mixed Biofuel for Pellet Production.

Engineering for Rural Development. 23 (24), 369-375.

[32] Lehtikangas, P. (2001). Quality properties of pelletised sawdust, logging residues

and bark. Biomass and Bioenergy. 20, 351–360.

[33] Van Dam, J. E. G., M. J. A. van den Oever, W. Teunissen, E. R. P. Keijsers, and A.

G. Peralta, (2004). Process for Production of High Density/High Performance

44

Binderless Boards from Whole Coconut Husk—Part 1: Lignin as Intrinsic

Thermosetting Binder Resin. Industrial Crops and Products. 19 (3), 207–216.

[34] Mani, S, Tabil, L. G., and Sokhansaj, S. An overview of Compaction of Biomass

Grinds, Powder Handling and Processing. Biomass Bioenergy. 2003; 15 (2): 160-

168

[35] Adapa, P. K., Tabil, L. G., Schoenau, G. J., Crerar, B. and Sokhansanj, S.

Compression characteristics of fractionated alfalfa grinds. Powder Handling &

Processing. 2002; 252‐ 259.

[36] Adapa P., Tabil L., and Schoenau G. (2009). Compaction characteristics of barley,

canola, oat and wheat straw, Biosystems Engineering, 104: 335‐344

[37] Tumuluru, J. S., Sokhansanj, S., Lim, C. J., Bi, X., Anthony, L., Staffan, M.,

Sowlati, T., and Ehsan, O. Quality of Wood P Pellets Produced in British

Columbia,” Applied Engineering in Agriculture 2010; 26 (6): 1013–1020

[38] Repsa, E. and Kronbergs, E. (2012). Friction affect on Common Reed Briquetting.

Engineering for Rural Development, 295 - 299

[39] Gurnham, F., and Masson, H.J. (1946). Expression of liquids from fibrous materials.

Industrial Engineering Chemistry. 38, (1946) 1309–1315.

[40] Jones, D. (1960). Fundamental Principles of Powder Metallurgy. Edward Arnold

Publishers Ltd., London, (1960). 242-370

[41] Tulumuru, J.S., Tabil, L.G., Song, Y., Iroba, K.L., and Meda, V. (2014). Grinding

energy and physical properties of chopped and hammer-milled barley, wheat, oat,

and canola straws, biomass and bioenergy. 60; 58-67

[42] Fell, J. T. and Newton, J. M. (1960). Determination of Tablet Strength by the

Diametral-Compression Test. Journal of Pharmaceutical Sciences. Volume 59,

Issue 5, Pages 688–691

45

Chapter 7 Collection of Papers

46

Paper I

Chemical and Mechanical Properties of Pellets Made

from Agricultural and Woody Biomass Blends

Noorfidza Y Haruna*, Muhammad T Afzalb

a Graduate student at UNB, affiliated with Universiti Teknologi PETRONAS,

Department of Chemical Engineering, 32610 Bandar Seri Iskandar,

Perak, Malaysia.

b University of New Brunswick, Department of Mechanical Engineering, 15 Dineen

Drive, E3B 5A3 Fredericton, New Brunswick, Canada.

Transactions of the ASABE, 58(4), 921-930

Abstract

Traditionally, the pellet industry depends heavily on forestry biomass. Large amounts of

agricultural residues are generated every year and have the potential to be used as

alternative feedstocks for pellet production. The aim of this study was to analyze chemical

(mineral content, proximate and ultimate analysis, thermogravimetric analysis), physical

(porosity, particle size) and mechanical (durability and strength) properties of agricultural

(reed canary grass, timothy, and switchgrass) and woody (spruce and pine) biomass

pellets and their blends. Agricultural and woody biomasses were blended in ratio of 25:50,

50:50, and 75:25 percentages. The biomass pellets were formed using a closed end

plunger-die assembly attached to an Instron machine. An increase in volatile and char

content of the blended biomass was observed using thermogravimetric analysis. This was

further confirmed by an increase in carbon and hydrogen content of the blended biomass.

A significant decrease in chloride was observed after blending. The strength of the

blended biomass pellets was inversely proportional to the particle size. The pellet

durability index (PDI) of pure agricultural biomass pellets was less than 90 which is well

below the PFI (Pellet Fuels Institute) standard requirement. This study revealed that using

a blend ratio of 50:50 it was possible to achieve PDI more than 95. Similarly, the pellet

strength of blended biomass was higher compared to individual agricultural biomass. In

conclusion, this study confirms that agricultural biomass can be mixed with forestry

biomass to develop an alternative feedstock for the pellet fuel industry.

Keywords: Blend; Chemical; Mechanical; Mineral; Pellets; Strength

47

Introduction

Interest in the production of biomass fuel pellets is increasing at global level for industrial

and domestic heating applications. Europe, the United States and Canada are the main

producers as well as consumers of fuel pellets (Lim, 2012; Chocci et al, 2011; Samuelsson

et al, 2009), however, North America is the largest producer (3.2 million tons in 2012).

Canada has had an increase in the export of fuel pellets of 25 percent year-over-year

(Canadian Biomass Energy, 2012). Most pellets are exported to Europe and Asia. China

is the second largest consumer of pellets after the United States. Forestry and agricultural

residues account for about 42% of the total Canadian biomass (Wood and Layzell, 2003).

According to the US Department of Energy, the total forest biomass feedstock available

in year 2011 for the USA was about 130 million dry tons and mill residues contributed

about 25 % of the total forest biomass (Billion ton report, USA). In Canada, the estimated

production of mill residues was under 11 million dry tons in 2009 (Bradley, 2010). Over

15 t ha-1 (of dry biomass) of reed canary grass is produced in Canada (Wrobel et al., 2009).

Further, about 25 Mtoe (Megatonne of oil equivalent) of agricultural residues is generated

annually in Canada (Helwig et al., 2002).

The pellet market heavily depends on woody biomass resources to fulfill the current

demand for pellet production. The market for fuel pellets is growing very rapidly due to

increase in demand for pellets for industrial and residential heating in Europe, Asia and

South America. Annual production of pellets is about 16 million metric tons, but

production is expected to be tripled to an estimated 46 million metric tons by year 2020

(Site Selection magazine, July 2013; Canadian Biomass Magazine, 2013). This means

other feedstock resources have to be exploited in order to meet the future demand.

48

Therefore, the pellet industry is looking for alternate biomass resources to meet future

demand. Additionally, abundant biomass resources in the form of agricultural residues

are available in Canada which can be exploited as a potential feedstock for pellet

production. On the other hand, pellets made from agricultural biomass are poor in quality

as compared to forestry biomass. This is due to the difference in physical and chemical

characteristics of agricultural and forestry biomass. For instance, agricultural biomass has

lower lignin content than forestry biomass. It is anticipated that blending would affect the

mechanical properties of densified biomass. Thus, in this study agricultural biomass was

blended with forestry biomass to investigate the unknown chemical and mechanical

properties after blending.

The quality of biomass pellets is usually defined by their physical, chemical and

mechanical properties. The Pellet Fuels Institute (PFI) has defined standards for pellets

such as physical, chemical and mechanical for residential and commercial-grade (Pellet

Fuels Institute, 2011), as shown in Table I.1. Basic chemical properties include heating

value, proximate and ultimate analysis, and mineral contents. On the other hand, pellets

should also have high mechanical strength and durability (Kaliyan and Morey, 2006).

Further, physical and chemical characteristics of the feedstock play an important role in

order to achieve the pellet quality as per the defined standards. Particle size has a major

influence on the mechanical properties of pellets. Particle size of less than 1 mm generally

provides pellets with greater strength because of stronger inter-particle bonding.

49

Table I.1: Pellet Fuels Institute (PFI) Standards for Pellet Fuel Quality

Fuel Property

PFI

Premium

PFI

standard PFI utility

Mandatory according to PFI

Bulk density 40.0 - 46.0 38.0 - 46.0 38.0 - 46.0

Diameter, inches 0.230 - 0.285 0.230 - 0.285 0.230 - 0.285

Pellet durability index ≥96.5 ≥95 ≥95

Fines, % ≤0.5 ≤1.0 ≤1.0

Ash, % ≤1.0 ≤2.0 ≤6.0

Length, % greater than 1.5

inches ≤1.0 ≤1.0 ≤1.0

Moisture, % ≤8.0 ≤10.0 ≤10.0

Chloride, ppm ≤300 ≤300 ≤300

Larger particles tend to develop fissures in the pellets (Serrano et al, 2011). It should be

noted that too small particles (less 150 µm) can create erosion of the pellet machine

components (Tumuluru et al. 2011). Pellets when compacted enable the particles to

remain intact while handling, and minimize the loss of fine mass when subjected to

mechanical or pneumatic agitation (Lehtikangas, 2001; Samson and Duxbury, 2000). In

previous studies it was reported (Jannasch, 2001; Zafari and Kianmehr, 2012) that the

hardness of switchgrass pellets increased from 2.3 to 6.3 kN when particle size was

decreased from 3.2 to 2.8 mm.

Agricultural and woody biomass shows much variation in chemical composition which

affects the combustion characteristics of the pellet. Further, blending of biomass can

increase thermochemical reactivity (Vamvuka and Sfakiotakis, 2011) as well as energy

density (Basmina et al. (2013). For this reason, it is hypothesized that blending biomass

can improve pellet fuel properties. Lignocellulosic components of biomass were reported

(Briggs et al, 1999) to strongly influence pellet quality. The degree of biomass pellet

50

strength is a function of the inter-particle bonding in which lignin and extractives usually

act as binding agent. However, biomass containing high minerals can limit the binding

properties of pellets (James, 2012). For instance, the presence of inorganic phosphate in

the biomass can affect the pellet durability (Wamsley et al, 2012).

Other pelletization process parameters, such as pressure, temperature, die geometry,

hammer mill size and moisture content, have been reported to influence the formation of

quality pellets (Lehtikangas, 2001; Zafari and Kianmehr, 2012; Wilson et al, 2011;

Dharmodaran and Afzal, 2013). This review concludes that chemical, physical and

process paramaters when combined can influence the performance of biomass and

blended biomass pellets. However, the knowledge of such effects is limited due to lack of

detailed chemical and mechanical characterization of individual and blend pellets.

Thus, the main objective of this study was to investigate the chemical, physical and

mechanical properties of agricultural and woody biomass pellets and their blends. The

blending ratio between agricultural and woody biomass was 25:75, 50:50, and 75:25.

Thermogravimetric analysis was carried out from room temperature to 800 °C to

determine the proximate analysis of individual biomass components and their blends.

Chemical properties included ultimate, mineral, and ash analysis in addition to heating

value of the blend biomass. Further, the strength and durability of biomass and their blend

pellets were analyzed as the mechanical properties.

51

Methodology

Sample Preparation

In this study, agricultural biomass such as reed canary grass, timothy hay and switchgrass

was selected and acquired from New Brunswick agricultural farms surrounding the

Fredericton area. The woody biomass (spruce and pine) were obtained from a sawmill

situated in the Fredericton, New Brunswick area. The raw biomass material was stored in

plastic bags under ambient temperature in a closed container. Prior to analysis, the

samples were dried in the oven at 105 oC for 24 h, and then ground using a Wiley mill

(Thomas Scientific, Swedesboro, NJ, USA) to less than 1 mm particle size. Table I.2

presents the individual biomass and their blend ratios.

Table I.2: Biomass and Blending Ratio.

Biomass Blending (%)

Spruce + RCG (S+RCG) 25S:75RCG 50S:50RCG 75S:25RCG

Spruce + Timothy Hay (S+H) 25S:75H 50S:50H 75S:25H

Spruce + Switchgrass (S+SW) 25S:75SW 50S:50SW 75S:25SW

Pine + RCG (P+RCG) 25P:75RCG 50P:50RCG 75P:25RCG

Pine + Timothy Hay (P+H) 25P:75H 50P:50H 75P:25H

Pine + Switchgrass (P+SW) 25P:75SW 50P:50SW 75P:25SW

Notation: RCG =Reed Canary Grass, H=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine

The ground biomass and their blends were stored in airtight containers for further analysis.

It should be noted that thermogravimetric analysis, ultimate and mineral analysis, and

heating value were conducted for 100% individual biomass and 50:50 wt% biomass

blends; whereas, the strength of biomass and its blend pellets was carried out for 100%

biomass, 25:50 wt%, 50:50 wt%, 75:25 wt%, biomass blends.

52

Pelletization

The process parameters for pelletizing the selected biomass in this study were; pressure

of 159 MPa, temperature at 80 °C and holding time of 5 s. The selection of temperature

and pressure in the present study was based on our previous detailed research

(Dharmodaran and Afzal, 2012) in which pellets showed lowest springback effect at 70

°C temperature and 159.3 MPa pressure. This would result in high density and good

mechanical properties such as durability. Further, the durability of pellets is significantly

influenced by the temperature (Li and Liu, 2000). For instance, temperature in the range

of 90 – 140 °C was used in pellet production by most of the researchers which requires

less pressure (<100 MPa) (Tumuluru et al, 2011). It is obvious that increase in temperature

would result in a low pressure requirement (Stelte et al., 2011). Lam et al. (2014) produced

pellets made from fir with pressure ranging from 63 to 190 MPa and temperature ranging

from 70 to 130 °C.

Prior to pelletization the biomass was stored in an airtight container under room condition

for about a week. The ground materials were sieved and pellets of three different types of

particle sizes (150 – 300 µm, 300 – 425 µm and 425 -600 µm) were formed. The moisture

content of the samples was about 10 % before pelletization. Pellets were made with the

help of a closed end die built in house to assess the influence of particle sizes and blend

ratio. The closed end die was 8 mm diameter and 202 mm in length (Figure I.1) and was

attached to an Instron testing unit, Model 1332 (Instron, Grove City PA, USA), which had

a control (Controller Instron Model # 8500P) on load and holding time with control

software (Instron Series 9). The temperature was controlled by the help of thermocouples

connected to a PID (proportional integral derivative) controller.

53

The load applied to produce pellets was 159 MPa at temperature of 80 °C, and the speed

of the piston was 0.43 mm/s with a hold time of 5 s for each run. For the blends with three

different ratios and three particle sizes, three replications were made leading to a total

number of 162 (3 particle size and blend ratio: 32 × 6 types of biomass × 3 replications)

tests. Approximately, 1.2 g of biomass sample was fed into the die for each run to produce

a single pellet.

Figure I.1: Schematic of Single Pelletization Setup (Dharmodaran and Afzal, 2012).

Pin connection to instron machine

crosshead

Piston

Cylinder (die) with 8 mm internal

diameter

Heating elements connected to PID

controller

Stopping blocks which can be removed

to take pellets

Pin connection to base of instron

machine

210 mm

202 mm

54

Proximate and Ultimate Analysis

Proximate analysis was determined twice using TGA Q500, (TA Instrument Inc., New

Castle DE, USA) equipment as per the ASTM E1641-04 method, while the ash content

was investigated as per NREL/TP-510-42622 method with 6 replications. The samples

with particle size of 150 µm were heated from room temperature to 800 °C with heating

rate of 20 °C/min. This heating rate was selected based on the highest volatiles and char

yield that it produced from spruce and pine (Yub Harun and Afzal, 2010), and from

agricultural biomass (miscanthus, poplar and rice husk) (Kok and Özgür, 2013). Nitrogen

was used as purge gas with a flow rate of 100 ml/min. Ultimate analyses for the tested

samples, which were repeated twice, were obtained using a LECO CHNS 932 elemental

analyzer as per the ASTM D5291 method. Element oxygen was calculated by difference.

The higher heating values (HHV) were measured using an Isoperibol Oxygen Bomb

Calorimeter (Parr Instrument Company, Moline IL, USA) according to the standard

method, ASTM D-5865.

Mineral Analysis

The mineral concentration was determined according to EPA (Environmental Protection

Agency) methods and procedures. The samples were prepared using MarsXpress

microwave (CEM Corporation, Matthews NC, USA) assisted digestion in nitric acid

according to EPA Method 3051. The resulting solutions were analyzed for trace elements

by Thermo X Series II ICP-MS (Thermo Fischer Scientific Inc., Waltham MA, USA)

according to EPA Method 200.8. Phosphorus was determined by Varian Vista AX ICP-

ES (Agilent Technologies Inc., Santa Clara CA, USA) according to EPA Method 200.7.

55

Silica was determined by ICP-ES on sodium peroxide fusions of samples ash. Chlorine

was determined calorimetrically on aqueous leaches of the samples using photometric

analyzer, AquaKem 250 (Thermo Fischer Scientific Inc., Waltham MA, USA) according

to standard method, APHA 4500-CL-E.

Mechanical Property

In order to determine the strength of individual biomass and blended pellets, an Instron

testing machine was used. Each pellet was tested for its compressive break force in axial

orientation. For each condition strength tests were carried out three times and the final

value represents the mean value of the strength. A two-way ANOVA statistical method

was performed by means of MINITAB software version 16.1.0 (Minitab Inc.,

Pennsylvania, USA) for analyzing the effect of blend ratio and particle size on the pellet

strength.

Results and Discussion

Proximate and TG Analysis

The weight loss for agricultural and woody biomass with respect to increase in

temperature is presented in Figure I.2a. The initial weight loss of about 2 to 3 wt% from

room temperature to 100 °C was due to removal of moisture content from the biomass.

The major weight loss (70 wt%) for woody biomass was in the temperature range of 225

°C to 375 °C. The initial degradation temperature (225 °C) and weight loss characteristics

for spruce and pine biomass were similar. The leftover char residue for woody biomass at

800 °C was also about 15 wt%. However, agricultural biomass showed significant

56

variation in weight loss profile. The major weight loss for RCG, switchgrass and timothy

hay were about 65 wt%, 78 wt%, and 66 wt%, respectively. The initial degradation

temperature for RCG (225 °C), switchgrass (225 °C) and timothy hay (180 °C) was also

found to vary. The final leftover char residue was about 15 wt%, 14 wt. %, 21 wt. %, 21

wt. % and 11 wt. % for spruce, pine, RCG, timothy hay, and switchgrass, respectively.

Low degradation temperature and weight loss for timothy hay might be due to its physical

and chemical characteristics. This indicates that the degradation characteristics of both

woody and agricultural biomass depend greatly on their chemical components and can

differ greatly.

A significant change in the weight loss profile of blended biomass was found as shown in

Figure I.2b. The initial degradation temperature (225 °C) was similar for all blend

biomass. The major weight loss for all blend biomass was found in the temperature range

between 225 °C and 400 °C. The only variation was in leftover char residue. The amount

of char residue left at 800 °C for S+RCG, S+H, S+SW, P+RCG, P+H, and P+SW was

about 15 wt%, 17 wt%, 14 wt%, 19 wt%, 15 wt%, and 20 wt%, respectively. The initial

degradation temperature for timothy blend biomass was 225 °C as compared to 180 °C

(Figure I.3a) when individual timothy hay was decomposed. The char residue of blend

biomass was in the range of 15.0 to 20.2 wt% (Figure I.2b) as compared to 10.5 to 21.0

wt% (Figure I.2a) for individual decomposed agricultural and woody biomass. Hence,

blending agricultural with woody samples can influence the thermal degradation

characteristics of the biomass and probably the chemical properties and heating values.

The difference in derivative thermogravimetric (DTG) curves for agricultural and woody

biomass in Figures I.3a and I.3b might be due to differences in their lignocellulosic

57

content. Therefore, woody biomass showed one major peak at 355 °C attributed due to

decomposition of cellulose and hemicellulose.

Figure I.2: TGA profile of agricultural and woody biomass; individual (a) and

blend 50%-50% (b)

The DTG curves of agricultural biomass depicted two distinguished peaks. The shoulder

peak corresponding to 230 °C, 305 °C, and 310 °C for timothy hay, RCG and switchgrass,

respectively might be due to decomposition of hemicellulosic materials. However, the

major peaks at 340 °C, 370 °C, and 360 °C for timothy hay, RCG and switchgrass,

respectively were due to degradation of the cellulosic component in biomass. This also

shows that woody biomass might contain higher cellulose content than agricultural

biomass. Magdziarz and Wilk, 2012 also reported one DTG peak for woody biomass and

two distinguished peaks for agricultural biomass.

DTG results of blended biomass in Figures 2.3c and 2.3d showed slight shoulder peaks

on the left side of the curve. Interestingly, the temperature corresponding to the major

58

derivative weight loss was similar for all blends at about 370 °C. Blended biomass showed

only one major peak similar to woody biomass attributed to decomposition of

hemicellulose and cellulose. The temperature where the major reaction mechanism takes

place for the individual agricultural biomass might change due to blending method. It

seems that the chemical content of woody biomass plays a more dominant role as

compared to agricultural biomass.

Figure I.3: DTG profiles of agricultural, woody biomass and their blends

(50%+50%).

Finally, Table I.3 shows the difference in proximate analysis of individual and blended

agricultural and woody biomass. Theoretically, the moisture content of blended biomass

should be an average of the individual agricultural and woody biomass, and

59

experimentally was higher than the agricultural biomass. Low moisture content can

reduce binding during the pelleting process. Volatile matter of agricultural biomass

ranged from 65 to 78 %, whereas the values for blends (69-73%) were similar to woody

biomass (70 %). Similarly, residual char of blended biomass was reduced due to

considerable reduction of mineral and ash content.

Table I.3: Proximate Analysis of Individual and Blend Biomass (50:50 wt%).

Biomass Moisture content,

wt.%

Volatile

product, wt.%

Residual char,

wt.% @ 800 °C

Spruce 8.0 70 15

Pine 9.3 70 14

RCG 6.4 65 21

Hay 6.9 66 21

Switchgrass 7.0 78 11

S+RCG 7.2 69 15

S+SW 7.7 73 14

S+H 7.3 70 17

P+RCG 7.5 70 19

P+SW 7.5 71 20

P+H 7.6 69 15

Notation: RCG =Reed Canary Grass, H=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine

Ultimate Analysis

The carbon and oxygen contents of individual and blended biomass were 45.0 to 47.2 %

and 46.6 to 48.8 wt%, respectively, as shown in Table I.4. Moreover, the heating value of

reed canary grass, timothy hay and switchgrass was lower as compared to woody biomass.

This might be due to higher carbon and hydrogen content in woody biomass. The

60

individual heating values of the biomass samples were averaged to provide the blend

heating values. For example, in case of S+RCG blend, the individual heating value of

spruce (19.15 MJ/kg) and RCG (18.61 MJ/kg) was averaged to give the blend heating

value of 18.88 MJ/kg. It was observed that the heating values calculated by averaging

were slightly higher than the measured values except in case of pine and hay blend which

was similar. The heating values of the biomass samples are the mean values of three

replications and the standard deviation of the individual and blend biomass are shown in

Table I.4. The heating value was found to improve as compared to individual agricultural

biomass after blending it with woody biomass. The heating value usually depends on the

chemical composition of the biomass material. In particular, carbon (C) and hydrogen (H)

plays an important role since these two are finally converted into carbon dioxide and water

during reaction (Senelwa and Sim, 1999). The C+H contents of the woody and blended

biomass were in the range of 51 to 53% as shown in Figure I.4.

Figure I.4: Carbon+hydrogen (C+H) versus Oxygen (O) Content for Agricultural

and Woody Biomass and Their Blends (50 wt%:50 wt%).

