development and characterization of new feedstock …
TRANSCRIPT
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
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
Results and Discussion ............................................................................................... 120
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
Results and Discussion ............................................................................................... 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
Results and Discussion ............................................................................................... 158
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
The authors appreciate the financial assistance from New Brunswick Department of
Agriculture, Aquaculture and Fisheries; New Brunswick Soil and Crop Improvement
Association; New Brunswick Agricultural Council; Agriculture and Agri-Food Canada;
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
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.
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
Proximate and Ultimate Analysis
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.
Results and Discussion
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
The authors appreciate the financial assistance from New Brunswick Department of
Agriculture, Aquaculture and Fishries; New Brunswick Soil and Crop Improvement
Association; New Brunswick Agricultural Council; Agriculture and Agri-Food Canada;
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
temperature, moisture content, and species for the conversion process of wood
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
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.
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.
Results and Discussion
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
RCG H SW S P S+RCG P+RCG S+H P+H S+SW P+SW
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
The authors appreciate the financial assistance from New Brunswick Department of
Agriculture, Aquaculture and Fishries; New Brunswick Soil and Crop Improvement
Association; New Brunswick Agricultural Council; Agriculture and Agri-Food Canada;
New Brunswick Innovation Foundation; and Universiti Teknologi PETRONAS,
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
temperature, moisture content, and species for the conversion process of wood
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
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.
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
RCG H SW S P S+RCG P+RCG S+H P+H S+SW P+SW
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
Proximate and Ultimate Analysis
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
Results and Discussion
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.
The authors appreciate the financial assistance from New Brunswick Department of
Agriculture, Aquaculture and Fishries; New Brunswick Soil and Crop Improvement
Association; New Brunswick Agricultural Council; Agriculture and Agri-Food Canada;
New Brunswick Innovation Foundation; and Universiti Teknologi PETRONAS,
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
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.
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.
Results and Discussion
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
Torrefaction 230 °C
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.
Figure V.4: Decomposition of Functional Groups at Torrefaction Temperature
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.
Torrefaction 260 °C
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.
Figure V.5: Decomposition of Functional Groups at Torrefaction Temperature
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.
Torrefaction 290 °C
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.
Figure V.6: Decomposition of Functional Groups at Torrefaction Temperature
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)
Spruce Pine SW Timothy Spruce Pine SW Timothy
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)
Spruce Pine SW Timothy Spruce Pine SW Timothy
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
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.
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
Results and Discussion
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
*S=spruce, P=pine, SW=switchgrass, T=timothy
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.
Figure VI.4: Decomposition of Functional Groups at Torrefaction Temperature
230 °C
Figure VI.5: Decomposition of Functional Groups at Torrefaction Temperature
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.
Figure VI.6: Decomposition of Functional Groups at Torrefaction Temperature
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.
*S=spruce, P=pine, SW=switchgrass, T=timothy
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
Concentration (%) at torrefaction temperature 230 °C
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
Concentration (%) at torrefaction temperature 260 °C
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
Concentration (%) at torrefaction temperature 290 °C
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).