Prediction of Heating and Ignition Properties of Biomass Dusts Using Near 
Infrared Spectroscopy 
 
by 
 
Jaskaran Dhiman 
 
 
 
 
A thesis submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Master of Science 
 
Auburn, Alabama 
August 02, 2014 
 
 
 
 
Keywords: biomass, bioenergy, combustible dusts, dust ignition, NIR 
spectroscopy, thermogravimetric analysis, differential scanning calorimetry 
 
 
Copyright 2014 by Jaskaran Dhiman 
 
 
Approved by 
 
Oladiran O. Fasina, Chair, Professor of Biosystems Engineering 
Brian K. Via, Associate Professor of Forest Products Development Center 
Sushil Adhikari, Associate Professor of Biosystems Engineering 
Timothy P. McDonald, Associate Professor of Biosystems Engineering 
 
 
 
  
ii 
 
Abstract  
Dusts (i.e. particles of size less than 500 ?m) are generated during handling and 
processing of biomass feedstock. Similar to damages that have been reported from 
ignition of dusts obtained from industries, ignited biomass dust may potentially cause fire 
and explosion in biorefinery plants that can result in human fatalities, serious injuries and 
substantial monetary loss. Control measures to prevent the heating and ignition of 
biomass dusts will play a critical role in development of safety guidelines and standards 
for bio-based industries. The research aims at quantifying and predicting (using NIR 
spectroscopy) the heating and ignition properties of dust from ten biomass feedstocks. 
Three different types of coals were also used for comparison purposes. The range of 
values obtained for these properties were 240?C-335?C (minimum hot surface ignition 
temperature, MIT), 266?C-448?C (temperature of onset of rapid volatilization, TORV), 
304?C-485?C (temperature of maximum rate of mass loss, TMML), 242?C-423?C 
(oxidation temperature, TOXY), 206?C-249?C (temperature of onset of rapid exothermic 
reaction, TRE) and 354?C-429?C (maximum temperature reached during exothermic 
reaction, TME). Coefficient of determination (R2) values for internal validation of prediction 
models developed using PCA on raw NIR spectral data for MIT, TORV, TMML, TOXY, 
TRE and TME were 0.994, 0.984, 0.963, 0.737, 0.931 and 0.901 respectively, whereas, 
first derivative NIR spectral data yielded R2 (calibration) for these properties as 0.976, 
0.964, 0.943, 0.798, 0.923 and 0.895 respectively.     
  Four different biomass dusts (eucalyptus, pine, sweetgum and switchgrass) were 
used to validate the prediction models externally. Coefficient of determination (R2) values 
  
iii 
 
for all the models was obtained less than 0.28. Poor performance of models under external 
validation was attributed to small sample sizes of the feedstocks that were used during 
building of prediction models. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
iv 
 
Acknowledgments 
 
 I would like to thank my parents, Dr. J.S. Dhiman and Mrs. Manjeet Kaur Dhiman 
for their unconditional support and care throughout my life. They were a great source of 
inspiration to me to study further. I will always be indebted to them. I would also like to 
thank my brother Mankaran Dhiman for encouraging me. I wish to thank my friend Aman 
for supporting me and being patient during my course of study.  
 I would like to express my sincere gratitude to my advisor Dr. Oladiran Fasina for 
his patience, motivation, guidance and continuous support of my Masters research. I am 
highly indebted by his guidance in writing of this thesis. I would like to thank my research 
committee members: Dr. Brian Via, Dr. Sushil Adhikari and Dr. Timothy McDonald for their 
encouragement, support and insightful comments. I would also like to thank Christian 
Brodbeck, Jonathan Griffith and James for their support in procurement of raw material 
for my research. I also wish to thank my lab mates: Oluwatosin, Gbenga, Gurdeep and 
Anshu for stimulating discussions and hours of working together. I am very thankful to my 
friends Jatinder, Raman, Gurjot, Jass, Deep, Manbir, Gurjeet and Roger for their 
encouragement. 
 I would like to thank Alabama Agricultural Experiment Station, United States 
Department of Agriculture (USDA) funded Southeast Integrated Biomass Supply System 
(IBSS) and Department of Energy (DOE) for providing infrastructure and funding for my 
research. 
  
v 
 
Table of Contents 
Abstract ........................................................................................................................... ii 
Acknowledgments .......................................................................................................... iv 
List of Tables .................................................................................................................. ix 
List of Figures ................................................................................................................ xv 
Chapter 1 Introduction ..................................................................................................... 1 
Chapter 2 Review of Literature ........................................................................................ 5 
2.1 Energy Overview ................................................................................................... 5 
2.2 Bioenergy .............................................................................................................. 6 
2.3 Biomass Logistics and Dust Generation ................................................................ 7 
2.4 Combustible Dust .................................................................................................. 9 
2.5 Hazardous Area ...................................................................................................10 
2.6 Dust Explosion .....................................................................................................11 
2.6.1 Primary and Secondary Dust Explosions .......................................................13 
2.6.2 Dust Explosion Characteristics ......................................................................13 
2.7 Dust Explosion Incidents ......................................................................................15 
2.8 Dust Ignition .........................................................................................................17 
2.8.1 Volatilization Properties .................................................................................18 
2.8.2 Exothermic Parameters .................................................................................19 
2.8.3 Minimum Hot Surface Ignition Temperature ...................................................20 
2.9 Factors Affecting Dust Ignition .............................................................................21 
2.9.1 Particle Size ..................................................................................................21 
2.9.2 Moisture Content ...........................................................................................23
  
vi 
 
2.9.3 Volatile Content .............................................................................................23 
2.9.4 Ash Content ...................................................................................................24 
2.10 NIR Spectroscopy (NIRS) ..................................................................................25 
2.10.1 Introduction ..................................................................................................25 
2.10.2 Analysis Techniques ....................................................................................27 
2.10.2.1 Principal Component Analysis ...............................................................27 
2.10.2.2 Partial Least Square Regression Analysis .............................................27 
2.10.3 Predictions Using NIRS ...............................................................................29 
Summary ....................................................................................................................... 34 
Chapter 3 Physical, Chemical and Heating and Ignition Properties of Biomass and Coal 
Dusts .............................................................................................................................36 
3.1 Abstract ................................................................................................................36 
3.2 Introduction ..........................................................................................................37 
3.3 Methods and Materials .........................................................................................42 
3.3.1 Raw Material .................................................................................................42 
3.3.2 Grinding and Dust Collection .........................................................................42 
3.3.3 Physical and Chemical Properties .................................................................45 
3.3.3.1 Moisture Content .....................................................................................45 
3.3.3.2 Bulk Density ............................................................................................45 
3.3.3.3 Particle Density .......................................................................................46 
3.3.3.4 Particle Size Distribution .........................................................................47 
3.3.3.5 Ash Content ............................................................................................49 
3.3.3.6 Volatile Matter Content ............................................................................51 
3.3.3.7 Energy Content .......................................................................................52 
3.3.4 Heating and Ignition Properties ......................................................................53 
3.3.4.1 Hot Surface Ignition Temperature ...........................................................53 
3.3.4.2 Volatilization Properties ...........................................................................55 
  
vii 
 
3.3.4.3 Exothermic Parameters ...........................................................................57 
3.3.5 Data Analysis ................................................................................................58 
3.4 Results and Discussion ........................................................................................59 
3.4.1 Physical and Chemical Properties .................................................................59 
3.4.1.1 Moisture Content .....................................................................................60 
3.4.1.2 Particle Size Distribution .........................................................................64 
3.4.1.3 Bulk Density ............................................................................................66 
3.4.1.4 Particle Density .......................................................................................67 
3.4.1.5 Ash Content ............................................................................................69 
3.4.1.6 Volatile Matter .........................................................................................73 
3.4.1.7 Energy Content .......................................................................................75 
3.4.2 Heating and Ignition Properties ......................................................................77 
3.4.2.1 Minimum Hot Surface Ignition Temperature ............................................77 
3.4.2.2 Volatilization Properties ...........................................................................82 
3.4.2.3 Exothermic Parameters ...........................................................................89 
3.5 Conclusion ......................................................................................................93 
Chapter 4 Prediction of Heating and Ignition Properties Using Near Infrared 
Spectroscopy (NIRS) .....................................................................................................94 
4.1 Abstract ................................................................................................................94 
4.2 Introduction ..........................................................................................................95 
4.3 Methods and Materials .........................................................................................98 
4.3.1 Raw Material .................................................................................................98 
4.3.2 Near Infrared Spectroscopy (NIRS) ...............................................................99 
4.3.3 Data Analysis .............................................................................................. 100 
4.4 Result and Discussion ........................................................................................ 102 
4.4.1 Raw NIR Spectra ......................................................................................... 102 
4.4.2 First Derivative NIR Spectra ........................................................................ 103 
  
viii 
 
4.4.4 Models for Prediction of Heating and Ignition Properties of Dusts ................ 108 
4.4.5 Model Elucidation ........................................................................................ 111 
4.4.6 External Validation ....................................................................................... 111 
4.5 Conclusion ......................................................................................................... 115 
Chapter 5 Summary and Future Recommendation ...................................................... 117 
5.1 Summary ........................................................................................................... 117 
5.2 Future Recommendation .................................................................................... 118 
References .................................................................................................................. 120 
Appendix A ? Initial Moisture Content and Physiochemical Properties of Dusts and 
Ground Material. .......................................................................................................... 133 
Appendix B ? Hot Plate Ignition Test, Thermogravimetric Analysis (TGA) and Differential 
Scanning Calorimetry (DSC) Results. .......................................................................... 147 
Appendix C ? SAS Codes for Tukey tests, correlation matrices and ANOVA results  (first 
objective) ..................................................................................................................... 156 
Appendix ? D SAS Codes and Results for Principal Component Analysis (PCA) on NIR 
Data for Internal Validation of Models .......................................................................... 187 
Appendix E ?Heating and Ignition Properties, NIR Spectra and SAS Code for Principal 
Component Analysis for Biomass Dusts Used for External Validation ......................... 217 
 
 
 
  
ix 
 
List of Tables 
 
Table 3.1 List and sources of biomass feedstocks and coals used in the study. ............43 
Table 3.2 Geometric mean diameter and geometric standard deviation values of ground 
(through 3.175 mm screen size) samples. .................................................................59 
Table 3.3 Geometric mean diameter and geometric standard deviation values of dust 
samples (passing through 437 ?m screen). ...............................................................60 
Table 3.4 Measured physical and chemical properties of ground biomass and coal 
samples ....................................................................................................................61 
Table 3.5 Measured physical and chemical properties of biomass and coal dust samples
 ..................................................................................................................................62 
Table 3.6 Minimum hot surface ignition temperature (MIT) of dust layer for all samples.
 ..................................................................................................................................80 
Table 3.7 Measured volatilization and exothermic properties of biomass and coal dust 
samples. ...................................................................................................................81 
Table 4.1 List of different biomass feedstock used for external validation of prediction 
models along with their sources. ...............................................................................99 
Table 4.2 Chemistry associated with influential wavenumbers derived from first 
derivative NIR spectra for dust samples (Schwanninger et al., 2011). ..................... 105 
Table 4.3 Calibration and validation statistics for prediction models developed using raw 
and first derivative NIR spectra. .............................................................................. 109 
Table 4.4 Chemistry/bond assignment for important wavelengths extracted from 
statistically significant principal components through regression analysis 
(Schwanninger et al., 2011). ................................................................................... 112 
Table 4.5 External validation statistics for performance of prediction models developed 
using raw NIR spectra. ............................................................................................ 115 
Table A.1 Initial moisture content of feedstock. ............................................................ 133 
Table A.2 Moisture content of dust samples. ............................................................... 134 
Table A.3 Bulk densities of dust samples. ................................................................... 135
  
x 
 
Table A.4 Particle densities for dust samples. ............................................................. 136 
Table A.5 Ash content values for dust samples. .......................................................... 137 
Table A.6 Volatile content values for dust samples. ..................................................... 138 
Table A.7 Energy content values for dust samples. ..................................................... 139 
Table A.8 Moisture content values for ground samples. .............................................. 140 
Table A.9 Bulk density values for ground samples....................................................... 141 
Table A.10 Particle density values for ground samples. ............................................... 142 
Table A.11 Ash content values for ground samples. .................................................... 143 
Table A.12 Volatile matter values for ground samples. ................................................ 144 
Table A.13 Energy content values for ground samples. ............................................... 145 
Table A.14 Moisture content values for feedstock, ground material and dust samples 
(biomass and coal) .................................................................................................. 146 
Table B.1 Temperature of rapid volatilization (TORV) for dust samples. ...................... 147 
Table B.2 Temperature of maximum rate of mass loss (TMML) values for dust samples.
 ................................................................................................................................ 148 
Table B.3 Oxidation temperature (TOXY) values for dust samples. ............................. 149 
Table B.4 Activation energy values for dust samples. .................................................. 150 
Table B.5 Temperature of rapid exothermic reaction (TRE) values for dust samples. .. 151 
Table B.6 Maximum temperature reached during exothermic reaction (TME) values for 
dust samples. .......................................................................................................... 152 
Table B.7 Exothermic energy values for dust samples. ............................................... 153 
Table B.8 Minimum hot surface ignition temperature (MIT) for dust samples. .............. 154 
Table B.9 Particle density of ash derived from sugarcane bagasse dust. .................... 155 
Table C.1 ANOVA results for Tukey test on moisture content (MC) for biomass dust. . 163 
Table C.2 ANOVA results for Tukey test on ash content (AC) for biomass dust. ......... 163 
Table C.3 ANOVA results for Tukey test on activation energy (AE) for biomass dust. . 164 
Table C.4 ANOVA results for Tukey test on bulk density (BD) for biomass dust. ......... 164 
Table C.5 ANOVA results for Tukey test on geometric mean diameter (dgw) for biomass 
dust. ........................................................................................................................ 165 
Table C.6 ANOVA results for Tukey test on energy content (EC) for biomass dust. .... 165 
Table C.7 ANOVA results for Tukey test on particle density (PD) for biomass dust. .... 166 
Table C.8 ANOVA results for Tukey test exothermic energy (Q) for biomass dust. ...... 166 
Table C.9 ANOVA results for Tukey test on maximum temperature reached during 
exothermic reaction (TME) for biomass dust. .......................................................... 167 
  
xi 
 
Table C.10 ANOVA results for Tukey test on temperature of maximum rate of mass loss 
(TMML) for biomass dust. ....................................................................................... 167 
Table C.11 ANOVA results for Tukey test on temperature of onset of rapid volatilization 
(TORV) for biomass dust. ........................................................................................ 168 
Table C.12 ANOVA results for Tukey test on oxidation temperature (TOXY) for biomass 
dust. ........................................................................................................................ 168 
Table C.13 ANOVA results for Tukey test on temperature of rapid exothermic reaction 
(TRE) for biomass dust. .......................................................................................... 169 
Table C.14 ANOVA results for Tukey test on volatile matter (VM) for biomass dust. ... 169 
Table C.15 ANOVA results for Tukey test on minimum hot surface ignition temperature 
(MIT) for biomass dust. ........................................................................................... 170 
Table C.16 ANOVA results for Tukey test on ash content (AC) for coal dust. .............. 170 
Table C.17 ANOVA results for Tukey test on activation energy (AE) for coal dust. ...... 171 
Table C.18 ANOVA results for Tukey test on bulk density (BD) for coal dust. .............. 171 
Table C.19 ANOVA results for Tukey test on geometric mean diameter (dgw) for coal 
dust. ........................................................................................................................ 172 
Table C.20 ANOVA results for Tukey test on energy content (EC) for coal dust. ......... 172 
Table C.21 ANOVA results for Tukey test on moisture content (MC) for coal dust. ...... 173 
Table C.22 ANOVA results for Tukey test on particle density (PD) for coal dust. ......... 173 
Table C.23 ANOVA results for Tukey test on exothermic energy (Q) for coal dust....... 174 
Table C.24 ANOVA results for Tukey test on maximum temperature reached during an 
exothermic energy (TME) for coal dust. ................................................................... 174 
Table C.25 ANOVA results for Tukey test on temperature of maximum rate of mass 
loss(TMML) for coal dust. ........................................................................................ 175 
Table C.26 ANOVA results for Tukey test on temperature of onset of rapid 
volatilization(TORV) for coal dust. ........................................................................... 175 
Table C.27 ANOVA results for Tukey test on oxidation temperature (TOXY) for coal dust.
 ................................................................................................................................ 176 
Table C.28 ANOVA results for Tukey test on temperature of rapid exothermic energy  
(TRE) for coal dust. ................................................................................................. 176 
Table C.29 ANOVA results for Tukey test on volatile matter (VM) for coal dust. .......... 177 
Table C.30 ANOVA results for Tukey test on minimum ignition temperature (MIT) for 
coal dust. ................................................................................................................ 177 
  
xii 
 
Table C.31 ANOVA results for Tukey test on moisture content (MC) for ground biomass.
 ................................................................................................................................ 178 
Table C.32 ANOVA results for Tukey test on bulk density (BD) for ground biomass. ... 178 
Table C.33 ANOVA results for Tukey test on particle density (PD) for ground biomass.
 ................................................................................................................................ 179 
Table C.34 ANOVA results for Tukey test on volatile matter (VM) for ground biomass.
 ................................................................................................................................ 179 
Table C.35 ANOVA results for Tukey test on ash content (AC) for ground biomass .... 180 
Table C.36 ANOVA results for Tukey test on energy content (EC) for ground biomass.
 ................................................................................................................................ 180 
Table C.37 ANOVA results for Tukey test on moisture content (MC) for ground coal. . 181 
Table C.38 ANOVA results for Tukey test on bulk density (BD) for ground coal. ......... 181 
Table C.39 ANOVA results for Tukey test on particle density (PD) for ground coal. ..... 182 
Table C.40 ANOVA results for Tukey test on volatile matter (VM) for ground coal. ...... 182 
Table C.41 ANOVA results for Tukey test on ash content (AC) for ground coal. .......... 183 
Table C.42 ANOVA results for Tukey test on energy content (EC) for ground coal. ..... 183 
Table C.43 Correlation matrix showing Pearson?s correlation coefficient and respective 
p-value for relation between all measured properties for all biomass dusts. ............ 184 
Table C.44 Correlation matrix showing Pearson?s correlation coefficient and respective 
p-value for relation between all measured properties for grassy biomass (Bermuda 
grass, corn stover, sugarcane bagasse and switchgrass) dusts. ............................. 185 
Table C.45 Correlation matrix showing Pearson?s correlation coefficient and respective 
p-value for relation between all measured properties for woody biomass (eucalyptus, 
pine and sweetgum) dusts. ..................................................................................... 186 
Table D.1 ANOVA results and model parameter estimates for MIT (raw data model). . 191 
Table D.2 ANOVA results and model parameter estimates for TORV (raw data model).
 ................................................................................................................................ 192 
Table D.3 ANOVA results and model parameter estimates for TMML (raw data model).
 ................................................................................................................................ 193 
Table D.4 ANOVA results and model parameter estimates for TOXY (raw data model).
 ................................................................................................................................ 194 
Table D.5 ANOVA results and model parameter estimates for TRE (raw data model). 195 
Table D.6 ANOVA results and model parameter estimates for TME (raw data model).196 
  
xiii 
 
Table D.7 Principal components (PC) obtained from raw NIR spectral data of biomass 
dusts used for PCA and internal validation of prediction models. ............................ 197 
Table D.8 Result of Tukey test on principal component values of different biomass dust 
samples for PC1...................................................................................................... 198 
Table D.9 Result of Tukey test on principal component values of different biomass dust 
samples for PC2...................................................................................................... 199 
Table D.10 Result of Tukey test on principal component values of different biomass dust 
samples for PC3...................................................................................................... 200 
Table D.11 Result of Tukey test on principal component values of different biomass dust 
samples for PC4...................................................................................................... 201 
Table D.12 Result of Tukey test on principal component values of different biomass dust 
samples for PC5...................................................................................................... 202 
Table D.13 Result of Tukey test on principal component values of different biomass dust 
samples for PC6...................................................................................................... 203 
Table D.14 Result of Tukey test on principal component values of different biomass dust 
samples for PC9...................................................................................................... 204 
Table D.15 Result of Tukey test on principal component values of different biomass dust 
samples for PC10. ................................................................................................... 205 
Table D.16 ANOVA results and model parameter estimates for MIT (first derivative data 
model). .................................................................................................................... 210 
Table D.17 ANOVA results and model parameter estimates for TORV (first derivative 
data model). ............................................................................................................ 211 
Table D.18 ANOVA results and model parameter estimates for TMML (first derivative 
data model). ............................................................................................................ 212 
Table D.19 ANOVA results and model parameter estimates for TOXY (first derivative 
data model). ............................................................................................................ 213 
Table D.20 ANOVA results and model parameter estimates for TRE (first derivative data 
model). .................................................................................................................... 214 
Table D.21 ANOVA results and model parameter estimates for TME (first derivative data 
model). .................................................................................................................... 215 
Table D.22 Principal components (PC) obtained from first derivative NIR spectral data of 
biomass dusts used for internal validation of prediction models. ............................. 216 
Table E.1 Minimum hot surface ignition temperature (MIT) values for biomass dusts 
used for external validation of prediction models. .................................................... 217 
  
xiv 
 
Table E.2 Temperature of onset of rapid volatilization (TORV) values for biomass dusts 
used for external validation of prediction models. .................................................... 218 
Table E.3 Temperature of maximum rate of mass loss (TMML) values for biomass dusts 
used for external validation of prediction models. .................................................... 218 
Table E.4 Oxidation temperature (TOXY) values for biomass dusts used for external 
validation of prediction models. ............................................................................... 219 
Table E.5 Temperature of rapid exothermic reaction (TRE) values for Biomass Dusts 
Used for External Validation of Prediction Models. .................................................. 219 
Table E.6 Maximum temperature reached during exothermic reaction (TME) values for 
Biomass Dusts Used for External Validation of Prediction Models. ......................... 220 
Table E.7 Principal components (PC) obtained from raw NIR spectral data of biomass 
dusts used for external validation of prediction models. .......................................... 225 
 
 
 
 
 
  
xv 
 
List of Figures 
 
Figure 2.1 Past, current and projected world energy consumption between 1990-2040 
(EIA, 2013). ................................................................................................................ 5 
Figure 2.2 Operational components of biomass supply chain (Mafakheri and Nasiri, 
2014). ......................................................................................................................... 8 
Figure 2.3 Dust generated during handling of biomass. Dust is generated as (a) material 
falls from one conveyor to another, (b) from conveyor to floor and (c) material falling 
into a silo (Wypch et al., 2005). .................................................................................. 9 
Figure 2.4 Dust explosion due to disturbance of a dust layer (Blair, 2012). ....................12 
Figure 2.5 Dust explosion pentagon (Kauffman, 1982). .................................................13 
Figure 2.6 Increase in rate of combustion with increasing surface area (Eckhoff, 2003).
 ..................................................................................................................................21 
Figure 2.7 Relationship between volatile content (dry basis) and ignition temperature for 
biomass and coal samples (Grotkjaer et al., 2003). ...................................................24 
Figure 2.8 Example of principal component score plot showing PC1 vs. PC2 (Vergnoux 
et al., 2009). ..............................................................................................................30 
Figure 2.9 Typical NIR spectra for three different Japanese plum fruit taken at start of 
fruit ripening (Louw and Theron, 2010). .....................................................................31 
Figure 2.10 Actual vs. predicted values for TA (a), TSS (b), firmness (c) and weight (d) 
for multi cultivar NIR model (Louw and Theron, 2010). ..............................................32 
Figure 3.1 Conventional oven used for moisture content determination. ........................44 
Figure 3.2 Air drying of sweetgum wood chips (a) and drying sugarcane bagasse in food 
dehydrator (b). ..........................................................................................................44 
Figure 3.3 Hammer mill used for grinding biomass feedstock (a) and vibratory screen 
used for collection of dust (b). ...................................................................................44 
Figure 3.4 Apparatus for measuring bulk density. ..........................................................47 
Figure 3.5 Gas pycnometer used for estimation of particle density. ...............................47 
Figure 3.6 Digital imaging particle size analyzer. ...........................................................49
  
xvi 
 
Figure 3.7 Muffle furnace used for ash content determination. ......................................50 
Figure 3.8 Volatile matter determination furnace. ..........................................................52 
Figure 3.9 Bomb calorimeter used for energy content determination. ............................53 
Figure 3.10 Apparatus to measure minimum hot surface ignition temperature (a), metal 
ring filled with dust sample before ignition (b), dust sample after ignition (c). .............54 
Figure 3.11 Thermogravimetric analyzer (TGA) used to measure volatilization properties.
 ..................................................................................................................................55 
Figure 3.12 Differential scanning calorimeter (DSC) equipment used to determine 
exothermic parameters of dust samples. ...................................................................58 
Figure 3.13 Moisture content before and after grinding biomass and coal. ....................63 
Figure 3.14 Particle size distribution of dust samples. ...................................................65 
Figure 3.15 Elongated particles in Bermuda grass dust (a) and switchgrass dust (b) 
samples. ...................................................................................................................65 
Figure 3.16 Bulk density of ground and dusts from biomass and coal............................67 
Figure 3.17 Particle densities of ground and dusts from biomass and coal. ...................68 
Figure 3.18 Particle density dependence on geometric mean particle size in case of all 
biomass dusts. ..........................................................................................................68 
Figure 3.19 Ash content of ground and dusts from biomass and coal. ...........................71 
Figure 3.20  Ash content dependence on geometric mean particle size (a) and particle 
density (b) for grassy biomass dusts. ........................................................................72 
Figure 3.21 Volatile matter of ground and dusts from biomass and coal. .......................74 
Figure 3.22 Effect of ash content on volatile matter content of grassy biomass dusts. ...75 
Figure 3.23 Energy content of ground and dusts from biomass and coal. ......................76 
Figure 3.24 Effect of ash content on energy content for all biomass dusts. ....................77 
Figure 3.25 Plots of temperature vs. time showing maximum temperature of no ignition, 
275?C (a) and minimum temperature of hot surface at which ignition occurred (MIT), 
280?C (b) for corn cobs sample. ................................................................................78 
Figure 3.26 Minimum hot surface temperature (MIT) dependence on ash content of 
grassy biomass dusts (a), bulk density of grassy biomass dusts (b), bulk density of 
woody biomass dusts (c) and volatile matter of grassy biomass dusts (d). ................82 
Figure 3.27 TGA mass loss curves for dust samples heated in air environment (a) and 
oxygen environment (b).............................................................................................84 
  
xvii 
 
Figure 3.28 Example of how TORV and TMML were estimated. Mass loss curves are for 
switchgrass dust sample heated in air atmosphere (a) and heated in oxygen 
atmosphere (b). .........................................................................................................85 
Figure 3.29 Example for determination of apparent activation energy (switchgrass dust 
sample). ....................................................................................................................88 
Figure 3.30 Effect of ash content on activation energy (a) and effect of volatile matter on 
activation energy (b) for grassy biomass dusts. .........................................................88 
Figure 3.31 Heat flow curves of biomass and coal dusts heated with DSC in air 
atmosphere. ..............................................................................................................90 
Figure 3.32 Example of how TRE and TME were estimated from heat flow vs. 
temperature curve for switchgrass sample when heated under air environment. .......90 
Figure 3.33 Effect of energy content on maximum temperature reached during 
exothermic reaction (TME) for woody biomass dusts. ...............................................92 
Figure 4.1 FT-NIR spectrophotometer used to collect spectral data of dust samples. .. 100 
Figure 4.2 NIR spectra showing average absorbance vs. wavenumber plot for different 
dusts. ...................................................................................................................... 103 
Figure 4.3 First derivative plot of NIR spectra for different dusts showing significant 
wavenumbers associated with peaks. ..................................................................... 104 
Figure 4.4 Principal component score plots for significant principal components viz. PC1 
vs. PC2 (a), PC3 vs. PC4 (b), PC5 vs. PC6 (c) and PC10 vs. PC9 (d) obtained from 
NIR raw spectral data of biomass dusts. ................................................................. 106 
Figure 4.5 Actual vs. predicted values for MIT (a), TORV (b), TMML (c), TOXY (d), TRE 
(e) and TME (f). ....................................................................................................... 110 
Figure 4.6 Eigenvector loading on NIR spectra showing wavenumber vs. eigenvectors 
for significant PC. Dashed line represents 95th percentile of eigenvector distribution.
 ................................................................................................................................ 113 
Figure 4.7 External validation results showing actual vs. predicted values for TORV (a), 
TMML (b), TOXY (c), TRE (d), TME (e) and MIT (f). ............................................... 114 
Figure E.1 NIR spectra showing average absorbance vs. wavenumber plot for biomass 
dusts used for external validation of prediction models. .......................................... 220 
  
1 
 
 Chapter 1 Introduction  
The United States and other countries rely heavily on fossil fuels. Since 
fossil fuels are fast depleting, shortage in the fuel supply could seriously jeopardize 
a nation?s economic and social well-being and national security. In 2013, more 
than 80% of U.S.?s total energy consumption were obtained from fossil fuels such 
as petroleum, natural gas and coal (EIA, 2013). In addition, United States also 
depends on petroleum imports from other countries in order to meet its energy 
requirements. In 2011, USA produced about 78 quadrillion Btu (quads) of energy 
but consumed about 96 quads of total energy. The energy deficit was met by 
importing petroleum fossil fuels (EIA, 2013). Due to environmental impacts, long 
term availability issues of fossil fuels and its effect on economic and national 
security, it is very important to derive energy from renewable sources such as 
biomass energy, solar energy, wind energy and geothermal energy. 
Out of the 8.8 quadrillion Btu of energy produced from renewable sources 
in the year 2012, energy produced from biomass accounts for about 50% (4.4 
quadrillion Btu) (EIA, 2013). The advantage that biomass has over other forms of 
renewable energy is that it is the only renewable resource that can be used to 
produce liquid fuels, chemicals and other products. However, biomass feedstocks 
have to be processed and handled before it can be converted into biofuels. 
Equipment such as mills, grinders, silos, hoppers and conveyors that are needed 
  
2 
 
to prepare, process and handle biomass also generate dust from the biomass. The 
National Fire Protection Association (NFPA, 2013) defines combustible dusts as 
?particles passing through a 500 ?m sieve which presents a dust fire or dust 
explosion hazard?.  According to Vijayraghavan (2004), more than 70% of dusts 
generated in process industries are combustible. Combustible dust, if ignited and 
when they are suspended in air can cause explosion (Eckhoff, 1996). Dust 
explosions cause damage to processing plants, injuries to personnel and in 
extreme cases, fatalities (Eckhoff, 2009). According to U.S. Chemical Safety and 
Hazard Investigation Board?s report (CSB, 2006), between 1980 and 2005 there 
were at least 281 dust fire and explosion incidents that resulted in 119 fatalities 
and at least 718 injuries in the United States. In 2011, there were seven fatalities 
due to dust explosions in the United States (BLS, 2013). In late June, 2011, the 
world?s largest biomass pellet factory in the state of Georgia, USA had dust 
explosion incident that led to shutting down of the plant for 1.5 months 
(Renewables International Magazine, 2011). In addition to structural damage dust 
explosions can result in loss of income by a plant due to down time and time 
required to repair the damaged portion of the plant (Sapko et al., 2000).  
The probability of dust causing explosion depends on the ignition of 
combustible dust. A dust with low minimum ignition temperature value will be more 
prone to ignition risks. Presence of hot surfaces such as surfaces of dryers, 
grinders or worn out bearings increase the chances of dust on these surfaces to 
ignite, thereby rendering the workplace hazardous and prone to dust explosion. 
Sparks, short circuit faults and electric arcs from electrical equipment as well as 
  
3 
 
electrical discharges may ignite suspended dust particles and cause explosions. 
Therefore, it is very important to study the factors that will lead to heating and 
ignition of biomass dusts in order to incorporate appropriate safety protocols during 
the preprocessing of biomass (Bilbao et al., 2002).  
The parameters that have been used to quantify the heating and ignition 
risks of dusts include minimum hot surface temperature for dust layer ignition, 
minimum temperature required for volatilization, temperature of rapid exothermic 
reaction, temperature of maximum rate of mass loss and temperature of oxidation 
(Ramirez et al., 2010; Hehar, 2013). The methods used to quantify these 
properties are time consuming and require the use of expensive pieces of 
equipment such as thermogravimetric analyzer (TGA) and differential scanning 
calorimeter (DSC). Near Infrared (NIR) spectroscopy has been used as a quick 
method of indirectly quantifying the properties of biological samples such as grain 
moisture content (Norris, 1964), dry matter content and fruit firmness (Nicolai et 
al., 2008), post-harvest quality of fruits (Bobelyn et al., 2010), moisture content, 
water activity and salt content of meat (Collell et al., 2011), quality control of potato 
chips (Shiroma and Rodriguez-Saona, 2009), taste characterization of fruits 
(Jamshidi et al., 2012) and proximate analysis and heating values of torrefied 
biomass (Via et al., 2013). Some of the advantages of NIR spectroscopy includes 
non-destructive measurement, ease of sample preparation, ability to be used by 
low skilled operator and high data/spectrum acquisition rates (Vergnoux et al., 
2009). 
  
4 
 
This research specifically aims at quantifying the heating and ignition risks 
of dust generated from biomass feedstocks and to develop NIR spectroscopy 
methodology to predict heating and ignition risks of biomass dust samples. NIR 
spectroscopy and principal component analysis (PCA) on the NIR spectral data 
will be employed for the prediction of heating and ignition parameters. To achieve 
this goal, the following specific objectives will be carried out on dust samples 
obtained from ten biomass feedstocks (Bermuda grass, corn cobs, corn stover, 
eucalyptus, loblolly pine, pecan shell, poultry litter, sugarcane bagasse, sweetgum 
and switchgrass) and three coal samples (bituminous coal, lignite coal, powder 
river basin (PRB) coal). 
1. Quantify the heating and ignition properties, and the physical and chemical 
properties of the biomass and coal dust samples. 
2. Predict the heating and ignition characteristics of biomass dusts using Near 
Infrared (NIR) spectroscopy. 
 
 
 
 
  
5 
 
Chapter 2 Review of Literature 
2.1 Energy Overview 
Energy plays a pivotal role in the development of human civilization and will 
continue to do so in the future. Also, growing population of the world would demand 
more energy in the future. The current world population is 7.2 billion and is 
projected to increase to 9.6 billion by the year 2050 (United Nations, 2014). The 
projection is that there will be a 56% increase in world energy consumption from 
year 2010 to 2040 (EIA, 2013) as depicted in figure 2.1.  
 
 
Figure 2.1 Past, current and projected world energy consumption between 1990-
2040 (EIA, 2013). 
0
200
400
600
800
1000
1990 2000 2010 2020 2030 2040
Wor
ld 
En
ergy
 Con
su
mp
tion
 
(qu
ad
rillio
n B
tu)
ProjectionHistory
  
6 
 
High energy consumption and dependence on fossil fuels in industrialized 
countries of the world pose global sustainability challenges (Weidenhofer et al., 
2013). Presently, United States has the highest per capita energy consumption. In 
the year 2013, U.S. consumed about 17.5% (96 quadrillion Btu) of the total world 
energy consumption (547 quadrillion Btu) (EIA, 2013) even though US is inhabited 
by only about 4.4% of the world population (USCB, 2014).  
Fossil fuels (coal, natural gas and petroleum) play a major role in meeting 
U.S. energy requirements. Out of 95 quads energy consumed in 2012, 78 quads 
were derived from fossil fuels (EIA, 2013). More than 67% of the petroleum 
consumed (34.58 quadrillion Btu) in the U.S. was imported. United States therefore 
relies heavily on petroleum imports to meet its energy requirements.  
In 2012, about 9.2% (8.8 quadrillion Btu) of the total energy consumed in 
United States was derived from renewable sources (EIA, 2013). About 50% (4.383 
quadrillion Btu) of this renewable energy amount was derived from biomass (EIA, 
2013).  
 
2.2 Bioenergy 
Bioenergy is energy derived from biomass which includes, but not limited to 
energy crops, agricultural crops, food, fiber, feed, forest products, aquatic plants, 
wood residues, industrial and residential waste, processing byproducts and non-
fossil organic material (ASABE S593.1, 2011). Biomass has been used as main 
source of energy in rural areas for centuries (Mafakheri and Nasiri, 2014). It is the 
fourth largest source of global energy accounting for 10% to 14% of global energy 
  
7 
 
consumption (Kheshgi et al., 2000; Parrika, 2004; Balat and Ayar, 2005; Demirbas, 
2005). Due to increasing interest in renewable and environment friendly energy 
sources, biomass has gained a lot of attention as a potential energy source. In 
addition, biomass is the only renewable resource of energy that can be converted 
into carbon based fuels and products. Bioenergy is also cleaner form of energy 
having lower impact on environment than energy derived from fossil fuels (McNew 
and Grif?th, 2005; Rajagapol at al., 2009).  
In United States, concerns about global climate change and air pollution 
has led to an increased interest in biomass as potential energy source because 
bioenergy is CO2 neutral and less polluting than fossil fuels (Cook and Beyea, 
2000). There is an abundance of biomass resources in United States. It is 
estimated that land resources of the U.S. will be capable of producing at least 1 
billion dry tons (0.91 billion dry metric tonnes) of biomass feedstock per year by 
the mid-21st century (Perlack et al., 2005). This quantity has been estimated to 
replace at least 30% of the nation?s petroleum consumption. Other benefits of 
using biomass include creation of employment opportunities, diversifying 
economic structure of rural communities and maintaining forest health (Mayfield et 
al., 2007; Gan and Smith, 2007). 
 
2.3 Biomass Logistics and Dust Generation 
Some of the challenges that will be faced by the emerging bioenergy 
industry for it to be viable, sustainable and mature includes supply logistics, a 
  
8 
 
continuous and large feedstock supply and residue handling (Tembo et al., 2003). 
Biomass supply logistics consists of biomass harvesting and collection, storage, 
transportation, pretreatment, storage, transport and energy conversion (figure 2.2).  
 