50.5

51.0

51.5

52.0

52.5

53.0

53.5

46.547.047.548.048.549.0

C+

H (

%)

O (%)

Agricultural

biomass

Woody and

blend biomass

61

The lower concentration of oxygen in the woody and blended biomass indicates lower

emission of NOx as compared to pure agricultural biomass. Among all blended biomass,

RCG + P has the lowest (C+H) content. However, the amount is higher than the individual

agricultural biomass. Higher C+H values as compared to O is essential for co-firing,

gasification and pyrolysis because it can lower the O concentration (or higher in C+H)

resulting in a reduction of NOx to nitrogen. The amount of nitrogen and sulfur content in

agricultural biomass was higher as compared to individual woody and blended biomass.

Typically, biomass materials have larger O/C and H/C ratios compared to coals (Senneca,

2011).

Ash Content

Tumuluru et al., (2010) reported that biomass with high ash content can affect the pellet

die causing erosion and consequently will affect the binding phenomena of the pellets.

Obviously, the ash content of blended biomass was found lower than the individual

agricultural biomass. This may help in reducing erosion of pellet dies.

62

Table I.4 :Mean Heating Value (HHV, MJ/kg) and Ultimate Analysis (Dry Basis) of Individual and Blend Biomass (50:50

wt%).

RCG H SW S P S+RCG S+H S+SW P+RCG P+H P+SW

HHV MJ/kg 18.61 17.58 18.20 19.15 19.75 18.56 18.07 18.46 19.00 18.68 18.44

mea. S.D ±0.02 ±0.06 ±0.17 ±0.05 ±0.06 ±0.04 ±0.06 ±0.14 ±0.16 ±0.10 ±0.08

HHV MJ/kg 18.88 18.37 18.68 19.18 18.67 18.98

cal. S.D ±0.01 ±0.01 ±0.06 ±0.04 ±0.06 ±0.11

C %

S.D

45.455

± 0.040

45.46

±0.061

45.033

±0.003

47.142

±0.043

46.798

±0.001

46.645

±0.019

46.568

±0.004

47.121

±0.004

45.983

±0.000

46.720

±0.006

47.208

±0.002

H %

S.D

5.891

±0.070

5.950

±0.004

5.981

±0.001

6.06

±0.044

6.140

±0.002

6.068

±0.016

6.012

±0.001

6.080

±0.017

6.022

±0.001

6.081

±0.003

6.120

±0.001

N %

S.D

0.174

±0.001

0.184

±0.008

0.119

±0.001

0.040

±0.001

0.039

±0.001

0.035

±0.001

0.040

±0.001

0.033

±0.001

0.036

±0.001

0.038

±0.002

0.027

±0.001

S %

S.D

0.041

±0.001

0.041

±0.001

0.028

±0.001

0.002

±0.000

0.002

±0.000

0.016

±0.001

0.018

±0.001

0.013

±0.001

0.017

±0.001

0.017

±0.001

0.010

±0.001

O %

S.D

48.439

±0.112

48.37

±0.075

48.840

±0.020

46.760

±0.087

47.020

±0.000

47.135

±0.037

47.358

±0.006

46.750

±0.016

46.912

±0.022

47.140

±0.013

46.644

±0.010

Ash % 5.34 4.06 3.61 0.00 0.00 2.07 1.51 1.53 1.54 1.62 1.55

S.D ±0.45 ±0.37 ±0.44 ±0.00 ±0.00 ±0.21 ±0.18 ±0.50 ±0.51 ±0.05 ±0.40 Notation: RCG =Reed Canary Grass, H=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine, mea.=measured, cal.=calculated, S.D=standard deviation

63

63

Minerals Content

The variation in mineral composition of selected agricultural and woody biomass was

high as shown in Table I.5. The biomass was stored in an airtight container as mentioned

earlier for about 1 year in our laboratory under room conditions prior to mineral content

analysis. Since ash content of woody biomass was very low, the mineral content in woody

biomass was much lower than the agricultural biomass. Hence, once the agricultural

biomass was blended with woody biomass, the mineral content was observed to decrease

as compared to individual agricultural biomass.

Table I.5: Mineral Contents (mg/kg) of Agricultural and Woody Biomass and

Their Blends (50:50 wt%).

Calcium Magnesium Phosphorus Potassium Silicon Sulfur Chloride

RCG 3550 599 1110 2970 9700 750 1300

H 3010 1600 968 8010 3510 880 2710

SW 3140 1080 477 656 5320 430 300

S 430 132 44 454 < 50 60 130

P 450 131 48 439 < 50 50 2590

S+RCG 2000 371 576 1680 4610 400 600

S+H 1180 497 287 2320 840 250 660

S+SW 950 377 141 529 1200 130 140

P+RCG 2020 373 604 1700 4790 410 50

P+H 1140 501 294 2330 820 260 610

P+SW 1030 396 165 553 1220 150 760

Note: RCG - Reed Canary Grass, H=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine

Among the agricultural biomass, reed canary grass and switchgrass showed a high amount

of silicon, while potassium was higher for hay. However, after blending reed canary grass

with woody biomass, the silicon content reduced to nearly half as compared to individual

reed canary grass. The alkaline minerals such as calcium, phosphorous, chloride, and

potassium are of primary importance. These minerals might not show any effects on

pelletization, but they can cause serious effect such as slagging and fouling to the thermal

64

system during combustion of biomass fuel. Nevertheless, silicate is believed to play a less

important role in combustion of biomass. Previous studies have also reported similar

results regarding the mineral content of agricultural and woody biomass (Vassilev et al,

2013; Barnes et al, 2014). It has also been shown that the process conditions, in particular

temperature, can severely affect the transformation of the mineral matter during

combustion of pellets (Senneca et al, 2004; Padersen et al, 2009). Even though blended

biomass is rich in calcium content, it may have a positive effect during combustion of

biomass pellets. For example, it can sustain the thermal operation and thus may prevent

agglomeration (Vassilev et al, 2013). It has also been reported that elimination of these

mineral elements and chlorine can improve the fusion temperature of the biomass ash

(Singh et al, 2011). Overall, the mineral content was significantly reduced in the blend

biomass.

Blended Biomass Pellet Strength

Blended biomass pellets were tested for their axial strength. The failure was measured at

the point of its breaking force (kN), as showed in Figure I.5. In general, small particle size

gave higher pellet strength with p value less than 0.001. Similar results were found in the

previous studies (Zafari and Kianmehr, 2004, Lam et al, 2008). The reason for higher

strength with smaller particle size might be due to higher contact surface area offered by

finer particles to form strong linkages or bonding between the biomass particles.

65

Figure I.5: Effect of Particle Size (mm) and Blend Ratio on Axial Strength (kN) of

Blend Pellets.

An interesting outcome of this study was that the strength of the pellet was higher when

the amount of agricultural biomass was lower in the blend pellets. For instance, the pellet

blended with 25% agriculture biomass and 75 % woody biomass showed higher strength

compared to other blends except for 50 % switchgrass blended with 50 % pine which

66

showed slightly higher strength than 25%+75% (switchgrass+pine) blend. Pellets made

from 50 % switchgrass and 50 % pine blend showed the highest strength (3.8 kN)

compared to all other blend pellets as shown in Figure I.5.

The strengths of pellets thus evaluated were analysed to select the blends for making

pellets in a pilot scale pellet mill. Although pellets made from a blend of 75 % forestry

biomass and 25 % agricultural biomass showed higher axial compressive strength

compared to other ratios, however the difference in axial strength of 50:50 blend was less

than 0.5 kN than 75:25 blend pellet. Since one of the main motivations of this study was

to increase the amount of agricultural biomass in the pellet, the ratio of 50:50 blend was

selected.

Porosity

This study also included the analysis of porosity using BET (Brunauer, Emmett and

Teller), for individual and blended biomass as presented in Figure 6. Lower particle size

(150 and 300 µm) might exhibit lower porosity than higher particle size (425 - 600 µm).

Therefore, pellets formed with smaller particles have higher strength than pellets made

from coarse particles due to their porosity characteristics. The porosity of the materials

depends on the type as well as the particle size of the biomass. The porosity of pure

forestry biomass (spruce and pine) was lower than the pure agricultural biomass (timothy

hay and swicthgrass) as shown in Figure 2.6. This indicates that the particle density of

forestry biomass is higher than the agricultural biomass (Tumuluru et al., 2010). The

porosity of blended biomass was lower than the pure agricultural biomass for larger

particle size (425-600 µm). Further, the strength of pellet was also found to depend on the

67

type of biomass materials used. This is because the chemical and mineral content vary

according to the type of biomass. Among the feedstocks selected, timothy hay showed the

highest porosity and chlorine content as depicted in Figure I.6 and Table I.4, respectively.

These physical (porosity) and chemical (chlorine) properties might have influence on the

degradation characteristics of the timothy hay.

Notation: RCG =Reed Canary Grass, T=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine

Figure I.6: Porosity of Agricultural and Forestry Biomass Blends (50:50 wt%) with

Particle Size ranging from 150-600 µm

Durability

Durability of blended biomass pellets produced by pellet mill was evaluated to meet the

PFI standards, and the results as presented in Figure I.7. Among the blends, switchgrass

pellets did not perform well in durability test since the PDI was below the minimum PDI

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

Poro

sity

(%

)

68

of 95. This might be due to presence of inorganic minerals such as high silica content in

switchgrass (Table I.4) which might result in poor inter-particle bonding. It is also

possible that the inorganic minerals and other extractives on the surface of the agricultural

biomass might form a layer, thus preventing the inner lignin content to expose and form

good inter-particle binding.

Figure I.7: Pellets Durability Index (PDI) of Biomass and Blended Biomass Pellets

(50:50 wt. %).

Kaliyan and Morey (2006) reported 90 % durability of switchgrass pellets but with

temperature of 110 °C. However, at present it is difficult to provide an exact reason behind

such low durability and it needs further detailed investigation. On the other hand, the PDI

of timothy hay blend pellets was higher than 95 which indicate good grade pellets and

RCG blend pellets just met the standard grade pellet by reaching the minimum PDI

requirement of 95.

0

10

20

30

40

50

60

70

80

90

100

PD

I

69

The influence of blending, particle size and moisture content on the axial strength of

pellets made from RCG, pine and their blends was analyzed statistically using analysis of

variance (ANOVA). From Table I.6, the axial strength of pellets is influenced by the

blending, particle size and moisture content with p < 0.05. The average axial stress of the

pellets increased as the proportion of pine in the reed canary grass and moisture content

increased. Overall, blending might help to improve the strength of pellets. The effects of

blending on the chemical, physical and mechanical properties of pellets were evident from

this study and can become the basis for blending agricultural residues with woody

biomass.

Table I.6: ANOVA Table for the Particle Size, Blend Ratio and Moisture Content

Affecting Anisotropic Strength Ratio of Pellets

Source of variables Sum of

squares (SS)

Degree of

freedom

(DF)

Mean

square

(MS)

F-value P-value

Particle size (mm) 0.1099 2 0.0550 7.385 0.024

RCG (%) in blend 0.0447 6 0.0074 25.299 0.000

Moisture content

%)

0.0053 18 0.0003 26.41 0.000

Total 0.15987 26

Conclusion

The effect of blending agricultural biomass with woody biomass on the physical, chemical

and mechanical properties of pellets was investigated. It was found that rising in carbon

and hydrogen with lessening in oxygen content signifies the change in the chemical

composition of biomass after blending. The mechanical properties such as strength and

70

durability of the blend pellet were found to depend on the physical properties such as

porosity, particle size and the chemical composition of the feedstock. For instance, the

strength of the blend pellet was inversely proportional to the particle size. The durability

of the blend pellets was higher than individual agricultural pellets and comparable to

woody pellets. Particle size is one of the most influential parameters affecting the strength

of the pellets. For example, blended biomass pellets made from smaller particle size (150

µm) and lower amounts of agricultural biomass (25 % and 50 %) in the blend showed

higher strength. Switchgrass blended with pine exhibited the highest strength among all

blended pellets as it has the lowest mineral content as compared to other blended biomass

pellets. In conclusion, blending of agricultural biomass with woody biomass can be

considered as one of the alternative options to increase the pellet production to meet the

future demand.

Acknowledgement


Agriculture, Aquaculture and Fisheries; New Brunswick Soil and Crop Improvement


New Brunswick Innovation Foundation, and Universiti Teknologi PETRONAS,

Malaysia.

References

Barmina, I., Lickrastina, A., Valdmanis, R., Zake, M., Arshanitsa, A., Solodovnik, V., &

Telysheva, G. (2013). Effects of biomass composition variations on gasification and

71

combustion characteristics. In Proc. 12th Intl. Sci. Conf. Engineering for Rural

Development (pp. 382-387). Jelgava, Latvia: Latvia University of Agriculture.

Barnes, H. M., Sanders, M. G., & Lindsey, G. B. (2014). Compression and bending

strength of steam-treated hardwood. BioResources, 9(1), 537-544.

Bradley, D. (2010). Canada report on bioenergy 2010. Ottawa, Ontario, Canada: Climate

Change Solutions. Retrieved from

www.canbio.ca/upload/documents/canada-report-on-bioenergy-2010-sept-15-2010.pdf.

Briggs, J. L., Maier, D. E., Watkins, B. A., & Behnke, K. C. (1999). Effect of ingredients

and processing parameters on pellet quality. Poultry Sci., 78(10), 1464-1471.

http://dx.doi.org/10.1093/ps/78.10.146.

Canadian Biomass Magazine. 2012. Available in Forest Business Network:

http://www.forestbusinessnetwork.com/27490/exportation-of-wood-pellets-from-north-

america-to-europe-reached-a-new-record-of-3-2-million-tons-in-2012/

Cocchi, M., Nikolaisen, L., Junginger, M., Sheng-Goh, C., Heinimö, J., Bradley, D., Hess,

R., Jacobson, J., Ovard, L.P., Thrän, D., Hennig, C., Deutmeyer, M., Schouwenberg, P.

P., & Marchal, D. (2011). Global wood pellets markets and trade studies. IEA Bioenergy

Task 40. Retrieved from

www.bioenergytrade.org/downloads/t40-global-wood-pellet-market-study_final_R.pdf .

Dhamodaran, A., & Afzal, M. T. (2012). Compression and springback properties of

hardwood and softwood pellets. BioResources, 7(3), 4362-4376.

Dhamodaran, A., & Afzal, M. T. (2013). Modeling and characterization of reed canary

grass pellet formation phenomenon. Intl. J. Renew. Sustain. Energy, 2(2), 63-73.

http://dx.doi.org/10.11648/j.ijrse.20130202.16.

72

Harun, N. Y., & Afzal, M. T. (2010). Thermal decomposition kinetics of forest residue.

J. Appl. Sci., 10(12), 1122-1127. Available at:

http://dx.doi.org/10.3923/jas.2010.1122.1127

Helwig, T., Jannasch, R., Samson, R., DeMaio, A., & Caumartin, D. (2002). Agricultural

biomass residue inventories and conversion systems for energy production in eastern

Canada. Final report prepared for Natural Resources Canada. Ste. Anne de Bellevue,

Quebec, Canada: Resource Efficient Agricultural Production (REAP). Retrieved from

www.reap-canada.com/online_library/feedstock_biomass/7-

Agricultural%20Biomass%20Residue%20Inventories%20and%20Conversion...Samson

%20et%20al.%202002.pdf .

James, A. K., Thring, R. W., Helle, S. & Ghuman, H. S. (2012). Ash Management

Review—Applications of Biomass Bottom Ash Energies. 5, 3856-3873.

Jannasch, R., Quan, Y., & Samson, R. (2001). A process and energy analysis of pelletizing

switchgrass. Ste. Anne de Bellevue, Quebec, Canada: Resource Efficient Agricultural

Production (REAP). Retrieved from

www.reap-canada.com/online_library/feedstock_biomass/11%20A%20Process.pdf .

Kaliyan, N., & Morey, R. V. (2006). Densification characteristics of corn stover and

switchgrass. Trans. ASABE, 52(3), 907-920. http://dx.doi.org/10.13031/2013.27380.

Kok, M. V., & Özgür, E. (2013). Thermal analysis and kinetics of biomass samples. Fuel

Proc. Tech., 106, 739-743. http://dx.doi.org/10.1016/j.fuproc.2012.10.010.

Lam, P. S., Sokhansaj, S., Bi, X., Lim, C. J., Naimi, L. J., Hoque, M., Mani, S., Womac,

A. R., Narayan, S., & Ye, X. P. (2008). Bulk density of wet and dry wheat straw and dry

particles. Appl. Eng. Agric., 24(3), 351-358. http://dx.doi.org/10.13031/2013.24490.

73

Lam, P. Y., Lam, P. S., Sokhansanj, S., Bi, X. T., Lim, C. J., & Melin, S. (2014). Effects

of pelletization conditions on breaking strength and dimensional stability of Douglas fir

pellet. Fuel, 117(B), 1085-1092. http://dx.doi.org/10.1016/j.fuel.2013.10.033.

Lehtikangas, P. (2001). Quality properties of pelletized sawdust, logging residues, and

bark. Biomass Bioenergy, 20(5), 351-360. http://dx.doi.org/10.1016/S0961-

9534(00)00092-1.

Li, Y. D. & Liu, H. (2000). High‐pressure densification of wood residues to form an

upgraded fuel. Biomass & Bioenergy. 19, 177‐186.

Lim, S. (2012). Update on the pellets industry in Malaysia: Global opportunity. Putrajaya,

Malaysia: Malaysia Biomass Industries Confederation. Retrieved from http://biomass-

sp.net/wp-content/uploads/2012/11/9-Stephen-Lim.pdf .

Magdziarz, A., & Wilk, M. (2013). Thermal characteristics of the combustion process of

biomass and sewage sludge. J. Thermal Anal. Calorimetry, 114(2), 519-529.

http://dx.doi.org/10.1007/s10973-012-2933-y.

ORNL. (2011). U.S. billion ton update: Biomass supply for bioenergy and bioproducts

industry. Energy efficiency and renewable energy. Oak Ridge, Tenn.: U.S. Department of

Energy, Oak Ridge National Laboratory. Retrieved from

www.eere.energy.gov/bioenergy/pdfs/billion_ton_update.pdf.

Padersen, K. H., Jensen, A. D., Berg, M., Olsen, L. H., & Dam-Johansen, K. (2009). The

effect of combustion conditions in a full-scale low-NOx coal-fired unit on fly ash

properties for its application in concrete mixtures. Fuel Proc. Tech., 90(2), 180-185.

http://dx.doi.org/10.1016/j.fuproc.2008.08.012.

74

PFI. (2011). Pellet Fuels Institute residential/commercial densified fuel QA/QC

handbook. Arlington, Va.: Pellet Fuels Institute. Retrieved from

www.pelletheat.org/assets/docs/qa-qc-handbook-november-2011.pdf.

Samson, R., & Duxbury, P. (2000). Assessment of pelletized biofuels. Ste. Anne de

Bellevue, Quebec, Canada: Resource Efficient Agricultural Production (REAP).

Retrieved from www.reap-

canada.com/online_library/feedstock_biomass/15%20Assessment%20of.PDF

Samuelsson, R., Thyrel, M., Sjostrom, M., & Lestander, T. A. (2009). Effect of

biomaterial characteristics on pelletizing properties and biofuel pellet quality. Fuel Proc.

Tech., 90(9), 1129-1134. http://dx.doi.org/10.1016/j.fuproc.2009.05.007.

Senelwa, K., & Sims, R. E. (1999). Opportunities for small-scale biomass electricity

systems in Kenya. Biomass Bioenergy, 17(3), 239-255. http://dx.doi.org/10.1016/S0961-

9534(99)00031-8.

Senneca, O. (2011). Chapter 16: Characterization of biomass as non-conventional fuels

by thermal techniques. In Progress Biomass Bioenergy Production. Rijeka, Croatia:

InTech. Retrieved from www.intechopen.com/books/progress-in-biomass-and-

bioenergy-production/characterization-of-biomass-as-non-conventional-fuels-by-

thermal-techniques

Senneca, O., Chirone, R., & Salatino, P. (2004). Oxidative pyrolysis of solid fuels. J.

Anal. Appl. Pyrolysis, 71(2), 959-970.

Serrano, C., Monedero, E., Lapuerta, M., & Portero, H. (2011). Effect of moisture content,

particle size, and pine addition on quality parameters of barley straw pellets. Fuel Proc.

Tech., 92(3), 699-706. http://dx.doi.org/10.1016/j.fuproc.2010.11.031.

75

Singh, S., Ram, L. C., Masto, R. E., & Verma, S. K. (2011). A comparative evaluation of

minerals and trace elements in the ashes from lignite, coal refuse, and biomass-fired

power plants. Intl. J. Coal Geol., 87(2), 112-120.

http://dx.doi.org/10.1016/j.coal.2011.05.006 .

Stelte, W., Holm, J. K., Sanadi, A. R., Barsberg, S., Ahrenfeldt, J., & Henriksen, U. B.

(2011). Fuel pellets from biomass: The importance of the pelletizing pressure and its

dependency on the processing conditions. Fuel, 90(11), 3285-3290.

http://dx.doi.org/10.1016/j.fuel.2011.05.011 .

Tumuluru, J. S., Wright, C. T., Kenney, K. L., & Hess, J. R. (2010). A technical review

of biomass processing: Densification, preprocessing, modeling, and optimization.

ASABE Paper No. 1009401. St. Joseph, Mich.: ASABE.

Tumuluru, J. S., Wright, C. T., Hess, J. R., & Kenney, K. L. (2011). A review of biomass

densification systems to develop uniform feedstock commodities for bioenergy

application. Biofuels Bioprod. Biorefining, 5(6), 683-707.

http://dx.doi.org/10.1002/bbb.324 .

Van Tilburg, M. (2013). More pellets, please! Site Selection Magazine July 2013.

Available at http://www.siteselection.com/issues/2013/jul/world-reports.cfm

Vamvuka, D., & Sfakiotakis, S. (2011). Combustion behaviour of biomass fuels and their

blends with lignite. Thermochimica Acta, 526(1-2), 192-199.

http://dx.doi.org/10.1016/j.tca.2011.09.021 .

Vassilev, S. V., Baxter, D., & Vassileva, C. G. (2013). An overview of the behaviour of

biomass during combustion: Part I. Phase-mineral transformations of organic and

inorganic matter. Fuel, 112, 391-449. http://dx.doi.org/10.1016/j.fuel.2013.05.043 .

76

Wamsley, K. G. S., Gehring, C. K., Corzo, A., Fontana, E. A., & Moritz, J. S. (2012).

Effects of inorganic feed phosphate on feed quality and manufacturing efficiency. J. Appl.

Poultry Res., 21(4), 823-829. http://dx.doi.org/10.3382/japr.2012-00542 .

Wilson, L., Yang, W., Blasiak, W., John, G. R., & Mhilu, C. F. (2011). Thermal

characterization of tropical biomass feedstocks. Energy Conv. Mgmt., 52(1), 191-198.

http://dx.doi.org/10.1016/j.enconman.2010.06.058 .

Wood, S. M., & Layzell., D. B. (2003). A Canadian biomass inventory: Feedstocks for a

bio-based economy. Final report prepared for Industry Canada. Kingston, Ontario,

Canada: BIOCAP Canada Foundation. Retrieved from

www.agrienvarchive.ca/bioenergy/download/BIOCAP_Biomass_Inventory.pdf .

WRI. (2013). Exportation of wood pellets from North America to Europe reached a new

record of 3.2 million tons in 2012. Tacoma, Wash.: Wood Resources International.