  
 
 
 
 
 
 
 
Biomass feedstocks have low energy density, high moisture content and 
are geographically dispersed. Therefore there is high logistics cost incurred to 
deliver them from farms/forest land to conversion plants (An and Searcy, 2012). 
For example, the cost of silage or bale logistics was estimated to vary between 
$40 Mg-1 and $60 Mg-1 (U.S. Department of Energy, 2010). In addition, biomass 
has to be processed before using it for energy production. Processing and handling 
of biomass involves unit operations such as grinding, milling, conveying and 
densifying that also generate dust. For example, dust is generated when bulk 
material such as biomass undergo freefall as shown in figure 2.3 (Wypych et al., 
2005). Dust generation while handling biomass can lead to dust fire or explosion 
and to health problems of workers (Khan et al., 2008). 
Biomass 
Harvesting Storage Transport Pretreatment 
Storage Transport Energy Conversion 
Figure 2.2 Operational components of biomass supply chain (Mafakheri 
and Nasiri, 2014). 
  
9 
 
Wood dusts was shown to cause allergic reactions (Hausen, 1981), nasal 
adenocarcinoma (Acheson et al., 1981) and pathological changes in the lungs of 
woodworkers (Michaels, 1967). Also, coal dust exposure can cause acute alveolar 
and interstitial inflammation that can lead to chronic pulmonary diseases (Pinho et 
al., 2004).  
 
 
 
Figure 2.3 Dust generated during handling of biomass. Dust is generated as (a) 
material falls from one conveyor to another, (b) from conveyor to floor and (c) 
material falling into a silo (Wypch et al., 2005). 
 
2.4 Combustible Dust 
Particle size is the main criteria used for defining dust (Amyotte et al., 2007). 
Previous editions of NFPA 654 defined combustible dust as material capable of 
passing through a U.S. No. 40 standard sieve (420 ?m) but 500 ?m (U.S. No. 35 
standard sieve) is now used as the new size criterion by NFPA 654 standard 
(NFPA, 2013). This is because particles of different shapes such as fiber 
(a) (b) (c) 
  
10 
 
segments, flat platelets and agglomerates cannot readily pass through a U.S. No. 
40 sieve, but may be combustible (Zalosh, 2005). Generally, combustible 
particulate solids with size greater than 500 ?m have surface area to volume ratios 
that are not sufficient to cause a dust explosion hazard and may therefore not 
contribute significantly to dust explosions (Calle et al, 2005; NFPA, 2013).  
2.5 Hazardous Area 
 The interior parts of a processing plant that have significant dust 
accumulation are classified as dust flash fire and/or dust explosion hazard area. 
This classification depends on the depth and mass of accumulated dust (NFPA, 
2013). The critical ?layer depth? of combustible dust that causes an area to be 
hazardous for dust explosion or dust fire prone is given by equation 2.1 (NFPA, 
2013). 
 ?? = 152527
160000 ?? (2.1) 
 
where,  
LD is layer depth (m), 
BD is bulk density (kg m-3)  
 
If the actual dust layer depth in an area is greater than the layer depth 
calculated in equation 2.1 (LD) then that area is considered as hazardous. Note 
that equation 2.1 is applicable to floor areas <1000 ft2 (92.9 m2) and for dust with 
bulk density <1201 kg/m3. 
  
11 
 
On a mass basis, the mass of dust in a processing plant should be less than 
the value calculated in equation 2.2 and 2.3 (NFPA, 2013) respectively to minimize 
the risk for dust explosion and dust fire hazard. 
 ?basic explosion =  0.004  ?floor  H    (2.2) 
where,  
Mbasic explosion is threshold dust mass (kg) based on building damage criterion, 
Afloor is enclosure floor area (m2) or 2000 m2, whichever is less, 
H is ceiling height (m) or 12 m, whichever is less. 
 ?basic fire =  0.02  ?floor (2.3) 
where,  
Mbasic fire is threshold dust mass (kg) based on personnel fire exposure criterion, 
Afloor is enclosure floor area (m2) or 2000 m2, whichever is less. 
 
2.6 Dust Explosion 
As discussed earlier, dust generated during handling of biomass can settle 
as dust layers on various sections of a processing plant such as on the floor, on 
process equipment and on storage structures. The accumulated combustible layer 
of dust can ignite and cause fire or explosion if these surfaces reach a critical high 
temperature. These dust layers may also be disturbed by an external event such 
as blowing air or mechanical action resulting in suspension of the dust to form dust 
  
12 
 
cloud. If the concentration of the dust cloud is high enough and in a confined space, 
dust explosion can occur when an ignition source such as electric spark, hot 
surface or a naked flame comes in contact with the cloud (Amyotte and Eckhoff, 
2010) (figure 2.4).  
 
Figure 2.4 Dust explosion due to disturbance of a dust layer (Blair, 2012). 
 
The three elements required for fire are fuel (in this case, dust), oxidizing 
medium (air) and heat (ignition source). Removal of any one of these would cause 
fire to cease. In the case of dust explosion, two other elements, dispersion or 
mixing and confinement are required along with the requirements for fire (figure 
2.5). 
 
 
  
13 
 
 
  
 
 
 
 
 
2.6.1 Primary and Secondary Dust Explosions 
Primary dust explosion occurs in units and equipment (milling, grinding, 
etc.) of a process industry when all the conditions of the explosion pentagon are 
met. However, shock or blast wave from such an explosion may disturb the dust 
accumulated outside of this equipment or unit by creating a suspension and 
eventually a secondary explosion (Amyotte and Eckhoff, 2010). A weak primary 
dust explosion can cause a very powerful secondary dust explosion depending 
upon the amount and extent of accumulated dust. The secondary explosion may 
further disturb the accumulated dust in other areas and lead to a series of dust 
explosions. This phenomenon is referred to as Domino effect in dust explosions 
(Abbasi and Abbasi, 2007). 
2.6.2 Dust Explosion Characteristics 
The ability of dust to cause explosion is determined by parameters such as 
minimum ignition energy (MIE), minimum ignition temperature (MIT), and minimum 
explosion concentration (MEC). The maximum explosive pressure (Pmax) and 
Fuel 
Mixin
g 
Confinement 
Oxidant 
Ignition 
Source Figure 2.5 Dust explosion pentagon (Kauffman, 1982). 
  
14 
 
maximum rate of explosion pressure rise ((dP/dt)max) (Nifuku et al., 2005; 
Cashdollar, 2000) are used to quantify the severity of the explosion that a dust fire 
can cause (Eckhoff, 2003). 
Minimum Ignition Energy (MIE) 
MIE is the minimum spark energy that can ignite the most ignition sensitive 
concentration of dust-air mixture (NFPA, 2013).  The value of MIE for combustible 
dust clouds range from 0.01 mJ to beyond 1 kJ (Eckhoff, 2002). Janes et al. (2008) 
found MIE for wood dust to lie between 45 and 58 mJ. They also reported MIE for 
crushed pea fiber, cocoa powder and coal to vary between 100-300 mJ, 300-1000 
mJ and >1000 mJ respectively.  
Minimum Ignition Temperature (MIT) 
MIT for a dust cloud is the minimum air temperature at which flame is 
observed when the dust particles are combusted (Benedetto et al., 2007). 
Minimum ignition temperature is measured in a BAM oven or Godbert-Greenwald 
furnace. For most dust clouds, the minimum ignition temperature values ranges 
from 420?C to 660?C (Zalosh, 2008). Minimum ignition temperature values of dusts 
from wheat flour, corn starch and rye dust at relative humidity of 30-90% are 410-
430?C, 410-450?C and 430-500?C respectively (Zalosh, 2008). Polka et al. (2012) 
found MIE for hop, nettle, barley, corn starch and sunflower hull dusts to be 460?C, 
500?C, 450?C, 460?C and 460?C respectively. 
Minimum Explosible Concentration (MEC) 
  
15 
 
Minimum explosible concentration (MEC) is defined as minimum 
concentration (mass of dust / volume of confined space) of explosible dust that is 
suspended in air and can support deflagration (a rapid burning slower than speed 
of sound) (NCDOL, 2012). MEC is typically measured with Hartmann tube 
apparatus or 20 liter sphere apparatus. Garcia-Torrent et al. (1998) measured 
minimum explosive concentration (MEC) of forest residue biomass including wood 
pieces and bark. At 6% moisture content, MEC as determined by Hartmann tube 
was 30 g/m3, and MEC value of 20 g/m3 was obtained from the 20 liter sphere 
apparatus. Amyotte et al. (2012), determined MEC of fibrous wood samples to be 
100 g/m3. When the samples were fractionated, those that passed through 35 
mesh sieve (<500 ?m) and 200 mesh sieve (<75 ?m) had MEC values of 30 and 
20 g/m3 respectively. 
  
2.7 Dust Explosion Incidents 
The total number of dust explosions that occurred in USA and Germany 
between 1900 - 1956 are reported to be 1120 (Theimer, 1973). Nearly half of these 
cases occurred in grain, flour and feed handling industries with about 392 
casualties and 1015 injuries. Financial losses were estimated to be over $75 
million.  
According to U.S. Chemical Safety and Hazard Investigation Board?s report 
(2006), there were at least 281 dust fire and explosion incidents that caused 119 
fatalities and over 718 injuries in the United States between 1980 and 2005. In 
  
16 
 
United Kingdom, a total of 571 dust explosion cases were reported in a period of 
ten (1968-1979) years (Lunn, 1992) that resulted in 247 fatalities and 324 non-fatal 
injuries. Between 1979 and 1988 in UK there were 36 dust explosions reported 
that lead to injuries and 123 dust explosions that did not cause injuries 
(Vijayaraghavan, 2004).  
On an average, 10.6 agricultural grain dust explosions are reported per year 
in the U.S. resulting in 1.6 deaths, 12.6 injuries and millions of dollars of damages 
(Schoeff, 2006). In February 2008, a major dust explosion and fire incident 
occurred at a sugar refinery in Georgia, USA, which claimed 14 lives, injured 38 
people and led to total destruction of the plant (CSB, 2013). Dust explosions are 
not only limited to agricultural and food processing facilities. Three combustible 
dust incidents occurred in a powdered iron producing facility (Hoeganaes 
Corporation) at Gallatin, TN over a period of 6 months that led to five fatalities and 
three injuries (CSB, 2013). In March 2011, two workers were killed and two 
seriously injured when fire and explosion occurred at Carbide Industries facility in 
Louisville, KY. This facility produced calcium carbide products. In April 2013, dust 
explosion caused a fire in two of the fuel storage silos of Koda Energy Plant (MN, 
USA) which continued for over a week. The facility uses wood chips, oat hulls and 
other organic materials to generate electricity (Biomass magazine, 2014). In 
August 2013, a combustible wood dust explosion ocuurred at ?Inferno Wood Pellet 
Inc.? ? a wood pellet manufacturing unit at East Providence, RI (USA) which 
partially demolished the building and injured a worker (OSHA, 2014). 
  
17 
 
More than 70% of powders in processing facilities are combustible. This is 
why most of the dust explosion accidents start in areas where powder processing 
equipment (e.g. mills, grinders, filters, driers, silos, hoppers and ducts) are installed 
and operated (Vijayraghavan, 2011).  
 
2.8 Dust Ignition 
A combustible dust is ignited first before it causes explosion. Requirements 
for ignition of combustible dust whether suspended in air or deposited on surfaces 
include air or oxidizing medium, dust in sufficient quantity and ignition source such 
as electrostatic discharge, electric current arc or spark, glowing ember, hot 
surface, welding slag, frictional heat or a naked flame (NFPA, 2013). Ignition 
occurs when the rate of slow fuel oxidation changes to a rate of rapid oxidation 
either of the volatiles or the solid matrix of the material (Grotkjaer et al., 2003). This 
results in a sudden rise of sample/fuel temperature (Haykiri-Acma, 2003). Ignition 
of biomass can be of three types, viz. homogeneous ignition, heterogeneous 
ignition and hetero-homogenous ignition. Chen et al. (1996) defined homogeneous 
ignition as ignition of the volatile matter released from the material, whereas 
heterogeneous ignition is ignition of actual particles of the material. Hetero-
homogeneous ignition is the simultaneous ignition of volatile matter and particles. 
Since, combustible dust layers when ignited can lead to dust explosion and 
associated financial losses and fatalities (Joshi et al., 2012), it is very important to 
  
18 
 
quantify the heating and ignition properties for combustible dusts. These properties 
are reviewed below. 
2.8.1 Volatilization Properties 
 Volatilization properties are measured by conducting thermal 
decomposition study (with TGA) on dust samples exposed to constant or 
programmed heating rates. A mass loss vs. temperature and mass loss rate vs. 
temperature curves are obtained and are used to estimate parameters such as 
temperature of onset of volatilization, temperature at maximum rate of mass loss 
and oxidation temperature. Temperature of onset of volatilization signifies the 
temperature at which significant release of volatiles starts as the material is being 
heated. Temperature at maximum rate of mass loss can be attributed to rapid 
release of volatile matter due to pyrolysis and thus gives an indication of reactivity 
of the sample. In air stream, TG analysis loss of mass due to thermal degradation 
of sample occurs over a range of temperature making it difficult to assign a single 
oxidation temperature to the sample. Thus, O2 stream TG analysis is used to obtain 
oxidation temperature which allows different dusts to be categorized based on their 
ignition risk (Ramirez et al., 2010). A dust with lower oxidation temperature value 
would be at a higher risk of ignition than a dust with higher oxidation temperature. 
 Ramirez et al. (2010), estimated temperature of onset of rapid volatilization 
(TORV) and temperature of maximum mass loss rate (TMML) for icing sugar to be 
212?C and 220 ?C respectively. For maize, wheat and barley TORV values were 
estimated at 268?C, 252?C and 242?C respectively, while their TMML values were 
estimated at 279?C, 283?C and 271?C. Both, TORV and TMML values for maize, 
  
19 
 
wheat and barley were more than that of icing sugar indicating that icing sugar has 
more ignition risk. Oxidation temperature (TOXY) values for icing sugar (239 ?C), 
was also smaller than those of maize (289 ?C), wheat (279 ?C) and barley (277 
?C). In addition, temperature of maximum mass loss rate for wheat straw in 20% 
O2 environment was found to lie between 220-270?C (Grotkjaer et al., 2003). Sahu 
et al. (2010) reported the temperature of maximum mass loss rate (TMML) for coal, 
sawdust and rice husk as 419.5?C, 417.3?C and 323.2?C respectively. Haykiri-
Acma (2003) measured the temperature of maximum mass loss rate (TMML) for 
sunflower shell, cozla seed, pine cone, cotton refuse and olive refuse as 300?C, 
262?C, 292?C, 325?C and 264?C respectively. 
2.8.2 Exothermic Parameters 
Exothermic parameters of samples are determined by measuring the heat 
flow of sample relative to inert reference material or empty crucible in a differential 
scanning calorimeter (DSC). Exothermic parameters are used to characterize the 
ignition characteristics of dust samples by obtaining maximum temperature 
reached during exothermic reaction (TME), temperature required for onset of rapid 
exothermic reaction (TRE) and exothermic energy from heat flow curves (Ramirez 
et al., 2010). Dust sample with lower TRE value would be easily ignited than the 
dust with higher TRE value. Samples with higher values of TME would promote 
secondary dust fire or ignition by providing sufficient energy for the reaction.  Dusts 
with higher exothermic energy release greater amount of energy in form of heat 
during ignition or explosion than samples with lower exothermic energy value. The 
  
20 
 
heat energy released could lead to secondary dust explosions or fires in a 
processing plant or facility leading to further destruction. 
Ramirez et al. (2010) estimated TME values for maize, wheat, barley and 
alfalfa as 386?C, 283?C, 311?C and 288?C respectively and TRE values for the 
given samples to be 242?C, 252?C, 257?C and 240?C respectively. Sahu et al. 
(2010), found TME values for coal, sawdust and rice husk as 423?C, 422?C and 
454?C respectively.  
2.8.3 Minimum Hot Surface Ignition Temperature 
One main cause of ignition of dust layers are hot surfaces that are 
commonly found in processing plants. Several studies have been conducted on 
the hot surface temperature requirements for ignition of dust layers (Park et al., 
2009; Janes et al., 2008; Joshi et al., 2012; Sweis, 1998). Hot surface minimum 
ignition temperatures for Pittsburgh seam coal, paper dust, Arabic gum powder 
and brass powder was measured to be 220?C, 360?C, 270?C and >400?C 
respectively (Park, 2006). Minimum ignition temperature of dust layer for icing 
sugar, maize grain dust, wheat grain dust, barley grain dust, alfalfa, bread-making 
wheat and soybean dust was reported to be 400?C, 420?C, 510?C, 480?C, 460?C, 
440?C and 560?C respectively (Ramirez et al., 2009). The hot surface ignition 
temperature is determined by ASTM E2021 standard (2010). A hot plate is used 
to heat the sample placed on a metal plate (200 mm in diameter and 20 mm thick 
) confined at the sides by a metal ring (12.7 mm in diameter and 10 mm high). 
Temperature of metal plate and sample is recorded by a data logger through 
  
21 
 
thermocouples. The minimum temperature of the hot plate that ignites the sample 
is the minimum hot surface ignition temperature. 
 
2.9 Factors Affecting Dust Ignition 
 
2.9.1 Particle Size  
 Dust particles are readily combustible because the surface area to volume 
ratio is significantly higher than the bulk material they are derived from (figure 2.6). 
Higher surface area to volume ratio means more oxygen interacts with the material 
during combustion. Oxygen is a prerequisite for combustion and combustion starts 
at surface of the particle. In addition, the energy required is smaller for smaller 
sized material since there is limited conductive heat transfer when compared to 
combustion of bulk solids (Eckhoff, 2003). 
 
Figure 2.6 Increase in rate of combustion with increasing surface area (Eckhoff, 
2003). 
 Particle size of dust particles also affects the dust burning rate and ignition 
front speed. Ryu et al. (2006), in their experiment with willow, miscanthus and pine 
  
22 
 
(M.C. <8%) showed that the burning rate for samples reduced from about 185 
kg/m3hr to about 165 kg/m3hr as the size of the samples increased from 5 mm to 
35 mm. Ignition front speed was also affected adversely with increase in particle 
size. It decreased from 0.93 m/hr to about 0.7 m/hr as the size increased from 5 
mm to 20 mm. Mass loss during ignition propagation also decreased from about 
87% to 80% as size increased from 5 mm to 35 mm. In summary, as the size of 
particles increases, burning rate and ignition front speed decreases.  
Chen et al. (1996) also showed that heterogeneous ignition temperature for 
Kaipin (bituminous) coal and Hongay coal samples shifted to higher temperatures 
as particle size increased. This is because as particle size increases, the rate of 
heating of particle surface becomes slower. For samples with sizes 105-149 ?m, 
350-500 ?m, 1410-2000 ?m, and 4000 ?m, heterogeneous ignition temperatures 
were found to be 500?C, 535?C, 600?C and 630?C respectively. Also, the ignition 
temperatures of Hongay coal samples of particle sizes 350-500 ?m, 1410-2000 
?m and 4000 ?m, were measured to be 515?C, 560?C and 625?C respectively.  
Vamvuka and Sfakiotakis (2011), showed that particle size had effect on 
ignition temperature and temperature of maximum mass loss rate for sewage 
sludge samples which were air dried, and separated into three different size 
fractions viz. <250 ?m, <500 ?m and <1000 ?m. Ignition temperature for the 
samples increased from 229?C to 242?C whereas, the temperature of maximum 
mass loss rate increased from 515?C to 527?C as the particle sieve size increased 
from <250 ?m to <1000 ?m respectively.  
  
23 
 
2.9.2 Moisture Content 
Shi and Chew (2011) in their study showed that moisture content had little 
effect on ignition temperatures for different types of wood samples (pine, beech, 
cherry, oak and maple wood). Correlation coefficient between moisture content 
and ignition temperature was obtained to be 0.12. Range of ignition temperatures 
for dry and wet woods (11% moisture) were recorded as 267-525?C and 264-558?C 
respectively. The ignition time however increased with increase in moisture content 
as more energy is required to reach ignition in a wet sample.  Range of ignition 
time for dry wood samples was obtained as 7-49 s whereas for wet wood samples 
it was 10-119 s. 
2.9.3 Volatile Content 
 Ignition temperature is dependent upon volatile matter content because 
released volatiles further fuels the ignition of unignited particles. Chen et al. (1995) 
found that ignition temperature of coal increases with decreasing volatile matter. 
Thus, they concluded that volatile matter is the most important factor affecting 
ignition.  Grotkjaer et al. (2003), measured the ignition temperature for poplar 
wood (volatile content: 75% d.b.) and eucalyptus wood (volatile content: 64% d.b.) 
to be 235?C and 285?C respectively. Using data for coal from previous studies 
(Tognotti et al., 1985; Zhang and Wall, 1994; Chen et al., 1995; Chen et al., 1996), 
they showed that ignition temperature decreases with increase in volatile matter 
(figure 2.7). 
  
24 
 
 
Figure 2.7 Relationship between volatile content (dry basis) and ignition 
temperature for biomass and coal samples (Grotkjaer et al., 2003). 
 
2.9.4 Ash Content 
Ash largely comprises of inorganic materials such as Al2O3, NaCl, SiO2, 
KCl, MgO and CaSO4, (Wang et al., 2011). Ash act as inhibitors to ignition as they 
are incombustible and therefore do not contribute towards ignition and act as heat 
sink. High ash content in biomass can cause ignition and combustion problems 
during bioenergy conversion processes such as gasification and pyrolysis 
(Demirbas, 2004). For example, the presence of alkali metals in biomass can 
cause fouling, slagging and ash agglomeration (Ryu et al., 2006). Liodakis et al. 
(2002), in their study with ground (0.3-0.5 mm) forest plant species (Pistacia 
lentiscus, Cupressus sempervirens, Olea europaea and Cistus incanus) showed 
that ignitability of samples decreases with increase in ash contents (i.e. the ignition 
delay increases). Samples of different species with ash contents of 3.51, 3.52, 
5.64, 5.06 and 4.41% (% mass dry basis) had ignition delay values at 500?C of 30 
  
25 
 
s, 36 s, 47 s, 48 s and no ignition respectively. Species with ash content 4.41-
5.64% were classified as least flammable whereas species with ash content 2.96-
3.52% were classified as most flammable species.  
Vuthaluru (2004) showed that addition of coal (ash content 9.7% on dry 
basis) to wheat straw (ash content 3.3% on dry basis) and wood waste (ash 
content 0.1% on dry basis) increased the temperature of onset of rapid mass loss. 
The temperature of onset of rapid volatilization for 0:100, 70:30, 90:10 and 100:0 
(blend coal:wheat straw mass/mass), were found to be 260?C, 316?C, 366?C and 
426?C respectively. Coal and wood waste blends (ratio 0:100, 70:30, 90:10 and 
100:0), had temperatures of onset of rapid volatilization to be 291?C, 345?C, 350?C 
and 426?C respectively. This shows that the presence of ash retards 
oxidation/ignition process of biomass. 
 
2.10 NIR Spectroscopy (NIRS) 
2.10.1 Introduction 
Near Infrared (NIR) spectroscopy is an inexpensive and quick method for 
predicting the concentration of the constituents of a sample (Foley et al., 1998). 
This spectroscopy method involves measuring the amount of light reflected from a 
sample within the wavelength range of 750 to 2500 nm (13333 cm-1 to 4000 cm-1) 
(Lu and Bailey, 2005). Amount of near infrared light reflected is a function of the 
chemical composition and microstructure of the material. Absorbance of the NIR 
radiation by a material is governed by Beer-Lambert?s law. Beer-Lambert law 
  
26 
 
relates the radiant power in a beam of electromagnetic radiation to length of path 
of beam travel within a material and concentration of the absorbing material 
(Swinehart, 1962) (equation 2.4). Some other advantages of NIRS include non-
destructive measurement; ease of sample preparation, ability to be used by low 
skilled operator and high data/spectrum acquisition rates (Vergnoux et al., 2009). 
 ? = ????10 ??
0
= ??? (2.4) 
Where A is absorbance (Absorbance units, Au), 
P is radiant power of reflected light, 
P0 is radiant power of incident light, 
a is absorptivity (coefficient), 
b is length of the beam in absorbing medium 
c is concentration of absorbing medium. 
The first application of NIRS was for measuring the moisture content of 
grains (Norris, 1964). NIRS is now widely used in many applications, such as 
measuring the solid content, firmness of fruits and post-harvest quality of fruits 
(Nicolai et al., 2008; Bobelyn et al., 2010), moisture content, water activity and salt 
content of meat (Collell et al., 2011), quality control of potato chips (Shiroma and 
Rodriguez-Saona, 2009), taste characterization of fruits (Jamshidi et al., 2012), 
proximate analysis and heating values of torrefied biomass (Via et al., 2013). 
Multivariate techniques such as principal component analysis (PCA) and partial 
least square (PLS) regression are used to analyze the complicated NIRS raw 
spectra. 
  
27 
 
Quantifying heating and ignition properties requires expensive equipment 
such as thermogravimetric analyzer (TGA) and differential scanning calorimeter 
(DSC) and is also time consuming. Thus, we attempted to use NIRS to develop 
prediction models for quick prediction of heating and ignition properties of biomass 
dusts. 
2.10.2 Analysis Techniques 
2.10.2.1 Principal Component Analysis 
Principal component analysis (PCA) is the most widespread multivariate 
statistical technique used in chemometrics (Brereton, 2007). PCA is used to obtain 
systematic variations in a given data set (Kettaneh et al., 2005) and for 
classification, description and interpretation of NIR spectral data (Vergnoux et al., 
2009). The PCA method involves the modeling of the variance or covariance 
structure of a given data set by obtaining principal components. This makes it 
possible to reduce large number of data points or variables to a few principal 
components that are then used for model development. Principal components are 
assumed to be independent of each other with no correlation amongst them.  
2.10.2.2 Partial Least Square Regression Analysis 
 Partial least square regression analysis is based on the relationship 
between signal intensity and sample properties (Martens, 1979).  PLS was first 
proposed by Herman Wold (a Swedish statistician) who used it as a tool for 
economic forecasting (Brereton, 2007). It takes into account the full spectral region 
rather than unique and isolated absorption bands (Vergnoux et al., 2009). The 
  
28 
 
algorithm involved in the analysis helps to mathematically correlate the spectral 
data and properties of the material while accounting for all significant factors (Liang 
and Kvalheim, 1996). Samples with known or measured properties are used to 
formulate a calibration model and the samples with unknown parameters are then 
employed in the model to calculate the unknown properties.  
The values for properties estimated using developed models are compared 
with real or actual values. The comparison gives rise to the standard error of 
prediction (RMSEP) which gives an idea of the performance of prediction (equation 
2.5). 
 
????? = ?(? (?? ???
?)2?
?=1
??? ) (2.5) 
Where, Ci is the actual value, 
 Ci? is the estimated value (as determined by calibration model),  
 dof is the ?degrees of freedom? value.  
The degrees of freedom value is usually equal to N-1 where N is the number 
of samples. Relative error of prediction (REP) (equation 2.6) is also a useful 
parameter which quantifies the ability of the model to predict (Vergnoux et al., 
2009).  
 ??? = 100 (?????
?? ) (2.6) 
Where,  ?? is the mean of known values. 
  
29 
 
2.10.3 Predictions Using NIRS 
Vergnoux et al. (2009) used NIRS to predict physiochemical (age, moisture, 
pH, composting time, temperature and organic carbon) and biochemical 
parameters (soluble fraction, hemicellulose, cellulose and lignin) of industrial 
compost. In principal component analysis, 100% of the spectral variation was 
explained by two principal components (PCs). PC1 accounted for 98% variation 
whereas PC2 accounted for the remaining 2%. PC1 vs. PC2 (spectral data) plot 
was used to distinguish between different stages of composting (figure 2.8). 
Loading graph for PC1 was in accordance with the drying of the compost as the 
spectrum resembled the water spectra. The compost dried with time and this trend 
was clearly shown in PC1 vs. PC2 plot. In the physiochemical and biochemical 
parameters PCA, PC1 explained 42% of the total variation and PC2 explained 17% 
of variation. Correlation loading plot was used to estimate the correlations between 
PCs and compost parameters to be predicted. Best results obtained for 
physiochemical and biochemical parameters were tabulated along with RMSEC, 
R2 calibration, RMSEP, R2 prediction and number of PCs used. Most of the models had 
good R2 values (above 0.90). It was concluded that NIRS and PCA can be used 
to estimate the stages of composting.  
Lu and Bailey (2005) performed experiment for prediction of soluble solids 
content (SSC) and firmness of apple as affected by postharvest storage. Samples 
(apple) were divided into three groups based on different post storage times. 
Actual vs predicted values were plotted for the model developed to predict SSC in 
these three groups. R2 values obtained for the three models were between 0.771 
  
30 
 
and 0.853 whereas standard error of validation (SEV) for these models was found 
to be between 0.42 and 0.55. R2 value obtained from model used to predict SSC 
from the samples of all the groups was 0.818 with SEV as 0.50. Model was also 
developed to predict firmness using NIR spectrometer. A good R2 value of 0.839 
was obtained with SEV as 4.81 showing that NIRS is capable of developing 
efficient prediction models.  
 
Figure 2.8 Example of principal component score plot showing PC1 vs. PC2 
(Vergnoux et al., 2009). 
FTNIR spectroscopy was used to develop prediction models for total soluble 
solids (TSS), total acidity (TA), sugar to acid ratio, firmness and weight for three 
South African plum cultivars - Pioneer, Laetitia and Angeleno (Louw and Theron, 
2010). Figure 2.9 shows absorbance vs. wavelength NIR spectra for the three 
cultivars. The TSS prediction models (single and all cultivars combined) performed 
well with coefficient of determination (R2) values ranging from 0.817 to 0.959 and 
root mean square error of prediction (RMSEP) values ranging from 0.4453 to 0.610 
(% brix). The TA, sugar to acid ratio, firmness and weight prediction models 
  
31 
 
performed well with R2 values of 0.608-0.830, 0.718-0.896, 0.623-0.791 and 
0.577-0.817 respectively. RMSEP values for the above parameters were found out 
to be 0.110-0.194, 0.608-1.590, 12.459-22.760 and 7.700-12.800 respectively.  
 
Figure 2.9 Typical NIR spectra for three different Japanese plum fruit taken at start 
of fruit ripening (Louw and Theron, 2010). 
 
The NIR spectra for different plums followed similar trend with difference in 
absorbance values at different wavelengths. This shows that similar materials will 
show similar NIR spectral trend. The peaks in the NIR spectral graph corresponded 
to the wavelengths associated with the specific chemical composition or 
components of the material which caused variation in the spectra of the three 
samples. For example, the peaks at 970, 1190, 1450 and 1940 nm (10309, 8403, 
6897 and 5155 cm-1) are due to pure water (Rambala et al., 1997) and peaks at 
970 and 1190 nm (10309 and 8403 cm-1) may also be due to sugar content 
  
32 
 
(Osborne et al., 1993). Thus, water and sugar content apart from other constituents 
of the fruit were the causes of variation in the spectra (figure 2.9). Actual vs 
predicted values for given properties were also plotted for all-cultivar model to 
validate the developed model (figure 2.10).  
 
Figure 2.10 Actual vs. predicted values for TA (a), TSS (b), firmness (c) and weight 
(d) for multi cultivar NIR model (Louw and Theron, 2010). 
Via (2013) developed models for predicting load capacity and deflection for 
oriented strand construction board in presence of phenol formaldehyde resin using 
NIRS. Models for prediction of load capacity were found to be better with R2 value 
of 0.69 than the model for deflection prediction which had R2 value of 0.59.  
Via et al. (2013) collected NIR spectra between 10000 and 4000 cm-1 and 
mid IR (FTIR) spectra at a different wavenumber range (4000 and 650 cm-1) for 
quick determination of proximate analysis and heating value of torrefied biomass 
  
33 
 
(pine, sweetgum and switchgrass). PCA technique was used to develop NIR and 
FTIR models. Actual vs. predicted values obtained from NIRS was also plotted for 
different parameters. R2, adjusted R2, RMSEC and RMSEP values obtained for 
NIR and FTIR models were tabulated and used to compare prediction efficiency of 
different models. It was concluded that NIR performed well for most of the 
multivariate models than FTIR. The R2 values obtained for prediction of moisture, 
ash, volatiles, fixed carbon and HHV of the samples were 0.85, 0.92, 0.99, 0.99 
and 0.92 respectively.   
 
 
 
 
  
34 
 
Summary 
 The United States and other countries rely heavily on fossil fuels which are 
available for a limited period of time and which have a negative impact on the 
environment. Thus, the focus is shifting towards renewable sources of energy such 
as biomass, solar energy and wind energy to meet the future world energy 
demand. The advantage that biomass has over other forms of renewable energy 
is that it is the only renewable source which can be used for producing liquid fuels, 
chemicals and other products. However, biomass have to be processed and 
handled before it could be converted into biofuels. Equipment such as mills, 
grinders, silos, hoppers and conveyors that are used to process and handle 
biomass can lead to dust generation. If dust is combustible, the presence of ignition 
source will ignite the dust that can result in a fire or explosion incident. It is therefore 
important to determine the ignition properties of biomass (e.g. minimum hot 
surface ignition temperature, temperature of onset of rapid volatilization, 
temperature of maximum rate of mass loss, oxidation temperature, temperature of 
rapid exothermic reaction and maximum temperature reached during exothermic 
reaction). Furthermore, study on physical properties of dust such as moisture 
content and particle size is also important as it affects dust ignition behavior and 
can be beneficial in designing dust removal and fire suppression systems. 
  
35 
 
Our review of literature shows that dusts with smaller particle are more 
susceptible to ignition. Similarly, dusts with high volatile content and low ash 
content have higher in terms of risk associated with ignition. Near infrared 
spectroscopy (NIRS) has been used to analyze and predict a wide range of 
properties for different materials in the past. It is a quick and inexpensive way to 
analyze effect of chemical and physical properties of biomass on different 
parameters. NIRS can be combined with statistical techniques such as principal 
component analysis (PCA) and partial least squares (PLS) regression analysis to 
develop prediction models for heating and ignition characteristics of biomass 
dusts. 
 
 
 
 
 
 
 
 
 
  
36 
 
Chapter 3 Physical, Chemical and Heating and Ignition Properties of 
Biomass and Coal Dusts 
3.1 Abstract 
 Plants and refineries utilizing biomass to produce energy have to 
preprocess, store and handle biomass several times using milling, grinding, sieving 
and conveying equipment.  These operations lead to dust generation and the dust 
settling on floor and equipment surfaces in the plant or refinery. Accumulated layer 
of dust on hot surfaces can ignite that can lead to fire and explosion. Therefore, 
this study was conducted to characterize heating and ignition properties of dusts 
from 10 biomass feedstocks and three types of coal. The range of values obtained 
for these properties were 240?C-335?C (MIT), 266?C-448?C (TORV), 304?C-
485?C (TMML), 242?C-423?C (TOXY), 206?C-249?C (TRE) and 354?C-429?C 
(TME). Physical and chemical properties of ground biomass and coal feedstock as 
well as dusts were also quantified. The effects of these properties on heating and 
ignition parameters of dusts was also studied. For grassy biomass, dusts with 
higher ash content had significantly higher MIT (p<0.0001). Also, grassy biomass 
dust with higher volatile matter have lower MIT (p=0.025). TORV and TMML of 
coal dusts were found to decrease with increase in their volatile contents. Grassy 
biomass dusts with higher ash content and lower volatile matter had significantly 
higher activation energy values (p<0.0001) whereas, woody biomass dusts with 
  
37 
 
higher energy content had significantly higher TME values (p=0.0013). Based on 
TORV and TMML values, biomass dusts were found to be at higher risk of ignition 
than coal dusts. Based on the exothermic energy values, most of the biomass 
dusts (all except poultry litter dust) are associated with higher release of energy 
during a dust explosion or fire event than coal dusts. 
 
3.2 Introduction 
 The world population is currently estimated to be about 7.2 billion and 
projected to increase to about 9.4 billion by the year 2050 (USCB, 2014). Since 
energy is needed for sustenance, energy consumption will increase 524 quads in 
2010 to 820 quads in 2040 because of this projected increase in population (EIA, 
2013). Most of the energy requirement of the world is currently met from 
nonrenewable fossil fuels (EIA, 2013) that also pose environmental challenges. To 
meet the increasing energy requirement, more focus is being given to renewable 
sources of energy such as wind energy, solar energy, hydroelectric power and 
bioenergy. This is partly responsible for the increase in consumption of renewable 
energy in U.S. from 3.0 quadrillion Btu in 1950 to 8.8 quadrillion Btu in 2012 (EIA, 
2013).  
Bioenergy is the energy derived from biomass which includes, but not 
limited to energy crops, agricultural crops, food, fiber, feed, forest products, aquatic 
plants, wood residues, industrial and residential waste, processing byproducts and 
non-fossil organic material (ASABE S593.1, 2011). Bioenergy is the only 
renewable source of energy that can supply the liquid fuels needed in the industrial 
  
38 
 
and transportation sectors. In 2012, about 50% (4.4 quadrillion Btu) of the 8.8 
quads of renewable energy consumed was derived from biomass (EIA, 2013).  
 Biomass supply logistics is one of the challenges that has limited its 
conversion to energy. Biomass logistics consists of biomass harvesting, collection, 
storage, transportation, pretreatment, and storage. Biomass is low in energy 
density, high in moisture content and has to be harvested from geographically 
dispersed locations. As a result there are high logistics costs involved in delivering 
biomass to conversion plants (An and Searcy, 2012). The various operations 
involved in the delivery logistics may lead to dust generation (Abbasi and Abbasi, 
2007; Eckhoff, 2003). Since biomass feedstocks are combustible (McKendry, 
2002), the dust generated from them can cause fire and explosion in process 
plants. Workers exposed to biomass dusts can also develop health problems 
(Khan et al., 2008).  
The National Fire Protection Association standard 654 defines combustible 
dusts as ?particles passing through a 500 ?m sieve which presents a dust fire or 
dust explosion hazard? (NFPA, 2013).  Accumulated layer of combustible dust on 
hot surfaces can ignite and cause fire hazard or explosion. These dust layers may 
also be disturbed by an external event such as blowing air or mechanical 
disturbance resulting in suspension of the dust to form dust cloud. If the 
concentration of the dust cloud surpasses a critical limit, and the dust cloud is 
formed in a confined space, dust fire and explosion can occur when an ignition 
source such as electric spark, hot surface or a naked flame comes in contact with 
the cloud (Amyotte and Eckhoff, 2010).  
  
39 
 
Particle size of dust plays an important role in dust ignition. Materials with 
less particle size has large surface area to volume ratio as compared to material 
with higher particle size which leads to more oxygen availability for combustion at 
the surface making the material more readily combustible (Eckhoff, 2003). As the 
dust particle size decreases in a layer of dust, its ignition temperature also 
decreases making the material easier to ignite (NFPA, 2013). Chen et al. (1996) 
found out that ignition temperatures of Kaipin (bituminous) coal and Hongay coal 
increased as particle size of the samples increased. Kaipin coal samples with size 
fractions of 105-149 ?m, 350-500 ?m, 1410-2000 ?m and 4000 ?m were found to 
have ignition temperatures of 500?C, 535?C, 600?C and 630?C respectively. 
Hongay coal samples with size fractions of 350-500 ?m, 1410-2000 ?m and 4000 
?m were found to have ignition temperatures as 515?C, 560?C and 625?C 
respectively.  
Chemical properties of dust material such as volatile matter and ash 
contents also affect ignition temperature. Volatile matter acts as fuel for 
combustion with ignition temperature decreasing as volatile content increases 
(Tognotti et al., 1985; Zhang and Wall, 1994; Chen and Mori, 1995; Chen et al., 
1996). Grotkjaer et al. (2003) measured the ignition temperatures of poplar wood 
(volatile content: 75% d.b.) and eucalyptus wood (volatile content: 64% d.b.) to be 
235?C and 285?C respectively - an increase in ignition temperature with decreasing 
volatile matter. Ash largely comprises of the inorganic material which do not 
contribute towards ignition and therefore act as heat sink and hindering the ignition 
process (Wang et al., 2011). Vuthaluru (2004) showed that addition of coal (ash 
  
40 
 
content 9.7% d.b.) to wheat straw (ash content 3.3% d.b.) and wood waste (ash 
content 0.1% d.b.) samples increased the temperature of onset of rapid mass loss. 
For 0:100, 70:30, 90:10 and 100:0 (blend coal:wheat straw mass/mass), the 
temperature of onset of rapid volatilization were found to be 260?C, 316?C, 366?C 
and 426?C respectively. Coal and wood waste blends (ratio 0:100, 70:30, 90:10 
and 100:0), had temperatures of onset of rapid volatilization to be 291?C, 345?C, 
350?C and 426?C respectively. This shows that the presence of ash retards 
oxidation/ignition process of biomass. 
 The study of volatilization properties of biomass is needed to understand 
the ignition behavior of biomass and coal. Thermal degradation of biomass 
includes release of moisture and volatiles followed by char oxidation. Gil et al. 
(2010) in their experiment with pine sawdust showed that temperature ranges for 
release of water, release and combustion of volatiles, and char oxidation are 25-
105?C, 196-364?C and 364-487?C respectively. The lower the temperature of 
release and combustion of volatiles in a biomass would be, the higher risk of its 
ignition. Ramirez et al. (2010), estimated temperature of onset of rapid 
volatilization (TORV) and temperature of maximum mass loss rate (TMML) for 
icing sugar to be 212?C and 220 ?C respectively. For maize, wheat and barley 
TORV values were estimated at 268?C, 252?C and 242?C respectively whereas 
TMML values were estimated at 279?C, 283?C and 271?C. Both, TORV and TMML 
values for maize, wheat and barley were more than that of icing sugar indicating 
that icing sugar has more ignition risk. Oxidation temperature (TOXY) values for 
  
41 
 
icing sugar (239 ?C), was also smaller than those of maize (289 ?C), wheat (279 
?C) and barley (277 ?C).  
Exothermic properties are also used to characterize the ignition behavior of 
combustible material. This includes determination of the maximum temperature 
reached during exothermic reaction (TME), temperature required for onset of rapid 
exothermic reaction (TRE) and exothermic energy from heat flow curves (Ramirez 
et al., 2010). Dust sample with lower TRE value would be easily ignited than the 
dust with higher TRE value. Samples with higher values of TME would promote 
secondary dust fire or ignition by providing sufficient energy for the reaction. Dusts 
with higher exothermic energy release greater amount of energy during ignition or 
explosion than samples with lower exothermic energy value. The heat energy 
released could lead to secondary dust explosions or fires in a processing plant or 
facility leading to further destruction. Ramirez et al. (2010) estimated TME values 
for maize, wheat, barley and alfalfa as 386?C, 283?C, 311?C and 288?C 
respectively and TRE values for the given samples to be 242?C, 252?C, 257?C and 
240?C respectively. 
Determination of heating and ignition properties of biomass dusts is 
important in order to quantify their ignition risk. It is also important to measure the 
physical and chemical properties of biomass dusts as they affect the heating and 
ignition behavior. Therefore, the objective of this study was to characterize the 
heating and ignition properties, and the physical and chemical properties of 
biomass dusts. 
 
  
42 
 
3.3 Methods and Materials 
3.3.1 Raw Material 
 The ten biomass feedstocks, and the three coal types used in this study are 
listed in table 3.1. Moisture content of the raw samples was measured (in 
triplicates) using ASTM E871-82 standard (ASTM E871-82, 2006).  About 2 g of 
raw sample was placed in a conventional oven at 105?C for 24 hours (VWR 
International model 1370FM, Sheldon Mfg., OR, USA) (figure 3.1). Sugarcane 
bagasse and sweetgum chips were dried prior to grinding because of their high 
moisture content. Sweetgum chips were air dried under a shed at Agricultural Land 
and Resource Management Center using fans for a week (figure 3.2a). Sugarcane 
bagasse was dried at 45?C with a food dehydrator (Excalibur Food Dehydrator, 
Sacramento, CA, USA) (figure 3.2b). Moisture contents of dried samples were also 
determined. Moisture content measurements are presented on wet basis. 
3.3.2 Grinding and Dust Collection 
The samples were ground with a hammer mill (C.S. Bell Co., model 10HBLPK, 
Tiffin, OH, USA) (figure 3.3 a) fitted with a 3.175 mm (1/8th inch) screen. Dust was 
obtained from the ground material by passing it through #35 market grade (437 
?m) screen using a vibratory sieve shaker (Kason Corp., model K30-2-8S, NJ,   
USA) (figure 3.3 b). This is the closest screen size to the NFPA 654 standard?s 
500 ?m size definition of dust (NFPA, 654).  Pecan shell and lignite coal samples 
obtained were already in dust form and were not ground using hammer mill. Dust 
collected from each feedstock was stored in three 80 oz. air tight containers for 
  
43 
 
further analysis that includes characterization of physical, chemical, heating and 
ignition properties and near infrared spectroscopy (NIRS) analysis. Ground 
material (1/8th inch hammer mill screen size) for biomass feedstock and coal 
samples was also stored in air tight containers for further analysis of physical and 
chemical properties. 
 
Table 3.1 List and sources of biomass feedstocks and coals used in the study. 
Sample Source 
Corn Stover Purdue University 
Corn Cobs Purdue University 
Sugarcane Bagasse Louisiana State University 
Sweetgum Auburn University 
Poultry Litter Department of Poultry Science, Auburn University 
Loblolly Pine South Alabama forests 
Eucalyptus Auburn University 
Bermuda Grass North Alabama 
Pulverized river bed (PRB) coal (sub-bituminous) National Carbon Capture Center, Wilsonville 
Bituminous Coal Alabama Power, Birmingham 
Lignite Coal National Carbon Capture Center, Wilsonville 
Pecan Shells Louisville Pecan Company 
Switchgrass E.V. Smith Research Station 
  
44 
 
 
Figure 3.1 Conventional oven used for moisture content determination. 
 
 
 
  
 
 
 
 
 
 
(a) 
 
                        (b) 
Figure 3.3 Hammer mill used for grinding biomass feedstock (a) and vibratory 
screen used for collection of dust (b). 
 
Figure 3.2 Air drying of sweetgum wood chips (a) and drying sugarcane bagasse 
in food dehydrator (b). 
(a) (b) 
  
45 
 
3.3.3 Physical and Chemical Properties  
3.3.3.1 Moisture Content 
 Moisture content of the ground material and dust collected from each 
biomass feedstock and coal type was measured using a moisture content analyzer 
(IR 200, Denver Instrument, Avrada, CO) according to ASTM E871-82 standard 
(ASTM E871-82, 2006). This involved about 2g sample on the sample pan of the 
analyzer and exposing the sample to temperature of 105?C until the balance of the 
moisture meter did not detect a change in sample mass greater than 0.05% per 
minute. The experiment was carried out in triplicates and equation 3.1 was used 
to estimate moisture content of samples. 
 
 ???? = ???? ? ?????
???
 ? 100 (3.1) 
 
where, 
????= moisture content - wet basis (%), 
???? = initial mass of sample (g), and 
???? = final mass of sample (g).  
3.3.3.2 Bulk Density 
 Bulk density was measured in triplicates with a bulk density apparatus 
(OHAUS, Burrows Co., Evanston, IL) (figure 3.4) on all ground and dust samples 
using the ASABE standard S269.5 (ASABE, 2012). All the samples were dried for 
  
46 
 
24 hours at 105?C to nullify the effect of moisture. Sample was poured from a set 
height of 610 mm (above top edge of container) through a funnel into a container 
of known volume (1137 ml). The mass of the dust sample required to fill the 
container was recorded and bulk density was calculated as follows:  
 
 ????? = ?? (3.2) 
 
Where, 
????? = bulk density (kg/m3), 
M = mass of sample (kg) and, 
V = volume of container (m3). 
 
3.3.3.3 Particle Density 
 A gas pycnometer (AccuPyc 1330, Micrometrics Instrument Corp., 
Norcross, GA, USA) (figure 3.5) was used to estimate average volume of the 
particles (based on three replications). The pycnometer estimates the volume of 
particles by passing helium gas through the chamber containing known mass of 
sample and measuring the change in pressure. Particle density was calculated as 
ratio of mass of dust sample to the volume obtained from pycnometer. All the 
samples were dried for 24 hours at 105?C before measuring the particle density to 
reduce the effect of moisture content. 
  
47 
 
 
Figure 3.4 Apparatus for measuring bulk density. 
 
 
Figure 3.5 Gas pycnometer used for estimation of particle density. 
 
3.3.3.4 Particle Size Distribution 
 A particle size analyzer that uses digital imaging system (CAMSIZER, 
model D-4278, Haan, Germany) (figure 3.6) was used to estimate volume based 
particle size distribution of the ground and dust sample. All samples (ground and 
  
48 
 
dusts) were dried for 24 hours at 105?C before analyzing the particle size 
distribution. Geometric mean diameter and geometric standard deviation was 
calculated according to ASABE S319.3 standard (ASABE, 2003) from the data 
obtained from the imaging system (equations 3.3 ? 3.5).  
 ??? = ????1 [? (????????)
?
?=1
? ????=1 ] 
(3.3) 
 
 
???? = [? ??(log ??
? ???? ???)2 ??=1
? ????=1 ]
1 2?
 (3.4) 
   
 ??? = 12 ??? [????1???? ?(????1????)?1] (3.5) 
 
Where,  
dgw is geometric mean diameter of particles (mm), 
Slog is geometric standard deviation of log-normal distribution by mass in base 10 
logarithm (dimensionless), 
Sgw is geometric standard deviation of particle diameter by mass (mm), 
n is the number of sieves + 1 (pan), 
di  = (di x di+1)1/2 , 
di is nominal sieve aperture size of the ith sieve (mm), 
di+1 is the nominal sieve aperture size of the ith + 1 sieve (mm), 
Mi is the mass of the sample retained on the ith sieve (g).  
 
  
49 
 
 
Figure 3.6 Digital imaging particle size analyzer. 
 
3.3.3.5 Ash Content 
 The ash content of the ground and dust samples was determined according 
to National Renewable Energy laboratory (NREL, 2005) laboratory analytical 
procedure. Porcelain crucibles were marked and kept in a muffle furnace 
(Thermoscientific, model F6020C, Dubue Iowa, USA) (figure 3.7) at 575?C for four 
hours. Crucibles were then removed and placed directly into the desiccator till they 
cooled down for about an hour. The mass of the dried crucibles was recorded and 
known amount of sample (within a range of 0.5-2 g) was placed into crucibles. 
Crucibles were placed back into the furnace set at 105?C and furnace was held at 
that temperature for 12 minutes. Furnace was then ramped to 250?C at 10?C/min 
and was held at this temperature for 30 minutes. Furnace was again ramped to 
575?C at 20?C/min and was held at that temperature for 180 minutes. Furnace was 
allowed to cool to 105?C and the samples were removed. Ash content was 
  
50 
 
calculated according to ASTM standard (ASTM D 3174-04, 2004) using equation 
3.6. Experiment was triplicated for each sample.  
 ??? = [??? ????
?
] ? ( 100100 ??
??
) (3.6) 
 
where, 
??? = estimated percentage of ash in biomass (%), 
??? = final mass of crucible with ash after completion of experiment (g), 
?? = mass of empty crucible (g) and, 
?? = initial mass of biomass sample (g), and 
Mwb = moisture content of biomass sample on wet basis (%) 
 
 
Figure 3.7 Muffle furnace used for ash content determination. 
  
51 
 
3.3.3.6 Volatile Matter Content 
 The volatile matter content of ground and dust samples of the biomass and 
coal samples were measured using ISO 562 standard (ISO 562, 2002). Crucibles 
and their lids were cleaned and placed in the volatile matter furnace (VMF 
Carbolite, model 10/6/3216P, England) (figure 3.8) that was at a temperature of 
900?C for an hour. They were then placed in a desiccator until they cooled down 
to room temperature. About 1?0.1 g of sample was added into each crucible. 
Crucibles with lids containing samples were placed in the furnace at 900?C for 
seven minutes. They were then taken out and allowed to cool in the desiccator. 
They were weighed again and volatile matter content was determined using 
equation 3.7 (ISO 562, 2002).  Three replications were performed for each dust 
sample. 
 ?? = [100 (?3 ??1)
?2 ??1 ] ? (
100
100 ? ???) (3.7) 
 
Where, 
??= volatile matter content on dry basis (%), 
?1= mass of empty crucibles and lid (g), 
?2= initial mass of crucible, lid and sample (g), 
?3= final mass of crucible, lid and sample (g) and, 
???= moisture content on (w.b. %) 
 
  
52 
 
 
Figure 3.8 Volatile matter determination furnace. 
 
3.3.3.7 Energy Content 
 A bomb calorimeter (IKA Works Inc., model C200, Wilmington, NC, USA) 
(figure 3.9) was used to estimate the energy content of the ground and dust 
samples. About 0.5 to 1 g of sample was compressed to a pellet form using a press 
(IKA Works Inc., model C21, Wilmington, NC, USA) and was weighed. The pellet 
was connected to the ignition wire using a thread and was placed in a pressurized 
(with oxygen) decomposition chamber. The decomposition chamber was placed 
inside the bomb calorimeter where the sample was combusted completely and 
energy content was recorded. Experiment was performed in triplicates. 
 
  
53 
 
 
Figure 3.9 Bomb calorimeter used for energy content determination. 
 
3.3.4 Heating and Ignition Properties 
3.3.4.1 Hot Surface Ignition Temperature 
 Hot surface ignition temperature of dust samples was determined according 
to ASTM E2021 standard (2010). A hot plate (VMWare, Thorofore, NJ, USA) 
(figure 3.10) was used to heat the dust sample and to estimate hot surface ignition 
temperature. Apparatus consisted of a hot plate, a cylindrical metal plate (200 mm 
in diameter and 20 mm thick), a metal ring (12.7 mm in diameter and 10 mm thick) 
and a bare wire type-K thermocouple wire (0.20 mm dia) connected to a 
datalogger. The metal plate rested on the hot plate and the metal ring was placed 
on top of the metal plate. The bare type thermocouple was installed through the 
holes in the metal ring at a height of 5 mm. Another K-type thermocouple 
(insulated, 0.25 mm dia) was also connected to the metal plate to measure its 
temperature. Temperatures from both thermocouples were recorded using a 
datalogger (Fuji Electric Systems Co., Ltd., Tokyo, Japan) at 5 s intervals. Desired 
temperature was set initially on the temperature controller of the hot plate. Once 
  
54 
 
the desired temperature was reached, the metal ring was filled with the dust 
sample. Sample was leveled with the top of metal ring. Temperature of the dust 
layer was monitored and recorded continuously. If ignition did not occur, a new 
dust sample layer is used and the test was repeated with a new set temperature 
5?C higher than the previous one. Tests were repeated until ignition occurred and 
was confirmed by running two more tests, at maximum temperature where ignition 
did not occur and at minimum temperature at which ignition occurred.  
 
  
 
Figure 3.10 Apparatus to measure minimum hot surface ignition temperature (a), 
metal ring filled with dust sample before ignition (b), dust sample after ignition (c). 
(a) 
(b) (c) 
 
 
Hot plate 
  
55 
 
3.3.4.2 Volatilization Properties 
 Temperature of onset of rapid volatilization (TORV), temperature of 
maximum rate of mass loss (TMML) and oxidation temperature (TOXY) were 
estimated using mass loss curves obtained from thermogravimetric analyzer 
equipment (TGA, model Pyris1, PerkinElmer, Shelton, CT, USA) (figure 3.11). 
About 5 mg of sample was heated in air and oxygen environments from 30?C to 
800?C at 20?C/min. TORV and TMML temperatures were obtained from mass loss 
data of samples heated in air environment while TOXY was estimated from mass 
loss data of samples heated in oxygen environment since a single oxidation 
temperature could not be obtained when sample is heated in air (Ramirez et al., 
2010). The desired temperatures were estimated with the software supplied by 
TGA equipment manufacturer. Experiments were conducted in triplicate.  
 
Figure 3.11 Thermogravimetric analyzer (TGA) used to measure volatilization 
properties. 
  
56 
 
Isoconversional method was applied to the air atmosphere mass loss data 
to estimate activation energies associated with biomass dust volatilization. This 
method has been widely used to estimate kinetic parameters during thermal 
decomposition of biomass (Park et al., 2009; Biagini et al., 2008; Ramirez et al., 
2010). The mass loss data was used to calculate conversion (?) at any time t, using 
equation 3.8 (SSCHE, 2009; White et al., 2011; Park et al., 2009). 
 ? = ?0 ???
0 ? ??
 (3.8) 
where, 
?0 is initial mass of the sample (mg), 
? is actual sample mass (mg), 
?? is residual mass at the end of TGA experiment (mg).  
Rate of decomposition (????) is a function of temperature and conversion 
(equation 3.9) (White et al., 2011; Friedman, 1963). 
 ???? = ?(?)?(?) (3.9) 
where, 
k(T) is a function of temperature, 
g(?) is a function of conversion. 
 Temperature dependent function is expressed by Arrhenius equation 
(equation 3.10) whereas the conversion function is given by equation 3.11 (White 
et al., 2011; Friedman, 1963). 
  
57 
 
 ?(?) = ? ???(????) (3.10) 
 ?(?) = (1??)? (3.11) 
Where, 
A is pre-exponential factor (1/s), 
E is activation energy (J/mol), 
R is universal gas constant (8.314 J K-1 mol-1). 
 Using natural logarithmic function equation 3.9 can be rewritten as equation 
3.12 (Ramirez et al., 2010). 
 ??(????) = ??? +? ? ln(1??)? ??? (3.12) 
A graph of ??(????) vs inverse of absolute temperature (1/T) was plotted. 
The activation energy (E) was obtained from the slope (???). 
3.3.4.3 Exothermic Parameters 
 A differential scanning calorimeter (TA Instruments, model Q200, New 
Castle, DE, USA) (figure 3.12) was used to estimate temperature of rapid 
exothermic reaction (TRE), maximum temperature reached during exothermic 
reaction (TME) and exothermic energy. This involved measuring heat flow required 
to heat about 5 mg of sample from 30?C to 550?C at the rate of 20?C/min. The 
sample was held at 550?C for ten minutes. A plot of heat flow vs. temperature was 
  
58 
 
obtained from the software provided by the manufacturer of the equipment. The 
exothermic reaction parameters were also obtained using the same software.  
 
Figure 3.12 Differential scanning calorimeter (DSC) equipment used to determine 
exothermic parameters of dust samples. 
 
3.3.5 Data Analysis 
 Tukey tests were performed on the measured physical, chemical, heating 
and ignition properties for biomass and coal dusts that were significantly different 
from others (95% significance level) using Statistical Analysis Systems 
(SAS,2009). Tukey test was also performed on measured physical and chemical 
properties for ground biomass and coal samples. Biomass dust samples were 
categorized into grassy biomass (switchgrass, Bermuda grass, corn stover and 
sugarcane bagasse) and woody biomass (eucalyptus, pine, sweetgum) groups 
because of similar nature of the samples in a group before carrying out correlation 
analysis for physical, chemical, heating and ignition properties of the dust samples. 
Graphs were plotted using Microsoft Excel (Microsoft Excel, Redmond, WA). Each 
experimental run was triplicated.  
  
59 
 
3.4 Results and Discussion 
3.4.1 Physical and Chemical Properties 
 Geometric mean diameter and geometric standard deviation values for 
biomass and coal ground samples and dust samples are given in table 3.2 and 
table 3.3 respectively. Values of other physical and chemical properties (bulk 
density, particle density, volatile matter, ash content and energy content) obtained 
for all the ground samples and dust samples are given in table 3.4 and table 3.5 
respectively. These results are discussed in more details below. 
 
Table 3.2 Geometric mean diameter and geometric standard deviation values of 
ground (through 3.175 mm screen size) samples. 
Sample Geometric mean size, d
gw (?m) 
Geometric Standard 
deviation, Sgw (?m) 
Bermuda grass 1013 680 
Bituminous coal 274 179 
Corn cobs 636 512 
Corn stover 568 481 
Eucalyptus 596 512 
Pine 846 668 
Poultry litter 526 524 
PRB coal 302 200 
Sugarcane 
bagasse 554 445 
Sweetgum 847 702 
Switchgrass 1074 590 
 
 
*Pecan shell and lignite coal samples were received in dust form 
  
60 
 
Table 3.3 Geometric mean diameter and geometric standard deviation values of 
dust samples (passing through 437 ?m screen). 
Sample Geometric mean size, dgw (?m) Geometric Standard deviation, S
gw (?m) 
Bermuda grass 526 520 
Bituminous coal 237 128 
Corn cobs 398 246 
Corn stover 337 208 
Eucalyptus 278 178 
Lignite coal 319 221 
Pecan shell 261 126 
Pine 516 349 
Poultry litter 204 126 
PRB coal 216 107 
Sugarcane bagasse 282 175 
Sweetgum 420 388 
Switchgrass 598 575 
 
3.4.1.1 Moisture Content 
 Moisture contents for the biomass feedstock before and after grinding 
operation are depicted in figure 3.13. Moisture contents of the ground samples was 
found to be significantly (?=0.05) less than initial moisture content except for 
bituminous coal, sugarcane bagasse and sweetgum. Loss of moisture can be 
attributed to the heat generated due to particle-particle and particle-hammer 
friction during the grinding process. Also, the higher surface area ground samples 
facilitated the drying process (Probst et al., 2013).  Probst et al. (2013) measured 
initial and post-grinding (hammer mill) moisture content for corn and corncobs 
samples. 
  
61 
 
Table 3.4 Measured physical and chemical properties of ground biomass and coal samples 
Sample Bulk Density (kg/m3) 
Particle 
Density 
(kg/m3) 
Volatile 
Matter  
(% db) 
Ash 
Content  
(% db) 
Energy Content 
(MJ/kg) 
Biomass Samples 
Bermuda grass 127.01?1.65f 1106.60?3.1h 84.51?0.48c,d 4.02?0.15d 19.42?0.06c,d 
Corn cobs 160.24?1.06d 1359.03?1.77e 88.57?0.85a,b 1.06?0.01e 19.25?0.05d 
Corn stover 93.61?1.35g 1267.00?2.51g 81.24?2.02d 7.58?1.59b 18.16?3.2e 
Eucalyptus 196.34?0.91b 1427.47?2.84d 88.36?1.36a,b 0.66?0.52e 19.80?0.06b,c 
Pine 184.44?1.37c 1465.47?2.08c 85.44?1.02b,c 0.48?0.13e 20.68?0.04a 
Poultry litter 271.10?2.70a 1501.27?1.86b 84.83?0.76c 9.43?0.15a 18.35?0.33e 
Sugarcane bagasse 92.20?1.44g 1535.27?5.12a 83.81?0.22c,d 5.84?0.52c 19.01?0.08d 
Sweetgum 166.42?0.71 d 1461.40?0.75c 89.86?1.75a 0.66?0.06e 19.88?0.24b 
Switchgrass 142.13?1.41e 1319.43?3.57f 84.16?0.74c,d 4.94?0.31c,d 19.13?0.14d 
Coal Samples 
Bituminous coal 651.74?1.67a 1385.03?1.88b 38.87?0.24b 10.47?0.95a 31.75?0.58a 
PRB coal 615.41?1.60b 1490.57?3.00a 53.71?0.54a 7.82?0.24a 27.11?0.07b 
*Superscripts with same letters in a column are not significantly different from each other (?=0.05)
  
62 
 
Table 3.5 Measured physical and chemical properties of biomass and coal dust samples 
Sample Bulk Density (kg/m3) Particle Density  (kg/m3) Volatile Matter  (% db) Ash Content  (% db) Energy Content (MJ/kg) 
Biomass Samples 
Bermuda grass 159.28?1.43f 1167.33?1.65h 78.61?0.85c,d 4.68?0.16d 19.14?0.06c 
Corn cobs 164.86?2.00f 1481.43?0.17e,f 81.51?1.51b 2.62?0.08e 19.08?0.05c 
Corn stover 126.52?1.76h 1500.90?4.56d 72.97?0.58e 14.33?0.26b 17.17?0.07e 
Eucalyptus 217.60?2.08c 1490.30?2.19d,e 81.27?0.14b,c 1.40?0.12f 19.51?0.06b 
Pecan shell 403.95?2.61a 1521.93?0.95c 64.93?0.19f 2.68?0.04e 20.31?0.17a 
Pine 173.06?3.43e 1471.87?3.80f 84.97?0.82a 1.10?0.14f 20.55?0.07a 
Poultry litter 360.74?3.71b 1539.23?1.51b 71.43?0.21e 13.74?0.07b 17.43?0.06e 
Sugarcane 
bagasse 112.34?1.74
i 1585.27?2.42a 72.41?1.18e 15.52?0.27a 16.69?0.07f 
Sweetgum 183.57?1.37d 1480.47?2.85e,f 86.82?0.55a 1.41?0.22f 19.71?0.07b 
Switchgrass 141.01?0.63g 1366.17?5.59g 77.86?0.73d 7.57?0.35c 18.77?0.12d 
Coal Samples 
Bituminous coal 651.75?3.88a 1406.20?13.98c 33.00?0.05b 10.27?1.81b 32.26?0.02a 
Lignite coal 503.40?3.34b 1672.20?2.38a 53.69?7.14a 20.84?0.16a 26.78?0.57b 
PRB coal 655.66?3.19a 1505.13?0.66b 47.06?0.53a 8.14?0.07b 27.41?0.16b 
*Superscripts with same letters in a column are not significantly different from each other (?=0.05)
  
63 
 
The author also reported that the mean moisture of corncobs with initial moisture 
contents of 10.04%, 14.65% and 20.13% (w.b.%), reduced to 8.53%, 9.19% and 
12.93% (w.b.%) respectively after grinding. We however found that samples with 
moisture content less than 6% before grinding showed a significant (?=0.05) 
increase in moisture content after grinding. This can be attributed to the 
hygroscopic nature of the material as material with low moisture content (less than 
atmospheric moisture content) tends to absorb moisture from the environment. 
Samples with initial moisture content less than 6% were the ones that were dried 
after receiving. 
 
Figure 3.13 Moisture content before and after grinding biomass and coal. 
 
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
Mo
istu
re 
co
nte
nt 
(%
 wb
)
Feedstock Ground material
*Pecan shell and lignite coal samples were received in dust form 
  
64 
 
3.4.1.2 Particle Size Distribution 
 Geometric mean diameter and geometric standard deviations obtained for 
the ground samples are given in table 3.2. Geometric mean particle size of ground 
material (3.175 mm screen size) for all the feedstock varied from 274 ?m 
(bituminous coal) to 1074 ?m (switchgrass) and as expected were greater than the 
geometric mean size for their respective dust samples (table 3.3).  
Particle size distributions for the different dust samples are shown in figure 
3.14. All the dust samples showed skewness in particle size distribution (log-
normal distribution) which is typically obtained for naturally occurring particle 
populations (Rhodes, 1998; Fasina, 2008). Geometric mean diameter and 
geometric standard deviations obtained for all the dust samples are given in table 
3.2. The geometric mean diameter of particles for the dust samples varied from 
204 ?m (poultry litter) to 598 ?m (switchgrass). The sieve size assigned by NFPA 
(NFPA, 2013), 500 ?m falls within this range. As expected, the geometric mean 
diameter for most dusts was smaller than the sieve size (437 ?m) used to 
fractionate them from ground material. However, in case of switchgrass and 
Bermuda grass, the geometric mean diameter of the dust samples was greater 
than that of sieve size as many long and thin particles managed to pass through 
the sieve (figure 3.15).  
  
65 
 
 
Figure 3.14 Particle size distribution of dust samples. 
 
 
Figure 3.15 Elongated particles in Bermuda grass dust (a) and switchgrass dust (b) 
samples. 
0
2
4
6
8
10
12
14
16
18
20
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Re
tai
ne
d (%)
Particle Size (mm)
Bermuda grass Bituminous coal
Corn cobs Corn stover
Eucalyptus Lignite coal
Pecan shell Pine
Poultry litter PRB coal
Sugarcane bagasse Sweetgum
Switchgrass
(a) 
(b) 
  
66 
 
3.4.1.3 Bulk Density 
Bulk density of ground material varied from 92.20 kg/m3 (sugarcane 
bagasse) to 651.74 kg/m3 (Bituminous coal) whereas for dust samples it varied 
from 112.34 kg/m3 (sugarcane bagasse) to 655.66 kg/m3 (PRB coal). Bulk density 
of ground Bermuda grass, corn stover, eucalyptus, poultry litter, PRB coal, 
sugarcane bagasse and sweetgum was found to be significantly (?=0.05) less than 
bulk density of corresponding dust samples for (figure 3.16). This is because larger 
particles have larger void spaces between them which occupy more volume 
whereas finer particles tend to rearrange themselves to fill the voids between 
coarser/larger particles resulting in more mass per unit volume and thus higher 
bulk density (Mani et al., 2003; Tabil, 1996). Also, higher ash content of dusts leads 
to higher bulk density because of higher particle density of ash (see section 
3.4.1.5). For example, the particle density of ash obtained from sugarcane 
bagasse dust sample was measured to be 2781.5?3.46 kg/m3. Bulk density of 
wheat straw grinds increased from 77 kg/m3 to 115 kg/m3 as geometric mean 
particle diameter decreased from 1.43 mm to 0.25 mm (Mani et al., 2004). Bulk 
density of ground corn cobs and corn cobs dust was not significantly different. This 
can be attributed to the heavier particles (which did not reduce in size below a 
certain size during hammer mill operation) in ground corn cobs sample which 
passed through the 3.175 mm hammer mill screen but could not pass through the 
437 ?m screen used to collect dust.  
  
67 
 
 
Figure 3.16 Bulk density of ground and dusts from biomass and coal. 
 
3.4.1.4 Particle Density 
 Particle density for ground material varied from 1106.6 kg/m3 (Bermuda 
grass) to 1535.3 kg/m3 (sugarcane bagasse) whereas for dust samples, it varied 
from 1167.33 kg/m3 (Bermuda grass) to 1585.27 kg/m3 (sugarcane bagasse). 
Mean particle density for ground material was found to be less than that of the 
dusts for all feedstocks (figure 3.17). This is because individual particles become 
less dense as the particle size increases (Littlefield et al., 2011) as there is 
reduction in porosity within a particle with decrease in particle size (Mani et al., 
2004). Particle densities of ground bituminous coal and pine were not significantly 
(?=0.05) different from their respective dust samples. Mean particle density for 
corn stover dust was measured to be 1500.90 kg/m3. Mani et al. (2006) reported 
the particle density of corn stover ground through hammer mill screen size of 0.8 
mm to be 1399.16 kg/m3.  
0
100
200
300
400
500
600
700
Bu
lk 
de
nsi
ty (k
g/
m3
)
ground material dust
*Pecan shell and lignite coal samples were received in dust form 
  
68 
 
 
Figure 3.17 Particle densities of ground and dusts from biomass and coal. 
 
 
Figure 3.18 Particle density dependence on geometric mean particle size in case of 
all biomass dusts. 
 
0
300
600
900
1200
1500
1800
Pa
rti
cle
 de
ns
ity
 (kg/
m3
) ground material dust
y = -0.6326x + 1702.1
R? = 0.5043
1000
1200
1400
1600
1800
150 250 350 450 550 650
Pa
rti
cle
 de
ns
ity
 (kg/
m3
)
Geometric mean diameter dgw (?m)
Bermuda grass Corn cobs Corn stover
Eucalyptus Pecan shell Pine
Poultry litter Sugarcane bagasse sweetgum
Switchgrass
*Pecan shell and lignite coal samples were received in dust form 
  
69 
 
The geometric mean particle diameter of biomass dusts had a significant 
effect on particle density (p<0.0001). Particle density was found to be higher in 
dusts with smaller geometric mean diameter (Pearson?s correlation coefficient, r=-
0.71) (figure 3.18).   This is because in large biomass particles, there are voids 
within the particles that do not contribute to the mass of the particle. As the particle 
size becomes small (due to grinding), internal particulate voids are reduced leading 
to increase in particle density (Mani et al., 2004; Esteban and Carrasco, 2006). In 
their experiment with raw pecan shells divided into three different sieve size 
fractions (>1.885 mm, 1.295-1.885 mm and <1.295 mm), Littlefield et al. (2011) 
showed that with decrease in size, particle density increased. Mani et al. (2006) 
also showed that particle density increased for ground wheat straw, barley straw, 
corn stover and switchgrass as the hammer mill screen size decreased from 3.2 
mm to 0.8 mm. Dusts with smaller particle sizes also have higher particle densities 
because of more ash content in finer fractions of material (Hehar, 2013), which 
have higher particle densities (1760 ? 2760 kg/m3) (Ghosal and Self, 1995) than 
biomass or coal dusts.   
3.4.1.5 Ash Content                              
Ash content of ground material varied from 0.48% d.b. (pine) to 10.47% d.b. 
(bituminous coal) whereas, ash content of the dust samples varied from 1.10% d.b. 
(pine) to 20.84% (lignite coal) where. Ash content of ground biomass samples was 
found to be significantly (?=0.05) less than the ash content of dusts except 
eucalyptus (figure 3.19) that had lower ash content in ground sample but was not 
significantly different from ash content of the dust. Ground biomass material had 
  
70 
 
less ash content than dusts because the inorganic content is more grindable and 
are easily separated into finer fractions from lignocellulosic structure of biomass 
during grinding (Hehar, 2013). Liu and Bi (2011) also obtained similar results with 
switchgrass sample milled to obtain size less than 1 mm. Ash content of 
switchgrass sample increased from 4.31% to 10.53% as the sieve size of the 
fraction decreased from >0.95 mm to <0.15 mm.  Ash content of ground bituminous 
coal and PRB coal was not significantly (?=0.05) different from their respective 
dust samples. This can be due to smaller geometric mean particle size of ground 
samples for these two coal types as compared to other ground samples.  
Ash content for woody biomass dusts - pine, eucalyptus, sweetgum (1.10% 
- 1.41%) was found to be lower than ash content in grassy biomass dusts ? 
Bermuda grass, corn stover, sugarcane bagasse and swithcgrass (4.68% - 
15.42%). This is similar to the results that have been documented for woody and 
grass like biomass. McKendry (2002) reported the ash contents of Danish pine and 
willow wood (woody biomass) to be 1.60% while Jenkins et al. (1998) measured 
the ash content of switchgrass to be 8.97%. Also, Cuiping et al. (2004) reported 
that the ash contents of corn stover are in the range of 4.33-21.91%. 
For grassy biomass, geometric mean diameter had a significant effect on 
ash content (p<0.0001). Ash content was found to be higher in grassy biomass 
dust samples with smaller geometric mean diameter (Pearson?s correlation 
coefficient, r=-0.91) (figure 3.20a). Also, for grassy biomass, ash content was 
found to be higher in dusts with higher particle density (Pearson?s correlation 
coefficient, r=0.97) (figure 3.20b). This is because ash consists of inorganic 
  
71 
 
mineral particles which have high particle densities (1760 ? 2760 kg/m3) (Ghosal 
and Self, 1995).  Also, particle density of ash from sugarcane bagasse sample 
used in this study was found out to be 2781.5?3.46 kg/m3 which is higher than 
particle densities of most biomass and coal dusts. 
 