Retrieved from www.woodworkingnetwork.com/wri .

Wrobel, C., Coulman, B. E., & Smith, D. L. (2009). The potential use of reed canarygrass

(Phalaris arundinacea L.) as a biofuel crop. Acta Agric. Scandinavica B, 59(1), 1-18.

http://dx.doi.org/10.1080/09064710801920230.

Zafari, A., & Kianmehr, M. H. (2014). Factors affecting mechanical properties of biomass

pellet from compost. Environmental Technology.

35 (4). DOI: 10.1080/09593330.2013.833639

Zafari, A., & Kianmehr, M. H. (2012a). Effect of raw materials properties and die

geometry on the density of biomass pellets from composted municipal solid waste.

BioResources, 7(4), 4704-4714. http://dx.doi.org/10.15376/biores.7.4.4704-4714 .

77

Zafari, A., & Kianmehr, M. H. (2012b). Effect of temperature, pressure, and moisture

content on durability of cattle manure pellet in open-end die method. J. Agric. Sci., 4(5),

203-208. http://dx.doi.org/10.5539/jas.v4n5p203 .

78

Paper II

Effect of particle size on mechanical properties of pellets made from

biomass blends




Perak, Malaysia.



Procedia Engineering, 148, 93 – 99

Abstract

Woody biomass is densified more oftenly in the form of pellets in order to improve its

physical and mechanical properties for better handling and storage. However, limited

research work has been conducted on the mechanical properties of pellets made from

agricultural and woody biomass blends. Two commonly available forestry biomasses,

spruce (S) and pine (P), and three agricultural biomasses, reed canary grass (RCG),

timothy hay (H) and swicthgrass (SW), were used to form pellets. An Instron attached

with an in-house built single unit pelletizer and temperature controlled die was employed

to produce a pellet. The aim of this study was to investigate the effect of particle size

(three different particles sizes; 150 – 300, 300 – 425 and 425 – 600 µm) and blending

(agricultural and woody biomass) on the mechanical properties of pellets (density and

intrinsic yield stress). For all biomasses, pellets made from lower particle size (150 – 300

µm) exhibited higher density (950 – 1178 kg/m3 for spruce and pine; 668 – 800 kg/m3 for

RCG, H and SW; 900 – 970 kg/m3 for blended biomass). The intrinsic yield stress

exhibited differences in values for individual forestry (40 MPa) and agricultural biomass

(27-48 MPa), however after blending the values converged closest to that value for

forestry biomass. In conclusion, blending low cost and abundant available agricultural

biomass with woody biomass would not only result in better mechanical properties but

would also help to expand the biomass feedstock resource for pellet fuel industry to meet

the growing demand in future.

Keywords: agricultural biomass; biomass pellets; axial stress; blended biomass; pellet’s

density

79

Introduction

Biomass densification help improve the physical and combustion properties of solid fuel

[1]. It has drawn attention due to its advantages in achieving uniform physical properties

over a wide range of raw biomasses which includes biomass from agricultural and forestry

resources [2]. Production of pellets and its demand has been rapidly increased in Canada,

America, Europe, and China in the past few years. It is estimated that the demand for

pellets will be tripled from 2012 to 2020, rising from 16 million to 46 million metric tons

per year [3]. As one of the main pellets producer, the Canadian pellet industry is looking

for alternative sources of feedstock. Agricultural residues can be one of the potential

alternative feedstock since it is abundantly available of about 25 Mtoe (Megatonne of oil

equivalent), annually [4, 5] and at low cost. Based on improvement in physical and

mechanical properties, biomass densification has been presented to significantly reduce

costs of transportation and storage losses [6].

Several studies have been reported on pelleting, focusing on material and process

parameters and mechanics of bonding formation [7, 8, 9, 13 and 15]. Both temperature

and applied pressure were found to be the influential parameters in determining the quality

of pellets. Material characteristics of different biomasses have also been explored for its

relationship with particle bonding in pellets made from beech, fir and straw [10]. They

analyzed the fracture surface of the pellet and it was found that pellets made from beech

and fir had a greater mechanical resistance than pellets made from straw. It is also

concluded that pellets made at temperature 100 °C had greater influence on mechanical

stability than pellets made at temperature 20 °C. The published work related to biomass

pelleting, has also been focused on economic analysis [6, 11, and 12]. Very limited

80

research work has been reported in the open literature about blending agricultural biomass

with woody biomass to produce pellets. It is hypothesized that blending agricultural

biomass with woody biomass might give direction in the development of new feedstock

and help expand biomass resource availability. The purpose of this study was to

investigate the effect of particle size on density and yield stress of pellets made from

selected individual agricultural and woody biomasses and their blends.

Methodology

Sample Preparation

In this study, the selected agricultural biomass such as reed canary grass, timothy hay and

switchgrass was used which was provided by New Brunswick Department of Agriculture.

The woody biomass (spruce and pine) was obtained from a local saw mill situated in New

Brunswick. The raw biomass materials placed in plastic bags and in closed container were

stored at room conditions. Prior to analysis, the samples were dried in the oven at 105 oC

for 24 h, and then ground using Wiley mill to less than 1 mm particle size. The ground

biomass were sieved according to ASTM standards (D 2013–72) to collect the particle

size in range of 150 – 300, 300 – 425 and 425 - 600 µm. The grinded biomass and their

blends were stored in airtight container prior to pelletizing and for further analysis.

Heating value, proximate, and ultimate analysis was conducted for individual and blended

biomass and is presented in Table II.1.

81

Table II.1: Agricultural and Woody Fraction

Agricultural Forestry Blends

RCG H SW S P S+RCG P+RCG S+H P+H S+SW P+SW

Prop

(wt. %) 100 100 100 100 100 50:50 50:50 50:50 50:50 50:50 50:50

Notes: RCG=Reed Canary grass, H=Timothy hay, SW=Switchgrass, S=Spruce and P=Pine


The ultimate analysis was performed using LECO CHNS 932 elemental analyzer as per

the ASTM D5291 method and was replicated twice. Element oxygen was determined by

difference. Proximate analysis was conducted twice using TG analyzer (Q500 TA

Instrument Inc.) as per ASTM E1641-04 method, while the ash content was investigated

as per NREL/TP-510-42622 method with 6 replications. The higher heating value (HHV)

of biomass was measured using bomb calorimeter (Isoperibol Oxygen Bomb Calorimeter)

according to the standard method, ASTM D-5865.

Pelletization Process

Prior to pelletization the ground biomass that was stored in an airtight container was

maintained under room conditions for about a week. The pellets were formed with the

adjusted moisture content of about 10 %. Pellets were made with the help of a closed end

die which was built in house. The closed end die with 8 mm diameter and 202 mm in

length as shown in Figure II.1 was attached to Instron SATEC machine which had a

control on the load and holding time. The temperature was controlled by the help of K-

type thermocouples connected to a PID controller.

82

Parameters such as load (159 MPa), temperature (80 °C), speed of the piston (0.43 mm/s)

and the hold time (5 s) for each run were maintained constant. The load and displacement

data collection frequency was 8 points/second.

Figure II.1: Schematic of Single Pelletization Set up [13 and 14].

The selection of pressure and temperature for this work were based on the previous

detailed studies [13, 15]. They found highest durability of pellets in the temperature range

between 80 – 85 °C. A total number of 330 (11 types of biomass × 3 particle sizes × 3

replications) experimental tests were performed based on number of biomass samples (5

individual and 6 blends), three different particle sizes and three replications.

Approximately, 1.2 g of biomass sample was fed into die in each run to produce a single

pellet. The energy required for densifying each biomass and blend biomass pellet was

Pin connection to instron machine crosshead

Piston

Cylinder (die) with 8 mm internal diameter

Heating elements connected to PID controller

Stopping blocks which can be removed to take pellets

Pin connection to base of instron machine

210 mm

202 mm

83

calculated using the plots of the force and displacement data recorded by the associated

software of the Instron machine.

Density of Pellets

Pellet size was measured with a digital caliper (0.01 mm resolution, model CD-56C,

Mitutoyo Corp., Aurora, Ill.). Reported values are the average of 10 measurements for

each biomass sample. True density of the pellets was calculated based on the ratio of mass

to the volume of single pellet.

Mechanical Strength

The mechanical strength of individual biomass and their blend pellets was measured using

Instron SATEC mMachine. Each pellet was tested for its compressive break force in an

axial orientation. The strength was tested for each condition and was repeated thrice to

present a mean value with the standard deviation.


Elemental Analysis

Carbon, hydrogen and oxygen are the main biomass elements which directly represents

the components of the organic part of fuel. Carbon is obviously representing foremost

contributions to overall heating value. Generally, agricultural biomass has slightly lower

carbon content (45 %) than woody (47 %) and blended biomass (46-47 %) as presented

in Table II.2. Harun and Afzal [14] have reported the similar results of elemental analysis

in previous study. The biomasses as well as their blends chosen in this study have its

84

benefits for combustion since almost half of constituents were carbon. It mainly

transforms to CO2, released to atmosphere during combustion [16].

The carbon contents are also directly related to lignin, cellulose and hemicellulose

contents in biomass. Lignin contributes to the mechanical properties of biomass pellets.

Since agricultural biomass is difficult to pelletize, blending it with forestry biomass

improves the densification process due to more lignin content. Fuel rich in lignin has the

total carbon and hydrogen content of higher than 50 % [17].

Table II.2 : Elemental Analysis (Dry Basis) of Individual and Biomass Blends

C H N S O (balance)

Individual Biomass (wt % ± S.D) (wt % ± S.D) (wt % ± S.D) (wt % ± S.D) (wt % ± S.D)

Reed canary grass

(RCG)

45.455±0.040 5.891±0.070 0.174±0.001 0.041±0.001 48.439±0.112

Timothy hay (H) 45.461±0.061 5.950±0.004 0.184±0.008 0.041±0.001 48.372±0.075

Switchgrass (SW) 45.033±0.003 5.981±0.001 0.119±0.001 0.028±0.001 48.840±0.020

Spruce (S) 47.142±0.043 6.060±0.044 0.040±0.001 0.002±0.000 46.760±0.087

Pine (P) 46.798±0.001 6.140±0.002 0.039±0.001 0.002±0.000 47.020±0.000

Blended biomass

(50 wt %:50 wt %)

RCG+S 46.645±0.019 6.068±0.016 0.035±0.001 0.016±0.001 47.135±0.037

RCG+P 45.983±0.000 6.022±0.001 0.036±0.001 0.017±0.001 46.912±0.022

H+S 46.568±0.004 6.012±0.001 0.040±0.001 0.018±0.001 47.358±0.006

H+P 46.720±0.006 6.081±0.003 0.038±0.002 0.017±0.001 47.140±0.013

SW+S 47.121±0.004 6.080±0.017 0.033±0.001 0.013±0.001 46.750±0.016

SW+P 47.208±0.002 6.120±0.001 0.027±0.001 0.010±0.001 46.644±0.010

Notes: C=carbon, H=hydrogen, N=nitrogen, S=sulfur, O=oxygen

85

Heating Value and Ash Analysis

Individual biomass and their blends were used for combustion testing. Calorific value and

ash content of the biomass fuel are provided in Fig 3.2 and Fig 3.3, respectively. Analyses

were performed according to ASTM standards. Forestry biomass is considered has higher

lignin according to its carbon and hydrogen percentage, calculated on a dry, ash-free basis.

The calorific values of spruce and pine were higher than the calorific values of reed canary

grass, timothy hay and switchgrass, as presented in Fig II.2. Blended fuel has shown that

the calorific values found were mostly at the average values between their parent biomass.

This has confirmed that blending agricultural biomass with woody biomass can improve

the heating value of agricultural biomass fuel.

Notes: RCG=reed canary grass, H=timothy hay, SW=switchgrass, S=spruce, P=pine

Figure II.2 : Calorific Values of Agricultural and Forestry Biomass, and Their

Blends

Combustion of spruce and pine can be of less difficulty since its ash content was only

found in traces. The average ash content of agricultural biomass, reed canary grass,

16

17

17

18

18

19

19

20

20

Hea

ting v

alue

(MJ/

kg)

86

timothy hay and switchgrass was obviously higher from 2 to 4 %, as presented in Fig II.3.

High ash in agricultural biomass was due to its high sulfur content as shown in Table 3.2.

Blended samples were prepared by mixing agricultural and woody biomass in the

proportion of 50:50 (agricultural wt%: woody wt%). The ash content was observed to be

less than 2 % in blended biomass, which is highly recommended and above standard

requirement according to national and international standard such as PFI standard (Pellet

Fuel Institute).

Notes: RCG=reed canary grass, H=timothy hay, SW=switchgrass, S=spruce, P=pine

Figure II.3 : Ash Content of Agricultural and Forestry Biomass, and Their Blends

Physical and Mechanical Characteristics

Effect of Particle Size on Pellet Density

The average diameter of these pellets was about 8.0 mm, and the average length ranged

from 23 to 29 mm. Usually, the bulk density of the biomasses differ according to the

0

1

2

3

4

5

Ash

per

centa

ge

87

particle size and moisture content [18 and 19]. However, the true density of biomass

pellets is significantly higher than the original bulk density of biomass. For example,

Colley et al. [20] reported that bulk density of ground switchgrass was 165.5 kg/m3, but

the density of pellets at 5% to 20% moisture content (wet basis) ranged from 536 to 708

kg/m3. In the present study, the density of pellets made from spruce and pine ranged from

943 to 1178 kg/m3. However, individual agricultural biomass showed lower pellet density

ranging from 584 to 799 kg/m3, as shown in Figure II.4, with mositure content 10%.

Figure II.4 : Pellet Density of Individual Agricultural and Woody Biomass at

Different Particle Sizes (150, 300 and 425 µm), Moisture Content 10%

(w.b).

The density of blended biomass pellets increased significantly (627 to 969 kg/m3) than

the individual agricultural biomass pellets (Figure II.5). Woody biomass particles might

have played an important role in filling the void between the particles to increase the inter-

particle bonding.

500

600

700

800

900

1000

1100

1200

100 150 200 250 300 350 400 450

Pel

let

den

sity

(kg/m

3)

Particle size (µm)

Spruce

Pine

RCG

Timothy hay

Switchgrass

88

Notation: RCG =Reed Canary grass, H=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine

Figure II.5 : Density of Biomass Pellets made from Blended Biomass for Particle Size

of 150, 300 and 425 µm

The results from this study agreed well with the previous work as reported by Arzola et

al. [21]. They found that the density of oil palm shell pellets decreased as the average

particle size was increased from 160 to 570 µm. In this study, spruce pellets showed the

highest density. Using the same amount of biomass (1.2 grams), the length of pellets made

from agricultural biomass was longer than the wood pellets. This might be due to the void

in agricultural biomass pellets which might have not easily filled while densifying.

Moreover, the inter-particle bonding in agricultural biomass is weaker [22 and 23] than

the inter-particle bonding in woody biomass. The lignin content, determined by means of

hydrolysis as per NREL 2008b standard, in spruce, pine, reed canary grass, timothy hay

and swicthgrass biomass was 23, 24, 17, 11 and 9 %, respectively. The density of biomass

pellet varied according to the lignin content of biomass and it was low for low lignin

500

600

700

800

900

1000

1100

1200

100 150 200 250 300 350 400 450

Pel

let

den

sity

(kg/m

3)

Particle size (µm)

S + RCG

P + RCG

S+H

P+H

S+SW

P+SW

89

content biomass. Therefore, there could be a strong relation between the density and the

lignin content of the biomass.

Effect of Particle Size on Pellet Yield Stress

Beside several other material and process parameters, particle size plays a major role in

pellet quality and density. It significantly contributes to the mechanical strength of

biomass pellets. During compaction, first layer of particles interacts with second layer of

particles by pushing each other and filling the gaps. As other layers of particles keep

moving and piled-up, particles resists each other and hence there is an increased inter-

particle stress. A pellet made from smaller particle size gains higher strength due to

decrease in gaps that have been filled more by the smaller particles more than the pellet

made from larger particle size. A change in particle size affects the yield strength of pellet

due to the level of applied stress to cause the particle boundary to failure. Theories that

have been proposed for the particle size dependence include the dislocation density model

[24], and Hall-Petch equation, which has been widely applied for metals [25],

nanocrystalline materials [26] and also biowaste eggshells [27]. It is difficult to develop

an accurate model for the yield strength of the pellets, however with the range of particle

size it is found that the axial yield stress, σy, is related to particle diameter, d, of the

selected biomass and their blends and it behaved according to the Hall-Petch equation:

where σi is the inherent compressive stress of biomass and k is a constant for a particular

material. This relation is valid for particle size hardening at low temperature (< melting

90

temperature, Tm). The effect of particle size on the axial stress of individual and blend

biomass pellets is presented in Figure II.6 and Figure II.7, respectively.

Notes: d=diameter, m=meter

Figure II.6 : Yield Stress Relating Particle Size of Individual Biomass Pellet

Notes: d=diameter, m=meter

Figure II.7 : Yield Stress Relating Particle Size of Blended Biomass Pellet

30

35

40

45

50

55

60

65

70

75

80

40 50 60 70 80 90

Ax

ial

stre

ss,

σy (

MP

a)

RCG

Timothy hay

Swicthgrass

Spruce

Pine

1

𝑑(𝑚−

1

2)

35

40

45

50

55

60

65

70

75

80

85

40 50 60 70 80 90

Ax

ial

stre

ss,

σy (

MP

a)

RCG+P

RCG+S

H+P

H+S

SW+P

SW+S

1

𝑑(𝑚−

1

2)

91

The particle size relationships with axial stress of pellet made from spruce, pine, reed

canary grass, timothy hay and switchgrass well fitted the Hall-petch equation.

Apparently, its intercept at σi, which is the inherent yield stress of spruce (40.5 MPa), pine

(41 MPa) and switchgrass (27.4 MPa) was found similar to the value of compressive yield

stress found in literature (spruce and pine were ranged 40-46, and switchgrass was 27.7

MPa) [28 and 29]. However, σi for RCG and timothy hay was difficult to find from

literature. Since σi for spruce, pine and swicthgrass fitted well with the value from

literature, it is therefore assumed that σi for RCG and timothy hay might also work well.

The model relating the particle size and the yield strength of the selected biomass and

their blend is presented in Table II.3.

Table II.3: Axial Stress and Particle Size Relationship using Hall-petch Equation, σy

= k/√D + σi, for Biomass Pellet, and Their Mean of Means Errors

Biomass Axial stress R2

Pine σy = 𝟔.𝟕𝟔𝟕𝟓

√𝒅 + 41.154 0.9084

Spruce σy = 𝟔.𝟔𝟗𝟐𝟓

√𝒅 + 40.466 0.9135

RCG σy = 𝟏.𝟕𝟓𝟓𝟕

√𝒅 + 47.834 0.9215

Timothy hay σy = 𝟑.𝟒𝟗

√𝒅 + 46.3 0.8755

Switchgrass σy = 𝟓.𝟔𝟕𝟕𝟓

√𝒅 + 27.398 0.8591

RCG+P σy = 𝟑.𝟗𝟑

√𝒅 + 44.51 0.9995

RCG+S σy = 𝟕.𝟑𝟔

√𝒅 + 38.863 0.9604

H+P σy = 𝟒.𝟕𝟕𝟓

√𝒅 + 40.648 0.9829

H+S σy = 𝟖.𝟖𝟓𝟐𝟓

√𝒅 + 36.639 0.9989

SW+P σy = 𝟔.𝟔𝟓𝟐𝟓

√𝒅 + 40.836 0.9558

SW+S σy = 𝟑.𝟒𝟒𝟕𝟓

√𝒅 + 44.191 0.8754

92

The pellets produced from smaller particle size typically showed much higher strength as

compared to larger particles. The yield stress of all blended biomass pellets was found to

be approximately 42-47 MPa at larger particle size (~400 µm) as presented in Figure II.7.

Generally, based on this relationship, blended biomass selected in this study has achieved

almost similar mechanical strength as that for woody biomass.

Conclusion

In this study the effect of particle size and blending of agricultural biomass with woody

biomass on the physical and mechanical properties of the pellets was investigated.

Biomass pellets made with smaller particle size showed higher yield stress and density.

The biomass pellets made from individual or only agricultural biomass are usually low in

density (584-799 kg/m3) depending on particle size. However, a significant improvement

in density (627-969 kg/m3) was achieved after blending them with woody biomass at ratio

50:50 (wt basis) and eventually met the pellet quality standard. Nevertheless, the

mechanical strength of the biomass pellet was also found to depend on the physical

characteristics of biomass such as particle size. The yield stress is in agreement with the

density that pellets made from smaller particles size are stronger than pellets made from

larger particle size. First and foremost, the mechanical strength of blended biomass pellets

was found comparable to that of woody biomass pellets. Blending agricultural biomass

with woody biomass to produce pellets can be one of the potential options for the pellet

industry. Since agricultural biomass are not only cheap and abundantly available, but

93

blending them with existing woody feedstock resources showed better improvement in

their mechanical and physical properties that also achieved pellet fuel quality standards.

Acknowledgements


Agriculture, Aquaculture and Fishries; New Brunswick Soil and Crop Improvement


New Brunswick Innovation Foundation; and Universiti Teknologi PETRONAS,

Malaysia.

References

[1] Obernberger, I.. and G. Thek. 2004. Physical characteristic and chemical

composition of densified biomass fuels with regard to their combustion behavior.

Biomass and Bioenergy, 27: 653–669.

[2] Bhattacharya, S. C. 2002. Biomass energy use and densification in developing

countries. Paper presented at the ‘‘Pellets 2002: The First World Conference on

Pellets’’, Stockholm, Sweden, 2–4 September.

[3] Taylor, R. E., Butzelaar, P., Leeuwen, G. V., Palmer, A., Keyes, J., Gimenez, C.

and B. MacDonald. 2013 Wood Pellet Market Outlook. In Wood Market

International Monthly Report. 18(1). Available at www. woodmarkets.com

[4] Gogolek, P. and F. Preto. 2000. Status and potential of energy from biomass in

Canada. CANMET Energy Technology Centre, Nepean, Ontario, K1A-1M1

94

[5] Helwig, T, R. Jannasch, R. Samson, A. DeMaio, and D. Caumartin. 2002.

Agricultural biomass residue inventories and conversion systems for energy

production in Eastern Canada Final Report Prepared for Natural Resources-Canada.

[6] Mani, S., L. Tabil, and S. Sokhansanj. 2006. Effects of compressive force, Particle

size, and moisture content on mechanical properties of biomass pellets from grasses.

Biomass and Bioenergy, 30: 648–654.

[7] Tumuluru, J. S., S. Sokhansanj, L. Tabil, M. Kashaninejad, and S. Bandyopadhyay.

2010. Effect of extrusion process conditions on the quality attributes of multi-

component biomass pellets. Food and Bioprocess Technology.

[8] Nielsen, N. P. K., D. J. Gardner, T. Poulsen, and C. Felby. 2009. Importance of


residues into fuel pellets. Wood and Fiber Science. 41: 414‐425.

[9] Holm, J. K., U. B. Henriksen, K. Wand, J. E. Hustad, and D. Posselt. 2007.

Experimental verification of novel pellet model using a single pelleter unit. Energy

& Fuels, 21: 2446‐2449.

[10] Wolfgang, S., K. H. Jens, R. S. Anand, B. Soren, A. Jesper, and B. H. Ulrik. 2011.

A study of bonding and failure mechanisms in fuel pellets from different biomass

resources. Biomass and Bioenergy 5; 910-918.