Figure 3.19 Ash content of ground and dusts from biomass and coal. 
Ash content of coal dusts were measured to vary between 8.14%-10.27% 
(dry basis). These values are higher than the ash contents (1.10% - 1.41%) of 
woody biomass ? an indication of higher amount of inorganic compounds in coal 
(Demirbas, 2004; Hehar, 2013). McKendry (2002) measured ash content of 
bituminous coal to be 8.0% which is comparable to the ash content value obtained 
in this study for bituminous coal (10.27% d.b.). Samaras et al. (1996) reported the 
ash content for lignite coal to be 22.1% (d.b.). This is also comparable to the ash 
content of lignite coal used in this study (20.84 d.b.%). 
0
2
4
6
8
10
12
14
16
18
Ash
 co
nte
nt 
(%
 db
)
ground material dust
*Pecan shell and lignite coal samples were received in dust form 
  
72 
 
Mean ash content (dry basis) for poultry litter was obtained as 13.74% which 
is comparable to the mean ash content of poultry litter (14.1 d.b.%) found by Tiqui 
and Tam (2000). Ash content of poultry litter was also found to be higher than that 
of woody biomass because poultry birds use up a significant amount of nutrients 
(carbohydrates, protiens, fat and minerals) from feed for bodily functions (Chiba, 
2014). Thus, mineral content of poyltry excreta is also high leading to higher ash 
content in poultry litter.  
 
 
Figure 3.20  Ash content dependence on geometric mean particle size (a) and 
particle density (b) for grassy biomass dusts. 
y = -0.0317x + 24.34
R? = 0.8251
4
6
8
10
12
14
16
200 300 400 500 600 700
As
h c
on
ten
t (% 
d.b
.)
Geometric mean diameter dgw (?m)
Bermuda grass
Corn stover
Sugarcane bagasse
Switchgrass
y = 0.0278x - 28.5
R? = 0.9314
0
4
8
12
16
20
1000 1100 1200 1300 1400 1500 1600 1700
As
h c
on
ten
t (% 
d.b
.)
Particle density (kg/m3)
Bermuda grass Corn stover
Sugarcane bagasse Switchgrass
(a) 
(b) 
  
73 
 
3.4.1.6 Volatile Matter 
Volatile content of ground biomass samples ranged from 81.24% (d.b.) 
(corn stover) to 89.86% (d.b.) (sweetgum) whereas ground bituminous coal and 
PRB coal had volatile content of 38.87% (d.b.) and 53.71% (d.b.) respectively. 
Mean volatile matter values for all dusts samples varied from 33.00% (d.b.) 
(bituminous coal) to 86.82% (d.b.) (sweetgum). Volatile content of coal dusts was 
in the range of 33.0 d.b.% (bituminous) to 53.69 d.b.% (lignite) which was lower 
than the volatile content range of biomass dusts that varied from 64.93 d.b.% 
(pecan shell) to 86.82 d.b.% (sweetgum). Volatiles are generated due to thermal 
decomposition of organic compounds present in biomass and coal. Since biomass 
have generally higher hydrogen to carbon and oxygen to carbon ratios than coal 
(Jenkins et al. 1998), the volatile matter of biomass is generally higher than that of 
coal.  
Mean values of volatile matter of ground material for all biomass and coal 
feedstocks were higher than volatile matter contents of respective dust samples 
(figure 3.21). This difference was found to be significant (?=0.05) for all samples 
except pine. Higher volatile matter in ground material as compared to dusts is due 
to higher ash content of dusts (Hehar, 2013; Gani and Naruse, 2007). Volatile 
matter of grass like biomass dusts (Bermuda grass, corn stover, sugarcane 
bagasse and switchgrass) was found to reduce significantly (p<0.0001) with 
increase in ash content (Pearson?s correlation coefficient, r=-0.94) (figure 3.22). 
Similar results were not obtained for woody biomass dusts because of their narrow 
range of ash content (1.10% - 1.41% d.b.). 
  
74 
 
Mean volatile matter value for bituminous coal dust was obtained to be 
33.00% (d.b.). These values are comparable to the volatile matter of bituminous 
coal that have been reported in literature: 35% (McKendry, 2002) and 28.33% 
(Cuiping et al., 2004). Jenkins et al. (1998) also found out volatile matter of 
switchgrass as 76.69%, which is comparable to the value for switchgrass used in 
this study (77.86% d.b.). Cuiping et al. (2004) found volatile matter for corn stover 
as 67.36% (d.b.) which is comparable to the value obtained in this study (72.97% 
d.b.).  
 
 
Figure 3.21 Volatile matter of ground and dusts from biomass and coal. 
0
10
20
30
40
50
60
70
80
90
100
Vo
latil
e m
att
er 
(%
 db
)
ground material dust
*Pecan shell and lignite coal samples were received in dust form 
  
75 
 
 
Figure 3.22 Effect of ash content on volatile matter content of grassy biomass dusts. 
3.4.1.7 Energy Content 
 Energy content of ground material varied from 18.16 MJ/kg d.b.(corn stover) 
to 31.75 MJ/kg d.b. (Bituminous coal) whereas, energy contents for dust samples 
varied from 16.69 MJ/kg d.b. (sugarcane bagasse) to 32.26 MJ/kg d.b.(bituminous 
coal). Energy contents of coal dust samples (27.41 MJ/kg ? 32.26 MJ/kg d.b.) were 
found to be higher than those of biomass dust samples (18.77MJ/kg - 20.55MJ/kg 
d.b.). This is because biomass has less carbon and more oxygen (Demirbas, 
2004). Material with more carbon will have higher energy content (Jenkins et al., 
1998). According to Jenkins (1989), each 1% increase in carbon concentration 
elevates the higher heating value by 0.39 MJ/kg. 
Mean value of energy content for all ground biomass samples was found to 
be higher than the respective dust samples (figure 3.23). This difference was 
significant (?=0.05) for all biomass feedstock except pine and sweetgum. Ground 
material had higher energy content than dusts due to higher ash content in dusts 
as compared to ground material. Mani et al. (2004) also confirmed that biomass 
y = -0.6041x + 81.822
R? = 0.8828
70
72
74
76
78
80
82
0 5 10 15 20
Vo
lati
le 
matte
r (% 
d.b
.)
Ash content (% d.b.)
Bermuda grass
Corn stover
Sugarcane bagasse
Switchgrass
  
76 
 
with higher ash content value has lower heating value. Similar results were also 
obtained by Ebling and Jenkins (1985) in their study on wheat and barley straws. 
Also, for biomass dusts, ash content had a significant (p<0.0001) effect on energy 
content of the samples. Samples with higher ash content were found to have lower 
energy content values (Pearson?s correlation coefficient, r=-0.95) (figure 3.24). 
Mean energy content of pine was found out as 20.55 MJ/kg which is comparable 
to the values obtained in earlier studies. Energy content for Danish pine and pine 
(Pinus tabulaeformis) samples was reported to be 21.2 MJ/kg (McKendry, 2002) 
and 19.38 MJ/kg (Cuiping et al., 2004) respectively. Mean energy content (dry 
basis) for switchgrass was found out as 18.77 MJ/kg which is comparable to values 
found out in earlier studies by McKendry (2002) (17.40 MJ/kg), Jenkins et al. 
(1998) (18.06 MJ/kg) and Mani et al. (2004) (17.61 MJ/kg).   
 
Figure 3.23 Energy content of ground and dusts from biomass and coal. 
0
5
10
15
20
25
30
35
En
erg
y c
on
ten
t (M
J/kg 
db
) ground material dust
*Pecan shell and lignite coal samples were received in dust form 
  
77 
 
 
Figure 3.24 Effect of ash content on energy content for all biomass dusts. 
 
3.4.2 Heating and Ignition Properties 
Minimum hot surface ignition temperature (MIT) values for biomass and 
coal dusts are given in table 3.6. Volatilization properties (TORV, TMML, TOXY 
and activation energy) and exothermic parameters (TRE, TME and exothermic 
energy) for biomass and coal dust samples are given in table 3.7.  
3.4.2.1 Minimum Hot Surface Ignition Temperature 
Hot surface ignition temperature of dust samples was determined according 
to ASTM E2021 standard (2010). Figure 3.25 shows a typical temperature profile 
obtained when a dust layer is heated on a hot plate. For this sample (corn cobs), 
there was no ignition because temperature of the dust layer increased to about 
y = -0.2152x + 20.237
R? = 0.9056
15
16
17
18
19
20
21
0 5 10 15 20
En
erg
y c
on
ten
t (M
J/kg)
Ash content (% d.b.)
Bermuda grass Corn cobs
Corn stover Eucalyptus
Pecan shell Pine
Poultry litter Sugarcane bagasse
sweetgum Switchgrass
  
78 
 
200?C and then became essentially constant. Also, the temperature of the dust 
layer did not cross the constant hot plate temperature profile (figure 3.25a). In 
figure 3.25b, there was ignition of the corn cobs dust sample with the temperature 
of the dust sample increasing at a rapid rate to about 575?C. The hot plate 
temperature that caused the ignition of dust sample was taken as the minimum hot 
surface ignition temperature (MIT).  
 
Figure 3.25 Plots of temperature vs. time showing maximum temperature of no 
ignition, 275?C (a) and minimum temperature of hot surface at which ignition 
occurred (MIT), 280?C (b) for corn cobs sample. 
0
100
200
300
400
500
600
0 250 500 750 1000 1250 1500
Te
mp
era
tur
e (
?C)
Time (s)
plate temp (?C)
dust temp  (?C)
0
100
200
300
400
500
600
0 250 500 750 1000 1250 1500
Te
mp
era
tur
e (
?C)
Time (s)
plate temp (?C)
dust temp (?C)
(a) 
(b) 
  
79 
 
MIT values were measured to be between 240?C (PRB coal and lignite coal) 
and 335?C (Bituminous coal). Bituminous coal had higher MIT than PRB or lignite 
coal because it has lower volatile content (33.00% wt. d.b.) than the other two coal 
dust samples. Volatile matter of coal is an important factor that affects ignition 
temperature. Higher volatile matter of coal would result in lower ignition 
temperatures (Miron and Lazzara, 1988). Hot surface minimum ignition 
temperatures for Pittsburgh seam coal was reported by Park (2006) to be 220?C. 
Reddy et al. (1998) measured the ignition temperatures for Prince and Pittsburgh 
coal samples to be 250?C and 270?C respectively.  
Similarly, lower ash content of grass like biomass significantly reduced the 
MIT (Pearson?s correlation coefficient, r=0.75, p=0.0049) (figure 3.26a). This is 
because ash content retards the oxidation process and acts as a heat sink 
(Vuthaluru, 2004; Hehar, 2013). This trend was not seen in case of woody biomass 
(eucalyptus, pine, sweetgum) since ash content values were very close to each 
other (1.10% - 1.41% db) for these samples, and were not significantly different.  
Grassy biomass dusts with higher volatile matter had significantly lower MIT values 
(Pearson?s correlation coefficient r=-0.64, p=0.025) (figure 3.26d). This is because 
volatiles escape the solid biomass fuel upon heating and ignite while in gas phase 
leading to ignition. This trend was not observed in case of woody biomass since 
the volatile matter values for them were not significantly different from each other.  
Bulk density of grass like biomass dusts had a significant effect on its MIT 
(p=0.0001). Samples with higher value of bulk density had lower MIT (Pearson?s 
  
80 
 
correlation coefficient, r=-0.89) (figure 3.26b). Similarly, MIT-bulk density 
relationship was also obtained for woody biomass (Pearson?s correlation 
coefficient r=-0.99, p<0.0001) (figure 3.26c). This is due to increase in the thermal 
conductivity of the dust layer with increase in bulk density (Bowes and Townshend, 
1962).  
Table 3.6 Minimum hot surface ignition temperature (MIT) of dust layer for all 
samples. 
Sample 
Temperature of no 
ignition of dust layer 
(?C) 
Minimum temperature 
of ignition of dust layer 
(?C) 
Bermuda grass 270 275 
Bituminous coal 330 335 
Corn cobs 275 280 
Corn stover 285 290 
Eucalyptus 280 285 
Lignite coal 235 240 
Pecan shell 260 265 
Pine 310 315 
Poultry litter 280 285 
PRB coal 235 240 
Sugarcane bagasse 300 305 
Sweetgum 300 305 
Switchgrass 290 295 
*The MIT values for duplicate runs were the same 
  
81 
 
Table 3.7 Measured volatilization and exothermic properties of biomass and coal dust samples. 
Sample TORV (?C) TMML (?C) TOXY (?C) Ea (kJ/mol) TRE (?C ) TME (?C ) 
Exothermic 
Energy 
(MJ/kg) 
Biomass Samples 
Bermuda grass 266.1?1.0g 312.4?1.6d 274.0?3.8d 48.3?0.6f 237.5?3.5b,c 379.3?11.7c,d 7.95?0.01a 
Corn cobs 276.52?0.62e,f 313.67?1.60d 276.51?4.02d 63.15?2.45d 235.2?1.0c 394.6?4.9a,b,c 8.15?0.11a 
Corn stover 277.5?1.1e 316.4?0.8d 286.1?1.7c,d 71.3?0.5c 239.7?0.4a,b,c 395.8?3.2a,b 7.78?0.24a 
Eucalyptus 291.88?0.88b 315.69?1.76d 285.87?4.23c,d 72.35?1.55c 243.5?1.0a,b,c 385.4?1.0b,c,d 7.52?0.22a,b 
Pecan shell 290.0?1.7c,d 331.2?4.7c 283.5?11.4c,d 55.6?0.3e 206.0?2.5d 407.7?2.6a 5.61?0.15c 
Pine 306.08?1.78a 348.75?2.09a 319.08?2.79a 64.43?0.11d 244.5?1.4a,b,c 401.1?3.3a,b 6.14?0.46c 
Poultry litter 272.8?0.5f 304.1?0.5e 277.8?0.4d 71.1?1.1c 241.2?4.6a,b,c 354.4?1.9e 4.48?0.26d 
Sugarcane 
bagasse 285.2?0.4
d 345.21?1.33a,b 309.57?6.34a,b 77.08?0.54b 245.3?4.7a,b 377.3?4.5d 7.55?0.73a,b 
Sweetgum 290.23?0.75b,c 338.8?1.5 b 297.2?2.2b,c 95.2?1.9a 248.9?3.1a 392.8?0.8a,b,c,d 6.57?0.05b,c 
Switchgrass 285.0?1.8d 329.76?0.49c 315.28?4.61a 61.6?1.3d 249.0?0.8a 389.0?1.6b,c,d 7.69?0.16a 
Coal Samples 
Bituminous coal 447.6?2.2a 485.1?4.4a 423.1?16.7a 91.6?6.3a 223.4?1.5b 395.6?2.2b 3.51?0.10a 
Lignite coal 311.6?4.5b 349.77?6.17c 241.80?3.33b 56.82?0.48b 221.2?2.4b 428.0?1.3a 5.16?0.87a 
PRB coal 317.3?1.6b 379.4?1.6b 258.7?18.4b 49.8?0.8b 244.9?2.4a 429.1?3.0a 4.64?0.86a 
*Superscripts with same letters in a column are not significantly different from each other (?=0.05)
  
82 
 
  
  
 
Figure 3.26 Minimum hot surface temperature (MIT) dependence on ash content of 
grassy biomass dusts (a), bulk density of grassy biomass dusts (b), bulk density of 
woody biomass dusts (c) and volatile matter of grassy biomass dusts (d).  
 
3.4.2.2 Volatilization Properties 
 TGA mass loss curves for all the dust samples when heated in air and 
oxygen atmosphere are depicted in figure 3.27. Figure 3.28 shows mass loss curve 
for switchgrass dust heated in air and oxygen atmospheres as an example of how 
y = 1.7892x + 272.41
R? = 0.5644
270
280
290
300
310
320
0 10 20
MI
T (
?C)
Ash content (% d.b.)
Bermuda grass Corn stover
Sugarcane bagasse Switchgrass
y = -0.5519x + 365.64
R? = 0.7923
270
280
290
300
310
320
100 150 200
MI
T (
?C)
Bulk density (kg/m3)
Bermuda grass Corn stover
Sugarcane bagasse Switchgrass
y = -0.6417x + 424.49
R? = 0.9727
270
280
290
300
310
320
160 180 200 220
MI
T (
?C)
Bulk density (kg/m3)
Eucalyptus Pine Sweetgum
y = -2.3709x + 470.16
R? = 0.4097
270
280
290
300
310
320
70 75 80 85
MI
T (
?C)
Volatile matter (% d.b.)
Bermuda grass Corn stover
Sugarcane bagasse Switchgrass
(a) (b) 
(c) (d) 
  
83 
 
temperature of onset of rapid volatilization (TORV), temperature of maximum rate 
of mass loss (TMML) and oxidation temperature (TOXY) were estimated. As 
shown in figure 3.28a, TORV was obtained by drawing tangents to the parts of the 
curve corresponding to the onset of significant mass loss of sample due to 
devolatilization. TMML was obtained from the peak of the mass loss rate curve 
(figure 3.28a). Similarly, the peak of mass loss rate curve when sample is heated 
under oxygen atmosphere gives the TOXY value (figure 3.28b). 
 The initial mass loss in all the dust samples when heated (in air and oxygen 
atmospheres) from ambient to about 110?C is due to the loss of moisture. Initial 
mass loss curve for lignite coal is prominent than the rest of the samples because 
of the high initial moisture content of lignite coal (29.4 w.b.%). After initial moisture 
loss from the sample, mass loss for biomass dust samples occurred in two parts. 
A rapid mass loss starts after 200?C (Vamvuka et al., 2003). After about 350?C, 
this major loss is followed by a slow rate of mass loss.  
 The maximum rate of mass loss in biomass generally occurred between the 
temperatures of 200?C to 400?C due to release of the volatiles (Jenkins et al., 
1998) from hemicellulose decomposition. From figure 3.27a, it can be seen that 
every biomass dust sample lost mass rapidly in this temperature range. The slow 
mass loss that typically occurs after the rapid mass loss has been attributed to the 
decomposition of cellulose in biomass (Yang et al., 2006). Lignin decomposes with 
difficulty in comparison to hemicellulose and therefore decomposes over a wide 
range of temperatures (Yang et al., 2006; Vamvuka et al., 2003; Gronli et al., 1999; 
Orfao et al., 1999; Sorum et al., 2001). However, since coal is not a lignocellulosic 
  
84 
 
material, mass loss in case of coal dust samples did not follow the two stage 
decomposition of biomass 
 
 
 
Figure 3.27 TGA mass loss curves for dust samples heated in air environment (a) 
and oxygen environment (b). 
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800
Ma
ss 
(%)
Temperature (?C)
bermuda grass Bituminous coal corn cobs corn stover
eucalyptus Lignite coal Pecan shell pine
poultry litter PRB coal sugarcane bagasse sweetgum
Switchgrass
0
10
20
30
40
50
60
70
80
90
100
0 100 200 300 400 500 600 700 800
Ma
ss 
(%)
Temperature (?C)
bermuda grass Bituminous coal corn cobs corn stover
eucalyptus Lignite coal Pecan shell pine
poultry litter PRB coal sugarcane bagasse sweetgum
Switchgrass
(b) 
  
85 
 
 
 
 
Figure 3.28 Example of how TORV and TMML were estimated. Mass loss curves are 
for switchgrass dust sample heated in air atmosphere (a) and heated in oxygen 
atmosphere (b). 
-0.8
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0
1
2
3
4
5
6
7
0 100 200 300 400 500 600 700 800
Der
ivati
ve 
mas
s (m
g/m
in)
Mass (m
g)
Temperature (?C)
Mass
Derivative Mass
TMML
TORV
-2.1
-1.9
-1.7
-1.5
-1.3
-1.1
-0.9
-0.7
-0.5
-0.3
-0.1
0.1
0
1
2
3
4
5
6
7
0 100 200 300 400 500 600 700 800
De
riv
ati
ve
 m
ass 
(m
g/m
in)
Ma
ss 
(m
g)
Temperature (?C)
Mass
Derivative Mass
TOXY
(a) 
(b) 
  
86 
 
  
 Data obtained from oxygen atmosphere decomposition was used to obtain 
the single point oxidation temperature (TOXY). The TG data showed rapid mass 
loss in the dust samples for both coal and biomass due to presence of excess 
oxygen for combustion (Ramirez et al., 2010). TOXY for biomass dusts varied from 
274.0?C (Bermuda grass) to 319.1?C (pine) whereas for coal dusts it varied from 
241.8?C (lignite coal) to 423.1?C (bituminous coal). TOXY value for bituminous 
coal was significantly higher than the other two coal samples because of its lower 
volatile content which acts as fuel and is easily oxidized as compared to carbon. 
Based on TOXY values it can be said that bituminous coal dust is at lower risk of 
ignition than lignite and PRB coal dusts.   
 Value of TORV of bituminous coal dust sample (447.6?C) was significantly 
higher than that of PRB coal dust (317.3?C) and lignite coal dust (311.6?C). This 
is due to the low volatile content of bituminous coal dust than the other two dust 
samples. TORV indicates the ease with which a material will release volatiles upon 
heating. The lower the TORV value, the more easily volatiles are released from 
the solid fuel matrix.  
 For coal dusts, TORV values (311.6?C - 447.6?C) were significantly higher 
than that of biomass dusts (266.1?C - 306.1?C) which again can be attributed to 
the low volatile content of the coal dust samples. Thus, based on TORV values, 
biomass dusts are at higher risk of ignition than coal dusts. According to 
Muthuraman et al. (2010), coal requires a higher temperature to release its 
  
87 
 
volatiles than compared to biomass like wood, which is in accordance to the 
findings in this study. 
 Similarly, TMML value for bituminous coal dust (485.1?C) was significantly 
higher than that of PRB coal (379.4?C) and lignite coal dusts (349.8?C). This again 
can be said due to low volatile content of bituminous coal than other types of coal. 
Also, TMML values for coal dusts (349.8?C - 485.1?C) was significantly higher than 
those of biomass dusts (304.1?C - 348.8?C) because of low volatile content of 
coals. This is because volatile release and ignition corresponds to maximum mass 
loss rate as compared to char oxidation (Jenkins et al., 1998). Thus, based on 
TMML values, biomass dust is at higher risk of ignition and dust explosion than 
coal dusts. 
 An example of determination of activation energy is shown in figure 3.29. 
Only a section of the conversion data corresponding to the maximum mass loss 
rate was used in estimation of activation energy (Ramirez et al., 2010). Activation 
energy value for grassy biomass dusts with higher ash content was found to be 
significantly higher (Pearson?s correlation coefficient r=0.96, p<0.0001) (figure 
3.30a). This is again because ash acts as heat sink and effects ignition as it hinders 
the oxidizer-fuel contact (Porteiro et al., 2010).  
 Grassy biomass dust samples with higher amount of volatile matter had 
significantly lower activation energy values (Pearson?s correlation coefficient r = -
0.89, p<0.0001) (figure 3.30b).  This is because of combustion of volatiles in gas 
phase as it escapes biomass particles upon heating. Higher volatile matter leads 
to lower ignition temperature (Grotkjaer et al., 2003).  
  
88 
 
 
Figure 3.29 Example for determination of apparent activation energy (switchgrass 
dust sample). 
  
 
  
 
Figure 3.30 Effect of ash content on activation energy (a) and effect of volatile matter 
on activation energy (b) for grassy biomass dusts. 
 
y = -7299.7x + 10.276
R? = 0.9475
-8
-7
-6
-5
-4
-3
-2
-1
0
0.0015 0.0017 0.0019 0.0021 0.0023 0.0025
ln(
?/d
t)
1/T (1/K)
y = 2.3216x + 40.115
R? = 0.93
40
50
60
70
80
0 5 10 15 20Act
iva
tion
 en
erg
y (kJ
/m
ol)
Ash content (% d.b.)
Bermuda grass Corn stover
Sugarcane bagasse Switchgrass
y = -3.3435x + 316.87
R? = 0.7975
40
50
60
70
80
90
70 75 80 85Act
iva
tion
 en
erg
y (kJ
/m
ol)
Volatile matter (% d.b.)
Bermuda grass Corn stover
Sugarcane bagasse Switchgrass
(a) (b)
 
??(????) = ??? + ? ? ln(1 ? ?)? ??? 
  
89 
 
3.4.2.3 Exothermic Parameters 
 DSC parameters (temperature of rapid exothermic reaction (TRE), 
maximum temperature reached during exothermic reaction (TME) and exothermic 
energy) obtained for all dust samples are tabulated in table 3.7. DSC heat flow vs. 
temperature curves for all the dust samples heated in air atmosphere are depicted 
in figure 3.31. Figure 3.32 shows DSC curve for switchgrass dust heated in air 
atmosphere as an example of how TRE and TME were estimated. The intersection 
point of the tangents drawn at parts of the curve corresponding to rapid increase 
in heat flow and end of rapid exothermic reaction gives the TRE and TME values 
respectively. Exothermic energy was obtained by integrating the area under part 
of the heat flow vs. time curve corresponding to exothermic reaction.  Negative 
value of heat flow in a DSC curve represents endothermic reaction taking place. 
An endothermic peak is observed for every dust sample at around 100?C due to 
the loss of moisture from the sample. This is the temperature at which water 
vaporizes upon heating. After the moisture loss, sample reacts with oxygen in 
presence of heat resulting in an exothermic reaction (positive heat flow) (Ramirez 
et al., 2010). Exothermic energy was estimated by calculating area under the DSC 
heat flow curve (calculated by the software provided along with the DSC 
equipment). 
 Rapid exothermic reaction occurs for all the dusts between temperatures of 
200?C and 250?C (figure 3.31). TRE for biomass dust varied from 206.1?C (pecan 
shell) to 248.96?C (switchgrass) whereas for coal dusts it varied from 221.2?C 
(lignite coal) to 244.9?C (PRB coal). In a separate study, Ramirez et al. (2010) 
  
90 
 
measured the TRE for bituminous and sub bituminous coal to be 240.0?C and 
220.0?C respectively. A value of 223.4?C was obtained as TRE for bituminous coal 
used in this study. 
 
Figure 3.31 Heat flow curves of biomass and coal dusts heated with DSC in air 
atmosphere. 
 
 
Figure 3.32 Example of how TRE and TME were estimated from heat flow vs. 
temperature curve for switchgrass sample when heated under air environment.  
-10
-5
0
5
10
15
0 100 200 300 400 500 600
He
at 
flo
w 
(W
/g
)
Temperature (?C)
bermuda grass bituminous coal corn cobs
corn stover eucalyptus lignite coal
pecan shell pine poultry litter
PRB coal sugarcane bagasse sweetgum
switchgrass
-4
-2
0
2
4
6
8
10
0 100 200 300 400 500 600
He
at 
flo
w 
(W
/g
)
Temperature (?C)
TME 
TRE 
  
91 
 
 Mean value of maximum temperature reached during an exothermic 
reaction (TME) for bituminous coal, lignite coal and PRB coal were measured to 
be 395.6?C, 428.0?C and 429.1?C respectively whereas, for biomass, it varied from 
354.4?C (poultry litter) to 407.7?C (pecan shell). Sahu et al. (2010) also found out 
DSC curve peak temperature (TME) for coal as 423.0?C. In general, TME was 
found to be higher in case of coal samples than biomass as coal had higher energy 
content than biomass dusts. For woody biomass, dust samples with higher energy 
content had significantly higher TME value (Pearson?s correlation coefficient 
r=0.89, p=0.0013) (figure 3.33).  
 Exothermic energy is the measure of quantity of heat evolved upon heating 
a material. Exothermic energy of the dust samples varied from 3.51 MJ/kg 
(bituminous coal) to 8.15 MJ/kg (corn cobs). The exothermic energy released 
during the DSC process is less than the calorific value (energy content) of the dust 
samples because of the incompletely burned volatiles (Jiricek et al., 2012). While 
measuring heating value, oxygen atmosphere was provided to the sample which 
led to complete burning of the sample whereas, samples were heated in air 
atmosphere during DSC experiments which lead to incomplete combustion (char 
residue).  This caused exothermic energy value to be less than the heating value 
of the dust samples.  
 Exothermic energy of bituminous coal was found to be less than that of 
biomass. This can be attributed to low volatile content of bituminous coal. Dust 
samples were heated only up to temperature of 550.0?C in DSC experiments, and 
from TG analysis of bituminous coal, it is evident (figure 3.27a) that the coal sample 
  
92 
 
still continued to show slight mass loss, thus release energy via oxidation process 
beyond 550.0?C, which was not accounted for in DSC experiment due to limitations 
of the equipment, resulting in low exothermic energy value. 
 
Figure 3.33 Effect of energy content on maximum temperature reached during 
exothermic reaction (TME) for woody biomass dusts. 
 
 Lignite and PRB coal dusts were found to have higher TME values (428.0?C 
and 429.1?C respectively) than biomass dusts (354.4?C ? 407.7?C). Although, TRE 
values for bituminous and lignite coal dusts is lower than that of all the biomass 
dusts except pecan shell, the exothermic energy of all biomass dusts except 
poultry litter was higher than that of coal dusts in a given range of temperature 
(30?C-550?C). Thus, biomass dusts in general are associated with higher 
destruction capability during a dust explosion than coal dusts. 
 
y = 13.24x + 129.34
R? = 0.792
380
385
390
395
400
405
19.2 19.4 19.6 19.8 20 20.2 20.4 20.6 20.8
TM
E (
?C)
Energy content (MJ/kg)
Eucalyptus Pine Sweetgum
  
93 
 
3.5  Conclusion 
 Physical and chemical properties were measured for ground material and 
dust samples. Heating and ignition parameters were also measured for dusts. 
Range of values obtained for MIT, TORV, TMML, TOXY, TRE and TME for all the 
dusts were 240.0?C-335.0?C, 266.1?C-447.6?C, 304.1?C-485.1?C, 274.0?C-
423.1?C, 206.1?C-249.0?C and 354.4?C-429.1?C respectively. Grassy biomass 
dusts with higher ash contents had significantly lower volatile matter (p<0.0001). 
All biomass dusts with higher ash content had significantly lower energy content 
(p<0.0001). Grassy and woody biomass dusts with higher bulk density had 
significantly lower MIT value (p=0.0001 and p<0.0001 respectively). Also, MIT 
values of grassy biomass with higher ash contents and lower volatile matter was 
significantly higher (p=0.0049 and p=0.025 respectively). Grassy biomass dusts 
with higher ash contents and lower volatile matter had significantly higher 
activation energy values (p<0.0001). Woody biomass dusts with higher energy 
content had significantly higher TME values (p=0.0013). Based on TORV and 
TMML values, biomass dusts are at higher risk of ignition than coal dusts. Based 
on TOXY values it can be said that bituminous coal dust is at lower risk of ignition 
than lignite and PRB coal dusts.  Based on the exothermic energy values, most of 
the biomass dusts (all except poultry litter dust) are associated with higher 
destruction capability during a dust explosion than coal dusts. 
 
  
 
94 
 
Chapter 4 Prediction of Heating and Ignition Properties Using Near Infrared 
Spectroscopy (NIRS) 
4.1 Abstract 
 Dusts (i.e. particles of size less than 500 ?m) are often generated during 
handling and processing of biomass feedstock. More than 70% of dusts generated 
in process industries are combustible (Vijayraghavan, 2004) and can lead to dust 
fire and/or explosion hazards if ignited. Fire and explosion due to dust ignition 
cause damage to plants or units and injuries to personnel and fatalities (Eckhoff, 
2009). In addition to structural damage, dust explosions can result in loss of 
income by a plant due to down time and time required to repair the damaged 
portion of the plant (Sapko et al., 2000). Thus, heating and ignition of biomass 
dusts plays a critical role in development of safety guidelines and standards for 
process industries handling biomass and coal. This research aims at developing 
near infrared spectroscopy (NIRS) models to predict the heating and ignition 
properties of dusts from ten biomass feedstocks. The heating and ignition 
properties predicted are minimum hot surface ignition temperature of dust layer 
(MIT), temperature of onset of rapid volatilization (TORV), temperature of 
maximum rate of mass loss (TMML), oxidation temperature (TOXY), temperature 
of rapid exothermic reaction (TRE) and maximum temperature reached during an 
exothermic reaction (TME). Principal component analysis (PCA) was used on NIR 
  
 
95 
 
spectral data for dusts to develop prediction models for heating and ignition 
parameters. Coefficient of determination (R2) values for internal validation of 
models developed using PCA on raw spectral data for MIT, TORV, TMML, TOXY, 
TRE and TME were 0.994, 0.984, 0.963, 0.737, 0.931 and 0.901 respectively, 
whereas, use of first derivative NIR spectral data yielded R2 for these properties 
as 0.976, 0.964, 0.943, 0.798, 0.923 and 0.895 respectively.  
Dusts from four biomass samples that were obtained from sources different 
than those used to develop models were used to validate the prediction models 
externally. Coefficient of determination (R2) values for all models was obtained less 
than 0.28. Poor performance of models under external validation was attributed to 
small sample sizes of the biomass feedstocks that were used during building of 
prediction models.  
4.2 Introduction 
Fossil fuels such as coal and petroleum products are non-renewable 
sources of energy even though all the countries in world rely mainly on fossil fuels 
for energy. In the year 2013, more than 80% of all the energy consumed in USA 
was derived from fossil fuels such as petroleum, coal and natural gas (EIA, 2013). 
Due to negative impact of fossil fuel extraction and usage on environment and its 
long term availability issues, a lot of focus is being given to obtaining energy from 
renewable sources such as solar energy, wind energy and bioenergy. The main 
advantage that biomass has over other renewable energy sources is that energy 
derived from biomass can be converted into liquid fuels, chemicals and products. 
  
 
96 
 
These fuels can be directly used in sectors such as transportation, industries and 
power generation. 
Biomass has to be preprocessed before it can be used to produce fuels 
chemicals and products. Preprocessing operations involves grinding, sieving, 
conveying and storage which could lead to dust generation. The National Fire 
Protection Association standard 654 defines combustible dusts as ?particles that 
pass through a 500 ?m sieve and are a dust fire or dust explosion hazard? (NFPA, 
2013).  Combustible dust, if ignited can cause fire hazard or dust explosion. Dust 
explosions lead to injuries and loss of life and property (Sapko et al., 2000; CSB, 
2006; Amyotte and Eckhoff, 2010). Ignition sources that are present in processing 
and biomass handling facilities include hot bearings, hot surfaces, flames and 
sparks from electric motors that can ignite dusts and thus cause fire or explosion 
hazard. Thus, knowledge of heating and ignition properties of combustible dusts is 
very important in order to incorporate safety measures in process industries and 
other facilities processing biomass. 
However, the methods used to quantify biomass heating and ignition 
properties are time consuming and require the use of expensive pieces of 
equipment such as thermogravimetric analyzer (TGA) and differential scanning 
calorimeter (DSC). NIRS can be used to develop prediction models for quick 
estimation of these properties. Near Infrared (NIR) spectroscopy has been used 
as a quick method of indirectly quantifying the properties of biological samples 
such as grain moisture content (Norris, 1964), dry matter content and fruit firmness 
(Nicolai et al., 2008), post-harvest quality of fruits (Bobelyn et al., 2010), moisture 
  
 
97 
 
content, water activity and salt content of meat (Collell et al., 2011), quality control 
of potato chips (Shiroma and Rodriguez-Saona, 2009), taste characterization of 
fruits (Jamshidi et al., 2012) and proximate analysis and heating values of torrefied 
biomass (Via et al., 2013). Some of the advantages of NIRS include non-
destructive measurement, ease of sample preparation, ability to be used by low 
skilled operator and high data/spectrum acquisition rates (Vergnoux et al., 2009).  
NIRS involves exposing a sample to near infrared light (light of wavelength 
750 to 2500 nm) (13333 cm-1 to 4000 cm-1) and measuring the amount of light 
reflected from the sample (Lu and Bailey, 2005) which is typically a function of 
chemical composition and microstructure of a sample (Vergnoux et al., 2009). 
Multivariate statistical techniques such as principal component analysis (PCA) or 
partial least squares (PLS) regression analysis are used to analyze the complex 
raw spectral data obtained from NIR equipment. PCA is the most widely used 
statistical approach used in chemometrics (Brereton, 2007). PCA involves 
modeling of variance or covariance structure of a given data set to reduce the 
number of variables to a fewer number of principal components. Principal 
components are independent of each other with no correlation amongst them. The 
objective of this study was to predict heating and ignition characteristics of biomass 
dusts using near infrared spectroscopy (NIRS). 
  
 
98 
 
4.3 Methods and Materials 
4.3.1 Raw Material 
Thirteen feedstock (10 biomass and 3 coal types) were obtained for this 
study from various sources as listed in table 3.1. Four other biomass samples were 
used for external validation of the prediction models (table 4.1). Wet biomass 
samples were either air dried or dried at low temperature (45?C) before they were 
further utilized for analysis. The samples were ground with a hammer mill (C.S. 
Bell Co., model 10HBLPK, Tiffin, OH, USA) (figure 3.3 a) fitted with a 3.175 mm 
(1/8 ?) screen. Dust was obtained from the ground material by passing it through 
#35 market grade (437 ?m) screen using a vibratory sieve shaker (Kason Corp., 
model K30-2-8S, NJ, USA) (figure 3.3 b). This is the closest screen size to the 
NFPA 500 ?m size definition of dust (NFPA, 2013).   
Pecan shell, lignite coal and switchgrass (external validation set) samples 
obtained were already in dust form and were not ground using hammer mill. Dust 
collected from each feedstock was stored in three 80 oz. air tight containers for 
further analysis and characterization of physical, chemical, heating and ignition 
properties which was performed in Chapter 3. Only heating and ignition properties 
were measured for samples used for external validation to check performance of 
developed prediction models. 
 
 
  
 
99 
 
Table 4.1 List of different biomass feedstock used for external validation of 
prediction models along with their sources. 
Biomass Source 
Eucalyptus Auburn University, AL 
Loblolly pine West Fraser Mills, AL 
Sweetgum Tuskegee (private forest), AL 
Switchgrass  University of Tennessee, TN 
 
4.3.2 Near Infrared Spectroscopy (NIRS) 
FT-NIR spectrophotometer (FT-NIR 100, PerkinElmer, Shelton, CT, USA) 
(figure 4.1) was used to collect absorbance vs. wavelength data for the dust 
samples. Glass plate sample holder was filled completely with dust and sample 
was exposed to NIR wavelength. The equipment performed 40 scans for each run. 
NIR spectra from each sample were collected in triplicates. Data was collected at 
2 cm-1 resolution scans and wavelengths ranging between 10000 and 4000 cm-1 
but was processed to 10 cm-1 resolutions as Statistical Analysis Systems (SAS, 
2009) was unable to process larger data matrices (Via et al., 2011). A standard 
reference check was performed after about every four readings for consistency. 
Absorbance vs. wavelength data was obtained from the software provided by the 
equipment manufacturer and the data was imported to Microsoft Excel (Microsoft 
Excel 2010, Redmond, WA, USA).  
  