[11] Sokhansanj, S. and A. F. Turhollow. 2004. Biomass densification ‐ Cubing

operations and costs for corn stover. Applied Engineering in Agriculture, 20: 495‐

499.

[12] Mani, S., S. Sokhansanj, X. Bi, and A. Turhollow. 2006. Economics of producing

fuel pellets from biomass. Applied Engineering Agricultural, 22: 421-426.

95

[13] Dhamodaran, A., and M. T. Afzal. 2012. Compression and Springback Properties

of Hardwood and Softwwod Pellets. BioResources. 7(3): 4362-4376.

[14] Harun, N. Y. and M. T. Afzal. 2015. Chemical and Mechanical Properties of Pellets

Made from Agricultural and Woody Biomass Blends. Transactions of the ASABE,

Vol. 58(4): American Society of Agricultural and Biological Engineers ISSN 2151-

0032 DOI 10.13031/trans.58.11027

[15] Kaliyan, N. and R. V. Morey. Densification characteristics of corn stover and

switchgrass. Transactions of the ASABE. 2009, 52, 907‐920

[16] Senelwa, K. and R. E. H. Sims. 1999. Fuel characteristics of short rotation forest

biomass. Biomass Bioenerg, 17: 127–140.

[17] Senneca, O., R. Chirone, P. Salatino, and L. Nappi. 2007. Patterns and kinetics of

pyrolysis of tobacco under inert and oxidative conditions. Journal of Analytical

Applied Pyrolysis, 79: 227–233.

[18] Lam, P. S., S. Sokhansaj, X. Bi, C. Lim, L. J. Naimi, M. Hoque, S. Mani, A. R.

Womac, X. P. Ye, and S. Narayan. 2008. Bulk density of wet and dry wheat straw

and dry particles. Applied Engineering Agricultural. 24: 351-358.

[19] Mani, S., L. Tabil, and S. Sokhansanj. 2006. Effects of Compressive Force, Particle

Size, and Moisture Content on Mechanical Properties of Biomass Pellets from

Grasses. Biomass and Bioenergy, 30: 648–54.

[20] Colley, Z., O. O. Fasina, D. Bransby, and Y. Y. Lee. 2006. Moisture effect on the

physical characteristics of switchgrass pellets. Transaction of ASABE, 49(6): 1845‐

1851

96

[21] Arzola, N., A. Gómez, and S. Rincón. 2012. The effects of moisture content, particle

size and binding agent content on oil palm shell pellet quality parameters. Ingeniería

E Investigación, 32(1): 24-29.

[22] Stelte, W., J. K. Holm, A. R. Sanadi, S. Barsberg, J. Ahrenfeldt, and U. B.

Henriksen. 2011. A study of bonding and failure mechanisms in fuel pellets from

different biomass resources. Biomass Bioenergy, 35(2): 910-918.

[23] Stelte, W., C. Clemons, J. K. Holm, R. A. Sanadi, R. A., Shang, L., Ahrenfeldt, J.,

and U. B. Henriksen. 2012. Fuel pellets from wheat straw: The effect of lignin glass

transition and surface waxes on pelletizing properties. Bioenergy Resources, 5(2):

450-458.

[24] Chia, K. H., K. Jung, and H. Conrad. 2005. Dislocation density model for the effect

of grain size on the flow stress of a Ti–15.2 at. % Mo -alloy at 4.2–650 K. Materials

Science and Engineering A, 409: 32–38.

[25] Hansen, N. 2004. Hall–Petch relation and boundary strengthening. Scripta

Materialia, 51: 801–806.

[26] Meyers, M. A., A. Mishra, and D. J. Benson. 2006. Mechanical properties of

nanocrystalline materials. Progress in Materials Science, 51: 427–556.

[27] Kamalanathana , P., S. Rameshan , L.T. Banga , A. Niakana , C.Y. Tana , J.

Purbolaksonoa , Hari Chandranb , and W.D. Teng. 2014. Synthesis and sintering of

hydroxyapatite derived from eggshells as a calcium precursor. Ceramics

International 40: 16349–16359.

[28] Green, D. W., J. E. Winandy, and D. E. Kretschmann. 1999. Mechanical

Properties of Wood in Wood handbook—Wood as an engineering material.

97

General Technical Report FPL–GTR–113. Madison, WI: U.S. Department of

Agriculture, Forest Service, Forest Products Laboratory. 463 p

[29] Yu, M. 2004. Ultimate Strength Characteristics of Switchgrass Stem Cross-Sections

at Representative Processing Conditions, Master Thesis, University of Tennessee,

Knoxville

98

Paper III

Process and Energy Analysis of Pelleting Agricultural and Woody

Biomass Blends




Perak, Malaysia.



Ready to be submitted

Abstract

Two important considerations with biomass energy system are that unprocessed biomass

has low energy density and high cost of transportation. The energy generated from the

biomass can be enhanced by densifying it into pellet form. However, limited research

work has been conducted on the energy requirement for the compaction of biomass, in

particular the agricultural biomass. Three agricultural biomasses (reed canary grass

(RCG), timothy hay (H) and switchgrass (SW)) and two woody biomasses spruce and

pine were selected for energy evaluation. The analysis was also conducted considering

three different particle sizes (150-300, 300-425, and 425-600 µm). An Instron machine

attached with an in-house built pelletizer unit was employed to produce a single pellet.

The aim of this study was to investigate the work of compaction required to pelletize the

biomass. The energy used for compacting ground biomass (spruce) with particle size 150

µm was found lower (2.07 kJ) as compared to the energy used for the biomass with

particle size 300 µm (2.24 kJ) and 425 µm (2.43 kJ). The energy required in compacting

ground reed canary grass, timothy hay and switchgrass was found lower (1.61, 1.97, and

1.68 kJ) than spruce (2.36 kJ) and pine (2.35 kJ) evaluated at 159 MPa load and at

temperature about 80 °C. After blending, the values were around 2 kJ with the pellet

quality approaching almost similar to pellet made from woody biomass. In conclusion,

blending helped to improve the quality of pellets lower the compaction energy

requirements.

Keywords: Biomass pellet; Energy of compaction; Agricultural biomass, Biomass blends

99

Introduction

Demand for sustainable alternative energy is growing caused by energy security and

global climate change [1]. The source of renewable energy from biomass feedstock

resources will be significant in the future energy mix [2, 3]. Traditionally, solid fuel

industry has been often dependent on forestry biomass. However, demand for pellet fuel

is increasing and biomass from forestry alone is insufficient. Biomass from agricultural

sector can support to fill the demand, but the pellet quality is low. The supply and

preprocessing of biomass is critical for an efficient, profitable and sustainable bioenergy

production [2, 3, 4]. Low energy density, especially of agricultural biomass [5] makes the

cost of feedstock logistic high. This cost is about 40-60 % of the total costs of biomass

energy production, and transportation cost alone represents 13 to 28 % [6].

Thermodynamic analysis of pellet formation can be utilized to accurately assess the work

of pelleting process for various types of biomasses. Design and development of a

mechanical compaction and thermal energy requirement for powdery and pharmaceutical

products have been widely studied in contrast to biomass compaction energy requirements

[7]. The sub-processes of compaction, unloading, and ejection are accounted in total

energy consumption. Compaction starts as the ground biomass flow under pressure.

Additional work may require due to additional compression during volume reduction of

the material, for instance brittle materials [8 and 9] may require more work than other

types of material. This event also produces significant resistance at the wall of the die and

shear stress at the edges of pellet, which can result in additional consolidation. Unloading,

decompression may increase the volume of the pellet as a result of elastic recovery. Then,

the ejection form of pellet is the net result of bond formation of the pelleting. This ejection

100

event imparts a significant stress on the pellet from unequal distribution of force

throughout the pellet. Perhaps, the work and total energy consumption for all the events

of compaction can be evaluated using compaction simulator that has hydraulic actuators

to control a uniform force during pelleting of biomass.

Biomass has a wide range in their physical, chemical and mechanical properties [10] that

has the effects on the thermodynamic and calorimetry analysis of biomasses to be pelleted.

The energy required to pelletize biomass was assessed using a single pelletizer unit that

attached to an Instron compaction machine. The objective of this study was to compare

the mechanical work and total energy requirement for pelleting individual woody and

agricultural biomasses, and their blends.

Materials and Methods

Raw Materials

The reed canary grass, switchgrass, and timothy hay used in this study was obtained from

farms near Fredericton, New Brunswick, Canada. The spruce and pine woodchips were

obtained from a sawmill in Fredericton area.

Chemical and Compositional Characterization

In order to understand the suitability of biomass for compaction, it is essential to know

the physical and chemical properties of biomass, which influence its behavior as a fuel.

These properties determine the density, stability and durability of the pellet. Physical

properties of interest include moisture content, particle size, bulk density, porosity and

thermal properties. Chemical properties of importance include the proximate and ultimate

101

analysis, and higher heating value. The physical properties are most important in the

explanation of the binding mechanisms of biomass compaction apart from the chemical

composition of the biomass [16].

Each sample was dried in an oven at 105 °C for 24 h and then ground in a Wiley mill

(Thomas Scientific, Swedesboro, NJ) to pass a 0.300 mm to 0.425 mm screen according

to ASTM Standard Practice E 1757-01 [11]. The heating value was measured in the Parr

6200 Isoperibol Oxygen Bomb Calorimeter (Parr Instrument Company, Moline, IL)

according to ASTM standard method D5865, and the ultimate analysis was determined

using an elemental analyzer (CHNS 932, LECO Corp., St Joseph, Mich) per ASTM

standard method D5291. The proximate analysis was determined using a

thermogravimetric analyzer (TGA Q500, TA Instrument, Inc., New Castle, Del.) per

ASTM method E1641-04, while ash content was analyzed in a muffle furnace according

to NREL/TP510-42622 [12].

Process of Compaction for Biomass Pellet

Ground biomasses for pelleting were chosen to encompass compaction behavior and

simulate all events of sub-processes of compaction, unloading and ejection. The

magnitude of work was evaluated during the pelleting of biomass using spruce, pine, reed

canary grass, switchgrass, timothy hay and their blends. The closed-end die pelletizer

attached to an Instron compaction machine with control software was used for compaction

simulation as described previously [11, 13].

102

Surface Fracture Analysis

Scanning Electron Microscopy (SEM) images of fracture surface for each pellet was

investigated to see the bonding characteristics between particles that were created through

solid bridges. The solid bridges were developed by the diffusion of molecules,

crystallization of components, and the chemical reaction or solidification of melted

components during the compaction (pelleting) process. The fracture surface analysis is

important to evaluate the compaction behavior especially for new pellets made from

blended biomass. The equipment used was JEOL JSM6400 Digital SEM equipped

with Geller dPict digital image acquisition software, Emitech K1250 Cryo-SEM

system, EDAX (Genesis) Energy Dispersive X-ray Analyzer, Gatan Microtest 5000

dynamic testing stage and Gatan ChromaCL Cathodoluminescence imaging system.


Chemical Analysis

The chemical property of fuel is the most important as it influences the method of

obtaining the thermal energy and compaction design. During compaction, the heat

resulting from friction softens the lignin which binds the ground biomass. The lignin

content of the used biomass should be as high. Thus, the carbon and hydrogen content

should be as high and oxygen content should be as low. The transformation of biomass

into pellets is to increase the energy density. For this reason the relevant chemical analysis

includes proximate and ultimate analyses which are presented in Table III.1. It seems that

the chemical content of woody biomass plays a more dominant role as compared to

agricultural biomass. The carbon and oxygen content of individual and blend biomass was

103

found in the range of 45.0 to 47.2 % and 46.6 to 48.8 wt%, respectively (Table III.1).

Moreover, the heating value of RCG, timothy hay and switchgrass was lower as compared

to woody biomass due to higher carbon and hydrogen content in woody biomass. Lignin

content in woody biomass is also in agreement with the calorific value when compared to

lignin content in agricultural biomass. The calorific value improved after blending

process.

The Pellet Fuel Institute (PFI) Standard establishes a maximum of 10 % moisture content

of pellets. In this study, all the selected biomasses (RCG, timothy hay, spruce and pine)

were analyzed for their moisture content after grinding and stored in air tight container

for 1 week. As can be seen in Table 4.1, the moisture content of RCG (6.4 %), timothy

hay (6.9 %) and switchgrass (7 %) are lower than spruce (8 %) and pine (9.3 %). This low

moisture content in agricultural biomass may be the reason for low quality of pellets as it

is too dry for compaction. During pelleting, heat is supplied to form a pellet and therefore

biomass requires certain amount of moisture to bind the particles so that the pellet is

compacted and retain its binding formation.

63

104

Table III.1: Fuel and Chemical Analysis (Dry Basis) of Individual and Blended Biomass


HHV MJ/kg 18.61 17.58 18.20 19.15 19.75 18.56 19.00 18.07 18.68 18.46 18.44

BTU/lb 7996 7561 7825 8232 8493 8017 8060 7886 7909 8929 8052

C

(wt. %,

dry

basis)

45.45 45.46 45.03 47.14 46.80 46.64 45.98 46.57 46.72 47.12 47.21

H 5.89 5.95 5.98 6.06 6.14 6.06 6.02 6.01 6.08 6.08 6.12

N 0.174 0.184 0.119 0.040 0.039 0.035 0.036 0.04 0.038 0.033 0.027

S 0.041 0.041 0.028 0.002 0.002 0.016 0.017 0.018 0.017 0.013 0.01

O 48.44 48.37 48.84 46.76 47.02 47.13 46.97 47.36 47.14 46.75 46.64

Lig. % 17 11 9 23 24 19 16 13 18 17 14

Ash (wt. %,

dry

basis)

5.34 4.06 3.61 traces traces 2.07 1.54 1.51 1.62 1.53 1.55

MC 6.4 6.9 7.0 8.0 9.3 7.2 7.5 7.3 7.6 7.7 7.5

VM 65 66 78 70 70 69 70 70 69 73 71

FC 17.84 18.51 11.70 21.33 20.70 21.57 20.94 21.03 21.33 17.77 19.96

Notation: RCG =Reed Canary grass, H=Timothy hay, SW=Switchgrass, S=Spruce, P=Pine

105

Work of Compaction

The work required to pelletize RCG, timothy hay, spruce and pine individually, as well as

their blends was evaluated by means of integrating the area under the compaction curve

and relaxation curve as shown in Figure III.1. The amount of work was determined using

the force versus displacement graph. Process a-b-c-d is represented by the compaction

curve (a-d), which the work required to stream the biomass into the die (a-b), to tamp

down the biomass at lower punch (pre-compaction zone b-c) and to consolidate the

biomass account the friction (c-d).

Figure III.1: Force-displacement Curves during Compaction of Biomass Pellet

Figure III.2 schematically illustrates the pelletizing process in a cylindrical die channel.

The figure shows that the compaction (pelletization) process can be separated into

component sequences of workflow (Workflow), pre-compression (Workcomp) and work of

friction (Workfric) of biomass particles. Therefore, the total work of compaction can be

obtained as the sum of the individual energy requirements for these components minus

the relaxation component (relieved from compaction).

relaxation curve

compaction curve

a

b

c

d

106

Figure III.2: Illustration of Compaction Processes for Energy Analysis. Process A-

B: Work of Flow, Workflow, Process B-C: Work of Compression,

Workcomp, And Process C-D: Work of Friction, Workfric.

The pressure of 159 MPa and temperature of 80 °C in the die caused the particles to bond

together probably due to natural adhesion processes. It should be noted that in this study

no adhesives were used while pelletizing biomass. The surface-to-surface bonding of the

particles may have contributed significantly to the strength of the pellets, and the extent

of the bonding might depend on the lignin and the moisture content of the biomass.

Figure III.3 illustrates the area under the plots of the force vs piston position. These were

used to calculate the energy required in the Workflow, Workcomp, and Workfric

measurements. For instance, the Workflow was calculated as the area under the plot in Fig.

III.3a whiles the biomass streaming in the die (process a-b in Fig. III.2). The Workcomp

was calculated as the area under the plot in Fig. III.3b while all biomass was tamped down

107

at lower pressure to escape the air between particles as presented by the process b-c in

Fig. III.2. Also, it was indicated by the curve 2-3 in Fig. III.1. Finally, the Workfric was

calculated as the area under the plot in Fig. III.3c while flattened biomass forming pellet

as process c-d in Figure III.2.

Figure III.3: Plot Force-displacement Behavior Used for Workflow, Workcomp, and

Workfric Measurement.

Effect of Particle Size on Work Needed to Pelletize Biomass and Their Blends

The effect of work of compaction of materials depends on particle surface area [17] and

thus the particle size. The work required to pelletize the woody biomass (spruce) at

different particle size is illustrated in Fig. III.4. Spruce showed larger phase area, and thus

larger specific surface area of the pellet made from particle size of 150-300 µm as

compared to 300-425 and 425-600 µm particle size. This indicates that less energy is

needed to pelletize spruce with smaller particle size. Similar trend was obtained for other

individual and blended biomasses.

108

Figure III.4: Load-displacement Profiles of Spruce with Particle Size 150 µm (ΣArea

=2.07 kJ), 300 µm (ΣArea = 2.24 kJ) and 425 µm (ΣArea = 2.43 kJ)

Determining the Work Needed to Pelletize Forestry, Agricultural and Blended

Biomass

The difference in the flow work, compression work and friction work components during

pelletization forestry, agricultural biomass as well as blended biomass (shown in Table

4.2) might be due to their difference in the chemical and physical constituents. High lignin

and moisture content (Table III.1) in spruce and pine biomass might be a cause of high

flow work since they might increase the viscous nature of the sample. A higher

compression force was required to pelletize pine than other biomasses. This could be

related due to presence of chemical extractives in pine [14], which are reported to act as

plasticizers [18]. Low friction during pelletizing agricultural biomass may be due to

bonding or interaction mechanism between pellet surfaces and die wall. Inter-particles

bonding in agricultural biomass are also weaker than the forestry biomass.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Lo

ad (

kN

)

Displacement (mm)

150 µm

300 µm

425 µm

lpellet

109

The energy required to pelletize individual biomass and their blends is shown in Table

III.2. The mechanical properties of the pellet can be related to the work needed to pelletize

biomass material. For instance, forestry biomass provided better mechanical property

pellet due to higher work done or energy consumed to pelletize them as compared to

agricultural biomass pellet. As the work or energy needed to pelletize the biomass material

increased, there is a greater possibility of enhanced inter-particle bonding between the

particles and less voids in the pellet. Further, the contribution of lignin content from the

forestry biomass to the blend pellet might also have increased the work needed to pelletize

the biomass material.

Table III.2: Results of Work of Flow, Work of Compression and Work of Friction of

Biomass and Blended Biomass

Wflow (J) Wcomp (J) Wfric (J) Wtotal (J)

Spruce 1.23 0.14 1.06 2.3596

Pine 1.23 0.27 0.92 2.3496

RCG 0.94 0.17 0.57 1.6096

Hay 1.18 0.18 0.68 1.9696

Switchgrass 0.98 0.05 0.72 1.6796

RCG+spruce 1.20 0.04 0.92 2.0896

RCG+pine 1.12 0.07 0.88 1.9996

Hay+spruce 1.15 0.02 0.94 2.0396

Hay+pine 1.17 0.04 1.01 2.1496

Switchgrass+spruce 1.02 0.13 0.75 1.8296

Swicthgrass+pine 1.11 0.06 0.81 1.9096

Fracture Analysis of Forestry and Agricultural Biomass Pellet

Figure III.5 shows the SEM images of microstructure fracture surface of pine, RCG and

timothy hay biomass pellets. From Figure III.5a, it was observed that pine showed strong

inter-particle bonding as compared to RCG (Figure III.5b) and timothy hay (Figure III.5c).

110

High moisture content (9.3 %) and high lignin content (24 %) in pine were responsible for

the strength of the pellet. Biomass with smaller particle size (150 µm) can cause the lignin

to squeeze out due to pressure, temperature and compaction during the pelleting process.

Smaller particles could make an ease for the particles to bind and compact as this required

lower amount of work (Figure III.4) to yield quality pellet as compared to larger particles.

The particle bonding in RCG (Figure III.5b) is most likely due to a combination of particle

entanglement and auto-adhesive surface [14 and 15]. Timothy hay and switchgrass

showed poor particle to particle bonding since adhesive surface was not found at any site

of fracture surface of the pellet. The images showed only entanglement of particles in

pellets made from timothy hay and switchgrass. This might show evidence that lignin

content in the biomass is responsible for mechanical properties of pellet.

This interpretation of bonding between particles at microscopic level has been reported

earlier [13 and 15]. Pressure, temperature and moisture promote adhesion by bonding with

a solid bridge between particles. Based on the adhesion theories, when the maximum

attractive force reaches the minimum potential energy, chemical bonding is established.

Materials in ground form also contribute to the particle bonding. The effectiveness of these

forces diminishes dramatically as the size of the particles or inter-particle distance

increased [16 and 17].

111

Figure III.5: SEM Images of Fracture Surfaces for Pellets Made from Particle Size 150

µm of: a) Pine (at magnification 100 µm), b) RCG (at magnification 100

µm), c) Timothy hay (at magnification 200 µm) and d) Timothy hay (using

Optical)

Conclusion

The effect of blending agricultural biomass with woody biomass on mechanical work of

compaction was investigated. It was found that blending agricultural biomass with woody

biomass not just resulted in better quality of pellet but also resulted in lower energy

requirement for compaction as compared to pelleting woody biomass alone. Thus,

112

blending of biomass can help reduce the energy requirement and the overall cost of pellet

production process.

Acknowledgements





Malaysia.

References

[1] Gan, J., and C. T., Smith. (2006). A comparative analysis of woody biomass and

coal for electricity generation under various CO2 emission reductions and taxes.

Biomass and Bioenergy. 30(4), 296-303.

[2] Somerville, C., H. Youngs, C. Taylor, S. C. Davis, and S. P. Long. (2010).

Feedstock for lignocellulosic biofuels. Science. 329, 790–792.

[3] Miao, Z., Y. Shastri, T. E. Grift, A. C. Hansen, and K. C. Ting. (2012).

Lignocellulosic biomass feedstock transportation alternatives, logistics, equipment

configuration and modelling. Biofuels Bioproducts Biorefinery. 6, 351–362.

[4] Buytaert, V., B. Muys, N. Devriendt, L. Pelkmans, J. G. Kretzschmar, and R.

Samson. (2011). Towards integrated sustainability assessment for energetic use of

biomass: A state of the art evaluation of assessment tools. Renewable and

Sustainable Energy Reviews. 15, 3918–3933.

[6] Humbird, D., R. Davis, L. Tao, C. Kinchin, D. Hsu, A. Aden, P. Schoen, J. Lukas,

B. Olthof, M. Worley, D. Sexton, and D. Dudgeon. (2011). Process design and

113

economics for biochemical conversion of lignocellulosic biomass to ethanol: dilute-

acid pretreatment and enzymatic hydrolysis of corn stover. National Renewable

Energy Laboratory.

[7] DeCrosta, M.T., Schwartz, J.B., Wigent, J.B., Marshall, K., (2000).

Thermodynamic analysis of compact formation; compaction, unloading, and

ejection: Part I. Design and development of a compaction calorimeter and

mechanical and thermal energy determinations of powder compaction, International

Journal Pharmaceutical. 198, 113–134

[8] DeCrosta, M.T., Schwartz, J.B., Wigent, J.B., Marshall, K. (2001). Thermodynamic

analysis of compact formation; compaction, unloading, and ejection: Part II.