 
100 
 
 
 
 
 
 
4.3.3 Data Analysis 
Data analysis consisted of developing linear regression models for 
prediction of heating and ignition parameters of dusts with principal components 
analysis (PCA) using SAS (2009) software. Prediction models were developed for 
each heating and ignition parameter based on raw spectral data and first derivative 
data using only the 10 biomass dusts listed in table 3.1. Absorbance vs. 
wavelength spectral data was imported to Microsoft Excel (Microsoft Excel 2010, 
Redmond, WA, USA) and average absorbance value corresponding to each 
wavelength were calculated. Raw spectral graph showing average absorbance vs. 
wavenumber and first derivative graph was plotted using Microsoft Excel. 
Important wavelengths (wavelengths corresponding to peaks) were obtained from 
first derivative spectral plot. Heating and ignition properties prediction models were 
developed only for biomass dusts since the properties (physical, chemical, heating 
and ignition) of coal dusts were significantly different thereby causing a leverage 
point during developing of the models. Tukey test was also performed on principal 
component values of each selected PC to check if difference among them is 
significant (?=0.05) for different dust samples.  
Figure 4.1 FT-NIR spectrophotometer used to collect spectral data of dust samples. 
  
 
101 
 
Raw and 1st-derivative spectra was standardized to a t-distribution with zero 
mean and standard deviation as 1, so that PCA could be performed by mean 
centering the spectral data. Principal component score plots and eigenvector 
loadings plots for significant PCs were obtained from the SAS results and were 
plotted using Microsoft Excel. Regression diagnostics such as coefficient of 
determination (R2), adjusted R2, root mean square error of calibration (RMSEC) 
and were employed to determine the best predictive model. A leave one out cross 
validation (LOOCV) strategy was used to validate the model using standard 
routines in SAS. For model validation, diagnostics such as root mean square error 
of prediction (RMSEP) was estimated form predicted sum of squares (PRESS) as 
shown in equation 4.2 (Via 2013). 
 ????? = ?(?
? ???)
2
?
?=1
 (4.2) 
 ????? = ??????
?  (4.3) 
where, 
N is number cases in validation data set, 
Yi is the actual value of ith sample, 
Yp is the predicted value of Y for ith sample. 
 Ten principal components (PC) were computed for each dust sample from 
NIR raw and first derivative spectra using SAS. Stepwise selection technique was 
chosen to decide the number of PC in a predictive model. Significant wavelengths 
associated with each PC were obtained from eigenvector loading graphs and the 
  
 
102 
 
wavelengths were considered significant if the peaks associated with 
wavenumbers exceeded the ?two standard deviation? mark along the loading 
distribution for a particular PC. Actual vs. predicted values were plotted for raw 
spectra models and first derivative models for each heating and ignition parameter. 
Calibration and validation statistics were tabulated. More information about various 
formulae involved in statistical analysis and their interpretation can be found in 
literature (Neter et al., 1996). 
 Similarly, principal component analysis was also performed for the four dust 
samples used for external validation (table 4.1). Principal component values thus 
obtained were used in the equations derived from prediction models to get 
predicted values for each heating and ignition property for these four dust samples. 
Plot of actual vs. predicted temperatures were plotted for each heating and ignition 
property and coefficient of determination (R2) values were obtained using Microsoft 
Excel. 
4.4 Result and Discussion 
4.4.1 Raw NIR Spectra  
 Figure 4.2 shows average absorbance vs. wavenumber plot for biomass 
dust samples used for developing prediction models and coal dusts. It can be seen 
that average absorbance vs. wavenumber curve for all the biomass dust samples 
follow a similar trend. Coal dust spectra are however different from that of biomass 
dust spectra. This can be attributed to the difference in chemical nature of biomass 
and coal. In addition the spectra of bituminous coal was different from that of PRB 
  
 
103 
 
and lignite coal. This can be due to the low volatile content (chemical nature) of 
bituminous coal (33.00%) as compared to that of PRB coal (47.06%) and lignite 
coal (53.69%) samples.  
 A baseline shift between the raw NIR spectra of dusts, corresponding to 
different absorbance values was observed (figure 4.2). This difference in 
absorbance values can be attributed to variation in densities of the material (Via et 
al., 2003; Via et al., 2010). According to Beer Lambert?s law, apart from absorptivity 
and length of NIR beam inside the sample, a material with higher density 
(concentration) will exhibit higher absorbance (Swinehart, 1962).  
 
Figure 4.2 NIR spectra showing average absorbance vs. wavenumber plot for 
different dusts. 
4.4.2 First Derivative NIR Spectra 
 Figure 4.3 shows the first derivative NIR spectral plots for the dust samples. 
First derivative analysis was carried out on the given samples so that the cause of 
0
0.2
0.4
0.6
0.8
1
1.2
1.4
3500450055006500750085009500
Av
erag
e A
bsor
ban
ce
 (A)
Wavenumber (cm-1)
Bermuda grass
Bituminous coal
Corn cobs
Corn stover
Eucalyptus
Lignite coal
Pecan shell
Pine
Poultry litter
PRB coal
Sugarcane bagasse
Sweetgum
Switchgrass
  
 
104 
 
variation in spectra can be deciphered. First derivative treatment to the NIR spectra 
narrows down the peaks associated with important wavenumbers which may be 
responsible for variation in spectra. It also helps removing the baseline shift 
between spectra of different dust samples (Breitkreitz et al., 2003). The major 
disadvantage of first derivative treatment is the addition of noise to the data (Moes 
et al., 2008). 
 
Figure 4.3 First derivative plot of NIR spectra for different dusts showing significant 
wavenumbers associated with peaks. 
  
 Wavenumbers pertaining to different peaks and chemistry associated with 
them are listed in table 4.2. Hemicellulose (corresponding to wavenumber 5264) 
(table 4.2) seems to have a greater impact on variation on NIR spectra of dusts 
since it corresponds to the maximum peak (figure 4.3).  
5264
7084
6654
5134
4664
4384
5794
-0.0035
-0.003
-0.0025
-0.002
-0.0015
-0.001
-0.0005
0
0.0005
0.001
0.0015
350045005500650075008500950010500
1st 
De
riv
ativ
e
Wavenumber (cm-1)
Bermuda grass
Bituminous coal
Corn cobs
Corn stover
Eucalyptus
Lignite coal
Pecan shell
Pine
Poultry litter
PRB coal
Sugarcane bagasse
Sweetgum
Switchgrass
  
 
105 
 
 
Table 4.2 Chemistry associated with influential wavenumbers derived from first 
derivative NIR spectra for dust samples (Schwanninger et al., 2011). 
Wavenumber 
(cm-1) Component 
4384 Cellulose (4392) 
4664 
Acetyl groups in Hemicellulose 
Lignin and extractives (4686) 
5134 Water (5220-5150) 
5264 Hemicellulose (5245) 
5794 Lignin (5795) 
6264 Cellulose (6257) 
6654 Cellulose (6660) 
7084 
Phenolic hydroxyl groups 
Lignin (7092) 
 
4.4.3 PCA Analysis 
  PC score plots were generated from raw spectral data by comparing 
various significant (?=0.05) PCs used in models to predict heating and ignition 
properties of dusts (Figure 4.4). PCs derived from raw spectra were used in this 
analysis.  
  
 
106 
 
 
Figure 4.4 Principal component score plots for significant principal components 
viz. PC1 vs. PC2 (a), PC3 vs. PC4 (b), PC5 vs. PC6 (c) and PC10 vs. PC9 (d) obtained 
from NIR raw spectral data of biomass dusts. 
 
 Generally, PC1 accounts for the baseline shift in the NIR spectra and thus 
it can be attributed to the difference in densities of the samples in the study. PC1 
failed to separate different biomass from each other statistically (?=0.05). PC2, 
however, was able to statistically separate (?=0.05) corn stover, eucalyptus and 
poultry litter from each other. This is due to distinct chemical natures of these 
materials. PC2 was unable to statistically (?=0.05) separate corn stover from 
sugarcane bagasse and switchgrass which may be due to similarities of chemical 
or physical nature for these dusts since these biomass are all grassy biomass. 
(a) (b) 
(c) (d) 
  
 
107 
 
Similarly it was unable to separate pecan shell, sweetgum, corn cobs and 
eucalyptus from each other. 
 It can be seen that PC4 was able to statistically (?=0.05) separate corn 
cobs, eucalyptus, pine, pecan shell and sweetgum from each other (figure 4.4b). 
Also, it is able to statistically (?=0.05) separate Bermuda grass, poultry litter, pecan 
shell and sweetgum dusts. Similarly, PC3 statistically (?=0.05) separates poultry 
litter, sugarcane bagasse, corn cobs and pine. However it fails to separate 
sugarcane bagasse, switchgrass, corn stover and Bermuda grass from each other. 
Chemistry associated with the significant PC which was the cause of separation 
between these biomass dusts will be discussed in a later section.  
 It is evident from figure 4.4c that PC6 was able to statistically (?=0.05) 
separate eucalyptus, sweetgum, pecan shell, corn stover and pine from each 
other. PC5 also statistically (?=0.05) separated eucalyptus, poultry litter and pecan 
shell. PC6 was however unable to separate switchgrass, poultry litter, sweetgum 
and sugarcane bagasse from each other. PC5 also was unable to statistically 
(?=0.05) separate these dusts from each other.  
 PC9 and PC10 were not able to effectively separate most of the dust 
samples from each other (figure 4.4d). PC9 only separated pine and sugarcane 
bagasse from Bermuda grass (?=0.05), whereas PC10 only managed to 
statistically (?=0.05) separate corn cobs from pine. Thus, dusts showed large 
variation in the chemistry associated with these PCs which were unable to 
separate different dusts.  
  
 
108 
 
 Variation in biomass dusts observed on a particular PC through PC score 
plots is due to the chemistry associated with that PC. Thus, PC analysis can also 
be an important analytical tool describing differences in materials/samples under 
study besides being helpful in prediction model building. 
 
4.4.4 Models for Prediction of Heating and Ignition Properties of Dusts 
 Models were developed for prediction of heating and ignition properties of 
biomass dusts using first derivative and raw NIR spectral data (table 4.3). 
Coefficient of determination (R2) values obtained for MIT, TORV, TMML, TRE and 
TME models derived from raw spectra were better than R2 values for models 
developed using first derivative data. However, R2 value for prediction of TOXY 
was better in case of first derivative model (0.798) than model developed using 
raw spectra (0.737). Number of PC used for first derivative models were either less 
than or equal to the number of PC used to develop raw spectral models for 
respective properties. Lower statistical performance was obtained from first 
derivative models. Similar conclusion was drawn by Via (2013) in his experiment 
to develop NIR models for load capacity and deflection of wood composites. 
 Highest RMSEC, PRESS and RMSEP, and lowest RPD values were 
obtained from TOXY raw spectra and first derivative spectra models (table 4.3). 
An effective RPD value should be between 1.5 to 2.5 or greater (Via, 2013). MIT, 
TORV and TMML models had RPD values greater than 2.5. This means that these 
models are adequate for internal validation. Except for TOXY, all the models 
developed using raw spectra had R2 values of >0.90.  
  
 
109 
 
Table 4.3 Calibration and validation statistics for prediction models developed 
using raw and first derivative NIR spectra. 
 
 Figure 4.5 shows actual vs. predicted values for heating and ignition 
properties of biomass dusts based on raw spectra NIR model. Actual vs. predicted 
value plots for MIT, TORV and TMML models were found to be linearly related. In 
case of TOXY model (figure 4.5d), spread of predicted values for given actual 
values was more as compared to other models. In case of TRE model, it can be 
seen that pecan shell dust has significantly (?=0.05) lower TRE values as 
compared to other biomass dusts. Pecan shell dust in this case has high leverage 
(influence) effecting the mode (figure 4.5e). Similarly, poultry litter was found to 
have significantly different TME value than the rest of the biomass dusts (figure 
4.5f).  
 
M I T R a w 10 1 1 . 0 0 0 0 . 9 9 4 0 . 9 9 1 1 . 3 9 3 1 . 7 6 9 8 . 3 3 3 9 0 . 7 3 1
F i r s t  d e r i v a t i v e 8 9 . 1 0 6 0 . 9 7 6 0 . 9 6 6 2 . 7 0 1 3 . 2 9 4 4 . 4 7 5 3 1 4 . 6 0 6
T OR V R a w 7 5 . 4 1 8 0 . 9 8 4 0 . 9 7 8 1 . 6 1 2 1 . 9 0 8 5 . 7 3 6 1 0 5 . 5 3 6
F i r s t  d e r i v a t i v e 6 8 . 0 1 5 0 . 9 6 4 0 . 9 5 5 2 . 3 1 9 2 . 6 1 7 4 . 1 8 1 1 9 8 . 6 1 1
T M W L R a w 8 7 . 8 8 5 0 . 9 6 3 0 . 9 4 9 3 . 3 8 1 4 . 1 3 3 3 . 6 0 5 4 9 5 . 4 4 4
F i r s t  d e r i v a t i v e 8 7 . 6 9 0 0 . 9 4 3 0 . 9 2 2 4 . 1 7 1 5 . 0 1 4 2 . 9 7 1 7 2 9 . 1 6 7
T OX Y R a w 8 9 . 8 2 4 0 . 7 3 7 0 . 6 3 7 1 0 . 2 1 3 1 1 . 7 1 1 1 . 4 4 7 3 9 7 7 . 1 5 8
F i r s t  d e r i v a t i v e 8 9 . 0 7 1 0 . 7 9 8 0 . 7 2 1 8 . 9 4 9 1 0 . 6 1 8 1 . 5 9 6 3 2 6 9 . 6 2 7
T R E R a w 8 9 . 1 9 1 0 . 9 3 1 0 . 9 0 4 3 . 8 1 9 4 . 9 3 4 2 . 5 0 0 7 0 5 . 9 4 3
F i r s t  d e r i v a t i v e 8 8 . 6 9 4 0 . 9 2 3 0 . 8 9 4 4 . 0 2 2 4 . 9 6 2 2 . 4 8 6 7 1 4 . 0 5 2
T M A X R a w 7 5 . 9 5 2 0 . 9 0 1 0 . 8 6 9 5 . 4 9 9 6 . 2 1 6 2 . 4 4 4 1 1 2 0 . 5 8 9
F i r s t  d e r i v a t i v e 7 6 . 7 2 2 0 . 8 9 5 0 . 8 6 2 5 . 6 5 2 6 . 6 3 1 2 . 2 9 1 1 2 7 5 . 2 6 2
R M S E P R PD PR E S SM o d e l T y p e C p R
2
A d j . R
2
R M S E CN u m b e r  o f  PCs
  
 
110 
 
 
 
Figure 4.5 Actual vs. predicted values for MIT (a), TORV (b), TMML (c), TOXY (d), 
TRE (e) and TME (f). 
  
 
111 
 
4.4.5 Model Elucidation 
 Table 4.4 shows all the significant wavelengths affecting heating and 
ignition properties of biomass dusts corresponding to the statistically significant 
principal components used for their prediction. The wavelengths were considered 
significant if the peaks associated with wavenumbers exceeded the ?two standard 
deviation? mark along the loading distribution for a particular PC (figure 4.6). 
 It can be seen that all of the significant PC are associated with either lignin, 
cellulose, hemicellulose or water. Based on PC score plots (figure 4.4) and table 
4.4, our summary is that the differences in PC score plots was due to difference in 
chemical composition of dusts.  
4.4.6 External Validation 
 External validation was performed on the selected models for six heating 
and ignition properties (TORV, TMML, TOXY, TRE, TME and MIT). Figure 4.7 
shows the actual vs. predicted temperature values for four biomass dust samples 
used for external validation. For prediction of TORV, TMML, TRE, TME and MIT, 
raw spectra based models were selected for validation purpose. Selection of 
models was based on their internal cross validation performance discussed in 
4.4.4. Although the models suggest a positive correlation between actual and 
predicted values for the six heating and ignition properties, coefficient of 
determination (R2) values obtained were all less than 0.28.  
 
  
 
112 
 
Table 4.4 Chemistry/bond assignment for important wavelengths extracted from 
statistically significant principal components through regression analysis 
(Schwanninger et al., 2011). 
Significant PCs Wavenumber Chemistry/bond assignment 
PC2 4534 Lignin (4546) 
 4724 Cellulose (4739) 
 5634 Cellulose (5618) 
  6344 Cellulose (6344) 
PC3 4424 Lignin (4411) 
 5174 Water (5220-5150) 
 5924 Lignin (5935) 
  6884 Lignin (6874) 
PC4 5024 Water (5051) 
 5344 ?? 
 6474 Cellulose (6472) 
  8174 Cellulose (8250-8160) 
PC5 4414 Lignin (4411) 
 4594 cellulose and hemicellulose (4591) 
 5734 Cellulose (5776) 
 5844 Hemicellulose (5848) 
  8164 Cellulose (8250-8160) 
PC6 4354 Cellulose (4365) 
 4414 Lignin (4411) 
 6014 Hemicellulose (6003) 
 6554 Cellulose (6520) 
  8144 Cellulose (8160) 
PC10 8134 Cellulose (8160) 
PC9 4344 Cellulose (4365) 
 5774 Cellulose (5776) 
 8204 Cellulose (8250-8160) 
  8354 Lignin (8370) 
  
 
113 
 
 
Figure 4.6 Eigenvector loading on NIR spectra showing wavenumber vs. 
eigenvectors for significant PC. Dashed line represents 95th percentile of 
eigenvector distribution. 
   
  
 
114 
 
 
Figure 4.7 External validation results showing actual vs. predicted values for TORV 
(a), TMML (b), TOXY (c), TRE (d), TME (e) and MIT (f). 
 
 
  
 
115 
 
Table 4.5 External validation statistics for performance of prediction models 
developed using raw NIR spectra. 
Model Number of PCs R2 Adj. R2 RMSEP RPD PRESS 
MIT 10 0.02 -0.07 21.14 0.73 4917.48 
TORV 7 0.13 0.04 14.36 0.55 2269.67 
TMWL 8 0.12 0.03 17.10 0.80 3217.99 
TOXY 8 0.03 -0.06 24.98 0.60 6861.60 
TRE 8 0.03 -0.06 12.23 0.39 1646.54 
TMAX 7 0.28 0.21 14.50 0.42 2311.26 
 
PRESS and RMSEP values for all the external validation models were high, 
resulting in low RPD values, ranging from 0.39 to 0.80 (table 4.5). The poor 
performance of external validation models is due to the variation between biomass 
of each type (from different sources) which is not accounted for while developing 
prediction models. 
    
4.5 Conclusion 
 In case of coal dusts, bituminous coal showed highest absorbance value 
because of its high bulk density which increases the concentration of material in 
the path of NIR light beam. Chemistry associated with influential wavenumbers 
(derived from first derivative spectra) tells us that the difference in chemical 
constituents of biomass such as lignin, cellulose, hemicellulose, water and other 
chemical groups are the main reason of variation in NIR spectra obtained. 
 All the PCs correspond to lignin, cellulose and/or hemicellulose content of 
the biomass and based on the difference between these constituents of samples, 
different PCs were able to separate some dusts from others in PC score plots. 
  
 
116 
 
 R2 values for internal validation of models were greater than 0.90 for MIT, 
TORV, TMML, TRE and TME models derived using raw NIR spectra for predicting 
heating and ignition parameters. For TOXY model, first derivative derived model 
yielded better R2 (0.798) than raw derivative based model (R2=0.737). 
 R2 values for external validation of all the models was less than 0.28. The 
poor performance of models when validated externally with biomass dusts from 
different sources can be attributed to the variation between biomass of each type 
(from various sources) which is not accounted for while developing prediction 
models.
  
 
117 
 
Chapter 5 Summary and Future Recommendation 
5.1 Summary 
 Most of the biomass feedstock showed reduction in moisture content after 
grinding and sieving. Bituminous coal, sugarcane bagasse and sweetgum 
however showed an increase in moisture content after grinding operation due to 
low initial moisture content and hygroscopic nature of the material. Bituminous coal 
had higher MIT as compared to other biomass dusts because of its low volatile 
content. TORV and TMML in coal dusts was found to decrease with increase in 
volatile content. This trend was not observed in biomass due to difference in 
structure of solid matrices of fuels. Ignition of volatiles corresponded to maximum 
rate of mass loss in biomass dusts. Based on oxidation temperature and activation 
energy values, lignite coal was found to have a very high risk of ignition. Bermuda 
grass, PRB coal, pecan shell, corn cobs, corn stover, eucalyptus and poultry litter 
were categorized as dusts having high risk of ignition. Switchgrass, pine and 
sugarcane bagasse were at medium risk of ignition whereas, bituminous coal and 
sweetgum were at low risk of ignition. High activation energy and TOXY value for 
bituminous coal can be due to its low volatile content and high carbon content.  
 Principal component analysis was performed on NIR spectra (raw and first 
derivative) to obtain models to predict heating and ignition parameters. PC 
  
 
118 
 
analysis on raw NIR spectra to build prediction models for MIT, TORV, TMML, 
TOXY, TRE and TME yielded R2 values in range of 0.737-0.994. Whereas, 
prediction models based on first derivative NIR spectra for the abovementioned 
properties yielded R2 values in range of 0.798-0.976. Except for TOXY models, all 
the prediction models based on raw NIR spectra performed better than models 
based on first derivative spectra. Significant principal components obtained 
corresponded mainly to the basic constituents of biomass viz., cellulose, lignin and 
hemicellulose. Cause of variation in the NIR spectra of biomass dusts was due to 
variation in these constituents of different biomass dusts.  
5.2 Future Recommendation  
 The study provides an insight into physical, chemical, heating and ignition 
properties of biomass and coal dusts. The study would be beneficial in setting 
guidelines for maximum permissible temperatures of process equipment in 
facilities handling biomass. Findings of this study can be incorporated in a standard 
against dust fire and explosion hazards. Use of NIR spectroscopy would provide a 
quick estimation of heating and ignition parameters. However, more robust 
prediction models can be developed using the procedure given in this study by 
incorporating samples from a number of sources for each biomass. Developed 
models can be validated externally to check for robustness.  
 The fluctuating prices of fossil fuels and its usage?s negative impact on 
environment is drawing a lot of attention on alternative fuels such as biomass. Liu 
et al. (2002) mentioned that co-combustion of biomass and coal could help in 
lowering NOx emission. Also, blending of different grades of biomass and/or coal 
  
 
119 
 
may help reduce flame stability problems and corrosion problems due to deposited 
ash (Jiricek et al., 2012). Thus, the study can also be performed on dusts from 
blends of biomass/coal fuels such as pine ? coal blends and other commonly used 
blends for power generation. Effect of different ratios of blended fuels on heating 
and ignition properties can be studied. 
    
  
 
120 
 
References 
Abbasi T., and S.A. Abbasi. 2007. Dust explosions?Cases, Causes, 
Consequences, and Control. J. Hazard. Mat. 140(1-2): 7-44. 
Acheson E.D., R.H. Cowdell, and E.H. Rang. 1981. Nasal cancer in England and 
Wales: An occupational survey, Brit. J. Ind. Med. 38: 218-224. 
Amyotte P.R., M.J. Pegg, F. I. Khan, M. Nifuku, and T. Yingxin. 2007. Moderation 
of dust explosions. J. Loss Prev. Proc. Ind. 20(4-6): 675-687. 
Amyotte P.R, and R.K. Eckhoff. 2010. Dust Explosion Causation, Prevention and 
Mitigation: An Overview. J. Chem. Health Safe. 17(1): 15-28. 
Amyotte P.R., C.T. Cloneya, F.I. Khan, and R.C. Ripley. 2012. Dust explosion risk 
moderation for ?occulent dusts. J. Loss Prev. Proc. Ind. 25: 862-869. 
An H., and S.W. Searcy. 2012. Economic and energy evaluation of a logistics 
system based on biomass modules. Biomass Bioenergy. 46: 190-202. 
ASABE Standard S319.3, 2003. Method of determining and expressing fineness 
of feed materials by sieving. American Society of Agricultural and Biological 
Engineers. St. Joseph, MI. 
ASABE S593.1. 2011. Terminology and Definitions for Biomass Production, 
Harvesting and Collection, Storage, Processing, Conversion and Utilization. 
American Society of Agricultural and Biological Engineers. St. Joseph, MI. 
ASABE S269.5. 2012. Densified Products for Bulk Handling ? Definitions and 
Method. American Society of Agricultural and Biological Engineers. St. 
Joseph, MI. 
ASTM E871-82 Reapproved. 2006. Standard Test Method for Moisture Analysis 
of Particulate Wood Fuels. American Society for Testing and Materials. West 
Conshohocken, PA. 
ASTM E2021-09. 2010. Standard Test Method for Hot-Surface Ignition 
Temperature of Dust Layers. American Society for Testing and Materials. 
West Conshohocken, PA. 
  
 
121 
 
ASTM D3174-04. 2004. Standard Test Method for Ash in the Analysis Sample of 
Coal and Coke from Coal. American Society for Testing and Materials. West 
Conshohocken, PA. 
Balat M., and G. Ayar. 2005. Biomass energy in the world, use of biomass and 
potential trends. Energy Sources. 27 (10): 931-940. 
Benedetto A., V.D. Sarli, and P. Russo. 2007. On the determination of the 
minimum ignition temperature for dust/air mixtures. Napoli, Italy. Istituto di 
Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche (CNR).  
Biagini E., A. Fantei, and L. Tognotti. 2008. Effect of the heating rate on the 
devolatilization of biomass residues. Thermochimica Acta. 472(1-2): 55-63. 
Bilbao R., J.F. Mastral, J.A. Lana, J. Ceamanos, M.E. Aldea, and M. Betran. A 
model for the prediction of the thermal degradation and ignition of wood 
under constant and variable heat ?ux. J. Anal. Appl. Pyrolysis. 62(1): 63-82. 
Biomass Magazine. 2014. Koda energy reopend with new design, Feb 14, 2014. 
Biomass Magazine. Grand Forks, ND. 
Blair A.S. 2012. Dust Explosion Primer. U.S. Chemical Safety and Hazard 
Investigation Board. Washington, DC. 
BLS. 2013. Bureau of Labor Statistics. United States Department of Labor. 
Available at  http://www.bls.gov/iif/oshwc/cfoi/cftb0266.pdf. Accessed 10 
March 2014. 
Bobelyn E., A. Serban, M. Nicu, J. Lammertyn, B.M. Nicolai., and W. Saeys. 2010. 
Postharvest quality of apple predicted by NIR-spectroscopy: Study of the 
effect of biological variability on spectra and model performance. Postharv. 
Biol. Tech. 55(3): 133-143. 
Bowes P.C., and S.E. Townshend. 1962. Ignition of combustible dusts on hot 
surfaces. Brit. J. Appl. Phys. 13: 105-114. 
Breitkreitz M.C., I.M. Raimundo, J. R. Rohwedder, C. Pasquini a, H.A. Dantas 
Filho, G.E. Jos? , and M.C. Ara?jo. 2003. Determination of total sulfur in 
diesel fuel employing NIR spectroscopy and multivariate calibration. Analyst. 
128: 1204-1207. 
Brereton R. 2007. Applied Chemometrics for Scientists. West Sussex, UK: John 
Wiley and Sons, Ltd. 
Buening-Pfaue H.2003. Analysis of water in food by near infrared spectroscopy. 
Food Chem. 82: 107-115. 
  
 
122 
 
Calle? S., L. Klaba, D. Thomas, L. Perrin, and O. Dufaud. 2005. Influence of the 
Size Distribution and Concentration on Wood Dust Explosion: Experiments 
and Reaction Modelling. Powder Tech. 157(1-3): 144-148. 
Cashdollar K.L. 2000. Overview of dust explosibility characteristics. J. Loss Prev. 
Proc. Ind. 13: 183-199. 
Chen Y., S. Mori, and S.P. Pan. 1995. Estimating the combustibility of various 
coals by TG-DTA. En. Fuels. 9: 71-74. 
Chen Y., S. Mori, and W.P. Pan. 1996. Studying the mechanisms of ignition of 
coal particles by TG-DTA. Thermochimica Acta. 275: 149-158. 
Chiba Lee. 2014. Poultry nutrition and feeding. In Animal Nutrition Handbook. 
410-425. Auburn, AL. 
Collell C., P. Gou, J. Arnau, I. Munoz, and J. Comaposada. 2012. NIR technology 
for on-line determination of superficial aw and moisture content during the 
drying process of fermented sausages. Food Chem. 135(3): 1750-1755. 
Cook J., and J. Beyea. 2000. Bioenergy in the United States: progress and 
possibilities. Biom. Bioen. 18(6): 441-455. 
CSB. 2006. Investigation Report ? Combustible Dust Hazard Study. Report No.     
2006-H-1. Washington, D.C.: U.S. Chemical Safety and Hazard 
Investigation Board. 
CSB. 2013. U.S. Chemical safety board. Available at www.csb.gov. Accessed 10 
February 2014. 
Cuiping L., W. Chuangzhia, Yanyongjieb, and H. Haitaoa. 2004. Chemical 
elemental characteristics of biomass fuels in China. Biomass Bioenergy. 
27(2): 119-130. 
Demirbas A. 2004. Combustion characteristics of different biomass fuels. Fuel 
Prog. En. Combust. Sci. 30(2): 219?230. 
Demirbas A. 2005. Potential applications of renewable energy sources, biomass 
combustion problems in boiler power systems and combustion related 
environmental issues. Prog. Energy. Combust. Sci. 31 (2): 171-192. 
Ebadat V. 2010. Dust explosion hazard assessment. J. Loss Prev. Proc. Ind. 
23(6): 907-912. 
Ebling J.M., and B.M. Jenkins. 1985. Physical and chemical properties of biomass 
fuels. Trans. ASAE. 28(3): 898-902. 
  
 
123 
 
Eckhoff R.K. 1991. Dust Explosions in the Porcess Industry. 3rd ed. Gulf 
Professional Publishing. Elsevier Science.  
Eckhoff R.K. 2002. Minimum ignition energy (MIE) ? a basic ignition sensitivity 
parameter in design of intrinsically safe electrical apparatus for explosive 
dust clouds. J. Loss Prev. Proc. Ind. 15(4).: 305-310. 
Eckhoff R. K. 2003. Dust explosions in the process industries. 3rd ed. Boston, 
Mass.: Gulf Professional Publishing/Elsevier. 
EIA. 2013. United States Energy Information Administration. Available at 
http://www.eia.gov. Accessed 28 Dec, 2013. 
Esteban L.S., and J.E. Carrasco. 2006. Evaluation of different strategies for 
pulverization of forest biomasses. Powder Tech. 166 (3): 139-151. 
Fasina O. O. 2008. Physical properties of peanut hull pellets. Biores. Tech. 99 (5): 
1259-1266. 
Field P. 1982. Handbook of Powder Technology. Vol. 4. Dust Explosions. Elsevier 
Science. 
Foley W.J., A. McIlwee, I. Lawler, L. Aragones, A. P. Woolnough, and N. Berdig. 
1998. Ecological applications of near infrared re?ectance spectroscopy: A 
tool for rapid, cost-effective prediction of the composition of plant and animal 
tissues and aspects of animal performance. Oecologia. 116:293?305. 
Friedman H.L. 1963. Kinetics of thermal degradation of char-foaming plactics from 
thermo-gravimetry?application to a phenolic resin. Polym. Sci. 6: 183-195. 
Gan J., and C.T. Smith. 2007. Co-benefits of utilizing logging residues for 
bioenergy production: the case for East Texas, USA. Biom. Bioen. 31(9): 
623-630. 
Gani, A., I. Naruse. 2007. Effect of cellulose and lignin content on pyrolysis and 
combustion characteristics for several types of biomass. Renewable Energy. 
32(4): 649?661.  
Garcia-Torrent J., E. Conde-Lazaro, C. Wilen, and A. Rautalin. 1998. Biomass 
dust explosibility at elevated initial pressures. Fuel. 77 (9/10): 1093-1097. 
Gil M.V., D. Cascal, C. Pevida, J.J. Pis and F. Rubiera. 2010. Thermal behaviour 
and kinetics of coal/biomass blends during co-combustion. Biores. Tech. 
101: 5601-5608. 
  
 
124 
 
Gronli M., M.J. Antal Jr., G. Varhegyi. 1999. A round-robin study of cellulose 
pyrolysis kinetics by thermogravimetry. Ind. Eng. Chem. Res. 38(6): 2238-
2244. 
Grotkjaer T., K.D. Johansen, A.D. Jensen, and P. Glarborg. 2003. An 
experimental study of biomass ignition. Fuel. 82(7): 825-833. 
Hausen B. 1981. Woods injurious to human health- A manual. Walter De Gruyter 
Inc. Berlin, German Federal Republic. 
Haykiri-Acma H. 2003. Combustion characteristics of different biomass materials. 
En. Cons. Mgmt. 44(1): 155-162. 
Hehar G.S. 2013. Physicochemical and Ignition Properties of Dust from Loblolly 
Pine Wood. MS Thesis. Auburn Ala.: Auburn University, Department of 
Biosystems Engineering.  
ISO 562. 2010. Hard coal and coke. Determination of volatile matter. International 
Organization for Standardization. Geneva, Switzerland. 
Jamshidi B., S. Minaei, E. Mohajerani, and H. Ghassemian. 2012. Reflectance 
Vis/NIR spectroscopy for nondestructive taste characterization of Valencia 
oranges. Comp. Elec. Agri. 85: 64-69. 
Janes A., D. Carson, A. Accorsi, J. Chaineaux, B. Tribouilloy, and D. Morainvillers. 
2008. Correlation between self-ignition of a dust layer on a hot surface and 
in baskets in an oven. J. Haz. Mat. 159 (2-3): 528-535. 
Jenkins B.M. 1989. Biomass handbook. In Physical properties of biomass, Ch 5.2. 
New York, NY.: Gordon & Breach, 
Jenkins B.M., L.L Baxter, T.R Miles Jr., and T.R Miles. 1998. Combustion 
properties of biomass. Fuel Proc. Tech. 54(1-3): 17-46. 
Jeske, A., and H. Beck. 1989. Evaluation of dust explosions in the Federal 
Republic of Germany. EUROPEX Newsletter. No 9: 2. 
Jiricek I., P. Rudasov, and T. Zemlov. 2012. A thermogravimetric study of the 
behaviour of biomass blends during combustion. Acta Polytechnica. 52(3): 
39-42. 
Joshi K.A., V. Raghavan, and A.S. Rangwala. 2012. An experimental study of 
coal dust ignition in wedge shaped hot plate configurations. Combust. Flame. 
159(1): 376-384. 
  
 
125 
 
Kauffman C.W. 1982.  Agricultural dust explosions in grain handling facilities. 
J.H.S. Lee, C.M. Guirao (Eds.). Fuel-air Explosions. University of Waterloo  
Press. Waterloo, Ontario, Canada: 305?347. 
Kettaneh N., A. Berglund, and S. Wold. 2005. PCA and PLS with very large data 
sets. Comp. Stat. Data An. 56: 101-8. 
Kheshgi H.S., R.C. Prince, and G. Marland. 2000. The potential of biomass fuels 
in the context of global climate change: focus on transportation fuels. Annual 
Rev. Energy Environ. 25: 199-244. 
Khan, N., M. Bradley, and R. Berry. 2008. Best Practice Guide for Handling of 
Biomass Fuels and Coal-Biomass Mixes. The Wolfson Centre for Bulk Solids 
Handling Technology. 
Liang Y.L., and O.M. Kvalheim. 1996. Robust methods for multivariate analysis ? 
a tutorial view. Chemometr. Intell. Lab. 32:1-10. 
Liodakis S., and A. Dimitrakopoulos. 2002. Ignition characteristics of forest 
species in relation to thermal analysis data. Thermochimica Acta. 390 (1-2): 
83-91. 
Littlefield B., O.O. Fasina, J. Shaw, S. Adhikari, and B. Via. 2011. Physical and 
flow properties of pecan shells?Particle size and moisture effects. Powder 
Tech. 212 (1): 173-180. 
Liu D.C., C.L. Zhang, T. Mi, B.X. Shen, and B. Feng. 2002. Reduction of N2O and 
NO emissions by co-combustion of coal and biomass. J. Inst. Energy. 75: 
81-85. 
Liu X., and X.T. Bi. 2011. Removal of inorganic constituents from pine barks and 
switchgrass. Fuel Proc. Tech. 92(7): 1273-1279. 
Louw E.D., and K.I. Theron. 2010. Robust prediction models for quality 
parameters in Japanese plums. Postharvest Biol. Tech. 58: 176?184. 
Lu R., and B.B. Bailey. 2005. NIR measurement of apple fruit soluble solids 
content and firmness as affected by postharvest storage. ASAE Annual 
International Meeting, Tampa, Fla. Paper No. 056070. 
Lunn G.A. 1992. Dust explosions in the process industries : by R. K. Eckhoff, 
published by Butterworth Heinemann, Oxford, 1991, 599 pp.; ISBN: 0-7506-
1109-x; ?80.00. Powder Tech. 73 (1): 97. 
  