Mechanical energy (work) and thermal energy (heat) determinations of compact

unloading and ejection. International Journal of Pharmaceuticals. 213, 45–62.

[9] Rowlings, C.E., Wurster, D.E., Ramsey, P.J., (1995). Calorimetric analysis of

powder compression: II. The relationship between energy terms measured with a

compression calorimeter and tableting behavior. International Journal

Pharmaceutical. 116, 191–200.

[10] Harun, N. Y. and Afzal, M. T. (2015). Chemical and mechanical properties of pellets

made from agricultural and woody biomass blends. Transactions of the ASABE. 58

(4), 921-930

[11] ASTM E 1757 - 01 “Standard Practice for Preparation of Biomass for

Compositional Analysis" in 2003 Annual Book of ASTM Standards, Volume 11.05.

Philadelphia, PA: American Society for Testing and Materials, International.

[12] NREL Laboratory Analytical Procedure "Determination of ash in biomass"

114

[13] Dhamodaran, A. and Afzal M. T. (2012). Compression and springback properties of

hardwood and softwood pellets. BioResources. 7(3), 4362-4376.

[14] Nielsen, N. P. K., Gardner, D.J., Poulsen, T. and Felby, C. (2009). Importance of


residues into fuel pellets. Wood and Fiber Science. 41 (4), 414–425.

[15] Gardner, D. J. (2006). Adhesion mechanisms of durable wood adhesive bonds In

Groom, L. H. and Stokke, D. D. editors. Characterization of the cellulosic cell wall.

Blackwell Publishing. Ames, IA. 254-265.

[16] An overview of compaction of biomass grinds. Available from:

https://www.researchgate.net/publication/230704670/ [accessed May 16, 2016].

[17] Bolhuis G. K., and Holzer A. W., (1996). Lubricant sensitivity. In Alderborn, G.,

and Nystrom, C., editors. Pharmaceutical Powder Compaction Technology, New

York, USA: Marcel Dekker. 517–560.

[18] Matsunaga, M., Obataya, E., Minato, K., and Nakatsubo, F. (2000). Working

mechanism of adsorbed water on the vibrational properties of wood impregnated

with extractives of pernambuco (Guilandina echinata Spreng.). Journal of Wood

Science, 46(2):122 – 129.

115

Paper IV

Combustion behavior and thermal analysis of agricultural and woody

biomass blends




Perak, Malaysia.



Advances in Environmental Biology, 9 (15), 34-40

Abstract

Blended biomass, a new feedstock for pellets production potentially is viable to support

the demand for wood pellets at present and in the future. Combustion behavior and

characteristics tests were carried out in thermogravimetry (TG) analyzer with an air flow

rate of 100 ml/min and heating rate of 20 °C/min. Heat flow and heat required for the

biomass reaction was measured using differential scanning calorimeter (DSC). The

purpose of this study was to investigate the combustion behaviour and thermal properties

of both individual and blend agricultural (reed canary grass, timothy hay and switchgrass)

and forestry biomass (spruce and pine). The TG results showed that the combustion

behavior of all the biomass samples including the blends were almost similar. Two main

stages of combustion reaction were presented by differential thermal analysis (DTA)

curve. The heat released from the blended biomass (6.94 – 9.26 kJ/kg) was higher than

the individual agricultural biomass (4.59 – 6.78 kJ/kg) but lower than individual spruce

(10.2kJ/kg) and pine (11.13kJ/kg). The result indicates that the reactivity of the individual

agricultural biomass material changed due to blending. Overall, blending can help to

increase the energy capacity and can improve the combustion characteristics.

Keywords: Biomass blends, agricultural biomass, combustion characteristics, biomass

energy capacity, TGA, DSC

116

Introduction

Global wood pellet demand is expanding rapidly and consumers, businesses and

regulators are looking for alternatives to fossil fuels [1]. Currently, pellet industries rely

heavily on forestry biomass, which could face threat from biomass power station in the

near future. Wood pellet demand is projected to grow from 23 million tons from year 2014

to 50 million tons in year 2024 [1]. This means new feedstock resources have to be

exploited in order to support the future demand. Therefore, pellet industry is now looking

for alternate biomass resources to increase the pellet production. Canada as one of the

main producers of pellets has increased the pellets fuel export by 25 percent every year

[2]. Wood and agricultural residues account about 42% of the total Canadian biomass [3].

About 25 Mtoe (Megatonne of oil equivalent) of agricultural residues is generated

annually in Canada [4]. Furthermore, agricultural biomass has been highly recommended

by industries as a feedstock for making pellet [5, 6 and 7]. This would help to increase the

production of pellets and obviously can fill the demand.

Generally, pellets made from agricultural biomass show poor combustion behavior due to

the chemical components specifically the inorganic components as compared to woody

biomass. Other than the combustion behavior, the inorganic mineral contents also affect

the thermal characteristics and emissions. For instance, Werther et al. [8] reported that

some agricultural residues have high alkali oxides and salts, which may lead to various

problems during combustion due to their low melting points. Few others have also

reported the ash related problem due to inorganic matters such as potassium, chlorine and

sodium in the biomass during combustion [9 and 10]. According to Demirbas [11]:

The high moisture and ash contents in biomass fuels can cause ignition and

combustion problems.

117

The melting point of the ash is usually low and it causes fouling and slagging

problems.

And due to the above problems it was anticipated that blending biomass with

higher quality coal will reduce flame stability problems, as well as minimize

corrosion effects.

Further, agricultural biomass has also been used as fuel for combustion but its use can be

limited due to low heating value as compared to woody biomass. Therefore, it was

anticipated that blending agricultural residues with woody biomass may improve the

combustion behavior. Furthermore, the studies related to blends are limited to biomass

with coal, biomass with plastic, and other materials. However, information on blending of

agricultural residues with woody biomass is very limited in the literature. Therefore, none

has investigated the combustion behavior of agricultural and woody biomass blends using

TGA and DSC techniques. Except, TGA-DSC analysis of mischantus and poplar [12, 13]

was found.

The present study focuses on the combustion behavior of the selected agricultural (Reed

Canary grass, timothy hay and switchgrass) residues and their blends with woody biomass

(pine and spruce). In this work, agricultural residues from reed canary grass, timothy hat

and swicthgrass were chosen as biomass feedstock since they are available in large

quantity in New Brunswick. A particular attention was also given to the mineral content,

TGA and DSC analysis, and heat release.

118

Methodology

Sample Preparation

Agricultural biomass such as reed canary grass, timothy hay and switchgrass was selected

and acquired from agricultural farm in New Brunswick (NB). The woody biomass (spruce

and pine) were obtained from a wood mill situated near Fredericton, NB. The raw biomass

materials were stored in a closed container at ambient conditions. Prior to analysis, the

samples were dried in the oven at 105 oC for 24 h, and then grinded using Wiley mill to

less than 1 mm particle size. The grinded materials were sieved according to ASTM

standards (D 2013–72) to collect the particle size in the range of 150 – 300 µm. The

grinded biomass and their blends were stored in airtight container for further analysis. It

should be noted, that heating value, proximate, and ultimate analysis was conducted for

individual and blended biomass as presented in Table IV.1.

Table IV.1: Biomass proportion.

Agricultural Forestry Blends


Proportion

(wt. %)

100 100 100 100 100 50:50 50:50 50:50 50:50 50:50 50:50

Notes: RCG=Reed Canary grass, H=Timothy hay, SW=Switchgrass, S=Spruce and P=Pine


The ultimate analysis was performed using LECO CHNS 932 elemental analyzer as per

the ASTM D5291 method and was repeated twice. Element oxygen was calculated by

difference. Proximate analysis was determined twice using TG analyzer (Q500 TA

Instrument Inc.) as per ASTM E1641-04 method, while the ash content was investigated

119

as per NREL/TP-510-42622 method with 6 replications. The higher heating value (HHV)

of a biomass was measured using a bomb calorimeter (Isoperibol Oxygen Bomb

Calorimeter) according to the standard method, ASTM D-5865.

Mineral Analysis

The mineral concentration was determined according to U.S. EPA methods and

procedures. The samples were prepared using microwave-assisted digestion (MarsXpress,

CEM Corp., Matthews, N.C.) in nitric acid according to EPA Method 3051. The resulting

solutions were analyzed for trace elements by ICP-MS (Thermo X Series II, Thermo

Fischer Scientific, Inc., Waltham, Mass.) according to EPA Method 200.8. Phosphorus

was determined by ICP-ES (Varian Vista AX, Agilent Technologies, Inc., Santa Clara,

Cal.) according to EPA Method 200.7. Silica was determined by ICP-ES on sodium

peroxide fusions of samples ash. Chlorine was determined calorimetrically on aqueous

leaches of the samples using a photometric analyzer (AquaKem 250, Thermo Fischer

Scientific, Inc., Waltham Mass.) according to APHA Standard Method 4500-CL-E.

Combustion Analysis

To evaluate the effect of blending and the type of biomass (agricultural and forestry) on

the combustion behavior, the samples were subjected to TG (TA Instruments, USA) and

DSC analyzer. The samples were heated from room temperature to 1000 °C with heating

rate of 20 °C/min. This heating rate was selected based on the highest volatiles and char

yield that it produced from spruce and pine [14], and from miscanthus, poplar and rice

husk [13]. Air was used as purge gas with a flow rate of 100 ml/min.

120


Chemical and Mineral Content of Individual and Blended Biomass

The heating value of blended biomass (18 – 19 MJ/kg) was between agricultural residues

(17 – 18 MJ/kg) and forestry biomass (19 - 20 MJ/kg), as presented in Table IV.2. There

was a slight improvement in the heating value of agricultural residues after blending them

with forestry residues. For instance, the heating value (15.64 MJ/kg) of the blends (50 %

fir, 30 % beech and 20 % wheat straw) was higher than wheat straw (14.7 MJ/kg) and

beech (15.00 MJ/kg) [15]. Barmina et al. [16] also found that the heating value of blend

biomass (19.1 MJ/kg) was slightly higher than the reed canary grass biomass (18.6

MJ/kg).

In comparison with spruce and pine, the selected agricultural biomass contained higher

proportion of oxygen, but less hydrogen and carbon (Table IV.2). These could be the

reason for lower heating value of agricultural biomass since the energy contained in

carbon-oxygen bonds is usually lower [17] than carbon-hydrogen bond. However, after

blending, the carbon content was increased by around 2% for RCG, 2.5 % for hay, and

4.5% for switchgrass, whereas hydrogen content increased by around 2% (averaged) for

all blend biomass. Higher oxygen content indicates that the biomass will have higher

thermal reactivity [18].

The moisture content of the blended biomass increased slightly, which might have

contributed from woody biomass. However, the moisture content of blend biomass was

comparable to that of bituminous coal [8]. Volatile matters in agricultural biomass ranged

from 65 to 78 %. After blending, its values were close to woody biomass (69-73 %). The

ash content in blend biomass reduced by about 62% (averaged) as compared to individual

121

agricultural residues and this could be due to considerable reduction in char content of

blend biomass.

Table IV.2: Heating Value, Ultimate and Proximate Analysis of Individual and

Blended Biomass.

RCG H SW S P

RCG

+S

RCG+

P

H+S H+P

SW+

S

SW+

P

HHV MJ/kg 18.61 17.58 18.20 19.15 19.75 18.56 19.00 18.07 18.68 18.46 18.44

C (wt. %,

dry

basis)

45.45 45.46 45.03 47.14 46.80 46.64 45.98 46.57 46.72 47.12 47.21

H 5.89 5.95 5.98 6.06 6.14 6.06 6.02 6.01 6.08 6.08 6.12

O 48.44 48.37 48.84 46.76 47.02 47.13 46.97 47.36 47.14 46.75 46.64

Ash (wt. %,

dry

basis)

5.34 4.06 3.61 traces traces 2.07 1.54 1.51 1.62 1.53 1.55

MC 6.4 6.9 7.0 8.0 9.3 7.2 7.5 7.3 7.6 7.7 7.5

VM 65 66 78 70 70 69 70 70 69 73 71

FC 17.84 18.51 11.70 21.33 20.70 21.57 20.94 21.03 21.33 17.77 19.96

Notes: RCG=reed canary grass, H=timothy hay, SW=switchgrass, S=spruce and P=pine

A particular attention was also given to analyze the individual and blend biomass for

mineral content. Table IV.3 shows that all blend biomass had a lower sulfur content (<

0.05%) that comply with the European pellet standard [15]. These characteristics are

favorable for combustion applications [20], as compared to coal.

All other chemical and mineral elements such as nitrogen, silver, cadmium, arsenic, lead,

chromium, nickel and zinc, were observed lower than the standard limits by European

Standards for wood pellets, EN 14961-2 (as shown in Table IV.3) in all biomass including

blend. The chlorine content was higher in individual agricultural residues and some blend

biomass than the standard limit of 0.03%, which is not favorable for thermal system since

it can cause agglomeration on the heating wall. Overall, switchgrass was found to have

higher nickel, copper, zinc content compared to other biomass including blend.

122

Table IV.3: Chemical Analysis of Biomass and Blended Biomass

RL RCG H SW S P

RCG+

S

RCG+

P

H+S H+P SW+S

SW+

P

Chlorine 0.03 0.13 0.27 0.03 0.01 0.26 0.06 0.01 0.07 0.06 0.01 0.08

Sulfur 0.05 0.08 0.09 0.04 0.01 0.01 0.04 0.04 0.03 0.03 0.01 0.02

Nitrogen 0.3 0.17 0.18 0.12 0.04 0.04 0.04 0.04 0.04 0.04 0.03 0.03

Silver 0.1 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02

Cadmium 0.5 0.02 0.02 0.08 0.22 0.11 0.12 0.06 0.17 0.08 0.19 0.09

Arsenic 1 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2 < 0.2

Lead 10 0.14 0.11 0.22 0.03 0.09 0.09 0.11 0.06 0.12 0.18 0.26

Chromium 10 0.2 0.3 0.4 0.4 < 0.2 0.3 < 0.2 0.4 < 0.2 0.4 < 0.2

Nickel 10 4.2 5.5 7.3 0.6 0.5 2.4 2.5 2 1.8 2.3 2.2

Copper 10 6.2 9.8 12.8 1.8 2.6 5.1 5 3.9 6.7 9.7 9.1

Zinc 100 12.2 9.3 13.9 8.5 10.1 10.4 11.4 8.7 8.5 10 10.9

RL=Relative Limits based on European Standards for Energy Pellet (Prvulovic et al., 2014)

RCG=reed canary grass, H=timothy hay, SW=switchgrass, S=spruce and P=pine

Combustion Behavior

The DSC and DTA curves revealed two different exothermic reaction regions as shown

in Figure IV.1. The curve obtained for each biomass was analyzed to determine the

relevant combustion parameters, such as peak temperatures, ignition temperature for

combustion, and burnout temperature. First region was attributed due to the combustion

of light volatile matters and the second region might be due to the combustion of heavy

volatile and fixed carbon [21]. Similar reaction regions were observed for rice husk using

TGA and DSC [22]. In this study, it is observed that the temperature range of each region

was different for each biomass species and blends (Figure IV.1).

123

Notes: RCG=reed canary grass, S=spruce and P=pine

Figure IV.1: DSC Curves of RCG, Spruce, Pine and Their Blends

The high volatile content in spruce and pine as presented in Table IV.2 was also confirmed

by the DTA curve (first region) in Figure IV.1 and DTG curve in Figure IV.2. Karampinis

et al. [12] characterized miscanthus and poplar using DSC and it revealed that these

biomasses are reactive fuels because of high volatile matter. As presented in Figure IV.1,

the combustion behavior of blended biomass (RCG+S and RCG+P) was found to follow

the combined effect of chemical composition present in agricultural residues (RCG) and

woody biomass (S and P). Since spruce and pine showed highest area under the curves of

heat flow profiles, they can be considered as highly exothermic or reactive biomass fuel.

On the other hand, the heat released by the blended biomass (RCG+S and RCG+P) was

higher than the individual agricultural (RCG) biomass.

Similarly, two main reaction regions during the combustion of RCG, spruce, pine and

their blends were observed from DTG profiles. The first region refers to devolatization of

-5

0

5

10

15

20

25

140 340 540

He

at flo

w (

W/g

)

Temperature (°C)

Spruce

Pine

RCG

RCG+S

RCG+P

Combustion of light

Combustion of heavy

124

biomass that took place at temperature around 200 °C due to combustion of light volatiles

present in the biomass. As observed in Figure IV.2, two ‘‘shoulder peaks” occur in the

first region of the DTG curve at around 220 ºC and 315 °C, respectively, for reed canary

grass. The same behavior was also found for other agricultural biomass (timothy hay and

switchgrass) [23]. This peak could be attributed due to the decomposition of hemicellulose

and cellulose, respectively. On the contrary, only one peak was observed in spruce and

pine. This could possibly be due to their lower hemicellulose content than cellulose, which

caused to merge the reaction mechanism [24]. Another reason could be due to delay in the

thermal decomposition of the hemicellulose [25]. The DTG result in Figure IV.2 shows

only one reaction peak for reed canary grass and woody biomass blend. This could be due

to greater contribution of woody biomass characteristics in the blend. Apparently, in first

reaction region the main reaction peak for woody biomass appeared earlier (335 °C) than

RCG (315 °C) and their blends (328 °C).

From Figure IV.2, the initial ignition temperature (180 °C) was similar in case of RCG

and it blends (RCG+S and RCG+P). For spruce and pine it was slightly higher (220 °C)

than woody biomass. The reactivity in combustion regions is proportional with the height

of DTG peak (Kok and Özgür, 2013). The DTG peaks as shown in Figure IV.2 for reed

canary grass was lower (0.7 %/°C) than reed canary grass blends (1.1 %/°C) and woody

biomass (1.5 %/°C). However, the degradation trend of reed canary grass blends followed

the degradation of woody biomass. Therefore, blending process is believed to improve the

reactivity as compared to individual agricultural biomass.

125

Notes: RCG=reed canary grass, S=spruce and P=pine

Figure IV.2: DTG curves of RCG, Spruce, Pine and Their Blends

For the second region, combustion of more complex and thermally stable structure and

formation of char took place. In this phase, DSC curves (Figure IV.1) corresponds to

decomposition of lignin. The curves indicated that the decomposition of woody biomass

(spruce and pine) was the highest as compared to blends, and individual agricultural

biomass. For other biomass samples, the change in ignition and burnout temperature of

biomass samples is presented in Table IV.3, and for the blended biomass is presented in

Figure IV.3. The combustion of woody biomass and switchgrass started at 220 °C

temperature (Table IV.4). However, combustion of reed canary grass and timothy hay

begin earlier (180 and 160 °C, respectively). High ignition temperature indicates higher

heat capacity of a biomass [13]. Combustion of heavy volatiles in all agricultural biomass

completed within 150 °C temperature difference. Compare to woody biomass, combustion

of heavy volatiles in agricultural biomass was within 80 °C temperature difference. This

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

140 240 340 440 540

DT

G (

%/°

C)

Temperature (°C)

Spruce

Pine

RCG

RCG+S

RCG+P

Maximum

degradation

Devolitization

Char

Ignition Burnout

126

could be due to higher ash and mineral content in agricultural biomass as presented in

Table IV.2 and IV.3, respectively. Similar observation was reported by Kok and Özgür

[13].

Table 6.4: Reaction Regions of Parent Biomass

Combustion of light

volatiles

Combustion of heavy

volatiles

Spruce (S) 220-400 400-470

Pine (P) 220-400 400-480

Reed canary grass (RCG) 180-370 370-520

Timothy hay (H) 160-370 370-510

Switchgrass (SW) 220-375 375-520

From Figure IV.3, the combustion temperature range for all blended biomass was almost

similar, which was around 180 – 500 °C. Irrespective of parent biomass as presented in

Table IV.3, the combustion characteristics of all blended biomass were uniform. The

ignition temperature, when the light volatiles starts to eliminate rapidly, was found almost

similar at 180 °C, for most blended biomass. However the peak was highest (1.64 %/°C

and 1.43 %/°C) for swicthgrass blends (SW+S and SW+P), and the lowest peak (1.1 %/°C)

was observed for reed canary grass blends (RCG+S and RCG+P) and timothy hay blends

(H+S and H+P). This could be due to the difference in the hemicellulose and cellulose

fractions [24] of blended biomass samples.

127

Notes: RCG=reed canary grass, H=timothy hay, SW=switchgrass, S=spruce and P=pine

Figure IV.3: a) DTG and b) DSC Curves of Blended Biomass.

Heat Release

Heat released during the combustion of biomass samples at heating rate of 20 °C/min was

calculated based on the area under DSC curve as presented in Figure IV.1 and Figure

IV.3b. The low heat release (4.5 – 6.7 kJ/kg) in agricultural biomass might be due to

breaking of low volatile chemical bound. Similar results were found by by Kok and Özgür

[13] for rice husk and miscanthus biomass. In woody biomass the chemical bonds are not

easily broken and this might result in the release of higher amount of heat (10 -12 kJ/kg)

than agricultural biomass. The heat released due to combustion of individual agricultural

and woody biomass as well as their blends is shown in Figure IV.4.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

140 240 340 440 540

DT

G (

%/°

C)

Temperature (°C)

-5

0

5

10

15

20

25

30

140 240 340 440 540

He

at flo

w (

W/g

)

Temperature (°C)

RCG+S

RCG+P

H+S

H+P

SW+S

SW+P

128

Figure IV.4: Heat Release of Biomass and Blended Biomass

Blending of agricultural with woody biomass improved the release of energy. Several

reason could be behind such result. For instance, increase in carbon and hydrogen content,

and contribution of lignin from woody biomass in the blend. As presented in Figure IV.4,

the energy released from the blend biomass was in the range of 7.0 to 10 kJ/kg.

Conclusions

Combustion behavior and thermal analysis of selected agricultural and woody biomass

and their blend was investigated in this study. The differences in thermal behavior of

agricultural and woody biomass were analyzed and the following conclusions were

derived:

The reactivity, combustion behavior and degradation rate peak of the agricultural

biomass changed after blending them with woody biomass.

0

2000

4000

6000

8000

10000

12000

He

at re

lease

(J/k

g)

Blended Biomass

129

Typically, agricultural biomass contains higher amount of inorganic minerals

which can result in higher emissions and problem during combustion process.

Blending agricultural residues with forestry biomass helped to reduce the sulfur,

copper, chlorine and ash content that can possibly meet the standard limit of

biomass fuel.

There was some improvement in heat released (6.94 to 9.26 kJ/kg) during

combustion for blended biomass as compared to individual agricultural biomass

(reed canary grass - 5.4 kJ/kg, timothy hay- 6.78 kJ/kg, and switchgrass - 4.59

kJ/kg).

Overall, blending of agricultural biomass with woody biomass can be considered as one

of the alternative options to increase the biomass fuel feedstock such as pellet production

to meet the future demand. Future works will be proposed to study the biomass availability

in Malaysia for industrial application and meeting pellets demand of neighbourhood

country.

Acknowledgements.





Malaysia.

References

[1] Walker, S. 2015. Wood Bioenergy Magazine. Available at:

www.woodbioenergy.com. Accessed 12 April 2015

130

[2] Exportation of wood pellets from North America to Europe reached a new record,

[Online]. Available at: http://www.forestbioenergyreview.com/pellets/item/127-

exportation-of-wood-pellets-from-north-america-to-europe-reached-a-new-record.