 
126 
 
Mafakheri F., and F. Nasiri. 2014. Modeling of biomass-to-energy supply chain 
operations: Applications, challenges and research directions. Energy Pol. 
67: 116-126. 
Mani S., L.G Tabil, and S. Sokhansanj. 2003. An overview of compaction of 
biomass grinds. Powder Handling Proc. 15 (3): 160?168.  
Mani S., L. G. Tabil, and S. Sokhansanj. 2004. Grinding performance and physical 
properties of wheat and barley straws, corn stover and switchgrass. Biomass 
Bioenergy. 27(4): 339-352. 
Martens H. 1979. Factor analysis of chemical mixtures. Anal. Chim. Acta. 112: 
423-42. 
Mayfield C.A., D. Foster, T. Smith, J. Gan, and S. Fox. 2007. Opportunities, 
barriers, and strategies for forest bioenergy and bio-based product 
development in the Southern United States. Biom. Bioen. 31(9): 631-637. 
McKendry P. 2002. Energy production from biomass (part 1): overview of 
biomass. Biores. Tech. 83(1): 37-46. 
McNew K., and D. Griffith. 2005. Measuring the impact of ethanol plants on local 
grain prices. Review. Agri. Eco. 27: 164-180. 
Michaels L. 1967. Lung changes in woodworkers. Can. Med. Assoc. J. 96: 1150. 
Miron Y., and C. P. Lazzara. 1988. Hot-surface ignition temperatures of dust 
layers. Fire Matls. 12: 115-126. 
Moes J.J., M.M. Ruijken, E. Gouta, H.W. Frijlink, and M.I. Ugwokea. 2008. 
Application of process analytical technology in tablet process development 
using NIR spectroscopy: Blend uniformity, content uniformity and coating 
thickness measurements. Int.J. Pharma. 357(1-2): 108-118. 
Moqbel S., D. Reinhart, and R.H. Chen. 2010. Factors influencing spontaneous 
combustion of solid waste. Waste Mgmt. 30 (8?9): 1600-1607.  
Muthuraman M., T. Namiokaa, K. Yoshikawab. 2010. A comparison of co-
combustion characteristics of coal with wood and hydrothermally treated 
municipal solid waste. Biores. Tech. 101(7): 2477-2482. 
NCDOL. 2012. A guide to combustible dusts. N.C. department of labor - 
Occupational Safety and Health Division. Available at 
http://www.nclabor.com/osha/etta/indguide/ig43.pdf. Accessed 11 February 
2014. 
  
 
127 
 
Neter J., M. Kunter, C. Nachtsheim, and W. Wasserman. 1996. Applied Linear 
Statistical Models. 4th Ed. Irwin, Boston, MA, USA. 
NFPA. 2013. Standard for Prevention of Dust Explosions from the Manufacturing, 
Processing and Handling of Combustible Particulate Solids. NFPA 654. 
Quincy, Mass.: National Fire Protection Association. 
NFPA. 2013. Recommended practice for the classification of combustible dusts 
and of hazardous locations for electrical installations in chemical process 
areas online. NFPA 499. Quincy, Mass.: National Fire Protection 
Association. 
Nicolai B.M., B.E. Verlinden, M. Desmet, S. Saevels, W. Sayes, K. Theron, R. 
Cubeddu, A. Pifferi, and A. Torricelli. 2008. Time-resolved and continuous 
wave NIR reflectance spectroscopy to predict soluble solids content and 
firmness of pear. Postharvest Biol. Tech. 47(1): 68-74. 
Nifuku M., H. Tsujita, K. Fujino, K. Takaichi, C. Barre, M. Hattori, S. Fujiwara, S. 
Horiguchi, and E. Paya. 2005. A study on ignition characteristics for dust 
explosion of industrial wastes. J. Electrostatics. 6(6-10): 455-462. 
Norris K. H. 1964. Reports on design and development of a new moisture meter. 
Agric. Eng. 45(7): 370-372. 
NREL. 2005. Determination of Ash in Biomass by Laboratory Analytical 
Procedure. National Renewable Energy Laboratory. NREL/TP-510-42622. 
Orfao J.J.M., F.J.A. Antunes, and J.L. Figueiredo. Pyrolysis kinetics of 
lignocellulosic materials: three independent reactions model. Fuel. 78: 349-
358. 
Osborne B.G., T. Fearn, and P.H. Hindle. 1993. Practical NIR spectroscopy with 
applications in food and beverage analysis. Longman Group United 
Kingdom. 2nd sub Ed. ISBN: 0-582-09946-3. 
OSHA. 2014. OSHA Regional News Release. Occupational Safety and Health 
Administration. Available at: 
https://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=NE
WS_RELEASES&p_id=25695. Accessed Jul 7, 2014. 
Park H. 2006. Hot surface ignition temperature of dust layers With and without 
Combustible additives. Master?s Thesis. Worcester Polytechnic Institute, 
Mass., USA. 
  
 
128 
 
Park H., A.S. Rangwala, and N.A. Dembsey. 2009. A means to estimate thermal 
and kinetic parameters of coal dust layer from hot surface ignition tests. J. 
Hazard. Mat. 168 (1): 145-155. 
Park Y.H., J. Kim, S.S. Kim, and Y.K. Park. 2009.  Pyrolysis 
characteristics and kinetics of oak trees using thermogravimetric analyzer 
and micro-tubing reactor. Biores. Tech. 100(1): 400-405. 
Parikka, M. 2004. Global biomass fuel resources. Biomass and Bioenergy. 27: 
613-620. 
Perlack, R. D., L. L. Wright, A. F. Turhollow, R. L. Graham, B. J. Stokes, and D. 
C. Erbach. 2005. Biomass as feedstock for a bioenergy and bioproducts 
industrythe technical feasibility of a billion-ton annual supply. United States 
Department of Energy. Oak Ridge, TN. 
Pinho R., F. Bonatto, M. Andrades, M. Frota, C.Ritter, F. Klamt, F. Dal-Pizzol, J.M. 
Uldrich-Kulczynski, and J.F. Moreiraa.2004. Lung oxidative response after 
acute coal dust exposure. Environ. Res. 96: 290-297. 
Polka M., Z. Salamonowicz, M. Wolinski, and B. Kukfisz. 2012. Experimental 
analysis of minimal ignition temperatures of a dust layer and clouds on a 
heated surface of selected flammable dusts. Procedia Engineering. Vol. 45: 
414-423. 
Porteir J., D. Patino, J. Collazo, E. Granada, J. Moran, and J.L. Miguez. 2010. 
Experimental analysis of the ignition front propagation of several biomass 
fuels in a fixed-bed combustor. Fuel. 89 (1): 26-35. 
Probst K.V., R. P. Kingsly Ambrose, R. L. Pinto, R. Bali, P. Krishnakumar, and K. 
E. Ileleji. 2013. The effect of moisture content on the grinding performance 
of corn and corncobs by hammermilling .Trans. ASABE.  56(3): 1025-1033. 
Rajagopal D., S. Sexton, G. Hochman, D. Roland-Holst, and D. Zilberman. 2009. 
Model estimates food-versus-biofuel trade-off. Cali. Agri. 63(4): 199-201. 
Rambla F.J., S. Garriguea, and M. de la Guardia. 1997. PLS-NIR determination 
of total sugar, glucose, fructose and sucrose in aqueous solutions of fruit 
juices. Analytica Chimica Acta. 344: 41-53. 
Ramirez A. 2009. Determination of parameters used to prevent ignition of stored 
materials and to protect against explosions in food industries. Pedro J. 
Aguadoc. 168(1): 115-120. 
  
 
129 
 
Ram?rez A., J. Garc?a-Torrent, and A. Tasc?n. 2010. Experimental determination 
of self-heating and self-ignition risks associated with the dusts of agricultural 
materials commonly stored in silos. J. Hazard. Mat. 175: 920-927. 
Reddy P. D., P. R. Amyotte, and M. J. Pegg. 1998. Effect of inerts on layer ignition 
temperatures of coal dust. combustion flame. 114(1?2):41-53. 
Renewables International Magazine. 2011. Following explosion, world?s largest 
pellet plant resumes operation. Available at 
http://www.renewablesinternational.net/following-explosion-worlds-largest-
pellet-plant-resumes-operation/150/515/31440/. Accessed 12 March 2014. 
Rhodes M. 1998. Introduction to Particle Technology. John Wiley and Sons. New 
York. 
Ryu C., Y.B. Yang, A. Khor, N.E. Yates, V.N. Sharifi, and J. Swithenbank. 2006. 
Effect of fuel properties on biomass combustion: Part I. Experiments?fuel 
type, equivalence ratio and particle size. Fuel. 85: 1039-1046. 
Sahu S.G., P. Sarkar, N. Chakraborty, and A.K. Adak. 2010. Thermogravimetric 
assessment of combustion characteristics of blends of a coal with different 
biomass chars. Fuel Proc. Tech. 91(3): 369-378. 
Samaras P., E. Diamadopoulos, and G.P. Sakellaropoulosa. 1996. The effect of 
mineral matter and pyrolysis conditions on the gasification of Greek lignite 
by carbon dioxide. Fuel. 75(9): 1108-1114. 
Sapko M.J., E.S. Weiss, K.L. Cashdollar, and I.A. Zlochower. 2000. Experimental 
mine and laboratory dust explosion research at NIOSH.  J. of Loss Prev. in 
Proc. Ind. 13 (3-5): 229-242. 
SAS. 2009. SAS User?s Guide: Statistics. Ver. 6a. Cary, N.C.: SAS Institute, Inc. 
Schoeff R.W. 2006. Agricultural Dust Explosions in 2005. Available at 
krex.kstate.edu/dspace/bitstream/handle/2097/4305/2005Dust_Explosions.
pdf?sequence=1. Accessed on 2 february 2014.  
Schwanninger M., J.C. Rodrigues, K. Fackler. 2011. A review of band 
assignments in near infrared spectra of wood and wood components. J. Near 
Infrared Spectroscopy. 19: 287-308. 
Shi l., and M. Chew. 2011. Influence of moisture on autoignition of woods in 
conecalorimeter. J. Fire Sc. 30(2): 158-169. 
  
 
130 
 
Shiroma C., and L. Rodriguez-Saona. 2009. Application of NIR and MIR 
spectroscopy in quality control of potato chips. J. Food Comp. Anal. 22(6): 
596-605. 
Sorum L., M.G. Gronli, and J.E. Hustad. 2001. Pyrolysis characteristics and 
Kinetics of municipal solid wastes. Fuel. 80 (9): 1217-1227. 
SSCHE. 2009. Proceedings: 36th International Conference of Slovak Society of 
Chemical Engineering. ISBN 978-80-227-3072-3. Slovak Society of 
Chemical Engineering. 
Stephenson A.H., T.A. McCaskey, and B.G. Ruffin. 1990. A survey of broiler litter 
composition and value as a nutrient resource potential. Biol. Wastes. 34(1): 
1-9. 
Sweis F.K. 1998. The effect of admixed material on the ignition temperature of 
dust layers in hot environments. J. Hazard. Mat. 63(1): 25-35. 
Swinehart D.F. 1962. The Beer-Lambert law. J. Chem. Educ. 39 (7): 333. 
Tabil L.G. Pelleting and binding characteristics of alfalfa. 1996. Ph.D thesis, 
Department of Agricultural and Bioresource Engineering, Saskatoon, SK 
Canada. University of Saskatchewan. 
Tembo G., F.M. Epplin, and R.L. Huhnke. 2003. Integrative investment appraisal 
of a lignocellulosic biomass-to-ethanol industry. J. Agr. Res. Econ. 28 (3): 
611?633. 
Theimer O.P.1973. Cause and prevention of dust explosions in grain elevators 
and flour mills. Powder tech. 8: 137-147. 
Tiqui S.M., and N.F.Y Tam. 2000. Fate of nitrogen during composting of chicken 
litter. Environ. Poll. 110(3): 535-541. 
Tognotti L., A. Malotti, I. Petarca, and S. Zanelli. 1985. Measurement of ignition 
temperature of coal particles using a thermogravimetric technique. Combust. 
Sci. Technol. 44: 15-28. 
United Nations. 2014. UN News Center. Available at 
http://www.un.org/apps/news/story.asp?NewsID=45165#.U0IGEUco4eg. 
Accessed 13 march 2014. 
U.S. Department of Energy. 2010. Biomass multi-year program plan. Available at 
http://www1.eere.energy.gov/biomass/pdfs/biomass_mypp_november2010
.pdf; 2010. Accessed 30 November 2013. 
  
 
131 
 
USCB. 2014. United States Census Bureau. Available at 
http://www.census.gov/population/international/data/worldpop/table_popula
tion.php. Accessed 2 February 2014.  
USDA. 2010. Agricultural Research Service. Chicken litter has advantages over 
conventional fertilizers. Available at 
www.sciencedaily.com/releases/2010/06/100623124254.htm. Accesed 10 
March, 2014. 
Vamvuka D., E. Kakaras, E. Kastanaki, and P. Grammelis. 2003. Pyrolysis 
characteristics and kinetics of biomass residuals mixtures with lignite. Fuel.  
82(15-17): 1949-1960. 
Vamvuka D., and S. Sfakiotakis. 2011. Combustion behaviour of biomass fuels 
and their blends with lignite. Thermochimica Acta. 526 (1-2): 192-199. 
Vergnoux A., M. Guiliano, Y.L. Dr?au, J. Kister, N. Dupuy, and P. Doumenq. 2009. 
Monitoring of the evolution of an industrial compost and prediction of some 
compost properties by NIR spectroscopy. Sci. Total Environ. 407(7): 2390-
2403. 
Via B.K., T.F. Shupe, L.H. Groom, M. Stine, and C. So. 2003. Multivariate 
modelling of density, strength and stiffness from near infrared spectra for 
mature, juvenile and pith wood of longleaf pine (pinus palustris). J Near 
Infrared Spectrosc. 11 (5):365?378 
Via B.K. 2010. Prediction of oriented strand board wood strand density by near 
infrared and Fourier transform infrared refplectance spectroscopy. J Near 
Infrared Spectrosc. 27: 491-498. 
 
Via B.K., O.O. Fasina, and H. Pan. 2011. Assessment of pine biomass density 
through midinfrared spectroscopy and multivariate modeling. Biores. 6 (1): 
807?822. 
Via B. 2013. Characterization and evaluation of wood strand composite load 
capacity with near infrared spectroscopy. Mat. Struct. 46: 1801-1810. 
Via B.K., S. Adhikari, and S. Taylor. 2013. Modeling for proximate analysis and 
heating value of torre?ed biomass with vibration spectroscopy. Biores. Tech. 
133: 1-8.  
Vijayraghavan G. 2004. Impact assessment, modeling and control of dust 
explosions in chemical process industries. MTech Thesis. Department of 
Chemical Engineering, Coimbatore Institute of Technology.  
  
 
132 
 
Vijayraghavan G. 2011. Emerging emergency due to dust explosions in process 
industry. J. Engg. Res. Stud. 2 (4): 193-198. 
Virk S.S., J.P. Fulton, O.O. Fasina, and T.P. McDonald. 2013. Capacitance and 
near-infrared techniques for the real-time moisture measurement of broiler 
litter. Biosys. Eng. 116: 357-367. 
Vuthaluru H.B. 2004. Investigations into the pyrolytic behaviour of coal/biomass 
blends using thermogravimetric analysis. Biores. Tech. 92(2): 187-195. 
Wang. C., Y. Wu, Q. Liu, and H. Yang. 2011. Study of the characteristics of the 
ashing process of straw/coal combustion. Fuel. 90: 2939-2944. 
White J.E., W.J. Catallo, and B.L. Legendre. 2011. Biomass pyrolysis kinetics: A 
comparitive critical review with relevant agricultural residue case studies. J. 
Anal. App. Pyrolysis. 91 (1): 1-33. 
Wiedenhofer D., M. Lenzen, and J.K. Steinberger. 2013. Energy requirements of 
consumption: Urban form, climatic and socio-economic factors, rebounds 
and their policy implications. Energy Pol. 63: 696-707. 
Windham W.R., F.E. Barton II, and J.A. Robertson. 1988. Moisture analysis of 
forage by near infrared reflectance spectroscopy: Preliminary collaborative 
study and comparison between Karl Fischer and oven drying reference 
methods. J. Assoc. Off. Anal. Chem. 71: 256-262. 
Wypych P., D. Cook, and P. Cooper. 2005. Controlling dust emissions and 
explosion hazards in powder handling plants. Chem. Eng. Proc.: Proc. Intes. 
44 (2): 323-326. 
Yang H., R. Yan, H. Chen, C. Zheng, D.H. Lee, and D.T. Liang. 2006. In-depth 
investigation of biomass pyrolysis based on three major components: 
hemicellulose, cellulose and lignin. Energy Fuels. 20: 388-393. 
Yang H., R. Yan, H. Chen, D.H. Lee, and C. Zheng. 2007. Characteristics of 
hemicellulose, cellulose and lignin pyrolysis. Fuel. 86(12-13): 1781-1788. 
Zalosh R. 2008. Dust Explosion Fundamentals: Ignition Criteria and Pressure 
Development. Wellesley, Mass.: National Fire Protection Association. 
Available at 
http://www.nfpa.org/assets/files/pdf/foundation%20proceedings/dust_explo
sion_fundamentals_ignition_criteria_and_pressure_d.pdf. Accessed on 25 
November 2012.  
Zhang D.K., and T.F. Wall. 1994. Ignition of coal particles?the influence of 
experimental technique. Fuel. 73: 1114-1119. 
  
 
133 
 
Appendix A ? Initial Moisture Content and Physiochemical Properties of 
Dusts and Ground Material. 
Table A.1 Initial moisture content of feedstock. 
Biomass  Moisture content  (% w.b.) Mean Moisture Content (% w.b.) Standard Deviation  (% w.b.) 
Bermuda grass 
11.28 
10.91 0.32 10.71 
10.74 
Bituminous coal 
2.52 
2.42 0.09 2.35 
2.39 
Corn cobs 
9.62 
9.45 0.19 9.24 
9.49 
Corn stover 
9.28 
9.59 0.55 9.27 
10.23 
Eucalyptus 
7.11 
6.96 0.13 6.86 
6.91 
Lignite coal 
29.36 
29.40 0.04 29.44 
29.41 
Pecan shell 
12.70 
12.71 0.27 12.98 
12.45 
Pine 
11.53 
11.74 0.23 11.70 
11.99 
Poultry Litter 
14.82 
14.81 0.02 14.79 
14.82 
PRB coal 
15.26 
15.25 0.02 15.23 
15.26 
Sugarcane Bagasse 
(after drying) 
5.71 
5.68 0.05 5.62 
5.71 
Sugarcane Bagasse 
(before drying) 
25.64 
24.77 1.49 23.05 
25.62 
Sweetgum 
 (After drying) 
5.50 
5.45 0.08 5.36 
5.50 
Sweetgum 
 (before drying) 
51.97 
51.58 0.68 50.79 
51.97 
Switchgrass 
12.08 
12.42 0.79 11.85 
13.32 
 
  
 
134 
 
Table A.2 Moisture content of dust samples. 
Sample Moisture Content (w.b.%) Mean Moisture Content (W.b.%) Standard Deviation (w.b.%) 
Bermuda grass 8.35 8.06 0.27 
 8.02   
  7.82     
Bituminous coal 2.98 2.96 0.02 
 2.96   
  2.95     
Corn cobs 6.88 7.40 0.65 
 8.13   
  7.18     
Corn stover 7.08 6.99 0.10 
 6.89   
  7.01     
Eucalyptus 7.18 6.80 0.33 
 6.63   
  6.57     
Lignite coal 29.36 29.40 0.04 
 29.44   
  29.41     
Pecan shell 12.70 12.71 0.27 
 12.98   
  12.45     
Pine 7.17 7.20 0.10 
 7.11   
  7.30     
Poultry litter 11.91 11.89 0.07 
 11.81   
  11.94     
PRB coal 13.79 13.79 0.02 
 13.77   
  13.81     
Sugarcane bagasse 6.11 6.06 0.05 
 6.06   
  6.01     
sweetgum 8.05 8.23 0.21 
 8.17   
  8.46     
Switchgrass 8.00 8.27 0.63 
 8.98   
  7.82     
  
 
135 
 
Table A.3 Bulk densities of dust samples. 
Sample Bulk Density (kg/m3) Mean Bulk Density (kg/m3) Standard Deviation (kg/m3) 
Bermuda grass 157.65 159.28 1.43 
 159.06   
  161.13     
Bituminous coal 648.09 651.75 3.88 
 657.12   
  650.04     
Corn cobs 164.64 164.86 2.00 
 167.41   
  162.53     
Corn stover 124.36 126.52 1.76 
 128.67   
  126.52     
Eucalyptus 215.08 217.60 2.08 
 220.18   
  217.55     
Lignite coal 499.02 503.40 3.34 
 507.12   
  504.05     
Pecan shell 401.50 403.95 2.61 
 402.80   
  407.56     
Pine 173.61 173.06 3.43 
 176.96   
  168.60     
Poultry litter 356.98 360.74 3.71 
 359.45   
  365.79     
PRB coal 655.32 655.66 3.19 
 659.72   
  651.93     
Sugarcane 
bagasse 111.65 112.34 1.74 
 114.73   
  110.64     
sweetgum 182.19 183.57 1.37 
 185.44   
  183.09     
Switchgrass 140.92 141.01 0.63 
 141.82   
  140.28     
 
 
  
 
136 
 
Table A.4 Particle densities for dust samples. 
Sample Particle Density (kg/m3) Mean Particle Density (kg/m3) Standard Deviation (kg/m3) 
Bermuda grass 1169.50 1167.33 1.65 
 1165.50   
  1167.00     
Bituminous coal 1392.50 1406.20 13.98 
 1425.40   
  1400.70     
Corn cobs 1481.20 1481.43 0.17 
 1481.60   
  1481.50     
Corn stover 1506.80 1500.90 4.56 
 1495.70   
  1500.20     
Eucalyptus 1493.20 1490.30 2.19 
 1487.90   
  1489.80     
Lignite coal 1674.40 1672.20 2.38 
 1673.30   
  1668.90     
Pecan shell 1520.70 1521.93 0.95 
 1523.00   
  1522.10     
Pine 1477.10 1471.87 3.80 
 1468.20   
  1470.30     
Poultry litter 1537.20 1539.23 1.51 
 1540.80   
  1539.70     
PRB coal 1505.70 1505.13 0.66 
 1504.20   
  1505.50     
Sugarcane 
bagasse 1582.00 1585.27 2.42 
 1587.80   
  1586.00     
sweetgum 1484.50 1480.47 2.85 
 1478.30   
  1478.60     
Switchgrass 1373.10 1366.17 5.59 
 1359.40   
  1366.00     
 
 
  
 
137 
 
Table A.5 Ash content values for dust samples. 
Sample 
Ash Content  
(% d.b.) 
Mean Ash Content (% 
d.b.) 
Standard deviation (% 
d.b.) 
Bermuda grass 4.80 4.68 0.16 
 4.46   
  4.79     
Bituminous coal 9.16 10.27 1.81 
 12.83   
  8.83     
Corn cobs 2.67 2.62 0.08 
 2.51   
  2.68     
Corn stover 13.97 14.33 0.26 
 14.45   
  14.58     
Eucalyptus 1.25 1.40 0.12 
 1.55   
  1.40     
Lignite coal 20.70 20.84 0.16 
 20.75   
  21.05     
Pecan shell 2.66 2.68 0.04 
 2.74   
  2.65     
Pine 1.11 1.10 0.14 
 1.27   
  0.93     
Poultry litter 13.71 13.74 0.07 
 13.67   
  13.84     
PRB coal 8.22 8.14 0.07 
 8.05   
  8.16     
Sugarcane 
bagasse 15.66 15.52 0.27 
 15.77   
  15.15     
sweetgum 1.62 1.41 0.22 
 1.11   
  1.49     
Switchgrass 8.06 7.57 0.35 
 7.26   
  7.40     
 
 
  
 
138 
 
Table A.6 Volatile content values for dust samples. 
Sample Volatile Matter  (% d.b.) Mean Volatile Matter  (% d.b.) Standard deviation  (% d.b.) 
Bermuda grass 78.70 78.61 0.85 
 79.61   
  77.52     
Bituminous coal 33.07 33.00 0.05 
 32.94   
  32.99     
Corn cobs 80.52 81.51 1.51 
 83.65   
  80.36     
Corn stover 73.69 72.97 0.58 
 72.95   
  72.26     
Eucalyptus 81.13 81.27 0.14 
 81.46   
  81.21     
Lignite coal 55.35 53.69 7.14 
 61.49   
  44.23     
Pecan shell 65.20 64.93 0.19 
 64.76   
  64.83     
Pine 84.21 84.97 0.82 
 84.59   
  86.10     
Poultry litter 71.24 71.43 0.21 
 71.31   
  71.73     
PRB coal 47.54 47.06 0.53 
 46.32   
  47.31     
Sugarcane bagasse 72.88 72.41 1.18 
 73.56   
  70.79     
sweetgum 87.50 86.82 0.55 
 86.81   
  86.15     
Switchgrass 78.86 77.86 0.73 
 77.54   
  77.17     
 
 
  
 
139 
 
Table A.7 Energy content values for dust samples. 
Sample Energy Content (MJ/kg) Mean Energy Content (MJ/kg) Standard deviation (MJ/kg) 
Bermuda grass 19.07 19.14 0.06 
 19.14   
  19.21     
Bituminous coal 32.26 32.26 0.02 
 32.23   
  32.29     
Corn cobs 19.05 19.08 0.05 
 19.04   
  19.15     
Corn stover 17.10 17.17 0.07 
 17.15   
  17.26     
Eucalyptus 19.57 19.51 0.06 
 19.43   
  19.53     
Lignite coal 25.99 26.78 0.57 
 27.32   
  27.02     
Pecan shell 20.07 20.31 0.17 
 20.40   
  20.47     
Pine 20.63 20.55 0.07 
 20.56   
  20.46     
Poultry litter 17.51 17.43 0.06 
 17.36   
  17.43     
PRB coal 27.34 27.41 0.16 
 27.63   
  27.25     
Sugarcane bagasse 16.65 16.69 0.07 
 16.79   
  16.64     
sweetgum 19.69 19.71 0.07 
 19.63   
  19.80     
Switchgrass 18.60 18.77 0.12 
 18.89   
  18.81     
 
 
 
  
 
140 
 
Table A.8 Moisture content values for ground samples. 
Sample Moisture content (% w.b.) Mean moisture content (% w.b.) Standard deviation (% w.b.) 
Bermuda grass 8.05 8.07 0.02 
 8.09   
  8.06     
Bituminous coal 2.90 2.90 0.02 
 2.91   
  2.88     
Corn cobs 8.86 8.85 0.02 
 8.82   
  8.86     
Corn stover 9.14 9.13 0.02 
 9.14   
  9.11     
Eucalyptus 6.44 6.45 0.01 
 6.44   
  6.46     
Pine 8.36 8.37 0.08 
 8.45   
  8.30     
Poultry litter 13.71 13.45 0.35 
 13.05   
  13.60     
PRB coal 15.05 15.01 0.04 
 15.00   
  14.98     
Sugarcane 
bagasse 5.94 6.10 0.15 
 6.13   
  6.23     
Sweetgum 8.37 8.58 0.21 
 8.78   
  8.60     
Switchgrass 8.00 8.01 0.02 
 8.01   
  8.03     
 
  
 
141 
 
Table A.9 Bulk density values for ground samples. 
Sample Bulk density (kg/m3) Mean bulk density (kg/m3) Stdev (kg/m3) 
Bermuda grass 128.66 127.01 1.65 
  125.36     
Bituminous coal 650.07 651.74 1.67 
  653.40     
Corn cobs 161.30 160.24 1.06 
  159.18     
Corn stover 92.26 93.61 1.35 
  94.97     
Eucalyptus 195.43 196.34 0.91 
  197.25     
Pine 183.07 184.44 1.37 
  185.80     
Poultry litter 268.40 271.09 2.70 
  273.79     
PRB coal 617.01 615.41 1.60 
  613.81     
Sugarcane 
bagasse 93.64 92.20 1.44 
  90.76     
Sweetgum 167.12 166.42 0.71 
  165.71     
Switchgrass 143.54 142.13 1.41 
  140.72     
 
 
 
 
 
 
  
 
142 
 
 
Table A.10 Particle density values for ground samples. 
Sample Particle density (kg/m3) Mean particle density (kg/m3) Stdev (kg/m3) 
Bermuda grass 1103.10 1106.60 3.10 
 1109.00   
  1107.70     
Bituminous coal 1384.10 1385.03 1.88 
 1387.20   
  1383.80     
Corn cobs 1357.00 1359.03 1.77 
 1360.20   
  1359.90     
Corn stover 1264.40 1267.00 2.51 
 1267.20   
  1269.40     
Eucalyptus 1426.30 1427.47 2.84 
 1430.70   
  1425.40     
Pine 1465.90 1465.47 2.08 
 1463.20   
  1467.30     
Poultry litter 1503.00 1501.27 1.86 
 1499.30   
  1501.50     
PRB coal 1493.50 1490.57 3.00 
 1487.50   
  1490.70     
Sugarcane 
bagasse 1539.30 1535.27 5.12 
 1529.50   
  1537.00     
Sweetgum 1462.20 1461.40 0.75 
 1460.70   
  1461.30     
Switchgrass 1322.20 1319.43 3.57 
 1315.40   
  1320.70     
 
  
 
143 
 
Table A.11 Ash content values for ground samples. 
Sample Ash content (% d.b.) Mean Ash content (% d.b.) Standard deviation (% d.b.) 
Bermuda grass 4.16 4.02 0.15 
 3.85   
  4.05     
Bituminous coal 10.42 10.47 0.95 
 9.56   
  11.45     
Corn cobs 1.08 1.06 0.01 
 1.05   
  1.05     
Corn stover 8.95 7.58 1.59 
 7.95   
  5.84     
Eucalyptus 1.00 0.66 0.52 
 0.93   
  0.06     
Pine 0.43 0.48 0.04 
 0.52   
  0.48     
Poultry litter 9.41 9.43 0.15 
 9.28   
  9.58     
PRB coal 7.57 7.82 0.24 
 7.85   
  8.04     
Sugarcane 
bagasse 5.49 5.84 0.52 
 6.43   
  5.58     
Sweetgum 0.73 0.66 0.06 
 0.60   
  0.66     
Switchgrass 5.30 4.94 0.31 
 4.73   
  4.79     
 
 
  
 
144 
 
Table A.12 Volatile matter values for ground samples. 
Sample Volatile matter (% d.b.) Mean volatile matter (% d.b.) Standard deviation (% d.b.) 
Bermuda grass 84.42 84.51 0.48 
 85.03   
  84.08     
Bituminous coal 39.12 38.87 0.24 
 38.64   
  38.84     
Corn cobs 88.21 88.57 0.85 
 89.54   
  87.97     
Corn stover 78.99 81.24 2.02 
 81.86   
  82.87     
Eucalyptus 87.47 88.36 1.36 
 87.69   
  89.92     
Pine 85.84 85.44 1.02 
 86.20   
  84.28     
Poultry litter 85.63 84.83 0.76 
 84.12   
  84.76     
PRB coal 53.58 53.71 0.54 
 53.24   
  54.30     
Sugarcane 
bagasse 83.85 83.81 0.22 
 83.57   
  84.01     
Sweetgum 91.86 89.86 1.75 
 89.10   
  88.61     
Switchgrass 84.52 84.16 0.74 
 84.66   
  83.31     
 
 
  
 
145 
 
Table A.13 Energy content values for ground samples. 
Sample Energy content (MJ/kg) Mean energy content (MJ/kg) Standard deviation (MJ/kg) 
Bermuda grass 19.42 19.42 0.06 
 19.35   
  19.48     
Bituminous coal 31.93 31.75 0.58 
 31.10   
  32.21     
Corn cobs 19.22 19.25 0.05 
 19.22   
  19.31     
Corn stover 18.16 18.16 0.01 
 18.17   
  18.14     
Eucalyptus 19.87 19.80 0.06 
 19.74   
  19.78     
Pine 20.63 20.68 0.04 
 20.69   
  20.71     
Poultry litter 18.60 18.35 0.33 
 17.97   
  18.47     
PRB coal 27.08 27.11 0.07 
 27.05   
  27.19     
Sugarcane 
bagasse 19.05 19.01 0.08 
 18.92   
  19.06     
Sweetgum 20.15 19.88 0.24 
 19.69   
  19.81     
Switchgrass 19.10 19.13 0.14 
 19.29   
  19.01     
  
 
146 
 
Table A.14 Moisture content values for feedstock, ground material and dust 
samples (biomass and coal) 
Moisture Content (% w.b.) 
Sample Feedstock Ground material Dust 
Biomass Samples 
Bermuda grass 10.91?0.32c 8.07?0.02e 8.06?0.27b,c 
Corn cobs 9.45?0.19d 8.85?0.02b,c 7.40?0.65b,c,d 
Corn stover 9.59?0.55d 9.13?0.02b 6.99?0.10d,e 
Eucalyptus 6.96?0.13e 6.45?0.01f 6.80?0.33d,e 
Pecan shell 12.71?0.27b ___ 12.71?0.27a 
Pine 11.74?0.23b,c 8.37?0.08d,e 7.20?0.10c,d 
Poultry litter 14.81?0.02a 13.45?0.35a 11.89?0.07a 
Sugarcane 
bagasse 5.68?0.05
f 6.10?0.15f 6.06?0.05e 
Sweetgum 5.45?0.08f 8.58?0.21c,d 8.23?0.21b 
Switchgrass 12.42?0.79b 8.01?0.02e 8.27?0.63b 
Coal Samples 
Bituminous coal 2.42?0.09c 2.90?0.02b 2.96?0.02c 
Lignite coal 29.40?0.04a ___ 29.40?0.04a 
PRB coal 15.25?0.02b 15.01?0.04a 13.79?0.02b 
*Superscripts with same letters in a column are not significantly different from each 
other (?=0.05). 
*Pecan shell and lignite coal samples were obtained in dust form. 
*Sweetgum and sugarcane bagasse samples were dried prior to measuring initial 
moisture content of the feedstock. 
 