Accessed 15 April 2013.

[3] Wood, S. M., and D. B. Layzell. 2003. A Canadian biomass inventory: feedstocks

for a bio-based economy. BIOCAP Canada Foundation Final Report Industry

Canada Contract #: 5006125:2003

[4] Helwig, T, R. Jannasch, R. Samson, A. DeMaio, and D. Caumartin. 2002.

Agricultural biomass residue inventories and conversion systems for energy

production in Eastern Canada Final Report Prepared for Natural Resources-Canada.

[5] Samson, R., S. Bailey, and C. Ho Lem. 2007. The Emerging Agro-Pellet Industry

in Canada Resource Final Report. Efficient Agricultural Production (REAP) –

Canada

[6] Thrän, D. and C. Hennig. 2011. Global Wood Pellet Industry Market and Trade

Study. In IEA Bioenergy Task 40: Sustainable Bioenergy Trade. 190pp

[7] BioDimension. 2012. West Kentucky Agricultural Biomass Pellet Report. Report

for West Kentucky Center for Emerging Technology, Murray State University, and

Memphis Bioworks Foundation. 55pp.

[8] Werther, J., Saenger, M., Hartge, E.U., Ogada, T. and Z. Siagi. 2000. Combustion

of agricultural residues. Progress in Engineering and Combustion Science, 26:

1 – 27

[9] Wang, K.S., Chiang, K.Y., Tsai, C.C,, Sun, C.J., Tsai, C.C., and K.L. Lin. 2001.

The effects of FeCl3 on the distribution of the heavy metals Cd, Cu, Cr, and Zn in a

simulated multimetal incineration system. Environment International. 26(4):257-63.

131

[10] Khan, A. A., de Jong, W., Jansens, P.J., and H. Spliethoff. 2009. Biomass

combustion in fluidized bed boilers: potential problems and remedies. Fuel

Processing Technology, 90(1): 21–50

[11] Demirbas, B. 2009. The global climate challenge: Recent trends in CO2 emissions

from fuel combustion. Energy Education Science Technology. 22:179–193.

[12] Karampinis, E., D. Vamvuka, S. Sfakiotakis, P. Grammelis, G. Itskos, and E.

Kakaras. 2012. Comparative study of combustion properties of five energy crops

and Greek lignite, Energy Fuels, 26(2): 869–878

[13] Kok, M.V. and E. Özgür. 2013. Thermal analysis and kinetic of biomass samples.

Fuel Processing Technology, 106: 739–743

[14] Harun, N. Y. and M. T. Afzal. 2010. Thermal Decomposition Kinetics of Forest

Residue. Journal Applied Science, 10(12): 1122-1127

[15] Prvulovic, S., Z. Gluvakov, J. Tolmac, D. Tolmac, M. Matic, and M. Brkic. 2014.

Methods for determination of biomass energy pellet quality. Energy fuels, 28:

2013−2018

[16] Barmina, I., A. Lickrastina, R. Valdmanis, M. Zake, A. Arshanitsa, V. Solodovnik,

and G. Telysheva. 2013. Effects of Biomass Composition Variations on Gasification

and Combustion Characteristics. Engineering for Rural Development, 282-287

[17] Munir, S., S. S. Daood, W. Nimmo, A. M. Cunliffe, and B. M. Gibbs. 2009. Thermal

analysis and devolatilization kinetics of cotton stalk, sugar cane bagasse and shea

meal under nitrogen and air atmospheres. Bioresource Technology, 100: 1413-1418

[18] Haykiri-Acma, H. and S. Yaman. 2008. Effect of co-combustion on the burnout of

lignite/biomass blends: A Turkish case study. Waste Management, 28: 2077-2084

132

[19] Vamvuka, D., E. Kakaras, E. Kastanaki, and P. Grammelis. 2003a. Pyrolysis

characteristics and kinetics of biomass residuals mixtures with lignite. Fuel, 82:

1949-1960

[20] Vamvuka, D., N. Pasadakis, E. Kastanaki, P. Grammelis, and E. Kakaras. 2003b.

Kinetic modeling of coal/agricultural by-product blends. Energy Fuel, 17: 549-558

[21] Biagini, E., F. Lippi, L. Petarca, and L. Tognotti. 2002. Devolatilization rate of

biomasses and coal-biomass blends: an experimental investigation. Fuel, 81: 1041-

1050

[22] Biagini, E., A. Fantei, and L. Tognotti. 2008. Effect of the heating rate on the

devolatilization of biomass residues. Thermochim Acta. 472:55–63. DOI:

10.1016/j.tca.2008.03.015.

[23] Harun, N. Y. and M. T. Afzal. 2015. Chemical and Mechanical Properties of Pellets

Made from Agricultural and Woody Biomass Blends. Transactions of the ASABE,

Vol. 58(4): American Society of Agricultural and Biological Engineers ISSN 2151-

0032 DOI 10.13031/trans.58.11027

[24] Shen, D. K, S. Gu, K. H. Luo, A. V. Bridgwater, and M. X. Fang. 2009. Kinetic

study on thermal decomposition of woods in oxidative environment. Fuel, 88: 1024–

1030

[26] Garcia-Maraver, A., L. C. Terron, M. Zamorano, and A. F. Ramos-Ridao. 2013.

Thermal events during the combustion of agricultural and forestry lopping residues.

In Materials and process for energy: communicating current research and

technological developments (A. Mendez-Vilas. Ed.)

133

Paper V

Torrefaction of Agriculture and Forestry Biomass using TGA-FTIR-

MS




Perak, Malaysia.



Dincer et al. (eds.), Progress in Exergy, Energy, and the Environment, 5 (77)

Abstract

The torrefaction of agriculture (switchgrass and timothy) and forestry (spruce and pine)

biomass was studied using simultaneously thermogravimetric analyzer (TGA) coupled

with Fourier transform infrared (FTIR), and mass spectrometer (MS). The chemical

functional groups present in the gases were identified by FTIR and the quantification of

gaseous products was determined using MS at different torrefaction temperatures ranging

from 200 to 290 oC. TG-FTIR and TG-MS techniques are paired to refine the

identification of gases. TGA results showed that the behaviour of the agricultural and

forestry biomass was not the same due to their composition variation. The decomposition

of switchgrass took place at a lower temperature than other biomass. Both switchgrass and

timothy have two peaks of degradation rate compared to only one peak present for forestry

biomass. The FTIR analysis indicated that most of the chemical compositions present in

the biomass are decomposed at torrefaction temperature of 290 oC. The mass

spectrometric analysis at torrefaction temperature 200 °C and 230 °C quantified the

degradation of combustible gases: CH4, C2H4, CO and O2 around 20-30%. Whilst at

torrefaction temperature 260 °C and 290 °C the degradation of combustible gases were

more than 30%. Moreover, all gaseous products evolved from the torrefaction of

agricultural and forestry biomasses were almost similar in characteristics, but vary in

proportions.

Keywords: Agricultural biomass; Forestry biomass; TG-FTIR; Torrefaction

134

Introduction

Canada is one of the countries in the world which generates abundant amount of biomass

from agriculture and forestry sector. Canada has about 42% of forest land and 6.8% of

agricultural land from which about 42% of residue is produced entirely from forestry and

agricultural (Wood and Layzell, 2001). Hence, there exists large potential to convert this

biomass into valuable products for various applications. Thermo-chemical treatment is an

effective method to handle the issue of this huge biomass resource. This process has great

advantages since it is flexible in feedstock, produces different fuels and is environmental

friendly.

Torrefaction is also emerging as one of thermo-chemical method to produce energy fuel

by improving the biomass properties (Acharya et al., 2012). The biomass is heated

between temperatures 200 to 300 °C in an inert condition.Torrefaction has been reported

(Acharya et al., 2012) to improve the hygroscopic behavior, decay resistivity, grindability, and

higher heating value. Moreover, it can help to reduce the greenhouse gas emissions. Assuming

net calorific value of 6.2 MWh/tonne, torrefied pellets (4,533,724 metric tons) is estimated

to save CO2 emissions of 21,081,816 tons, if the former is replaced by traditional pellets

(5,736,549 metric tons) (Melin, 2012). Despite this saving in emissions, the behavior of

biomass torrefaction still needs serious research attention.

With regard to this, TGA–FTIR technique is the most widely used method to observe the

thermal decomposition of materials under various environments. Understanding about the

kinetics and in depth analysis of mass loss with the type of release of gas products (Fasina

and Littlefield, 2012) is essential for the fundamental knowledge, design, operation and

control (Miranda et al., 2007) of torrefaction equipment and process. Therefore, the

influence of parameters on chemical composition of the torrefied product has been

135

extensively studied. In torrefaction, the lignin network is modified (Alen et al, 2002)

(Tjeerdsma and Militz, 2005) (Nguila et al, 2007), and hemicelluloses are strongly

degraded (Nuopponen et al, 2004) leading to formation of carbonaceous material within

the wood (Nguila et al, 2007). The anhydrous mass loss during torrefaction could be a

reliable and accurate indicator to predict dimension stability and decay resistivity of pellet

fuel (Hakkou et al, 2006) (Welzcher et al, 2007).

Some studies were found related to torrefaction of biomass using TG analysis to determine

the torrefaction kinetic of cellulose, hemicellulose, lignin and xylan (Chen and Kuo,

2011). The same author has also conducted torrefaction of various biomasses (bamboo,

willow, coconut shell and wood) using TG analysis (Chen and Kuo, 2010). Another study

also reported which was focused on torrefaction of deciduous wood (beech and willow),

coniferous wood (larch) and straw using TG analyser (Prins et al., 2006). Very recently, a

study on torrefaction of wheat straw but using TGA/DSC configuration mode was

performed (Shang et al., 2013). From this and other literature survey, none has attempted

to torrefy biomass using TGA coupled with FTIR. Since TGA coupled with FTIR can be

advantageous for continuous online gas analysis, the torrefaction thermal decomposition

can be studied in depth. It was also difficult to find torrefaction of agriculture (switchgrass

and timothy) and forestry (spruce and pine) biomass using TG-FTIR-MS technology.

Keeping this in view, the study contributes new knowledge about the characteristics of the

gas evolved during torrefaction of various types of biomass. The main objective was to

determine the torrefaction behavior as well as simultaneously investigate the chemical

functional groups and quantification of the gas product. TG-FTIR-MS analyses of gas

were carried out at different temperature of the torrefied biomass. Comparison of

torrefaction of agriculture and forestry biomass is presented in this study.

136

Materials and Method

Four types of biomass were selected for torrefaction. Two were agriculture (switchgrass

and timothy) and other two were forestry (spruce and pine). The agricultural biomass

samples were provided by NB Department of Agriculture. The forestry biomass chips of

size ranging from 10 to 50 mm were acquired from local sawmill. Prior to the analysis,

biomasses were first dried in the oven at 105 oC for 24 h, and then grounded to about 500

μm particle size.

TG-FTIR Experiment

A Nicolet 6700 FTIR Thermo Scientific connected to TGA Q500 TA Instrument was used

to perform torrefaction experiments. Approximately 10 mg of sample was loaded in a

TGA crucible. The temperature was initially raised to 105 oC at heating rate of

20 oC/min and held for 5 min in order to remove moisture. At second stage the samples

were torrefied using same heating rate to a desired temperature. The inert atmosphere was

maintained using nitrogen gas at flow rate of 100 ml/min. Vapors released from biomass

torrefaction were subjected to the FTIR spectrometer through a transfer tube, which is

heated to about 200oC to prevent the condensation of vapors on the tube wall. The IR

spectra were recorded with a temporal resolution about 2 s. The resolution of the collected

spectra was set to be 1 cm−1 and the spectral range was set from 4000 to 400 cm−1. At the

carrier gas flow rate of 100 ml/min, it took about 50 s for the vapors to reach the FTIR

cell from the thermogravimetric analyzer. Therefore, there was time delay of about 50 s.

137

TG-MS Experiment

The volatile products generated from TGA that was first scanned by FTIR spectrometer

to detect the functional group, and then were furthered transferred to the ionization source

of the mass spectrometer, SRS RGA 200, in order to determine specific chemical

compound. The transfer line was heated and maintained at a temperature of 200ºC to

prevent the condensation of the volatile gases released during the torrefaction process.

The mass spectrometer is operated at electron energy of 70 eV. The gases were analysed

by using the RGA software. A scan of the m/z was carried out from 1 to 100 amu to

determine which m/z has to be followed during the TG experiments. The intensities of 10

selected ions (m/z = 15, 16, 27, 28, 29, 32, 43, 44, 94, and 96) were monitored with the

thermogravimetric parameters. An absolute quantification in this work was performed for

each monitored gases: CH4, CO2, CO, C2H4, NH3, O2, C6H6O, CH2O, C2H4O2 and

CH5O3P. The concentrations of CH4, CO2, CO, C2H4, NH3, O2, C6H6O, CH2O, C2H4O2

and CH5O3P were directly deduced from the partial pressure of gases by using Eq.1.

[Ai] = (Pi / PT) * 1 000 000 (1)

Where, [Ai] is the concentration of the gas A, Pi is the partial pressure for m/z = i

representing of gas A and PT is the total pressure for all species i.


Proximate Analysis

Table V.1 shows the proximate analysis of the biomass used in this study. Moisture

content, volatile matter and fixed carbon were determined using TGA analysis. However,

ash content was investigated as per ASTM Standard Method E1755-01. Low moisture

138

content in the present biomass suggests them as a good candidate for torrefaction process.

Switchgrass showed highest amount of volatile matter and lowest fixed carbon content

compared to other biomass. Agricultural biomass showed high amount of ash content

compared to forestry biomass. The same comparison in the ash content was reported for

agricultural biomass and woody biomass (Raimie et al., 2012).

Table V.1: Proximate Analysis Data of Selected Biomass Materials (% Weight Dry)

Spruce Pine Switchgrass Timothy

Moisture (%) 2.92 4.63 2.23 2.82

Volatile matter (%) 73.75 73.05 68.46 79.55

Fixed Carbon (%) 23.33 22.32 25.70 13.57

Ash (%) trace trace 3.61 4.06

TG-DTG Analysis

Fig. V.1 depicts the thermogravimetric curve for different biomasses. The degradation

reactivity for forestry biomass (spruce and pine) was almost similar to each other.

However the degradation behavior of switchgrass differed from other biomass. The initial

decrease in the weight loss of biomass was attributed to removal of moisture content up

to temperature 110 °C. The initial degradation temperature for spruce, pine and timothy

was about 210 °C. Whilst for switchgrass it was about 175 °C. This shows that the

decomposition of switchgrass took place earlier compared to other biomass.

139

Figure V.1: Thermo-Diagram of Biomass in Nitrogen Atmosphere, the Loss of

Moisture, Volatiles and Char

From the TGA, the final degradation temperature for spruce, pine, switchgrass and

timothy was about 375, 375, 360 and 380 °C, respectively. Between temperatures 200 to

500 °C higher decomposition of cellulose and hemicellulose might have occurred, while

small amount of lignin degradation is expected to take place. This observation was

reported similarly for various woody biomasses (Chen and Kuo, 2010). No significant

weight loss was observed after temperature of 500 °C. The residual or biochar content for

spruce, pine, switchgrass and timothy was about 16, 15, 21 and 11 wt %, respectively.

Apparently, switchgrass showed a highest total weight loss at temperature 800 °C. With

respect to the torrefaction temperature of about 300 °C, switchgrass showed highest

weight loss of about 30 wt% compared to spruce (17 wt%), pine (20 wt%) and timothy

(20 wt%). This shows that switchgrass can be an attractive raw material for torrefaction.

According to previous studies (Antal, 1983) and (Mansaray and Ghaly, 1998), the thermal

140

decomposition of hemicellulose, cellulose and lignin occurs at temperatures ranging from

150-350 °C, 275–350 °C, and 250-500 °C, respectively.

Similarly, the decomposition of lignocellulosic materials in four biomasses is presented

in Fig. V.2. The figure shows that agricultural biomass has two peaks as contrast to one

peak for the forestry biomass. The first peak is attributed due to degradation of

hemicellulose. However, second peak represent the decomposition of cellulose. The first

and the second peak for switchgrass were found at temperature of about 230 and 330 °C,

respectively. In case of timothy it was found at temperature of about 310 and 355 °C,

respectively. The DTG peaks for forestry biomass (spruce and pine) was at temperature

350 °C.

Figure V.2: DTG Curves of Biomass.

From this observation, it can be concluded that the hemicellulose content in the softwood

biomass is less compared to agricultural biomass. A kinetic study on willow and

141

agricultural biomass was found that willow has less hemicellulose content compared to

reed canary grass and wheat straw (Bridgeman et al., 2008).

FTIR Analysis

Fig. V.3 to V.6 illustrates the possible chemical functional groups of the torrefied biomass.

The FTIR spectra shown in Fig. V.3-6 correspond to the gas analysed at different

torrefaction temperature of 200, 230, 260 and 290 °C, respectively. Overall, the peaks for

different functional groups became more sharp and visible when the temperature was

increased from 200 to 290 °C. This shows that at higher temperature around 300 °C more

degradation of biomass took place. The band around 3600–3400 cm−1 is apparently due

to hydroxyl (O-H) groups, which was found in all torrefied biomass. The C-H stretch band

around 2970-2780 cm-1 is the indication of an organic compound. A sharp band around

2300-2200 cm-1wave number is due to the formation of CO2 during torrefaction of

biomass. It is observed that the two major bands in the region of 1740–1710 cm−1and

1250–1220 cm−1, might be because of C–O and C–O–C functional groups, respectively.

Organic phosphates (P–O) functional groups can be found in range of 1350–1250 cm-1

stretch band, and the presented of aliphatic phosphates (P-O-C) functional groups is more

obvious at around 1050-990 cm-1 than organic phosphates (P-O) functional groups. A very

sharp peak around 730-550 cm-1 might be due to acid chlorides (C-Cl) functional groups

present (John Coates, 2000)].

Torrefaction 200 °C

Same procedure was employed for all torrefaction condition; the change of transmittance

intensities indicates a variation in the gas concentration (Bassiliakis et al, 2001). Fig. V.3

142

shows, pine and swicthgrass are more reactive than timothy and spruce at torrefaction

temperature 200 ºC, particularly for O-H, C-C and C-O-C of functional groups. C-H and

O=C=O functional groups in all biomass, which are obviously methane (CH4), and carbon

dioxide (CO2) and carbon monoxide (CO), significantly transmit as they are reactive

between 150 to 300 ºC (Tihay and Gillard, 2010). However, different biomass gives out

CO2 and CO at different concentration as shown in Fig. V.3 to Fig. V.6. CO2 from spruce

is the lowest in concentration compared to pine, switchgrass, and timothy at torrefaction

temperature 200 ºC. Switchgrass shows significant C-O-P and O-P groups than the others.

Figure V.3: Decomposition of Functional Groups at Torrefaction Temperature

200 °C

Organochlorine compounds (C-Cl functional groups) are normally present in plants and

wood. Its reactivity varies enormously, but most are relatively inert. However, this

functional group has shown different concentration from different biomass. The intensity

at torrefaction temperature 200 ºC is at the highest in pine compared to switchgrass and

timothy, and seemingly none from spruce.

143


As at torrefaction temperature 200 ºC, most functional groups significantly transmitted at

torrefaction 230 ºC but at different level of intensity. Fig. V.4 shows, the chemical

compounds in switchgrass are the most reactive compared to the one in pine, spruce and

timothy.


230 °C

O-H, C-C and C-O-C of functional groups in particular, an increase in the intensity were

transmitted from torrefaction temperature 230 ºC. The release of CO2 at torrefaction

temperature 230 ºC is about the same as at torrefaction temperature 200 ºC for pine,

switchgrass and timothy. However, the CO2 release from spruce has increased at this

torrefaction temperature.

At torrefaction temperature 200 ºC, as well as torrefaction temperature 230 ºC, the

phosphate compounds released significantly only from switchgrass, and apparently none

from the other biomass. Organochlorine compounds (C-Cl functional groups) are released

from all biomass at this torrefaction temperature, but the transmittance still at different

144

intensity. Pine and switchgrass released almost of this compound at torrefaction

temperature 230 ºC.


From Fig. V.5, pine and timothy were found has less reactive as compared to spruce and

switchgrass. The transmittance intensity for spruce has shown significant increased at

torrefaction temperature 260 ºC from the torrefaction temperature 200 °C and 230 °C.


260 °C

The intensities of all the functional groups of chemicals for spruce were just a slight lesser

than switchgrass. This can be anticipated that the degradation of spruce, pine and timothy

were engaged at later temperature than that of switchgrass. This might be almost

hemicellulose and partially cellulose from switchgrass easily released at lower torrefaction

temperature (200 °C, 230 °C and 260 °C) due to less lignin content to bind them from

devolatilization and depolymerisation (Tumuluru et al., 2010) (Demibras, 2009) (Mohan

et al.,2006). Lignin decompose from temperature 280 ºC and difficult to dehydrate. Thus

145

it converts to more char and loose the covalent bond to cellulose or hemicelluloses

(Tumuluru et al., 2010). The releases of CO2 from switchgrass and spruce at torrefaction

temperature 260 ºC were higher than that of lower torrefaction temperature (200 and 230

ºC). Switchgrass was also identified giving off the gases (functional groups O=C=O) the

highest among the selected biomasses at torrefaction temperature 260 °C. The phosphate

compounds were gradually degraded at this torrefaction temperature from torrefaction

temperature 200 °C and 230 °C for all biomass. Organochlorine compounds (C-Cl

functional groups) released from spruce and switchgrass significantly greater than that

from pine and timothy.


It is clearly shown from Fig. V.6, the transmittance intensities of all the selected functional

groups were significantly high which means almost gases and organic compound were

degraded at torrefaction temperature 290 ºC.


290 °C

146

The released intensities of O-H, C-H, O=C=O, C-C, C-O-C, and O-P functional groups

were found greater at torrefaction temperature 290 °C than that of at lower torrefaction

temperature. The lignocellulosic biomass has converted to carbonaceous materials due to

generous hemicellulose and cellulose was degraded at this torrefaction temperature. The

organochlorine compounds (C-Cl functional groups) and gases (O=C=O functional

groups) released from swicthgrass were still significantly greater than that of spruce, pine

and timothy.

Gas quantification

FTIR and mass spectrometric analyses have advantages and disadvantages. Using

information provided by the FTIR spectrum is possible to have an identification of the

gases emitted by each biomass. However, quantification of each emitted gases require

specific tool or software to measure. TG-FTIR and TG-MS techniques are therefore

complementary since they can refine the identification of gases. Table V.2 shows the

concentration of each gas during torrefaction of the selected biomass as the results from

the TG-MS. To compare the evolution of gas emissions during torrefaction, the gas

composition was calculated as a function of temperature. The comparison is made upon

the gas concentration released by the biomass, which the yields of the gases were

calculated for the four range of torrefaction temperature (Table V.2). From the table all

the four stages of temperature, the degradation of gases are mainly composed on O2, CO2,

and CO, and then are followed by C2H4, NH3, CH2O and CH4. Carbolic acid, acetic acid

and phosphonic acid have the lowest degradation during the torrefaction. Switchgrass has

degraded carbon dioxide and carbon monoxide the highest compared to spruce, pine, and

timothy. These results are in agreement with the FTIR spectrum, which the functional

147

group showed among the highest of transmittance for switchgrass. The C-H functional

group in the FTIR spectrum has depicted about the same transmittance intensities for all

the biomass. The mass spectrum analysis found for these same intensities was attributed

due to the methane (CH4) generated from switchgrass, timothy and pine, and ethylene

(C2H4) generated from spruce. Simple gas molecule O2 does not have infrared spectra,

which cannot be scanned by FTIR. MS has this advantage on O2 that has shown in this

study the amount composed in biomass is as high as other main composition such as H2O

(depicted in FTIR spectrum). Volatile products released from biomass was identified H2O,

CO, CO2 as the main composition (Tihay and Gillard, 2010), which was found the same

in this study including O2. Pine has generated the oxygen at the highest concentration as

compared to switchgrass, timothy and spruce.

Table V.2: Concentration of Selected Gaseous Products Evolved from Torrefaction

of Spruce, Pine, Switchgrass and Timothy

Temp.

(oC)

Methane (mg CH4/ g biomass) Carbon dioxide (mg CO2 / g biomass)

Spruce Pine SW Timothy Spruce Pine SW Timothy

Residence

time 30

mins.

200 0.015 0.02 0.023 0.024 0.017 0.036 0.02 0.043

230 0.018 0.023 0.027 0.028 0.019 0.042 0.028 0.05

260 0.021 0.026 0.032 0.032 0.025 0.05 0.039 0.058

290 0.024 0.03 0.036 0.036 0.031 0.063 0.061 0.068

*Total of 81

mins. 30-800 0.07 0.088 0.110 0.105 0.171 0.284 0.339 0.341

Temp.

(oC)

Carbon monoxide (x 10-2 g CO / g biomass) Ethylene (mg C2H4 / g biomass)


Residence

time 30

minutes

200 0.074 0.11 0.143 0.121 0.041 0.02 0.017 0.02

230 0.087 0.128 0.166 0.141 0.047 0.022 0.02 0.024

260 0.098 0.146 0.189 0.16 0.054 0.025 0.023 0.028

290 0.111 0.164 0.214 0.18 0.06 0.028 0.026 0.031

*Total of 81

mins. 30-800 0.309 0.454 0.575 0.511 0.167 0.074 0.081 0.092

Temp.

(oC)

Ammonia (mg NH3 / g biomass) Oxygen (mg O2 / g biomass)


200 0.035 0.041 0.04 0.042 0.181 0.324 0.258 0.231

230 0.04 0.046 0.046 0.049 0.209 0.36 0.299 0.269

148

Residence

time 30

minutes

260 0.046 0.051 0.053 0.055 0.236 0.393 0.34 0.305

290 0.052 0.057 0.06 0.063 0.262 0.428 0.381 0.342

*Total of 81

mins. 30-800 0.152 0.16 0.18 0.183 0.589 0.916 0.896 0.854

Phenol (mg C6H6O / g biomass) Formaldehyde (mg CH2O / g biomass)

Temp.

(oC) Spruce Pine SW Timothy Spruce Pine SW Timothy

Residence

time 30

minutes

200 0.006 0.007 0.007 0.009 0.034 0.026 0.027 0.027

230 0.007 0.008 0.008 0.011 0.044 0.03 0.032 0.032

260 0.008 0.009 0.009 0.013 0.046 0.034 0.037 0.036

290 0.009 0.01 0.01 0.014 0.047 0.039 0.043 0.042

*Total of 81

mins. 30-800 0.023 0.028 0.029 0.035 0.155 0.125 0.135 0.135

Acetic Acid (mg C2H4O2 / g biomass) Phosphonic Acid (mg CH5O3P / g biomass)

Temp.

(oC) Spruce Pine SW Timothy Spruce Pine SW Timothy

Residence

time 30

minutes

200 0.006 0.008 0.009 0.009 0.005 0.007 0.007 0.009

230 0.007 0.01 0.01 0.011 0.006 0.008 0.008 0.01

260 0.008 0.011 0.012 0.013 0.007 0.009 0.009 0.012

290 0.009 0.013 0.014 0.014 0.008 0.01 0.011 0.013

*Total of 81

mins. 30-800 0.037 0.045 0.05 0.049 0.024 0.028 0.031 0.035

*Note: The continuous torrefaction from room temperature to 800 °C.

Generally, the concentration of the selected gas degradation emitted during torrefaction is

around 20 to 35% of the total concentration for all biomass, but vary in the propositions

when compare between biomass and stages of torrefaction temperatures. For the

continuous torrefaction range of temperature, the degradation of combustible gases (CH4,

CO, C2H2, O2) released by switchgrass (1.662 mg/gswitchgrass) is more than timothy (1.562

mg/gtimothy), pine (1.532 mg/gpine) and spruce (1.135 mg/gspruce). The degradation of

combustible gases during torrefaction temperature 200 °C, 230 °C, and 260 °C by pine

showed the highest (0.474 mg/gpine, 0.533 mg/gpine, and 0.590 mg/gpine, respectively) as

compared to switchgrass (0.441 mg/gswitchgrass, 0.512 mg/gswitchgrass and 0.584 mg/gswitchgrass,

respectively), timothy (0.396 mg/gtimothy, 0.462 mg/gtimothy and 0.525 mg/gtimothy,

respectively) and spruce (0.311 mg/gspruce, 0.361 mg/gspruce, and 0.409 mg/gspruce,

149

respectively). Whilst, for torrefaction temperature 290 °C, the degradation of combustible

gases has similar trend as for continuous torrefaction range of temperature that is

switchgrass took place the highest (0.657 mg/gswitchgrass) as compared to pine (0.650

mg/gpine), timothy (0.587 mg/gtimothy) and spruce (0.457 mg/gspruce). For all species, the

percentages of combustible gases may increases significantly after 300 °C, and thus, loose

the energetic content of the biomass.

Conclusion

Four different types of Canadian biomasses were torrefied using TGA-FTIR analyser and

TGA-Mass spectrometry. Forestry and agriculture biomass showed single and two DTG

peaks respectively during toreffaction. It was observed that the species had different

evolution patterns, which indicates the presence of different chemical functional groups

within the biomass samples. In general, TG-FTIR analysis showed that the torrefaction

behavior of forestry (spruce and pine) biomass were similar. Of all the biomass,

switchgrass experienced considerable weight loss at lower torrefaction temperature. Pine,

timothy and spruce was found to have similar degradation behaviour in terms of chemical

functional groups when torrefied at lower temperature (200 and 230 °C). Degradation of

combustible gases: CH4, CO, C2H2 and O2, during torrefaction at lower temperature: 200

°C and 230 °C is around 20–30 %, whilst at temperature 260 °C and 290 °C is around 30-

45%. Most of the chemical compositions were decomposed at higher torrefaction

temperature of 290 oC. Overall, the gaseous products evolved during torrefaction of

forestry and agricultural biomass showed similar FTIR spectra patterns and in agreement

with mass spectrometric, MS, with regard to the degradation of combustible gases.

Torrefaction temperature played a key role in formation of these gaseous products.

150

Acknowledgment

The authors appreciate the financial assistance from New Brunswick Soil and Crop

Improvement Association, New Brunswick Agricultural Council, Agriculture and Agri-

Food Canada and Universiti Teknologi Petronas, Malaysia.

References

Acharya, B., I. Sule, and A. Dutta. 2012. A review on advances of torrefaction

technologies for biomass processing. Biomass Conversion and Biorefinery 2: 349-369.

Alen, R., R. Kotilainen and A. Zaman. 2002. Thermochemical behavior of Norway spruce

(Piceaabies) at 180-225oC. Wood Science Technology 36:163-171.

Antal, M. J. 1983. Biomass pyrolysis: a review of the literature. Part I – carbonydrate

pyrolysis. Advances in Solar Energy 11: 61–111.

Bassilakis, R., R. M. Carangelo, and M. A. Wójtowicz. 2001. TGFTIR analysis of biomass

pyrolysis. Fuel 80: 1765-1786.

Bridgeman, T. G., J. M. Jones, I. Shield, and P. T. Williams. 2008. Torrefaction of reed

canary grass, wheat straw and willow to enhance solid fuel qualities and combustion

properties. Fuel 87: 844–856.

Chen, W-H. and P-C. Kuo. 2012. Isothermal torrefaction kinetics of hemicellulose,

cellulose, lignin and xylan using thermogravimetric analysis. Energy 36: 6451–6460.

Chen, W-H. and P-C. Kuo. 2010. A study on torrefaction of various biomass aaterials and

its impact on lignocellulosic structure simulated by a thermogravimetry. Energy 35: 2580–

2586.

Demirbas, A. 2009. Pyrolysis Mechanisms of Biomass Materials. Energy Sources Part A:

Recovery, Utilization, and Environmental Effects 31(13): 1186–1193.

Hakkou, M., M. Pétrissans, P. Gérardin, and A. Zoulalian. 2006. Investigations of the

reasons for fungal durability of heat-treated Beech wood. Polymer Degradation Stability

91:393-397.

151

John, C. 2000. Interpretation of Infrared spectra, a practical approach. In Encyclopedia of

Analytical Chemistry, 10815–10837, by R.A. Meyers (Ed.) Chichester: Ó John Wiley &

Sons Ltd.

Mansaray, K.G. and A.E. Ghaly. 1998. Thermal degradation of rice husks in nitrogen

atmosphere. Bioresource Technology 65:13–20.

Melin, S. 2012. Development of the Canadian bulk pellet market. Wood Pellet Association

of Canada REF to FII Agreement #11/12-050.

Miranda, R., C. Sosa-Blanco, D. Bustos-Martinez, and C. Vasile. 2007. Pyrolysis of textile

wastes. I. Kinetics and yields. Journal of Analytical and Applied Pyrolysis 80: 489–495.

Mohan, D., C. U. Pittman, and P.H. Steele. 2006. Pyrolysis of Wood/Biomass for Bio‐oil:

A Critical Review. Energy & Fuels 20 (3): 848–889.

Nguila-Inari, G., M. Pétrissans and P. Gérardin. 2007. Chemical reactivity of heat treated

wood. Wood Science Technology 41:157-168.

Nguila-Inari, G., S. Mounguengui, S. Dumarcay, M. Pétrissans, and P. Gérardin. 2007.

Evidence of char formation during wood heat treatment by mild pyrolysis. Polymer

Degradation and Stability 92:997-1002.

Nuopponen, M., T. Vuorinen, S. Jamsa, and P. Viitaniemi. 2004. Thermal modifications

in softwood studied by FT-IR and UV resonance Raman spectroscopies. Journal of Wood

Chemical Technology 24(1):13-26.

Oladiran, F. and B. Littlefield. 2012. TG-FTIR analysis of Pecan shells thermal

decomposition. Fuel Processing Technology 102: 61–66.

Prins, M. J., K. J. Ptasinski and F. J. J. G. Janssen. 2006. Torrefaction of Wood Part 1:

Weight loss kinetics. Journal of Analytical and Applied Pyrolysis 77: 28–34

Raimie, H., H. Ibrahima, L. I. Darvella, M. Jenny, M. Jonesa, and A. William. 2012.

Physicochemical characterisation of torrefied biomass. Journal of Analytical and Applied

Pyrolysis doi:10.1016/j.jaap.2012.10.004

Shang, L., J. Ahrenfeldt, J. K. Holm, S. Barsberg, R. Zhang, Y. Luo, H. Egsgaard, and U.

B. Henriksen. 2013. Intrinsic kinetics and devolatilization of wheat straw during

torrefaction. Journal of Analytical and Applied Pyrolysis 100: 145–152.

152

Tihay, V. and P. Gillard. 2010. Pyrolysis gases released during the thermal decomposition

of three Mediterranean species. Journal of Analytical and Applied Pyrolysis 88: 168–174.

Tjeerdsma, B. and H. Militz. 2005. Chemical changes in hydrothermal treated wood: FTIR

analysis of combined hydrothermal and dry heat-treated wood. HolzalsRoh- und Werkstoff

63:102–111.

Tumuluru, J. S., S. Sokhansanj, C. T. Wright, and R, D. Boardman. 2010. Biomass

Torrefaction Process Review and Moving Bed Torrefaction System Model Development

INL publication 08,

URL:http://www.inl.gov/technicalpublications/Documents/4737111.pdf [cited 18 March,

2013].

Welzbacher, C., C. Brischke, and A. Rapp. 2007. Influence of treatment temperature and

duration on selected biological, mechanical, physical and optical properties of thermally

modified wood. Wood Material Science Engineering 2:66-76.

Wood, S. M. and D. B. Layzell. 2003. A Canadian Biomass Inventory: Feedstocks for a

Bio-based Economy. BIOCAP Canada Foundation Final Report Industry Canada

Contract # 5006125.

153

Paper VI

Characteristics of Product Gas from Torrefied Biomass Blends




Perak, Malaysia.



International Journal of Chemical and Environmental Engineering, 4(4), 266-272

Abstract

The torrefaction of biomass blends was studied using thermogravimetric analyzer (TGA)

coupled with Fourier transform infrared (FTIR), and mass spectrometer (MS). The

chemical functional groups present in the gases were identified by FTIR and the

quantification of gaseous products was determined using MS at different torrefaction

temperatures ranging from 200 to 290 oC. The thermal degradation characteristics of

biomass blends were almost similar to the parent biomass. FTIR analysis indicated that

biomass blends had similar functional groups but their decomposition was largely at

higher torrefaction temperature of 290 °C. Major product gas (CH4, CO, C2H2, O2)

detected by MS during torrefaction was in the range of 28-30 % (v/v). The composition

of product gas generated from torrefied biomass depends on types of biomass blended and

temperature.

Keywords: Agricultural biomass; Biomass blends; Torrefaction; Mass spectrometry;

Fourier transform infrared spectroscopy

154

Introduction

Canada is one of the countries in the world which generates abundant amount of biomass

from agriculture and forestry sector. Canada has about 42% of forest land and 6.8% of

agricultural land from which about 42% of residue is produced entirely from forestry and

agricultural [1]. Hence, there exists large potential to convert this biomass into valuable

products for various applications. Thermo-chemical treatment is an effective method to

improve the handling of this huge biomass resource. This process has advantages since it

can convert variety of feedstock to value-added products and biofuels.

Torrefaction as one of thermo-chemical method is an emerging process to produce energy

and fuel by improving the biomass properties [2]. In this process, biomass is heated

between temperatures 200 to 300 °C in an inert condition. Torrefaction has been reported

[2] to improve the hygroscopic behavior, decay resistivity, grindability, and higher heating

value. Moreover, it can help to reduce the greenhouse gas emissions. Assuming net

calorific value of 6.2 MWh/tonne, torrefied pellets (4,533,724 metric tons) are estimated

to save CO2 emissions of 21,081,816 tons, if the former is replaced by traditional pellets

(5,736,549 metric tons) [3]. Despite this saving in emissions, the torrefaction behavior of

biomass still needs to be understood.

Various assessments of the potential biomass show that the paramount opportunities for

biomass production lie on agricultural activities [4] for energy production. Perennial

agricultural crops which are associated with higher biomass yields have better net energy

balance and less environmental impact [5, 6]. Reed Canary grass, switchgrass, timothy

hay are among which show better environmental performance than traditional annual

crops. A study showed that switchgrass can displace approximately 8 – 10 times more

greenhouse gas (GHG) emissions compared to corn [7]. However the adaptability of

155

perennial grasses to useful energy efficiency largely depends on the technology and

economics of converting them.

The differences in the lignocellulosic contents and other chemical compounds such as

alkali metal oxide and halogen compound are the main drawback for effective biomass

processing. Fuel from woody biomass has better combustion characteristics. It is

important for the biomass from agricultural land to have comparative chemical

composition similar to biomass from forest in order to increase its value and utilization in

energy production, which may be achieved by blending of both wood and agricultural

biomass.

TGA coupled with FTIR-MS is the most widely used method to observe the thermal

decomposition of materials under various environments. Understanding about the kinetics

and in depth analysis of mass loss with the type of gas products released [8] is essential

for the fundamental knowledge, design, operation and control [9] of torrefaction

equipment and process. Therefore, the influence of process parameters on chemical

composition of the torrefied product has been extensively studied. In torrefaction, the

lignin network is modified [10, 11, 12], and hemicelluloses are strongly degraded [13]

leading to formation of carbonaceous material within the wood [14]. The mass loss during

torrefaction is an indicator to predict the quality and characteristics of biofuel [15, 16].

Some studies were found on torrefaction of biomass using TG analysis to determine the

torrefaction kinetic of cellulose, hemicellulose, lignin and xylan [17, 18] and of various

biomasses (bamboo, willow, coconut shell and wood). Another study reported torrefaction

of deciduous wood (beech and willow), coniferous wood (larch) and straw using TG

analyser [19]. Very recently, a study on torrefaction of wheat straw but using TGA/DSC

configuration mode was performed [20]. From the literature, torrefaction of biomass was

156

mostly carried out using TGA and FTIR individually. The work on torrefying biomass

using coupled TGA-FTIR-MS is very scarce. There was also no literature found on

torrefaction of mixed agriculture (switchgrass and timothy) and forestry (spruce and pine)

biomass using TG-FTIR-MS. Keeping this in view, it is anticipated to contribute new

knowledge about the characteristics of the gas evolved during torrefaction of selected

biomass blends. The product gas was analyzed at different temperatures.

Materials and Method

Four types of biomass were selected for torrefaction. Two were agriculture (switchgrass

and timothy) and other two were forestry (spruce and pine). The timothy and switchgrass

bales were provided by NB Department of Agriculture. The woody biomass was acquired

from a local sawmill. Prior to the analysis, biomasses were first dried in the oven at 105

oC for 24 h, and then grounded using Wiley mill to about 500 μm particle size. The

biomass blending ratio was kept constant at 1:1 weight percent. Proximate analysis was

determined using TGA analysis, and the ash content was investigated as per ASTM

Standard Method E1755-01.

TG-FTIR-MS Experiment

The selected biomass was torrefied in TGA Q500 TA Instrument. Approximately 10 mg

of sample was loaded in a TGA crucible. The temperature was initially raised to 105 oC

at heating rate of 20 oC/min and held for 5 min in order to remove moisture. Further the

temperature was ramped to torrefaction temperature using same heating rate. The inert

atmosphere was maintained using nitrogen gas at a flow rate of 100 ml/min. Vapors

released from torrefied biomass were passed to the FTIR spectrometer (Nicolet 6700 FTIR

Thermo Scientific) through a transfer tube, which was heated to about 200 oC to prevent

157

the condensation of vapors on the tube wall. The IR spectra were recorded with a temporal

resolution about 2 s. The resolution of the collected spectra was set to be 1 cm−1 and the

spectral range was set from 4000 to 400 cm−1. At the carrier gas flow rate of 100 ml/min,

it took about 50 s for the vapors to reach the FTIR cell from the thermogravimetric

analyzer with a time delay of about 50 s.

After scanning through FTIR the volatile products were further transferred to the

ionization source of the mass spectrometer (SRS RGA 200) in order to determine the

concentration of product gas. The transfer line was heated and maintained at a temperature

of 200ºC to prevent the condensation of the volatile gases released during the torrefaction

process. The mass spectrometer is operated at electron energy of 70 eV. The gases were

analysed by using the RGA software. A scan of the m/z was carried out from 1 to 100 amu

to determine which m/z has to be followed during the TG experiments. The intensities of

10 selected ions (m/z = 15, 16, 27, 28, 29, 32, 43, 44, 94, and 96) were monitored with the

thermogravimetric parameters. An absolute quantification in this work was performed for

each monitored gas: CH4, CO2, CO, C2H4, NH3, O2, C6H6O, CH2O, C2H4O2 and CH5O3P.

The concentrations of CH4, CO2, CO, C2H4, NH3, O2, C6H6O, CH2O, C2H4O2 and CH5O3P

were directly deduced from the partial pressure of gases by using (1).

[Ai] = (Pi / PT) * 1 000 000 (1)

Where, [Ai] is the concentration of the gas A, Pi is the partial pressure for m/z = i

representing of gas A and PT is the total pressure for all species i.

158


Proximate Analysis

Table VI.1 shows the proximate analysis of individual and blended biomass. Biomass

blends showed lower moisture content compared to original biomass suggesting them as

a good candidate for torrefaction process. Timothy showed highest amount of volatile

matter and lowest fixed carbon content compared to other selected biomass. However,

agricultural biomass contains high amount of ash. Similar results were reported in

literature [21, 22] for agricultural and forestry biomass. The proximate data of the forestry

biomass (spruce and pine) was almost similar.

Blending two or more biomasses can increase the supply chain of the materials as well as

the economy for the farmers. Based on proximate analysis (Table VI.1), blending

agricultural and forestry biomass at ratio 1:1 improves the chemical properties of

individual biomass. Particularly, timothy blend with pine has shown significant increase

in fixed carbon. The blended biomass showed lower ash content compared to individual

agricultural biomass. Similarly, the volatile content of all selected biomass were also

influenced by blending.

159

Table VI.1: Proximate Analysis of Selected Biomass and Their Blends (% Weight

Dry)

Moisture Volatile matter Fixed Carbon Ash

(wt %)

Spruce 2.92 73.75 23.33 trace

Pine 4.63 73.05 22.32 trace

Switchgrass (A) 2.23 68.46 25.70 3.61

Timothy (B) 2.82 79.55 13.57 4.06

Blends (1:1)

Spruce + (A) 1.39 71.07 26.01 1.53

Spruce + (B) 1.51 76.08 20.90 1.51

Pine + (A) 1.20 72.25 25.00 1.55

Pine + (B) 1.23 76.66 20.49 1.62

TG-DTG Analysis

Fig. VI.1 shows the thermogravimetric curve for selected biomass. The thermal

degradation profile of spruce and pine was almost similar unlike the switchgrass and

timothy. However, the degradation behavior of blended biomass was close to that of

forestry biomass. The initial decrease in the weight loss (1-5 wt %) of biomass up to

temperature 150 °C was attributed to removal of moisture content. The initial degradation

temperature for spruce, pine, and blended samples (spruce-timothy and pine-timothy) was

about 210 °C. Whilst for switchgrass and its blend (spruce-switchgrass and pine-

switchgrass) was about 175 °C. The degradation of switchgrass and its blends was found

to be earlier than other selected biomass and their blends. Moreover, the final degradation

160

temperature for spruce, pine, switchgrass and timothy was about 375 °C, 375 °C, 360 °C

and 380 °C, respectively and that for blended biomass was about 375 °C.

*S=spruce, P=pine, SW=switchgrass, T=timothy

Figure VI.1: Thermo-diagram of Mass Loss, Volatiles and Char

Between temperatures 200 to 500 °C higher decomposition of cellulose and hemicellulose

might have occurred, while small amount of lignin degradation could have taken place.

Similar observation was also reported by Chen [17]. No significant weight loss was

observed after temperature 500 °C. The residual or biochar for spruce, pine, switchgrass

and timothy was about 16, 15, 21 and 11 wt %, respectively at 800 °C, while blended

biomass showed 14 – 17 wt%. However, at torrefaction temperature of 300 °C,

switchgrass showed highest weight loss of about 30 wt% as compared to spruce (17 wt%),

pine (19 wt%), timothy (20 wt%), pine-switchgrass (21 wt%), spruce-switchgrass (23

wt%), pine-timothy (17 wt%) and spruce-timothy (17 wt%). The weight loss of timothy

0

10

20

30

40

50

60

70

80

90

100

0 100 200 300 400 500 600 700 800

Wei

ght

loss

(%

)

Temperature ( °C)

S

P

SW

T

S+SW

P+SW

S+T

P+T

moistur

volatile matter

moisture

161

blends was lower than switchgrass blends at torrefaction temperature

300 °C.