  
 
147 
 
Appendix B ? Hot Plate Ignition Test, Thermogravimetric Analysis (TGA) 
and Differential Scanning Calorimetry (DSC) Results. 
Table B.1 Temperature of rapid volatilization (TORV) for dust samples. 
Sample TORV (?C) Mean TORV (?C) Standard deviation (?C) 
Bermuda grass 267.06 266.12 0.97 
 266.51   
  264.78     
Bituminous coal 444.53 447.60 2.18 
 448.91   
  449.35     
Corn cobs 276.38 276.52 0.62 
 277.34   
  275.83     
Corn stover 279.03 277.47 1.14 
 276.36   
  277.01     
Eucalyptus 290.66 291.88 0.88 
 292.25   
  292.72     
Lignite coal 314.17 311.64 4.52 
 305.29   
  315.46     
Pecan shell 287.85 286.96 1.73 
 284.54   
  288.49     
Pine 308.07 306.08 1.78 
 306.43   
  303.74     
Poultry litter 272.88 272.79 0.52 
 272.11   
  273.37     
PRB coal 315.18 317.34 1.57 
 318.86   
  317.98     
Sugarcane bagasse 285.48 285.17 0.40 
 284.61   
  285.43     
sweetgum 289.77 290.23 0.75 
 291.28   
  289.63     
Switchgrass 285.82 285.05 1.82 
 282.54   
  286.79     
  
 
148 
 
Table B.2 Temperature of maximum rate of mass loss (TMML) values for dust 
samples. 
Sample TMML (?C) Mean TMML (?C) Standard deviation  (?C) 
Bermuda grass 310.34 312.42 1.61 
 312.67   
  314.25     
Bituminous coal 482.56 485.12 4.36 
 481.54   
  491.26     
Corn cobs 312.12 313.67 1.60 
 315.88   
  313.02     
Corn stover 316.63 316.41 0.83 
 317.29   
  315.30     
Eucalyptus 313.25 315.69 1.76 
 317.33   
  316.48     
Lignite coal 353.71 349.77 6.17 
 341.06   
  354.55     
Pecan shell 327.57 331.21 4.74 
 337.90   
  328.15     
Pine 348.43 348.75 2.09 
 351.45   
  346.36     
Poultry litter 303.55 304.07 0.49 
 304.73   
  303.94     
PRB coal 380.91 379.39 1.62 
 377.15   
  380.10     
Sugarcane bagasse 343.33 345.21 1.33 
 346.00   
  346.30     
sweetgum 336.89 338.80 1.49 
 339.00   
  340.52     
Switchgrass 330.10 329.76 0.49 
 329.07   
  330.12     
 
 
 
 
  
 
149 
 
Table B.3 Oxidation temperature (TOXY) values for dust samples. 
Sample TOXY (?C) Mean TOXY (?C) Standard deviation  (?C) 
Bermuda grass 269.61 274.02 3.80 
 278.88   
  273.58     
Bituminous coal 399.62 423.10 16.72 
 437.31   
  432.37     
Corn cobs 280.77 276.51 4.02 
 271.12   
  277.63     
Corn stover 288.13 286.12 1.74 
 283.88   
  286.36     
Eucalyptus 281.41 285.87 4.23 
 284.64   
  291.56     
Lignite coal 246.22 241.80 3.33 
 238.18   
  241.00     
Pecan shell 267.92 283.52 11.42 
 287.69   
  294.94     
Pine 315.87 319.08 2.79 
 318.70   
  322.68     
Poultry litter 278.36 277.77 0.42 
 277.53   
  277.43     
PRB coal 280.44 258.68 18.42 
 260.19   
  235.40     
Sugarcane bagasse 305.74 309.57 6.34 
 318.51   
  304.46     
sweetgum 300.11 297.22 2.20 
 296.78   
  294.77     
Switchgrass 318.44 315.28 4.61 
 308.77   
  318.64     
 
 
 
 
 
  
 
150 
 
Table B.4 Activation energy values for dust samples. 
Sample Activation Energy (kJ/mol) Mean Activation Energy (kJ/mol) Standard deviation (kJ/mol) 
Bermuda grass 47.77 48.25 0.60 
 49.09   
  47.88     
Bituminous coal 98.90 91.56 6.29 
 92.24   
  83.54     
Corn cobs 64.75 63.15 2.45 
 65.02   
  59.69     
Corn stover 70.67 71.26 0.52 
 71.16   
  71.95     
Eucalyptus 74.45 72.35 1.55 
 71.84   
  70.75     
Lignite coal 57.24 56.82 0.48 
 56.16   
  57.07     
Pecan shell 55.86 55.55 0.31 
 55.67   
  55.13     
Pine 64.28 64.43 0.11 
 64.53   
  64.49     
Poultry litter 69.50 71.06 1.11 
 72.02   
  71.65     
PRB coal 48.87 49.77 0.79 
 49.66   
  50.80     
Sugarcane bagasse 77.51 77.08 0.54 
 76.32   
  77.42     
sweetgum 97.81 95.16 1.88 
 94.02   
  93.65     
Switchgrass 60.69 61.64 1.27 
 60.79   
  63.42     
  
 
151 
 
Table B.5 Temperature of rapid exothermic reaction (TRE) values for dust 
samples. 
Sample TRE (?C) Mean TRE (?C) Standard deviation (?C) 
Bermuda grass 240.73 237.53 3.47 
 239.16   
  232.71     
Bituminous coal 222.44 223.43 1.50 
 225.56   
  222.30     
Corn cobs 235.87 235.16 1.04 
 233.68   
  235.92     
Corn stover 239.79 239.67 0.43 
 239.09   
  240.13     
Eucalyptus 244.54 243.54 1.00 
 242.17   
  243.90     
Lignite coal 221.74 221.20 2.42 
 223.86   
  218.00     
Pecan shell 205.37 206.05 2.52 
 203.36   
  209.43     
Pine 243.06 244.53 1.44 
 246.48   
  244.06     
Poultry litter 246.71 241.21 4.56 
 241.38   
  235.55     
PRB coal 245.03 244.90 2.41 
 241.89   
  247.78     
Sugarcane bagasse 240.04 245.29 4.73 
 244.32   
  251.51     
sweetgum 249.88 248.86 3.10 
 252.05   
  244.66     
Switchgrass 248.00 248.96 0.77 
 249.00   
  249.89     
 
 
 
  
 
152 
 
Table B.6 Maximum temperature reached during exothermic reaction (TME) values 
for dust samples. 
Sample TME (?C) Mean TME (?C) Standard deviation (?C) 
Bermuda grass 374.61 379.27 11.73 
 367.81   
  395.39     
Bituminous coal 392.59 395.60 2.17 
 396.64   
  397.58     
Corn cobs 395.50 394.56 4.88 
 400.00   
  388.17     
Corn stover 400.17 395.83 3.21 
 392.50   
  394.83     
Eucalyptus 386.66 385.37 0.99 
 385.19   
  384.25     
Lignite coal 428.38 428.00 1.29 
 429.36   
  426.27     
Pecan shell 409.67 407.65 2.59 
 404.00   
  409.28     
Pine 403.06 401.12 3.29 
 396.49   
  403.82     
Poultry litter 352.44 354.38 1.87 
 353.81   
  356.90     
PRB coal 430.92 429.11 3.04 
 431.57   
  424.83     
Sugarcane bagasse 374.40 377.25 4.48 
 373.78   
  383.58     
sweetgum 392.30 392.82 0.83 
 394.00   
  392.17     
Switchgrass 390.77 388.98 1.55 
 389.18   
  386.98     
  
 
153 
 
Table B.7 Exothermic energy values for dust samples. 
Sample Exothermic energy (MJ/kg) Mean Exothermic energy (MJ/kg) Stdev (MJ/kg) 
Bermuda grass 7.96 7.95 0.01 
 7.94   
  7.94     
Bituminous coal 3.63 3.51 0.10 
 3.39   
  3.51     
Corn cobs 8.07 8.15 0.11 
 8.30   
  8.08     
Corn stover 7.83 7.78 0.24 
 7.46   
  8.04     
Eucalyptus 7.21 7.52 0.22 
 7.69   
  7.67     
Lignite coal 4.09 5.16 0.87 
 6.22   
  5.17     
Pecan shell 5.47 5.61 0.15 
 5.81   
  5.55     
Pine 6.75 6.14 0.46 
 5.64   
  6.02     
Poultry litter 4.48 4.48 0.26 
 4.80   
  4.17     
PRB coal 4.48 4.64 0.86 
 3.68   
  5.77     
Sugarcane 
bagasse 8.35 7.55 0.73 
 6.59   
  7.71     
sweetgum 6.63 6.57 0.05 
 6.53   
  6.54     
Switchgrass 7.46 7.69 0.16 
 7.85   
  7.75     
 
 
  
 
154 
 
Table B.8 Minimum hot surface ignition temperature (MIT) for dust samples. 
Sample MIT (?C) Mean MIT (?C) Standard deviation (?C) 
Bermuda grass 275.00 275.00 0.00 
 275.00   
  275.00     
Bituminous coal 335.00 335.00 0.00 
 335.00   
  335.00     
Corn cobs 280.00 280.00 0.00 
 280.00   
  280.00     
Corn stover 290.00 290.00 0.00 
 290.00   
  290.00     
Eucalyptus 285.00 285.00 0.00 
 285.00   
  285.00     
Lignite coal 240.00 240.00 0.00 
 240.00   
  240.00     
Pecan shell 265.00 265.00 0.00 
 265.00   
  265.00     
Pine 315.00 315.00 0.00 
 315.00   
  315.00     
Poultry litter 285.00 285.00 0.00 
 285.00   
  285.00     
PRB coal 240.00 240.00 0.00 
 240.00   
  240.00     
Sugarcane 
bagasse 305.00 305.00 0.00 
 305.00   
  305.00     
sweetgum 305.00 305.00 0.00 
 305.00   
  305.00     
Switchgrass 295.00 295.00 0.00 
 295.00   
  295.00     
  
 
155 
 
Table B.9 Particle density of ash derived from sugarcane bagasse dust. 
 
Sample Particle Density (kg/m3) Mean particle density (kg/m3) Stdev (kg/m3) 
Sugarcane 
bagasse ash 
2785.40 
2781.50 3.46 2778.80 
2780.30 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
156 
 
Appendix C ? SAS Codes for Tukey tests, correlation matrices and 
ANOVA results 
 (first objective)
 
SAS code for Tukey test on physical, chemical, heating and ignition properties. 
/******************************** 
 * Author: Jaskaran Dhiman      * 
 * Research work                * 
 * Date: 01/30/2014             * 
 ********************************/ 
 
/*ANOVA analysis*/ 
 
%MACRO anova; 
ods listing close; 
proc import datafile=&inputfile 
out=&dataset 
replace; 
range="&range"; 
run; 
proc sort data=&dataset 
out=&dataset; 
by Biomass; 
run; 
ods html file=&outfile; 
title &titleanova; 
proc anova data=&dataset; 
class Biomass; 
model &prop=Biomass; 
means Biomass/tukey; 
run; 
title; 
title &title2; 
proc univariate data=&dataset; 
var &prop; 
run; 
title; 
ods html close; 
ods listing; 
%MEND anova; 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Moisture 
Content(MC)'; 
%LET titleanova='MC'; 
%LET inputfile='I:\Desktop\MCbiomass.xls'; 
  
 
157 
 
%LET dataset=work.MCanova; 
%LET range=A1:B31; 
%LET prop=MC; 
%LET outfile='I:\Desktop\MCbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Ash Content(AC)'; 
%LET titleanova='AC'; 
%LET inputfile='I:\Desktop\ACbiomass.xls'; 
%LET dataset=work.ACanova; 
%LET range=A1:B31; 
%LET prop=AC; 
%LET outfile='I:\Desktop\ACbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Bulk Density(BD)'; 
%LET titleanova='BD'; 
%LET inputfile='I:\Desktop\BDbiomass.xls'; 
%LET dataset=work.BDanova; 
%LET range=A1:B31; 
%LET prop=BD; 
%LET outfile='I:\Desktop\BDbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Temperature of 
Maximum Mass Loss(TMML)'; 
%LET titleanova='TMML'; 
%LET inputfile='I:\Desktop\TMMLbiomass.xls'; 
%LET dataset=work.TMMLanova; 
%LET range=A1:B31; 
%LET prop=TMML; 
%LET outfile='I:\Desktop\TMMLbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Temperature of Rapid 
Exothermic Reaction(TRE)'; 
%LET titleanova='TRE'; 
%LET inputfile='I:\Desktop\TREbiomass.xls'; 
%LET dataset=work.TREanova; 
%LET range=A1:B31; 
%LET prop=TRE; 
%LET outfile='I:\Desktop\TREbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Volatile Matter 
(VM)'; 
%LET titleanova='VM'; 
%LET inputfile='I:\Desktop\VMbiomass.xls'; 
%LET dataset=work.VManova; 
%LET range=A1:B31; 
%LET prop=VM; 
%LET outfile='I:\Desktop\VMbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Oxidation 
Temperature(TOXY)'; 
%LET titleanova='TOXY'; 
  
 
158 
 
%LET inputfile='I:\Desktop\TOXYbiomass.xls'; 
%LET dataset=work.TOXYanova; 
%LET range=A1:B31; 
%LET prop=TOXY; 
%LET outfile='I:\Desktop\TOXYbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Temperature of Onset 
of Rapid Volatalization (TORV)'; 
%LET titleanova='TORV'; 
%LET inputfile='I:\Desktop\TORVbiomass.xls'; 
%LET dataset=work.TORVanova; 
%LET range=A1:B40; 
%LET prop=TORV; 
%LET outfile='I:\Desktop\TORVbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Maximum Temperature 
reached during exothermic reaction (TME)'; 
%LET titleanova='TME'; 
%LET inputfile='I:\Desktop\TMEbiomass.xls'; 
%LET dataset=work.TMEanova; 
%LET range=A1:B40; 
%LET prop=TME; 
%LET outfile='I:\Desktop\TMEbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE ? Exothermic Energy 
(Q)'; 
%LET titleanova='Q'; 
%LET inputfile='I:\Desktop\Qbiomass.xls'; 
%LET dataset=work.Qanova; 
%LET range=A1:B31; 
%LET prop=Q; 
%LET outfile='I:\Desktop\Qbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Particle Density 
(PD)'; 
%LET titleanova='PD'; 
%LET inputfile='I:\Desktop\PDbiomass.xls'; 
%LET dataset=work.PDanova; 
%LET range=A1:B31; 
%LET prop=PD; 
%LET outfile='I:\Desktop\PDbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Energy Content 
(EC)'; 
%LET titleanova='EC'; 
%LET inputfile='I:\Desktop\ECbiomass.xls'; 
%LET dataset=work.ECanova; 
%LET range=A1:B31; 
%LET prop=EC; 
%LET outfile='I:\Desktop\ECbiomass.htm'; 
%anova 
 
  
 
159 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Geometric Mean 
Diameter (dgw)'; 
%LET titleanova='dgw'; 
%LET inputfile='I:\Desktop\dgwbiomass.xls'; 
%LET dataset=work.dgwanova; 
%LET range=A1:B31; 
%LET prop=dgw; 
%LET outfile='I:\Desktop\dgwbiomass.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Activation Energy 
(AE)'; 
%LET titleanova='AE'; 
%LET inputfile='I:\Desktop\AEbiomass.xls'; 
%LET dataset=work.AEanova; 
%LET range=A1:B31; 
%LET prop=AE; 
%LET outfile='I:\Desktop\AEbiomass.htm'; 
%anova 
%LET title2='Descriptive Statistics: UNIVARIATE ? Minimum Ignition 
Temperature of dust layer (MIT)'; 
%LET titleanova='MIT'; 
%LET inputfile='I:\Desktop\MITbiomass.xls'; 
%LET dataset=work.MITanova; 
%LET range=A1:B31; 
%LET prop=MIT; 
%LET outfile='I:\Desktop\MITbiomass.htm'; 
%anova 
%LET title2='Descriptive Statistics: UNIVARIATE - Moisture 
Content(MC)'; 
%LET titleanova='MC'; 
%LET inputfile='I:\Desktop\MCcoal.xls'; 
%LET dataset=work.MCanova; 
%LET range=A1:B10; 
%LET prop=MC; 
%LET outfile='I:\Desktop\MCcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Ash Content(AC)'; 
%LET titleanova='AC'; 
%LET inputfile='I:\Desktop\ACcoal.xls'; 
%LET dataset=work.ACanova; 
%LET range=A1:B10; 
%LET prop=AC; 
%LET outfile='I:\Desktop\ACcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Bulk Density(BD)'; 
%LET titleanova='BD'; 
%LET inputfile='I:\Desktop\BDcoal.xls'; 
%LET dataset=work.BDanova; 
%LET range=A1:B10; 
%LET prop=BD; 
%LET outfile='I:\Desktop\BDcoal.htm'; 
%anova 
  
 
160 
 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Temperature of 
Maximum Mass Loss(TMML)'; 
%LET titleanova='TMML'; 
%LET inputfile='I:\Desktop\TMMLcoal.xls'; 
%LET dataset=work.TMMLanova; 
%LET range=A1:B10; 
%LET prop=TMML; 
%LET outfile='I:\Desktop\TMMLcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Temperature of Rapid 
Exothermic Reaction(TRE)'; 
%LET titleanova='TRE'; 
%LET inputfile='I:\Desktop\TREcoal.xls'; 
%LET dataset=work.TREanova; 
%LET range=A1:B10; 
%LET prop=TRE; 
%LET outfile='I:\Desktop\TREcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Volatile Matter 
(VM)'; 
%LET titleanova='VM'; 
%LET inputfile='I:\Desktop\VMcoal.xls'; 
%LET dataset=work.VManova; 
%LET range=A1:B10; 
%LET prop=VM; 
%LET outfile='I:\Desktop\VMcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Oxidation 
Temperature(TOXY)'; 
%LET titleanova='TOXY'; 
%LET inputfile='I:\Desktop\TOXYcoal.xls'; 
%LET dataset=work.TOXYanova; 
%LET range=A1:B10; 
%LET prop=TOXY; 
%LET outfile='I:\Desktop\TOXYcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Temperature of Onset 
of Rapid Volatalization (TORV)'; 
%LET titleanova='TORV'; 
%LET inputfile='I:\Desktop\TORVcoal.xls'; 
%LET dataset=work.TORVanova; 
%LET range=A1:B40; 
%LET prop=TORV; 
%LET outfile='I:\Desktop\TORVcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Maximum Temperature 
reached during exothermic reaction (TME)'; 
%LET titleanova='TME'; 
%LET inputfile='I:\Desktop\TMEcoal.xls'; 
%LET dataset=work.TMEanova; 
%LET range=A1:B40; 
  
 
161 
 
%LET prop=TME; 
%LET outfile='I:\Desktop\TMEcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE ? Exothermic Energy 
(Q)'; 
%LET titleanova='Q'; 
%LET inputfile='I:\Desktop\Qcoal.xls'; 
%LET dataset=work.Qanova; 
%LET range=A1:B10; 
%LET prop=Q; 
%LET outfile='I:\Desktop\Qcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Particle Density 
(PD)'; 
%LET titleanova='PD'; 
%LET inputfile='I:\Desktop\PDcoal.xls'; 
%LET dataset=work.PDanova; 
%LET range=A1:B10; 
%LET prop=PD; 
%LET outfile='I:\Desktop\PDcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Energy Content 
(EC)'; 
%LET titleanova='EC'; 
%LET inputfile='I:\Desktop\ECcoal.xls'; 
%LET dataset=work.ECanova; 
%LET range=A1:B10; 
%LET prop=EC; 
%LET outfile='I:\Desktop\ECcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Geometric Mean 
Diameter (dgw)'; 
%LET titleanova='dgw'; 
%LET inputfile='I:\Desktop\dgwcoal.xls'; 
%LET dataset=work.dgwanova; 
%LET range=A1:B10; 
%LET prop=dgw; 
%LET outfile='I:\Desktop\dgwcoal.htm'; 
%anova 
 
%LET title2='Descriptive Statistics: UNIVARIATE - Activation Energy 
(AE)'; 
%LET titleanova='AE'; 
%LET inputfile='I:\Desktop\AEcoal.xls'; 
%LET dataset=work.AEanova; 
%LET range=A1:B10; 
%LET prop=AE; 
%LET outfile='I:\Desktop\AEcoal.htm'; 
%anova 
%LET title2='Descriptive Statistics: UNIVARIATE ? Minimum Ignition 
Temperature of dust layer (MIT)'; 
%LET titleanova='MIT'; 
  
 
162 
 
%LET inputfile='I:\Desktop\MITcoal.xls'; 
%LET dataset=work.MITanova; 
%LET range=A1:B10; 
%LET prop=MIT; 
%LET outfile='I:\Desktop\MITcoal.htm'; 
%anova 
 
proc import datafile='C:\Users\jzd0028\Desktop\Combined_woody.xls' 
out=work.corr1 
replace; 
range="A1:P10"; 
run; 
ods html file='C:\Users\jzd0028\Desktop\woody_ corr.htm'; 
proc corr data=worK.corr1; 
var AC AE BD dgw EC MC PD Q TME TMML TORV TOXY TRE VM MIT; 
RUN; 
proc import datafile='C:\Users\jzd0028\Desktop\Combined_grassy.xls' 
out=work.corr2 
replace; 
range="A1:P13"; 
run; 
ods html file='C:\Users\jzd0028\Desktop\grassy_corr.htm'; 
proc corr data=worK.corr2; 
var AC AE BD dgw EC MC PD Q TME TMML TORV TOXY TRE VM MIT; 
RUN; 
proc import datafile='C:\Users\jzd0028\Desktop\Combined_biomass.xls' 
out=work.corr3 
replace; 
range="A1:P31"; 
run; 
ods html file='C:\Users\jzd0028\Desktop\biomass_corr.htm'; 
proc corr data=worK.corr3; 
var AC AE BD dgw EC MC PD Q TME TMML TORV TOXY TRE VM MIT; 
RUN; 
proc import datafile='C:\Users\jzd0028\Desktop\Combined_all.xls' 
out=work.corr4 
replace; 
range="A1:P40"; 
run; 
ods html file='C:\Users\jzd0028\Desktop\all_corr.htm'; 
proc corr data=worK.corr4; 
var AC AE BD dgw EC MC PD Q TME TMML TORV TOXY TRE VM MIT; 
RUN; 
 
 
 
 
  
 
163 
 
 
Table C.1 ANOVA results for Tukey test on moisture content (MC) for biomass 
dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 130.0453486 14.4494832 126.77 <.0001 
Error 20 2.2795512 0.1139776   
Corrected Total 29 132.3248998    
 
R-Square Coeff Var Root MSE MC Mean 
0.982773 4.038353 0.337606 8.359982 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 9 130.0453486 14.4494832 126.77 <.0001 
 
 
Table C.2 ANOVA results for Tukey test on ash content (AC) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 931.0711157 103.4523462 1804.35 <.0001 
Error 20 1.1467003 0.0573350   
Corrected Total 29 932.2178159    
 
R-Square Coeff Var Root MSE AC Mean 
0.998770 3.680139 0.239447 6.506475 
 
Source DF Anova SS Mean Square F Value  Pr > F 
Biomass 9 931.0711157 103.4523462 1804.35  <.0001 
 
 
  
 
164 
 
 
Table C.3 ANOVA results for Tukey test on activation energy (AE) for biomass 
dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 4442.711331 493.634592 208.25 <.0001 
Error 20 47.407565 2.370378   
Corrected Total 29 4490.118896    
 
R-Square Coeff Var Root MSE AE Mean 
0.989442 2.264383 1.539603 67.99217 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 4442.711331 493.634592 208.25 <.0001 
 
 
 
Table C.4 ANOVA results for Tukey test on bulk density (BD) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 264035.6623 29337.2958 3826.42 <.0001 
Error 20 153.3405 7.6670     
Corrected Total 29 264189.0028       
 
R-Square Coeff Var Root MSE BD Mean 
0.999420 1.355370 2.768939 204.2940 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 264035.6623 29337.2958 3826.42 <.0001 
 
 
 
 
  
 
165 
 
 
 
Table C.5 ANOVA results for Tukey test on geometric mean diameter (dgw) for 
biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 0.64206673 0.07134075 Infty <.0001 
Error 20 0.00000000 0.00000000   
Corrected Total 29 0.64206673    
 
R-Square Coeff Var Root MSE dgw Mean 
1.000000 0 0 0.377688 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 0.64206673 0.07134075 Infty <.0001 
 
 
 
Table C.6 ANOVA results for Tukey test on energy content (EC) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 47454058.42 5272673.16 451.31 <.0001 
Error 20 233658.47 11682.92   
Corrected Total 29 47687716.89    
 
R-Square Coeff Var Root MSE EC Mean 
0.995100 0.573826 108.0876 18836.30 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 47454058.42 5272673.16 451.31 <.0001 
 
 
 
  
 
166 
 
 
 
Table C.7 ANOVA results for Tukey test on particle density (PD) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 371614.3737 41290.4860 3016.62 <.0001 
Error 20 273.7533 13.6877     
Corrected Total 29 371888.1270       
 
R-Square Coeff Var Root MSE PD Mean 
0.999264 0.253318 3.699685 1460.490 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 371614.3737 41290.4860 3016.62 <.0001 
 
 
Table C.8 ANOVA results for Tukey test exothermic energy (Q) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 39.06237438 4.34026382 29.64 <.0001 
Error 20 2.92818643 0.14640932     
Corrected Total 29 41.99056081       
 
R-Square Coeff Var Root MSE Q Mean 
0.930266 5.511725 0.382635 6.942195 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 39.06237438 4.34026382 29.64 <.0001 
 
 
 
  
 
167 
 
Table C.9 ANOVA results for Tukey test on maximum temperature reached during 
exothermic reaction (TME) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 6044.605963 671.622885 20.65 <.0001 
Error 20 650.636933 32.531847   
Corrected Total 29 6695.242897    
 
R-Square Coeff Var Root MSE TME Mean 
0.902821 1.471066 5.703670 387.7237 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 6044.605963 671.622885 20.65 <.0001 
 
 
 
Table C.10 ANOVA results for Tukey test on temperature of maximum rate of mass 
loss (TMML) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 6316.689137 701.854349 116.32 <.0001 
Error 20 120.679933 6.033997   
Corrected Total 29 6437.369070    
 
R-Square Coeff Var Root MSE TMML Mean 
0.981253 0.754431 2.456419 325.5990 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 6316.689137 701.854349 116.32 <.0001 
 
 
 
 
  
 
168 
 
Table C.11 ANOVA results for Tukey test on temperature of onset of rapid 
volatilization (TORV) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 3430.537680 381.170853 183.11 <.0001 
Error 20 41.633867 2.081693   
Corrected Total 29 3472.171547    
 
R-Square Coeff Var Root MSE TORV Mean 
0.988009 0.508343 1.442807 283.8253 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 3430.537680 381.170853 183.11 <.0001 
 
 
 
Table C.12 ANOVA results for Tukey test on oxidation temperature (TOXY) for 
biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 7556.354363 839.594929 21.85 <.0001 
Error 20 768.614267 38.430713   
Corrected Total 29 8324.968630    
 
R-Square Coeff Var Root MSE TOXY Mean 
0.907674 2.119424 6.199251 292.4970 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 7556.354363 839.594929 21.85 <.0001 
 
 
 
 
  
 
169 
 
Table C.13 ANOVA results for Tukey test on temperature of rapid exothermic 
reaction (TRE) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 4185.010013 465.001113 40.71 <.0001 
Error 20 228.428133 11.421407   
Corrected Total 29 4413.438147    
 
R-Square Coeff Var Root MSE TRE Mean 
0.948243 1.413560 3.379557 239.0813 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 4185.010013 465.001113 40.71 <.0001 
 
 
 
Table C.14 ANOVA results for Tukey test on volatile matter (VM) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 1245.119098 138.346566 145.28 <.0001 
Error 20 19.045808 0.952290   
Corrected Total 29 1264.164907    
 
R-Square Coeff Var Root MSE VM Mean 
0.984934 1.262799 0.975854 77.27705 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 1245.119098 138.346566 145.28 <.0001 
 
 
 
  
 
170 
 
Table C.15 ANOVA results for Tukey test on minimum hot surface ignition 
temperature (MIT) for biomass dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 9 6300.000000 700.000000 Infty <.0001 
Error 20 0.000000 0.000000   
Corrected Total 29 6300.000000    
 
R-Square Coeff Var Root MSE MIT Mean 
1.000000 0 0 290.0000 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 9 6300.000000 700.000000 Infty <.0001 
 
 
 
Table C.16 ANOVA results for Tukey test on ash content (AC) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 277.3325531 138.6662766 83.86 <.0001 
Error 6 9.9211421 1.6535237   
Corrected Total 8 287.2536952    
 
R-Square Coeff Var Root MSE AC Mean 
0.965462 9.829077 1.285894 13.08255 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 277.3325531 138.6662766 83.86 <.0001 
 
 
 
  
 
171 
 
Table C.17 ANOVA results for Tukey test on activation energy (AE) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 3002.795344 1501.397672 74.27 <.0001 
Error 6 121.293156 20.215526   
Corrected Total 8 3124.088500    
 
R-Square Coeff Var Root MSE AE Mean 
0.961175 6.806950 4.496168 66.05261 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 3002.795344 1501.397672 74.27 <.0001 
 
 
 
Table C.18 ANOVA results for Tukey test on bulk density (BD) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 45207.75978 22603.87989 1243.39 <.0001 
Error 6 109.07517 18.17920     
Corrected Total 8 45316.83495       
 
R-Square Coeff Var Root MSE BD Mean 
0.997593 0.706376 4.263707 603.6030 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 45207.75978 22603.87989 1243.39 <.0001 
 
 
 
 
  
 
172 
 
Table C.19 ANOVA results for Tukey test on geometric mean diameter (dgw) for 
coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 0.37376761 0.18688380 5.05E15 <.0001 
Error 6 0.00000000 0.00000000   
Corrected Total 8 0.37376761    
 
R-Square Coeff Var Root MSE dgw Mean 
1.000000 1.61018E-6 6.08337E-9 0.377808 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 0.37376761 0.18688380 5.05E15 <.0001 
 
 
 
Table C.20 ANOVA results for Tukey test on energy content (EC) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 53987555.20 26993777.60 154.00 <.0001 
Error 6 1051700.64 175283.44   
Corrected Total 8 55039255.84    
 
R-Square Coeff Var Root MSE EC Mean 
0.980892 1.453015 418.6687 28813.79 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 53987555.20 26993777.60 154.00 <.0001 
 
 
 
 
 
  
 
173 
 
Table C.21 ANOVA results for Tukey test on moisture content (MC) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 1060.026849 530.013424 659672 <.0001 
Error 6 0.004821 0.000803   
Corrected Total 8 1060.031669    
 
R-Square Coeff Var Root MSE MC Mean 
0.999995 0.184235 0.028345 15.38536 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 1060.026849 530.013424 659672 <.0001 
 
 
 
Table C.22 ANOVA results for Tukey test on particle density (PD) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 108455.0756 54227.5378 537.93 <.0001 
Error 6 604.8467 100.8078     
Corrected Total 8 109059.9222       
 
R-Square Coeff Var Root MSE PD Mean 
0.994454 0.657155 10.04031 1527.844 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 108455.0756 54227.5378 537.93 <.0001 
 
 
 
 
 
  
 
174 
 
Table C.23 ANOVA results for Tukey test on exothermic energy (Q) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 4.26432203 2.13216101 2.84 0.1355 
Error 6 4.50297828 0.75049638     
Corrected Total 8 8.76730030       
 
R-Square Coeff Var Root MSE Q Mean 
0.486389 19.51531 0.866312 4.439140 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 4.26432203 2.13216101 2.84 0.1355 
 
 
 
Table C.24 ANOVA results for Tukey test on maximum temperature reached during 
an exothermic energy (TME) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 2173.450689 1086.725344 139.64 <.0001 
Error 6 46.695000 7.782500   
Corrected Total 8 2220.145689    
 
R-Square Coeff Var Root MSE TME Mean 
0.978968 0.668081 2.789713 417.5711 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 2173.450689 1086.725344 139.64 <.0001 
 
 
 
 
 
  
 
175 
 
Table C.25 ANOVA results for Tukey test on temperature of maximum rate of mass 
loss(TMML) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 30375.20747 15187.60373 508.69 <.0001 
Error 6 179.13773 29.85629   
Corrected Total 8 30554.34520    
 
R-Square Coeff Var Root MSE TMML Mean 
0.994137 1.349958 5.464091 404.7600 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 30375.20747 15187.60373 508.69 <.0001 
 
 
 
Table C.26 ANOVA results for Tukey test on temperature of onset of rapid 
volatilization(TORV) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 35483.50442 17741.75221 1284.01 <.0001 
Error 6 82.90487 13.81748   
Corrected Total 8 35566.40929    
 
R-Square Coeff Var Root MSE TORV Mean 
0.997669 1.035835 3.717187 358.8589 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 35483.50442 17741.75221 1284.01 <.0001 
 
 
 
 
  
 
176 
 
Table C.27 ANOVA results for Tukey test on oxidation temperature (TOXY) for coal 
dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 60189.54442 30094.77221 95.53 <.0001 
Error 6 1890.18427 315.03071   
Corrected Total 8 62079.72869    
 
R-Square Coeff Var Root MSE TOXY Mean 
0.969552 5.765338 17.74910 307.8589 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 60189.54442 30094.77221 95.53 <.0001 
 
 
 
Table C.28 ANOVA results for Tukey test on temperature of rapid exothermic 
energy  (TRE) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 1027.495556 513.747778 73.79 <.0001 
Error 6 41.772467 6.962078   
Corrected Total 8 1069.268022    
 
R-Square Coeff Var Root MSE TRE Mean 
0.960934 1.147983 2.638575 229.8444 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 1027.495556 513.747778 73.79 <.0001 
 
 
 
 
  
 
177 
 
Table C.29 ANOVA results for Tukey test on volatile matter (VM) for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 669.6700713 334.8350356 13.06 0.0065 
Error 6 153.8719318 25.6453220   
Corrected Total 8 823.5420031    
 
R-Square Coeff Var Root MSE VM Mean 
0.813158 11.35916 5.064121 44.58184 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 669.6700713 334.8350356 13.06 0.0065 
 
 
 
Table C.30 ANOVA results for Tukey test on minimum ignition temperature (MIT) 
for coal dust. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 18050.00000 9025.00000 Infty <.0001 
Error 6 0.00000 0.00000   
Corrected Total 8 18050.00000    
 
R-Square Coeff Var Root MSE MIT Mean 
1.000000 0 0 271.6667 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 18050.00000 9025.00000 Infty <.0001 
 
 
 
  
 
178 
 
 
Table C.31 ANOVA results for Tukey test on moisture content (MC) for ground 
biomass. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 8 106.3450000 13.2931250 609.47 <.0001 
Error 18 0.3926000 0.0218111   
Corrected Total 26 106.7376000    
 
R-Square Coeff Var Root MSE MC Mean 
0.996322 1.725974 0.147686 8.556667 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 8 106.3450000 13.2931250 609.47 <.0001 
 
 
 
Table C.32 ANOVA results for Tukey test on bulk density (BD) for ground biomass. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 2 4.26432203 2.13216101 2.84 0.1355 
Error 6 4.50297828 0.75049638     
Corrected Total 8 8.76730030       
 
R-Square Coeff Var Root MSE Q Mean 
0.486389 19.51531 0.866312 4.439140 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 2 4.26432203 2.13216101 2.84 0.1355 
 
 
 
 
  
 
179 
 
 
Table C.33 ANOVA results for Tukey test on particle density (PD) for ground 
biomass. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 8 439688.9474 54961.1184 6643.16 <.0001 
Error 18 148.9200 8.2733     
Corrected Total 26 439837.8674       
 
R-Square Coeff Var Root MSE PD Mean 
0.999661 0.208046 2.876340 1382.548 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 8 439688.9474 54961.1184 6643.16 <.0001 
 
 
Table C.34 ANOVA results for Tukey test on volatile matter (VM) for ground 
biomass. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 8 182.0291098 22.7536387 16.89 <.0001 
Error 18 24.2525601 1.3473644   
Corrected Total 26 206.2816699    
 
R-Square Coeff Var Root MSE VM Mean 
0.882430 1.355351 1.160760 85.64276 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 8 182.0291098 22.7536387 16.89 <.0001 
 
 
 
 
  
 
180 
 
 
Table C.35 ANOVA results for Tukey test on ash content (AC) for ground biomass 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 8 268.8139148 33.6017394 94.20 <.0001 
Error 18 6.4204212 0.3566901   
Corrected Total 26 275.2343360    
 
R-Square Coeff Var Root MSE AC Mean 
0.976673 15.50393 0.597235 3.852156 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 8 268.8139148 33.6017394 94.20 <.0001 
 
 
Table C.36 ANOVA results for Tukey test on energy content (EC) for ground 
biomass. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 8 14.45738444 1.80717306 78.39 <.0001 
Error 18 0.41498679 0.02305482   
Corrected Total 26 14.87237123    
 
R-Square Coeff Var Root MSE EC Mean 
0.972097 0.786828 0.151838 19.29751 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 8 14.45738444 1.80717306 78.39 <.0001 
 
 
 
 
  
 
181 
 
 
Table C.37 ANOVA results for Tukey test on moisture content (MC) for ground 
coal. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 1 220.0992667 220.0992667 287086 <.0001 
Error 4 0.0030667 0.0007667   
Corrected Total 5 220.1023333    
 
R-Square Coeff Var Root MSE MC Mean 
0.999986 0.309256 0.027689 8.953333 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 1 220.0992667 220.0992667 287086 <.0001 
 
 
Table C.38 ANOVA results for Tukey test on bulk density (BD) for ground coal. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 1 1319.727674 1319.727674 247.14 0.0040 
Error 2 10.680060 5.340030     
Corrected Total 3 1330.407733       
 
R-Square Coeff Var Root MSE BD Mean 
0.991972 0.364733 2.310850 633.5730 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 1 1319.727674 1319.727674 247.14 0.0040 
 
 
 
 
 
  
 
182 
 
 
Table C.39 ANOVA results for Tukey test on particle density (PD) for ground coal. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 1 16705.92667 16705.92667 2660.89 <.0001 
Error 4 25.11333 6.27833     
Corrected Total 5 16731.04000       
 
R-Square Coeff Var Root MSE PD Mean 
0.998499 0.174270 2.505660 1437.800 
 
Source DF Anova SS Mean Square F Value Pr > F 
Biomass 1 16705.92667 16705.92667 2660.89 <.0001 
 
 
Table C.40 ANOVA results for Tukey test on volatile matter (VM) for ground coal. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 1 59.5350000 59.5350000 2.32 0.2023 
Error 4 102.5933333 25.6483333   
Corrected Total 5 162.1283333    
 
R-Square Coeff Var Root MSE PD Mean 
0.367209 0.364403 5.064418 1389.783 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 1 59.53500000 59.53500000 2.32 0.2023 
 
 
 
 
 
  
 
183 
 
 
Table C.41 ANOVA results for Tukey test on ash content (AC) for ground coal. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 1 1.39011120 1.39011120 2.98 0.1595 
Error 4 1.86763650 0.46690913   
Corrected Total 5 3.25774770    
 
R-Square Coeff Var Root MSE AC Mean 
0.426709 6.837435 0.683307 9.993623 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 1 1.39011120 1.39011120 2.98 0.1595 
 
 
Table C.42 ANOVA results for Tukey test on energy content (EC) for ground coal. 
Source DF Sum of Squares Mean Square F Value Pr > F 
Model 1 32.30221089 32.30221089 190.84 0.0002 
Error 4 0.67705021 0.16926255   
Corrected Total 5 32.97926110    
 
R-Square Coeff Var Root MSE EC Mean 
0.979470 1.398088 0.411415 29.42700 
 
Source DF Anova SS Mean Square F Value Pr > F 
Sample 1 32.30221089 32.30221089 190.84 0.0002 
 
 
 
 
 
  
 
184 
 
Table C.43 Correlation matrix showing Pearson?s correlation coefficient and 
respective p-value for relation between all measured properties for all biomass 
dusts. 
 