The derivative weight loss of individual and blended biomass is presented in Fig. VII.2.

The agricultural biomass showed two distinct peaks compared to one peak in forestry

biomass. The first peak is generally attributed to decomposition of hemicellulose and

second due to decomposition of cellulose. The first and the second peak for switchgrass

were found at temperature 230 and 330 °C, respectively and in case of timothy it was at

temperature 310 and 355 °C, respectively.

The DTG peak for forestry biomass (spruce and pine) was at temperature 350 °C.

Detection of single peak for forestry biomass might be due to overlapping reactivity of

hemicellulose and cellulose. A kinetic study on willow and agricultural biomass was

found that willow has less hemicellulose content compared to reed canary grass and wheat

straw [25]. According to a previous report [24], the thermal decomposition of

hemicellulose, cellulose and lignin occurs at temperatures in the range of 150-350 °C,

275–350 °C, and 250-500 °C, respectively.

Interestingly, biomass blends showed two distinct DTG peaks as shown in Fig. VI.2.

However, the reactivity of switchgrass blends was observed to be lower than timothy

blends. It is also clear from the result that blending the biomass can alter the degradation

rate of the materials. The DTG peaks for all blended biomass was at temperature 360 °C.

The temperature at which the first DTG peak occurred for switchgrass (230 °C) and

timothy (310 °C) blends was almost similar to their individual parent agricultural biomass.

63

162


Figure VI.2: DTG Curves of Biomass

FTIR Analysis

Fig. VI.3-6 illustrates the possible chemical functional groups present in the torrefied

biomass. The FTIR spectra were collected at different torrefaction temperatures (200, 230,

260 and 290 °C). Overall, the chemical functional groups detected in both individual and

blend biomass were similar in pattern. The intensity of the peaks for different chemical

functional groups increased with the rise in temperature from 200 to 290 °C. This shows

that at higher torrefaction temperature (290-300 °C) the rate of biomass degradation

increases. The band around 3600–3400 cm−1 is due to hydroxyl (O-H) groups, which was

0.0

0.2

0.4

0.6

0.8

1.0

1.2

30 120 210 300 390 480 570 660 750

Der

ivat

ive

Wei

gh

t lo

ss (

%/

°C)

Temperature ( °C)

Spruce (S)

Pine (P)

Switchgrass

(Sw)Timothy (T)

0.750

(S+SW)

0.752

(P+SW)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

30 120 210 300 390 480 570 660 750

Der

ivat

ive

wei

gh

t lo

ss (

%/

°C)

Temperature (° C)

SW

S+SW

P+SW

1.085 (S+T)

1.072 (P+T)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

30 120 210 300 390 480 570 660 750

Der

ivat

ive

wei

gh

t lo

ss (

%/

°C)

Temperature (° C)

T

S+T

P+T

165

163

found in all torrefied biomass. The C-H stretch band around 2970-2780 cm-1 indicates the

presence of an organic compound or methane. A sharp band around wave number 2300-

2200 cm-1 is due to the formation of CO2 and CO generated during torrefaction of biomass.

It was observed that two major bands in the region of 1740–1710 cm−1 and 1250–1220

cm−1, were because of C–O and C–O–C functional groups corresponding to alkyl and

aromatic rings, respectively. Organic phosphates (P–O) functional groups can be found in

range of 1350–1250cm-1 stretch band, and aliphatic phosphates (P-O-C) functional group

is more obvious at around 1050-990 cm-1 than organic phosphates (P-O) functional

groups. A very sharp peak around 730-550 cm-1 might be due to carbon chlorides (C-Cl)

compound [26].

The change in transmittance intensities indicates a variation in the gas concentration [27].

Fig. VI.3 shows the FTIR spectra of blended biomass at temperature 200 °C. Very

minimal or no formation of methane (C-H functional group) was seen from blended

biomass at temperature of 200 °C. Carbon dioxide (CO2), and carbon monoxide (CO)

emits in the temperature range of 150 to 300 ºC [28].

Figure VI.3: Decomposition of Functional Groups at Torrefaction Temperature

200 °C

400100016002200280034004000wave number (cm-1)

S+T S+SW P+SW P+T

O-H C-H

CO2, CO

C-C C-O-C O-P

C-Cl

164

The intensity of the product gases detected at temperature 200 °C from blended biomass

was very small. The intensity of the product gases released from blended biomass

increased as the torrefaction temperature was increased from 200 to 290 °C. At 260 °C,

more alkyl (C-C) chemical compounds were released because of decomposition of

cellulosic materials in biomass. Hence, the reaction of hemicellulose and cellulose can

play a major role in torrefaction temperature (230 to 300 °C). In contrast to this, lignin is

expected to contribute less in the torrefaction reaction mechanism since it decomposes at

higher temperature.


230 °C


260 °C

400100016002200280034004000Wave number (cm -1)

S+T S+SW P+SW P+T

O-HC-H

CO2, CO

C-C C-O-C

400100016002200280034004000Wave number (cm -1)

S+T S+SW P+SW P+T

O-HC-H

165

In general, the intensity of the chemical compounds from the biomass torrified at

temperature 200 ºC, 230 ºC and 260 ºC was not significant because of lower reaction

activity. However, once the temperature reached above 260 ºC, i.e. 290 ºC, major

transformation in the intensity of chemical compounds was observed as shown in

Fig. VI.6. Thus, higher torrefaction temperature can result in total decomposition of

biomass with release of heavy volatile compounds.


290 °C

Analysis of Product Gas by MS

Fig. VI.7 presents the concentration of gases from blended biomass from 200 °C to

500 °C. The concentration of product gases was depended on types of biomass blended

and the temperature. Oxygen, carbon monoxide and carbon dioxide was found to be the

major gas released from torrified biomass. The presence of high concentration of oxygen

in the product gas might be due to breaking of highly oxygenated chemical compounds

present in the biomass in the form of hemicellulose and cellulose.

400100016002200280034004000

Wave number (cm -1)

S+T S+SW P+SW P+T

166

The concentration of CO2 increases beyond temperature 300 °C and it almost become

constant at 500 °C. At any temperature, spruce and timothy blend released higher amount

of CO2. CO was found in range of 10 to 12 % for torrefaction temperature of 200 °C and

300 °C. The amount of CH4 increased gradually with increase in temperature. During

torrefaction (temperature around 300 °C) no major product gases (methane, ethylene,

phenol, etc.) are evolved due to low degradation temperature of biomass.


Figure VI.7: Product Gases from Selected Biomass and Their Blends at

Temperature from 200 °C to 500 °C

Table VI.2 shows the concentration of main gases evolved from blended biomass at

different torrefaction temperatures. It is clear that the concentration of O2, CO, and CO2

0

4

8

12

16

20

S+SW S+T P+SW P+T

200 °C Phenol

Ethylene

Methane

0

4

8

12

16

20

S+SW S+T P+SW P+T

300 °CPhenol

Ethylene

Methane

0

4

8

12

16

20

S+SW S+T P+SW P+T

400 °CPhenol

Ethylene

Methane

0

4

8

12

16

20

S+SW S+T P+SW P+T

500 °CPhenolEthyleneMethaneCarbon dioxide

167

gases is highest followed by formaldehyde, ammonia, methane, and ethylene. Acetic acid,

carbolic acid, and phosphonic acid showed the lowest concentration. Tihay and Gillard

[28] identified H2O, CO, CO2 as the main gases released from biomass. Since at higher

temperature the oxygen reacts with other chemical compounds to form high molecular

weight chemicals due to the pyrolysis reactions between the vapors, the concentration of

oxygen for all blended biomass decreased with increase in torrefaction temperature.

Pine and timothy biomass blend showed higher amount of CO2, methane, acetic acid than

other blends. On other hand, highest amount of ethylene was observed in case of spruce

and timothy blends while switchgrass and spruce blends gave the highest concentration

of phenol. The C-H functional group in the FTIR spectrum confirms the presence of

methane (CH4) and ethylene (C2H4) gases from the terrified biomass. The total

concentration of product gases (CH4, CO, C2H2, O2) generated during torrefaction was

about 28 – 30 % (v/v). The difference in concentration of gases was not significant for

different types of blended biomass. This indicates that blending different biomass can

have almost similar degradation characteristics. The concentration of product gases

increases significantly after temperature 300 °C, and thus, the energy content of the

biomass may be lost.

168

Table VI.2: Concentration of Gases Generated during Torrefaction for Blended

Biomass

Concentration (%) at torrefaction temperature 200 °C

O2 CO CO2 CH2O NH3 CH4 C2H4 C2H4O2 C6H6O H3PO3

P+T 15.80 11.20 7.77 3.27 2.89 1.22 0.29 0.29 0.24 0.23

P+SW 17.76 11.93 2.39 5.01 2.03 0.10 0.09 0.13 0.39 0.30

S+T 16.24 11.71 3.30 3.53 2.32 0.65 0.37 0.09 0.30 0.18

S+SW 17.10 12.30 1.70 3.73 1.72 0.07 0.02 0.16 0.40 0.34


P+T 15.80 11.20 7.87 3.34 2.95 1.30 0.32 0.32 0.23 0.21

P+SW 17.75 11.63 1.84 3.67 1.46 0.09 0.07 0.09 0.30 0.23

S+T 16.05 11.6 3.82 3.62 2.36 0.74 0.42 0.15 0.26 0.14

S+SW 17.00 12.20 1.86 3.80 1.76 0.11 0.02 0.12 0.37 0.31


P+T 15.60 11.10 8.09 3.46 3.02 1.44 0.35 0.39 0.21 0.19

P+SW 16.75 11.54 1.58 2.60 1.04 0.16 0.06 0.04 0.21 0.16

S+T 15.97 11.52 4.41 3.77 2.43 0.82 0.44 0.2 0.25 0.14

S+SW 16.90 12.30 2.33 3.93 1.80 0.15 0.05 0.09 0.37 0.31


P+T 15.40 10.90 8.63 3.72 3.24 1.61 0.44 0.55 0.19 0.17

P+SW 16.39 11.43 1.98 2.54 1.00 0.46 0.09 0.06 0.18 0.14

S+T 15.64 11.38 5.24 3.97 2.5 0.95 0.47 0.31 0.26 0.16

S+SW 16.60 12.10 3.10 4.10 1.86 0.29 0.07 0.01 0.38 0.32

Conclusion

Four different types of Canadian agricultural and forestry biomass blends were torrefied

using TG-FTIR-MS method. Agricultural and blended biomass showed two distinct DTG

reaction peaks as compared to forestry biomass. The total concentration of product gases

(CH4, CO, C2H2, O2) generated during torrefaction is about 28 – 30 % (v/v). The major

chemical transformation in the blended biomass takes place at torrefaction temperature of

169

290 °C. Thus, in order to retain the energetic value of torrified biomass the temperature

should be kept below 290 °C. In conclusion, both temperature and biomass blends were

found to be the key factors influencing the concentration and formation of product gases.

Acknowledgment

The authors appreciate the financial assistance from New Brunswick Soil and Crop

Improvement Association, New Brunswick Agricultural Council, Agriculture and Agri-

Food Canada and Universiti Teknologi Petronas, Malaysia.

References

[1] Wood, S. M. and Layzell, D. B. (2003) A Canadian biomass inventory: feedstocks

for a bio-based economy. BIOCAP Canada Foundation Final Report Industry

Canada Contract #: 5006125.

[2] Acharya, B., Sule, I., and Dutta, A. (2012). A review on advances of torrefaction

technologies for biomass processing. Biomass Conversion and Biorefinery,

2, 349-369.

[3] Melin, S. (2012). Development of the Canadian bulk pellet market. Wood Pellet

Association of Canada REF to FII Agreement #:11/12-050:2012

[4] Berndes, G., Hoogwijk, M., and Van den Broek, R. (2003). The contribution of

biomass in the future global energy supply: a review of 17 studies. Biomass and

Bioenergy, 25, 1–28.

[5] Börjesson, P. I. I. (1996). Energy analysis of biomass production and transportation.

Biomass and Bioenergy, 11, 305–318.

170

[6] Biewinga E.E., and Van der Bijl, G. (1996). Sustainability of energy crops in Europe

– a methodology developed and applied. Utrecht, CLM: Centre for Agriculture and

Environment, 234

[7] Samson R., (2008). Analyzing biofuel options: Greenhouse Gas mitigation

efficiency and costs. Resource Efficiency Agricultural Production (REAP), Canada.

[8] Fassina, O., and Littlefield, B. (2012). TG-FTIR analysis of Pecan shells thermal

decomposition. Fuel Processing Technology. 102, 61–66.

[9] Miranda, R., Sosa-Blanco, C., Bustos-Martinez, D., and Vasile, C. (2007). Pyrolysis

of textile wastes. I. Kinetics and yields. Journal of Analytical and Applied Pyrolysis,

80, 489–495.

[10] Alen, R., Kotilainen, R., and Zaman, A. (2002). Thermochemical behavior of

Norway spruce (Piceaabies) at 180-225oC. Wood Science Technology. 36, 163-171.

[11] Tjeerdsma, B. and Militz, H. (2005). Chemical changes in hydrothermal treated

wood: FTIR analysis of combined hydrothermal and dry heat-treated wood.

HolzalsRoh-und Werkstoff. 63, 102–111.

[12] Nguila-Inari, G., Pétrissans, M., and Gérardin, P. (2007). Chemical reactivity of

heat treated wood. Wood Science Technology. 41, 157-168.

[13] Nuopponen, M., Vuorinen, T., Jamsa, S., and Viitaniemi, P. (2004). Thermal

modifications in softwood studied by FT-IR and UV resonance Raman

spectroscopies. Journal of Wood Chemical Technology. 24(1), 13-26.

[14] Nguila-Inari, G., Mounguengui, S., Dumarcay, S., Pétrissans, M., and Gérardin, P.

(2007). Evidence of char formation during wood heat treatment by mild pyrolysis.

Polymer Degradation and Stability. 92, 997-1002.

171

[15] Hakkou, M., Pétrissans, M., Gérardin, P., and Zoulalian, A. (2006). Investigations

of the reasons for fungal durability of heat-treated Beech wood. Polymer

Degradation Stability, 91, 393-397.

[16] Welzbacher, C., Brischke, C., and Rapp, A. (2007). Influence of treatment

temperature and duration on selected biological, mechanical, physical and optical

properties of thermally modified wood. Wood Material Science Engineering,

2, 66-76.

[17] Chen, W-H., and Kuo, P-C. (2012). Isothermal torrefaction kinetics of

hemicellulose, cellulose, lignin and xylan using thermogravimetric analysis.

Energy. 36, 6451–6460.

[18] Chen, W-H. and Kuo, P-C. (2010). A study on torrefaction of various biomass

aaterials and its impact on lignocellulosic structure simulated by a

thermogravimetry Energy, 35, 2580–2586

[19] Prins, M. J., Ptasinski, K. J., and Janssen, F. J. J. G. (2006). Torrefaction of Wood

Part 1: Weight loss kinetics. Journal of Analytical and Applied Pyrolysis,

77: 28–34

[20] Shang, L., Ahrenfeldt, J., Holm, J. K., Barsberg, S., Zhang, R., Luo, Y., Egsgaard,

H., and Henriksen, U. B. (2013). Intrinsic kinetics and devolatilization of wheat

straw during torrefaction. Journal of Analytical and Applied Pyrolysis, 100, 145–

152.

[21] Raimie, H., Ibrahima, H., Darvella, L. I., Jenny, M., Jonesa, M., and William, A.

(2012). Physicochemical characterisation of torrefied biomass. Journal of

Analytical and Applied Pyrolysis, doi:10.1016/j.jaap.2012.10.004

172

[22] James, A.K., Thring, R.W., Helle, S., and Ghuman, H.S. (2012). Ash Management

Review—Applications of Biomass Bottom Ash. Energies, 5, 3856-3873.

[23] Antal, M. J. (1983). Biomass pyrolysis: a review of the literature. Part I –

carbonydrate pyrolysis. Advances in Solar Energy, 11, 61–111.

[24] Mansaray, K.G., and Ghaly, A. E. (1998). Thermal degradation of rice husks in

nitrogen atmosphere. Bioresource Technology 65, 13–20.

[25] Bridgeman, T. G., Jones, J. M., Shield, I., and Williams, P. T. (2008). Torrefaction

of reed canary grass, wheat straw and willow to enhance solid fuel qualities and

combustion properties. Fuel, 87, 844–856.

[26] John, C. (2000). Interpretation of Infrared spectra, a practical approach. In R. A.

Meyers (Ed.), Encyclopedia of Analytical Chemistry (10815–10837) Chichester: Ó

John Wiley & Sons Ltd.

[27] Bassilakis, R., Carangelo, R. M., and Wójtowicz, M. A. (2001). TG-FTIR analysis

of biomass pyrolysis. Fuel, 80, 1765-1786.

[28] Tihay, V., and Gillard, P. (2010). Pyrolysis gases released during the thermal

decomposition of three Mediterranean species. Journal of Analytical and Applied

Pyrolysis, 88, 168–174.

173

Curriculum Vitae

Candidate’s full name: Noorfidza Yub Harun

Universities attended:

University of New Brunswick, 2012-2016, PhD, Mechanical Engineering.

University of New Brunswick, 2012, Diploma in University Teaching.

University of New Brunswick, 2009-2011, MSc, Forest Engineering.

University of Sheffield, 1999-2001, MSc, Environmental Energy Engineering.

National University of Malaysia, BEng, Chemical and Process Engineering.

Publications:

Harun, N. Y. and Afzal, M. T. (2016): Effect of Particle Size on Mechanical Properties

of Pellets Made from Biomass Blends. Procedia Engineering, 148, 93–99

Harun, N. Y. and Afzal, M. T. (2015): Combustion Behavior and Thermal Analysis of

Agricultural and Woody Biomass Blends. Advances in Environmental Biology, 9 (15),

34-40

Harun, N. Y. and Afzal, M. T. (2015): Chemical and Mechanical Characteristics of

Biomass Pellets made from Agricultural and Woody Biomass Blends. Transactions of

the ASABE, 58(4), 921-930

Noorfidza Yub Harun and Muhammad T. Afzal, “Torrefaction of Agriculture and

Forestry Biomass Using TGA-FTIR-MS”, Dincer et al. (eds.), Progress in Exergy,

Energy, and the Environment, DOI 10.1007/978-3-319-04681-5_77, # Springer

International Publishing Switzerland 2014

Harun, N. Y. and Afzal, M. T. (2013): Characteristics of Product Gas from Torrefied

Biomass Blends. International Journal of Chemical and Environmental Engineering,

4(4), 266-272

N. Yub Harun and M.T. Afzal, “Kinetic Characteristics of Pyrolysis Balsam Fir

Needles”, Journal of Material and Science Engineering, 5(2011) 489-503

N.Yub Harun, and M.T. Afzal, “Thermal Decomposition Kinetics of Forest Residue”,

Journal of Applied Sciences, 10(12): 1122-1127, 2010

N.Yub Harun, M.T. Afzal and M.T. Azizan, “TGA Analysis of Rubber Seed Kernel”,

CSC Journal - International Journal of Engineering (IJE), Volume (3): Issue (6), 639-

652 (2010) ISSN (Online) :1985-2312



174

N.Yub Harun, M.T. Afzal & N. Shamsudin, “Reactivity Studies of Sludge and Biomass

Combustion”, CSC Journal - International Journal of Engineering (IJE), Volume (3):

Issue (4), 413-425 (2009) ISSN (Online) :1985-2312

Conference Presentations:

Harun, N. Y. and Afzal, M. T. (2016). Effect of Particle Size on Mechanical Properties of

Pellets Made from Biomass Blends. 4th International Conference on Process Engineering

and Advanced Materials (ICPEAM 2016), Kuala Lumpur, Malaysia, 15-17 August 2016

Harun, N. Y. and Afzal, M. T. (2015). Combustion Behavior and Thermal Analysis of

Agricultural and Woody Biomass Blends. International Postgraduate Conference

Agriculture, Biology and Environment (IPCABE 2015), Jakarta, Indonesia, 31 July – 1

August 2015

Harun, N. Y. and Afzal, M. T. (2014). Chemical and Mechanical Properties of Pellets

Made from Agricultural and Woody Biomass Blends. 2014 ASABE and CSBE/SCGAB

Annual International Meeting Paper, Montreal, Canada. July 13-17, 2014. Paper Number:

1908650

Harun, N. Y. and Afzal, M. T. (2013). Characteristics of Product Gas from Torrefied

Biomass Blends. 2nd International Renewable Energy and Environment Conference

(IREEC-2013), Kuala Lumpur, Malaysia, 4-6 July, 2013.

Harun, N. Y. and Afzal, M. T. (2013). Torrefaction of Agriculture and Forestry Biomass

Using TGA-FTIR-MS. Proceeding of the Sixth International Exergy, Energy and

Environment Symposium (IEEES-6), Italy, 1-4 July 2013, p317-325.

N.Yub Harun, and M.T. Afzal, “Thermal Decomposition Kinetics of Forest Residue”,

Accepted by International Conference of Process Engineering and Advanced

Material/Symposium of Malaysian Chemical Engineers (ICPEAM2010/ SOMCHE2010)

on June 15-17, 2010

N.Y. Harun, M.T. Afzal, “Pretreatment of Bagasse to Improve Fuel Quality via

Torrefaction”, Canadian Society for Bioengineering Annual General Meeting and

Technical Conference (CSBE-SCGAB July 12-15, 2009), Prince-Edward-Island, Canada.

N.Y. Harun, M.T. Afzal and N. Shamsuddin, “Reactivity Studies of Sludge and Biomass

Combustion”, Canadian Society for Bioengineering Annual General Meeting and

Technical Conference (CSBE-SCGAB July 12-15, 2009), Prince-Edward-Island, Canada.

N. Y. Harun, C. S. Y. Thor, M. Z. O. Zakaria, S. Yusop, A. S. Azmi, H. Zabiri, Prediction

of CO Forecasting in Ipoh Using Neural Network, Asia Pacific Conference on Ecological

Modeling (ECOMOD 2007).

175

N. Y. Harun, A. Ahmad, A.A Omar, A. S. Azmi, Leachability Studies of Lead and

Strength in Solidified/Stabilised Petroleum-based Sludge Waste, Asia Pacific Conference

on Ecological Modeling (ECOMOD 2007).

H. Zabiri, M. Ramasamy, D. D. Doanh, N. M. Ramli, S. Yusop, N. Y. Harun, Oscillation

detection in multivariable systems using principal component analysis, Symposium of

Malaysian Chemical Engineers (SOMChE 2007).

N.Y. Harun, N. Maarof, Effects of Fly Ash Addition On The Properties of Concrete and

Mortar, International Conference on Environmental (CENV 2006).

N.Y. Harun, K.M. Sabil, A. Adnan, F. Taha, Evaluation Of Oil Palm Shell And Rice

Husks Activated Carbons As Potential Adsorbent For Methylene Blue (Mb) Dye

Removal, Symposium of Malaysian Chemical Engineers (SOMChE 2004).

S. Yusup, S. L. Mustapa, A.A Omar, N.Y. Harun, Comparison of Continuous Stirred Tank

Reactor and Batch Reactor Performance for Saponification Process, Symposium of

Malaysian Chemical Engineers (SOMChE 2004).