 
 
 
 
AC AE BD dgw EC MC PD EXO T M E T M M L T O R V T O XY T R E VM M I T
AC 1 0 . 1 2 2 6 6 - 0 . 1 3 2 6 8 - 0 . 3 6 7 4 9 - 0 . 9 5 1 6 6 - 0 . 0 3 0 5 4 0 . 2 9 2 4 2 - 0 . 0 4 9 4 6 - 0 . 5 3 5 3 2 - 0 . 1 7 0 1 4 - 0 . 4 3 6 0 1 0 . 0 2 1 5 0 . 1 8 7 5 4 - 0 . 5 5 6 6 3 0 . 0 9 9 3 4
AC 0 . 5 1 8 5 0 . 4 8 4 6 0 . 0 4 5 7 < . 0 0 0 1 0 . 8 7 2 7 0 . 1 1 6 9 0 . 7 9 5 2 0 . 0 0 2 3 0 . 3 6 8 7 0 . 0 1 6 0 . 9 1 0 2 0 . 3 2 1 0 . 0 0 1 4 0 . 6 0 1 5
AE 0 . 1 2 2 6 6 1 - 0 . 1 7 4 6 3 - 0 . 2 8 0 4 6 - 0 . 2 2 0 4 - 0 . 2 4 7 8 9 0 . 5 6 2 9 6 - 0 . 1 2 1 7 2 - 0 . 1 1 2 2 6 0 . 2 9 7 9 6 0 . 3 2 3 6 0 . 2 4 3 7 2 0 . 4 8 9 6 5 0 . 3 5 0 4 6 0 . 5 8 7 4 6
AE 0 . 5 1 8 5 0 . 3 5 6 0 . 1 3 3 3 0 . 2 4 1 9 0 . 1 8 6 6 0 . 0 0 1 2 0 . 5 2 1 7 0 . 5 5 4 8 0 . 1 0 9 8 0 . 0 8 1 1 0 . 1 9 4 3 0 . 0 0 6 0 . 0 5 7 6 0 . 0 0 0 6
BD - 0 . 1 3 2 6 8 - 0 . 1 7 4 6 3 1 - 0 . 5 6 6 5 6 0 . 2 7 1 4 5 0 . 9 1 9 8 5 0 . 2 6 7 2 6 - 0 . 7 8 5 4 2 - 0 . 1 0 1 1 8 - 0 . 2 6 5 6 4 - 0 . 0 4 3 3 2 - 0 . 3 9 3 2 4 - 0 . 6 6 8 6 1 - 0 . 5 3 2 6 4 - 0 . 5 5 5 7 9
BD 0 . 4 8 4 6 0 . 3 5 6 0 . 0 0 1 1 0 . 1 4 6 8 < . 0 0 0 1 0 . 1 5 3 4 < . 0 0 0 1 0 . 5 9 4 7 0 . 1 5 6 0 . 8 2 0 2 0 . 0 3 1 6 < . 0 0 0 1 0 . 0 0 2 4 0 . 0 0 1 4
d g w - 0 . 3 6 7 4 9 - 0 . 2 8 0 4 6 - 0 . 5 6 6 5 6 1 0 . 3 7 1 2 9 - 0 . 3 5 7 6 9 - 0 . 7 1 0 1 1 0 . 4 3 4 8 7 0 . 3 0 0 0 7 0 . 2 7 5 8 8 0 . 1 2 8 1 0 . 4 1 9 0 1 0 . 3 5 9 9 9 0 . 5 6 4 3 9 0 . 3 1 0 0 1
d g w 0 . 0 4 5 7 0 . 1 3 3 3 0 . 0 0 1 1 0 . 0 4 3 4 0 . 0 5 2 3 < . 0 0 0 1 0 . 0 1 6 3 0 . 1 0 7 2 0 . 1 4 0 . 4 9 9 9 0 . 0 2 1 2 0 . 0 5 0 7 0 . 0 0 1 2 0 . 0 9 5 5
EC - 0 . 9 5 1 6 6 - 0 . 2 2 0 4 0 . 2 7 1 4 5 0 . 3 7 1 2 9 1 0 . 1 9 7 6 6 - 0 . 2 6 7 4 7 - 0 . 1 3 2 1 9 0 . 5 8 1 2 0 . 2 6 6 9 6 0 . 5 1 1 1 2 0 . 0 8 1 9 - 0 . 3 0 9 7 7 0 . 4 0 2 - 0 . 0 9 2 9 5
EC < . 0 0 0 1 0 . 2 4 1 9 0 . 1 4 6 8 0 . 0 4 3 4 0 . 2 9 5 1 0 . 1 5 3 0 . 4 8 6 2 0 . 0 0 0 8 0 . 1 5 3 8 0 . 0 0 3 9 0 . 6 6 7 0 . 0 9 5 7 0 . 0 2 7 7 0 . 6 2 5 2
MC - 0 . 0 3 0 5 4 - 0 . 2 4 7 8 9 0 . 9 1 9 8 5 - 0 . 3 5 7 6 9 0 . 1 9 7 6 6 1 0 . 0 9 1 5 6 - 0 . 7 3 8 9 8 - 0 . 1 0 9 4 8 - 0 . 2 6 5 5 3 - 0 . 2 1 2 5 3 - 0 . 3 7 3 1 1 - 0 . 6 4 1 9 6 - 0 . 5 6 8 0 3 - 0 . 5 6 0 4 4
MC 0 . 8 7 2 7 0 . 1 8 6 6 < . 0 0 0 1 0 . 0 5 2 3 0 . 2 9 5 1 0 . 6 3 0 4 < . 0 0 0 1 0 . 5 6 4 7 0 . 1 5 6 1 0 . 2 5 9 5 0 . 0 4 2 3 0 . 0 0 0 1 0 . 0 0 1 1 0 . 0 0 1 3
PD 0 . 2 9 2 4 2 0 . 5 6 2 9 6 0 . 2 6 7 2 6 - 0 . 7 1 0 1 1 - 0 . 2 6 7 4 7 0 . 0 9 1 5 6 1 - 0 . 3 8 3 5 6 0 . 0 2 0 2 6 0 . 2 5 5 9 7 0 . 3 9 9 9 9 0 . 1 7 9 4 4 - 0 . 0 9 1 2 5 - 0 . 2 7 0 7 3 0 . 2 5 3 6 9
PD 0 . 1 1 6 9 0 . 0 0 1 2 0 . 1 5 3 4 < . 0 0 0 1 0 . 1 5 3 0 . 6 3 0 4 0 . 0 3 6 4 0 . 9 1 5 4 0 . 1 7 2 2 0 . 0 2 8 5 0 . 3 4 2 7 0 . 6 3 1 5 0 . 1 4 7 9 0 . 1 7 6 1
E X O - 0 . 0 4 9 4 6 - 0 . 1 2 1 7 2 - 0 . 7 8 5 4 2 0 . 4 3 4 8 7 - 0 . 1 3 2 1 9 - 0 . 7 3 8 9 8 - 0 . 3 8 3 5 6 1 0 . 2 4 8 3 - 0 . 0 2 4 8 5 - 0 . 1 7 5 3 0 . 0 0 6 3 3 0 . 2 6 1 5 4 0 . 3 0 2 3 3 0 . 0 1 5 1 8
E X O 0 . 7 9 5 2 0 . 5 2 1 7 < . 0 0 0 1 0 . 0 1 6 3 0 . 4 8 6 2 < . 0 0 0 1 0 . 0 3 6 4 0 . 1 8 5 8 0 . 8 9 6 3 0 . 3 5 4 2 0 . 9 7 3 5 0 . 1 6 2 7 0 . 1 0 4 4 0 . 9 3 6 5
T M E - 0 . 5 3 5 3 2 - 0 . 1 1 2 2 6 - 0 . 1 0 1 1 8 0 . 3 0 0 0 7 0 . 5 8 1 2 - 0 . 1 0 9 4 8 0 . 0 2 0 2 6 0 . 2 4 8 3 1 0 . 4 5 0 2 4 0 . 4 8 3 9 3 0 . 2 0 2 - 0 . 3 8 9 9 2 0 . 1 2 4 6 7 - 0 . 0 0 0 2 3
T M E 0 . 0 0 2 3 0 . 5 5 4 8 0 . 5 9 4 7 0 . 1 0 7 2 0 . 0 0 0 8 0 . 5 6 4 7 0 . 9 1 5 4 0 . 1 8 5 8 0 . 0 1 2 5 0 . 0 0 6 7 0 . 2 8 4 4 0 . 0 3 3 2 0 . 5 1 1 6 0 . 9 9 9
T M M L - 0 . 1 7 0 1 4 0 . 2 9 7 9 6 - 0 . 2 6 5 6 4 0 . 2 7 5 8 8 0 . 2 6 6 9 6 - 0 . 2 6 5 5 3 0 . 2 5 5 9 7 - 0 . 0 2 4 8 5 0 . 4 5 0 2 4 1 0 . 7 4 9 2 3 0 . 7 9 8 5 4 0 . 1 0 7 9 5 0 . 2 0 9 2 1 0 . 6 7 1 6 1
T M M L 0 . 3 6 8 7 0 . 1 0 9 8 0 . 1 5 6 0 . 1 4 0 . 1 5 3 8 0 . 1 5 6 1 0 . 1 7 2 2 0 . 8 9 6 3 0 . 0 1 2 5 < . 0 0 0 1 < . 0 0 0 1 0 . 5 7 0 2 0 . 2 6 7 2 < . 0 0 0 1
T O R V - 0 . 4 3 6 0 1 0 . 3 2 3 6 - 0 . 0 4 3 3 2 0 . 1 2 8 1 0 . 5 1 1 1 2 - 0 . 2 1 2 5 3 0 . 3 9 9 9 9 - 0 . 1 7 5 3 0 . 4 8 3 9 3 0 . 7 4 9 2 3 1 0 . 6 9 1 6 0 . 1 3 2 3 1 0 . 3 6 9 4 1 0 . 6 1 1 9 4
T O R V 0 . 0 1 6 0 . 0 8 1 1 0 . 8 2 0 2 0 . 4 9 9 9 0 . 0 0 3 9 0 . 2 5 9 5 0 . 0 2 8 5 0 . 3 5 4 2 0 . 0 0 6 7 < . 0 0 0 1 < . 0 0 0 1 0 . 4 8 5 8 0 . 0 4 4 5 0 . 0 0 0 3
T O X Y 0 . 0 2 1 5 0 . 2 4 3 7 2 - 0 . 3 9 3 2 4 0 . 4 1 9 0 1 0 . 0 8 1 9 - 0 . 3 7 3 1 1 0 . 1 7 9 4 4 0 . 0 0 6 3 3 0 . 2 0 2 0 . 7 9 8 5 4 0 . 6 9 1 6 1 0 . 4 2 6 0 1 0 . 2 6 7 9 5 0 . 7 7 6 2 2
T O X Y 0 . 9 1 0 2 0 . 1 9 4 3 0 . 0 3 1 6 0 . 0 2 1 2 0 . 6 6 7 0 . 0 4 2 3 0 . 3 4 2 7 0 . 9 7 3 5 0 . 2 8 4 4 < . 0 0 0 1 < . 0 0 0 1 0 . 0 1 8 9 0 . 1 5 2 3 < . 0 0 0 1
T R E 0 . 1 8 7 5 4 0 . 4 8 9 6 5 - 0 . 6 6 8 6 1 0 . 3 5 9 9 9 - 0 . 3 0 9 7 7 - 0 . 6 4 1 9 6 - 0 . 0 9 1 2 5 0 . 2 6 1 5 4 - 0 . 3 8 9 9 2 0 . 1 0 7 9 5 0 . 1 3 2 3 1 0 . 4 2 6 0 1 1 0 . 6 3 2 9 0 . 7 2 8 6 9
T R E 0 . 3 2 1 0 . 0 0 6 < . 0 0 0 1 0 . 0 5 0 7 0 . 0 9 5 7 0 . 0 0 0 1 0 . 6 3 1 5 0 . 1 6 2 7 0 . 0 3 3 2 0 . 5 7 0 2 0 . 4 8 5 8 0 . 0 1 8 9 0 . 0 0 0 2 < . 0 0 0 1
VM - 0 . 5 5 6 6 3 0 . 3 5 0 4 6 - 0 . 5 3 2 6 4 0 . 5 6 4 3 9 0 . 4 0 2 - 0 . 5 6 8 0 3 - 0 . 2 7 0 7 3 0 . 3 0 2 3 3 0 . 1 2 4 6 7 0 . 2 0 9 2 1 0 . 3 6 9 4 1 0 . 2 6 7 9 5 0 . 6 3 2 9 1 0 . 5 5 3 7 7
VM 0 . 0 0 1 4 0 . 0 5 7 6 0 . 0 0 2 4 0 . 0 0 1 2 0 . 0 2 7 7 0 . 0 0 1 1 0 . 1 4 7 9 0 . 1 0 4 4 0 . 5 1 1 6 0 . 2 6 7 2 0 . 0 4 4 5 0 . 1 5 2 3 0 . 0 0 0 2 0 . 0 0 1 5
M IT 0 . 0 9 9 3 4 0 . 5 8 7 4 6 - 0 . 5 5 5 7 9 0 . 3 1 0 0 1 - 0 . 0 9 2 9 5 - 0 . 5 6 0 4 4 0 . 2 5 3 6 9 0 . 0 1 5 1 8 - 0 . 0 0 0 2 3 0 . 6 7 1 6 1 0 . 6 1 1 9 4 0 . 7 7 6 2 2 0 . 7 2 8 6 9 0 . 5 5 3 7 7 1
M I T 0 . 6 0 1 5 0 . 0 0 0 6 0 . 0 0 1 4 0 . 0 9 5 5 0 . 6 2 5 2 0 . 0 0 1 3 0 . 1 7 6 1 0 . 9 3 6 5 0 . 9 9 9 < . 0 0 0 1 0 . 0 0 0 3 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 1 5
Pe a r s o n  C o r r e l a ti o n  C o e ffi c i e n ts ,  N  =  3 0
Pr o b  >  | r |  u n d e r  H 0 :  R h o = 0
  
 
185 
 
Table C.44 Correlation matrix showing Pearson?s correlation coefficient and 
respective p-value for relation between all measured properties for grassy 
biomass (Bermuda grass, corn stover, sugarcane bagasse and switchgrass) 
dusts. 
 
 
 
 
 
 
AC AE BD dgw EC MC PD EXO T M E T M M L T O R V T O XY T R E VM M I T
AC 1 0 . 9 6 4 3 9 - 0 . 9 5 9 2 7 - 0 . 9 0 8 3 3 - 0 . 9 8 8 6 5 - 0 . 8 7 3 9 3 0 . 9 6 5 0 7 - 0 . 2 6 1 7 5 0 . 1 3 2 2 2 0 . 5 4 4 2 7 0 . 5 7 5 8 8 0 . 3 2 5 2 2 0 . 1 1 8 4 1 - 0 . 9 3 9 5 9 0 . 7 5 1 2 7
AC < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 0 2 < . 0 0 0 1 0 . 4 1 1 2 0 . 6 8 2 1 0 . 0 6 7 3 0 . 0 5 0 1 0 . 3 0 2 3 0 . 7 1 4 < . 0 0 0 1 0 . 0 0 4 9
AE 0 . 9 6 4 3 9 1 - 0 . 9 8 8 5 8 - 0 . 7 9 5 2 6 - 0 . 9 4 1 8 7 - 0 . 8 1 8 1 0 . 9 9 6 4 8 - 0 . 2 6 7 6 0 . 1 1 8 0 6 0 . 6 8 5 1 3 0 . 7 5 7 9 5 0 . 5 3 3 7 7 0 . 3 3 7 1 2 - 0 . 8 9 3 0 3 0 . 8 8 2 0 9
AE < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 2 < . 0 0 0 1 0 . 0 0 1 1 < . 0 0 0 1 0 . 4 0 0 4 0 . 7 1 4 8 0 . 0 1 3 9 0 . 0 0 4 3 0 . 0 7 3 9 0 . 2 8 3 9 < . 0 0 0 1 0 . 0 0 0 1
BD - 0 . 9 5 9 2 7 - 0 . 9 8 8 5 8 1 0 . 8 2 3 1 3 0 . 9 5 5 2 9 0 . 8 4 6 8 - 0 . 9 8 9 3 4 0 . 2 5 6 2 5 - 0 . 0 4 1 3 6 - 0 . 7 3 0 9 5 - 0 . 7 4 4 1 9 - 0 . 5 2 1 0 5 - 0 . 3 4 2 3 2 0 . 8 8 9 6 4 - 0 . 8 9 0 1 3
BD < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 1 < . 0 0 0 1 0 . 0 0 0 5 < . 0 0 0 1 0 . 4 2 1 4 0 . 8 9 8 4 0 . 0 0 6 9 0 . 0 0 5 5 0 . 0 8 2 4 0 . 2 7 6 1 0 . 0 0 0 1 0 . 0 0 0 1
d g w - 0 . 9 0 8 3 3 - 0 . 7 9 5 2 6 0 . 8 2 3 1 3 1 0 . 9 4 4 3 9 0 . 9 2 2 0 4 - 0 . 7 8 9 2 7 0 . 1 6 5 2 1 0 . 0 4 8 1 - 0 . 3 9 0 2 2 - 0 . 2 3 8 8 2 0 . 0 0 3 7 7 0 . 1 6 0 0 8 0 . 9 1 1 6 6 - 0 . 5 0 5 1
d g w < . 0 0 0 1 0 . 0 0 2 0 . 0 0 1 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 2 3 0 . 6 0 7 9 0 . 8 8 2 0 . 2 0 9 8 0 . 4 5 4 7 0 . 9 9 0 7 0 . 6 1 9 2 < . 0 0 0 1 0 . 0 9 3 9
EC - 0 . 9 8 8 6 5 - 0 . 9 4 1 8 7 0 . 9 5 5 2 9 0 . 9 4 4 3 9 1 0 . 9 1 3 7 - 0 . 9 4 1 9 2 0 . 2 3 0 6 3 - 0 . 0 5 3 2 5 - 0 . 5 5 5 8 1 - 0 . 5 2 8 2 8 - 0 . 2 7 2 4 2 - 0 . 0 9 3 9 9 0 . 9 4 0 4 5 - 0 . 7 2 6 0 5
EC < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 0 . 4 7 0 8 0 . 8 6 9 4 0 . 0 6 0 6 0 . 0 7 7 5 0 . 3 9 1 7 0 . 7 7 1 4 < . 0 0 0 1 0 . 0 0 7 5
MC - 0 . 8 7 3 9 3 - 0 . 8 1 8 1 0 . 8 4 6 8 0 . 9 2 2 0 4 0 . 9 1 3 7 1 - 0 . 8 0 6 4 5 0 . 2 6 2 8 3 0 . 1 1 9 6 - 0 . 5 9 0 2 - 0 . 4 0 1 2 - 0 . 2 2 5 8 9 0 . 0 1 2 4 5 0 . 8 5 8 9 5 - 0 . 6 3 2 4 5
MC 0 . 0 0 0 2 0 . 0 0 1 1 0 . 0 0 0 5 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 1 5 0 . 4 0 9 2 0 . 7 1 1 2 0 . 0 4 3 4 0 . 1 9 6 1 0 . 4 8 0 2 0 . 9 6 9 4 0 . 0 0 0 3 0 . 0 2 7 3
PD 0 . 9 6 5 0 7 0 . 9 9 6 4 8 - 0 . 9 8 9 3 4 - 0 . 7 8 9 2 7 - 0 . 9 4 1 9 2 - 0 . 8 0 6 4 5 1 - 0 . 3 0 4 0 7 0 . 1 3 8 7 2 0 . 6 9 0 9 3 0 . 7 6 5 7 2 0 . 5 4 5 6 5 0 . 3 4 2 0 4 - 0 . 8 8 0 9 3 0 . 8 8 8 4 7
PD < . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 2 3 < . 0 0 0 1 0 . 0 0 1 5 0 . 3 3 6 6 0 . 6 6 7 2 0 . 0 1 2 8 0 . 0 0 3 7 0 . 0 6 6 5 0 . 2 7 6 5 0 . 0 0 0 2 0 . 0 0 0 1
E X O - 0 . 2 6 1 7 5 - 0 . 2 6 7 6 0 . 2 5 6 2 5 0 . 1 6 5 2 1 0 . 2 3 0 6 3 0 . 2 6 2 8 3 - 0 . 3 0 4 0 7 1 0 . 0 8 0 9 3 - 0 . 3 7 2 1 8 - 0 . 3 0 5 2 1 - 0 . 4 4 6 1 2 - 0 . 2 8 2 8 8 0 . 1 0 9 5 3 - 0 . 3 4 6 0 4
E X O 0 . 4 1 1 2 0 . 4 0 0 4 0 . 4 2 1 4 0 . 6 0 7 9 0 . 4 7 0 8 0 . 4 0 9 2 0 . 3 3 6 6 0 . 8 0 2 6 0 . 2 3 3 5 0 . 3 3 4 7 0 . 1 4 6 0 . 3 7 3 0 . 7 3 4 7 0 . 2 7 0 5
T M E 0 . 1 3 2 2 2 0 . 1 1 8 0 6 - 0 . 0 4 1 3 6 0 . 0 4 8 1 - 0 . 0 5 3 2 5 0 . 1 1 9 6 0 . 1 3 8 7 2 0 . 0 8 0 9 3 1 - 0 . 2 8 2 5 5 0 . 0 8 8 4 8 - 0 . 0 5 3 1 7 - 0 . 0 7 8 9 3 - 0 . 2 1 1 5 8 - 0 . 0 2 7 9 3
T M E 0 . 6 8 2 1 0 . 7 1 4 8 0 . 8 9 8 4 0 . 8 8 2 0 . 8 6 9 4 0 . 7 1 1 2 0 . 6 6 7 2 0 . 8 0 2 6 0 . 3 7 3 6 0 . 7 8 4 5 0 . 8 6 9 7 0 . 8 0 7 4 0 . 5 0 9 2 0 . 9 3 1 3
T M M L 0 . 5 4 4 2 7 0 . 6 8 5 1 3 - 0 . 7 3 0 9 5 - 0 . 3 9 0 2 2 - 0 . 5 5 5 8 1 - 0 . 5 9 0 2 0 . 6 9 0 9 3 - 0 . 3 7 2 1 8 - 0 . 2 8 2 5 5 1 0 . 8 1 3 5 6 0 . 8 1 6 8 4 0 . 6 1 8 7 3 - 0 . 4 5 9 1 5 0 . 9 1 5 1 6
T M M L 0 . 0 6 7 3 0 . 0 1 3 9 0 . 0 0 6 9 0 . 2 0 9 8 0 . 0 6 0 6 0 . 0 4 3 4 0 . 0 1 2 8 0 . 2 3 3 5 0 . 3 7 3 6 0 . 0 0 1 3 0 . 0 0 1 2 0 . 0 3 2 0 . 1 3 3 2 < . 0 0 0 1
T O R V 0 . 5 7 5 8 8 0 . 7 5 7 9 5 - 0 . 7 4 4 1 9 - 0 . 2 3 8 8 2 - 0 . 5 2 8 2 8 - 0 . 4 0 1 2 0 . 7 6 5 7 2 - 0 . 3 0 5 2 1 0 . 0 8 8 4 8 0 . 8 1 3 5 6 1 0 . 9 2 3 0 3 0 . 7 6 0 9 6 - 0 . 4 3 6 3 6 0 . 9 3 6 0 1
T O R V 0 . 0 5 0 1 0 . 0 0 4 3 0 . 0 0 5 5 0 . 4 5 4 7 0 . 0 7 7 5 0 . 1 9 6 1 0 . 0 0 3 7 0 . 3 3 4 7 0 . 7 8 4 5 0 . 0 0 1 3 < . 0 0 0 1 0 . 0 0 4 0 . 1 5 6 1 < . 0 0 0 1
T O X Y 0 . 3 2 5 2 2 0 . 5 3 3 7 7 - 0 . 5 2 1 0 5 0 . 0 0 3 7 7 - 0 . 2 7 2 4 2 - 0 . 2 2 5 8 9 0 . 5 4 5 6 5 - 0 . 4 4 6 1 2 - 0 . 0 5 3 1 7 0 . 8 1 6 8 4 0 . 9 2 3 0 3 1 0 . 7 7 3 5 5 - 0 . 1 5 2 7 2 0 . 8 3 2 5 7
T O X Y 0 . 3 0 2 3 0 . 0 7 3 9 0 . 0 8 2 4 0 . 9 9 0 7 0 . 3 9 1 7 0 . 4 8 0 2 0 . 0 6 6 5 0 . 1 4 6 0 . 8 6 9 7 0 . 0 0 1 2 < . 0 0 0 1 0 . 0 0 3 2 0 . 6 3 5 6 0 . 0 0 0 8
T R E 0 . 1 1 8 4 1 0 . 3 3 7 1 2 - 0 . 3 4 2 3 2 0 . 1 6 0 0 8 - 0 . 0 9 3 9 9 0 . 0 1 2 4 5 0 . 3 4 2 0 4 - 0 . 2 8 2 8 8 - 0 . 0 7 8 9 3 0 . 6 1 8 7 3 0 . 7 6 0 9 6 0 . 7 7 3 5 5 1 - 0 . 0 6 9 4 0 . 6 2 7 2 3
T R E 0 . 7 1 4 0 . 2 8 3 9 0 . 2 7 6 1 0 . 6 1 9 2 0 . 7 7 1 4 0 . 9 6 9 4 0 . 2 7 6 5 0 . 3 7 3 0 . 8 0 7 4 0 . 0 3 2 0 . 0 0 4 0 . 0 0 3 2 0 . 8 3 0 3 0 . 0 2 9
VM - 0 . 9 3 9 5 9 - 0 . 8 9 3 0 3 0 . 8 8 9 6 4 0 . 9 1 1 6 6 0 . 9 4 0 4 5 0 . 8 5 8 9 5 - 0 . 8 8 0 9 3 0 . 1 0 9 5 3 - 0 . 2 1 1 5 8 - 0 . 4 5 9 1 5 - 0 . 4 3 6 3 6 - 0 . 1 5 2 7 2 - 0 . 0 6 9 4 1 - 0 . 6 4 0 1
VM < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 0 1 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 0 3 0 . 0 0 0 2 0 . 7 3 4 7 0 . 5 0 9 2 0 . 1 3 3 2 0 . 1 5 6 1 0 . 6 3 5 6 0 . 8 3 0 3 0 . 0 2 5
M IT 0 . 7 5 1 2 7 0 . 8 8 2 0 9 - 0 . 8 9 0 1 3 - 0 . 5 0 5 1 - 0 . 7 2 6 0 5 - 0 . 6 3 2 4 5 0 . 8 8 8 4 7 - 0 . 3 4 6 0 4 - 0 . 0 2 7 9 3 0 . 9 1 5 1 6 0 . 9 3 6 0 1 0 . 8 3 2 5 7 0 . 6 2 7 2 3 - 0 . 6 4 0 1 1
M I T 0 . 0 0 4 9 0 . 0 0 0 1 0 . 0 0 0 1 0 . 0 9 3 9 0 . 0 0 7 5 0 . 0 2 7 3 0 . 0 0 0 1 0 . 2 7 0 5 0 . 9 3 1 3 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 0 8 0 . 0 2 9 0 . 0 2 5
Pe a r s o n  C o r r e l a ti o n  C o e ffi c i e n ts ,  N  =  1 2
Pr o b  >  | r |  u n d e r  H 0 :  R h o = 0
  
 
186 
 
Table C.45 Correlation matrix showing Pearson?s correlation coefficient and 
respective p-value for relation between all measured properties for woody 
biomass (eucalyptus, pine and sweetgum) dusts. 
 
AC AE BD dgw EC MC PD EXO T M E T M M L T O R V T O XY T R E VM M I T
AC 1 0 . 4 7 6 5 1 0 . 4 6 7 5 9 - 0 . 5 1 7 4 8 - 0 . 6 0 4 0 7 0 . 0 6 6 8 5 0 . 5 2 0 2 5 0 . 4 4 8 3 3 - 0 . 6 9 1 5 8 - 0 . 4 2 2 7 3 - 0 . 6 2 1 8 7 - 0 . 5 8 1 1 7 0 . 0 0 7 4 6 - 0 . 1 3 3 0 1 - 0 . 4 8 4 6 8
AC 0 . 1 9 4 7 0 . 2 0 4 4 0 . 1 5 3 6 0 . 0 8 4 9 0 . 8 6 4 3 0 . 1 5 1 1 0 . 2 2 6 1 0 . 0 3 9 0 . 2 5 7 0 . 0 7 3 8 0 . 1 0 0 8 0 . 9 8 4 8 0 . 7 3 3 0 . 1 8 6 1
AE 0 . 4 7 6 5 1 1 - 0 . 0 4 9 5 8 - 0 . 1 3 8 3 1 - 0 . 5 4 8 0 3 0 . 8 2 5 8 3 0 . 2 1 8 0 9 0 . 0 1 7 9 9 - 0 . 2 6 0 4 9 - 0 . 0 3 3 2 9 - 0 . 7 5 5 3 1 - 0 . 4 0 8 2 5 0 . 6 9 8 0 6 0 . 5 6 0 0 8 - 0 . 0 6 0 1 8
AE 0 . 1 9 4 7 0 . 8 9 9 2 0 . 7 2 2 7 0 . 1 2 6 6 0 . 0 0 6 1 0 . 5 7 2 9 0 . 9 6 3 4 0 . 4 9 8 4 0 . 9 3 2 2 0 . 0 1 8 6 0 . 2 7 5 3 0 . 0 3 6 5 0 . 1 1 6 8 0 . 8 7 7 8
BD 0 . 4 6 7 5 9 - 0 . 0 4 9 5 8 1 - 0 . 9 7 4 9 3 - 0 . 7 8 9 3 - 0 . 5 2 9 9 8 0 . 8 7 1 2 0 . 8 7 9 9 5 - 0 . 9 2 0 5 9 - 0 . 9 6 7 2 6 - 0 . 5 8 0 7 2 - 0 . 8 6 4 7 8 - 0 . 3 1 9 8 4 - 0 . 8 3 0 7 5 - 0 . 9 8 6 2 6
BD 0 . 2 0 4 4 0 . 8 9 9 2 < . 0 0 0 1 0 . 0 1 1 4 0 . 1 4 2 2 0 . 0 0 2 2 0 . 0 0 1 8 0 . 0 0 0 4 < . 0 0 0 1 0 . 1 0 1 1 0 . 0 0 2 6 0 . 4 0 1 4 0 . 0 0 5 5 < . 0 0 0 1
d g w - 0 . 5 1 7 4 8 - 0 . 1 3 8 3 1 - 0 . 9 7 4 9 3 1 0 . 8 8 9 7 9 0 . 3 5 8 3 8 - 0 . 9 2 5 6 6 - 0 . 8 8 4 7 2 0 . 9 4 3 5 8 0 . 9 8 5 1 5 0 . 7 3 4 6 4 0 . 9 3 3 0 . 2 1 2 2 7 0 . 7 1 2 1 1 0 . 9 9 6 8 8
d g w 0 . 1 5 3 6 0 . 7 2 2 7 < . 0 0 0 1 0 . 0 0 1 3 0 . 3 4 3 6 0 . 0 0 0 3 0 . 0 0 1 5 0 . 0 0 0 1 < . 0 0 0 1 0 . 0 2 4 2 0 . 0 0 0 2 0 . 5 8 3 5 0 . 0 3 1 4 < . 0 0 0 1
EC - 0 . 6 0 4 0 7 - 0 . 5 4 8 0 3 - 0 . 7 8 9 3 0 . 8 8 9 7 9 1 - 0 . 0 4 4 1 5 - 0 . 8 2 8 4 8 - 0 . 7 3 6 9 7 0 . 8 8 9 9 5 0 . 8 2 9 8 4 0 . 9 4 2 7 9 0 . 9 3 6 3 4 - 0 . 1 6 3 4 5 0 . 3 2 1 6 7 0 . 8 5 2 9 2
EC 0 . 0 8 4 9 0 . 1 2 6 6 0 . 0 1 1 4 0 . 0 0 1 3 0 . 9 1 0 2 0 . 0 0 5 8 0 . 0 2 3 5 0 . 0 0 1 3 0 . 0 0 5 6 0 . 0 0 0 1 0 . 0 0 0 2 0 . 6 7 4 3 0 . 3 9 8 6 0 . 0 0 3 5
MC 0 . 0 6 6 8 5 0 . 8 2 5 8 3 - 0 . 5 2 9 9 8 0 . 3 5 8 3 8 - 0 . 0 4 4 1 5 1 - 0 . 2 5 9 7 1 - 0 . 4 4 9 7 9 0 . 2 5 2 6 0 . 4 3 9 4 - 0 . 3 4 4 6 6 0 . 0 4 5 2 2 0 . 6 5 1 7 2 0 . 8 2 4 0 2 0 . 4 2 6 9 2
MC 0 . 8 6 4 3 0 . 0 0 6 1 0 . 1 4 2 2 0 . 3 4 3 6 0 . 9 1 0 2 0 . 4 9 9 8 0 . 2 2 4 5 0 . 5 1 2 0 . 2 3 6 7 0 . 3 6 3 7 0 . 9 0 8 0 . 0 5 7 2 0 . 0 0 6 3 0 . 2 5 1 8
PD 0 . 5 2 0 2 5 0 . 2 1 8 0 9 0 . 8 7 1 2 - 0 . 9 2 5 6 6 - 0 . 8 2 8 4 8 - 0 . 2 5 9 7 1 1 0 . 9 0 1 1 7 - 0 . 8 3 3 4 4 - 0 . 9 3 4 2 5 - 0 . 7 1 4 8 3 - 0 . 8 9 8 6 5 - 0 . 1 6 1 8 4 - 0 . 6 2 1 6 2 - 0 . 9 1 7 4 7
PD 0 . 1 5 1 1 0 . 5 7 2 9 0 . 0 0 2 2 0 . 0 0 0 3 0 . 0 0 5 8 0 . 4 9 9 8 0 . 0 0 0 9 0 . 0 0 5 3 0 . 0 0 0 2 0 . 0 3 0 4 0 . 0 0 1 0 . 6 7 7 4 0 . 0 7 3 9 0 . 0 0 0 5
E X O 0 . 4 4 8 3 3 0 . 0 1 7 9 9 0 . 8 7 9 9 5 - 0 . 8 8 4 7 2 - 0 . 7 3 6 9 7 - 0 . 4 4 9 7 9 0 . 9 0 1 1 7 1 - 0 . 7 5 8 9 4 - 0 . 8 8 5 4 4 - 0 . 5 4 4 3 - 0 . 7 9 7 8 8 - 0 . 3 7 9 8 9 - 0 . 7 1 2 5 6 - 0 . 8 8 9 4
E X O 0 . 2 2 6 1 0 . 9 6 3 4 0 . 0 0 1 8 0 . 0 0 1 5 0 . 0 2 3 5 0 . 2 2 4 5 0 . 0 0 0 9 0 . 0 1 7 7 0 . 0 0 1 5 0 . 1 2 9 8 0 . 0 1 0 . 3 1 3 2 0 . 0 3 1 2 0 . 0 0 1 3
T M E - 0 . 6 9 1 5 8 - 0 . 2 6 0 4 9 - 0 . 9 2 0 5 9 0 . 9 4 3 5 8 0 . 8 8 9 9 5 0 . 2 5 2 6 - 0 . 8 3 3 4 4 - 0 . 7 5 8 9 4 1 0 . 8 8 6 1 8 0 . 7 7 0 8 9 0 . 9 0 9 9 4 0 . 0 7 1 4 4 0 . 6 0 5 6 0 . 9 2 9 9 9
T M E 0 . 0 3 9 0 . 4 9 8 4 0 . 0 0 0 4 0 . 0 0 0 1 0 . 0 0 1 3 0 . 5 1 2 0 . 0 0 5 3 0 . 0 1 7 7 0 . 0 0 1 5 0 . 0 1 5 0 . 0 0 0 7 0 . 8 5 5 1 0 . 0 8 3 9 0 . 0 0 0 3
T M M L - 0 . 4 2 2 7 3 - 0 . 0 3 3 2 9 - 0 . 9 6 7 2 6 0 . 9 8 5 1 5 0 . 8 2 9 8 4 0 . 4 3 9 4 - 0 . 9 3 4 2 5 - 0 . 8 8 5 4 4 0 . 8 8 6 1 8 1 0 . 6 6 0 2 4 0 . 8 8 4 4 0 . 2 7 5 0 2 0 . 7 5 8 7 2 0 . 9 9 1 0 6
T M M L 0 . 2 5 7 0 . 9 3 2 2 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 5 6 0 . 2 3 6 7 0 . 0 0 0 2 0 . 0 0 1 5 0 . 0 0 1 5 0 . 0 5 2 9 0 . 0 0 1 5 0 . 4 7 3 9 0 . 0 1 7 8 < . 0 0 0 1
T O R V - 0 . 6 2 1 8 7 - 0 . 7 5 5 3 1 - 0 . 5 8 0 7 2 0 . 7 3 4 6 4 0 . 9 4 2 7 9 - 0 . 3 4 4 6 6 - 0 . 7 1 4 8 3 - 0 . 5 4 4 3 0 . 7 7 0 8 9 0 . 6 6 0 2 4 1 0 . 8 6 3 5 7 - 0 . 3 0 0 5 8 0 . 0 6 4 1 1 0 . 6 8 0 5 2
T O R V 0 . 0 7 3 8 0 . 0 1 8 6 0 . 1 0 1 1 0 . 0 2 4 2 0 . 0 0 0 1 0 . 3 6 3 7 0 . 0 3 0 4 0 . 1 2 9 8 0 . 0 1 5 0 . 0 5 2 9 0 . 0 0 2 7 0 . 4 3 1 9 0 . 8 6 9 8 0 . 0 4 3 6
T O X Y - 0 . 5 8 1 1 7 - 0 . 4 0 8 2 5 - 0 . 8 6 4 7 8 0 . 9 3 3 0 . 9 3 6 3 4 0 . 0 4 5 2 2 - 0 . 8 9 8 6 5 - 0 . 7 9 7 8 8 0 . 9 0 9 9 4 0 . 8 8 4 4 0 . 8 6 3 5 7 1 0 . 0 3 1 1 8 0 . 5 1 3 6 0 . 9 0 7 9 7
T O X Y 0 . 1 0 0 8 0 . 2 7 5 3 0 . 0 0 2 6 0 . 0 0 0 2 0 . 0 0 0 2 0 . 9 0 8 0 . 0 0 1 0 . 0 1 0 . 0 0 0 7 0 . 0 0 1 5 0 . 0 0 2 7 0 . 9 3 6 5 0 . 1 5 7 3 0 . 0 0 0 7
T R E 0 . 0 0 7 4 6 0 . 6 9 8 0 6 - 0 . 3 1 9 8 4 0 . 2 1 2 2 7 - 0 . 1 6 3 4 5 0 . 6 5 1 7 2 - 0 . 1 6 1 8 4 - 0 . 3 7 9 8 9 0 . 0 7 1 4 4 0 . 2 7 5 0 2 - 0 . 3 0 0 5 8 0 . 0 3 1 1 8 1 0 . 6 6 7 1 8 0 . 2 6 8 1 5
T R E 0 . 9 8 4 8 0 . 0 3 6 5 0 . 4 0 1 4 0 . 5 8 3 5 0 . 6 7 4 3 0 . 0 5 7 2 0 . 6 7 7 4 0 . 3 1 3 2 0 . 8 5 5 1 0 . 4 7 3 9 0 . 4 3 1 9 0 . 9 3 6 5 0 . 0 4 9 6 0 . 4 8 5 4
VM - 0 . 1 3 3 0 1 0 . 5 6 0 0 8 - 0 . 8 3 0 7 5 0 . 7 1 2 1 1 0 . 3 2 1 6 7 0 . 8 2 4 0 2 - 0 . 6 2 1 6 2 - 0 . 7 1 2 5 6 0 . 6 0 5 6 0 . 7 5 8 7 2 0 . 0 6 4 1 1 0 . 5 1 3 6 0 . 6 6 7 1 8 1 0 . 7 6 1 9 2
VM 0 . 7 3 3 0 . 1 1 6 8 0 . 0 0 5 5 0 . 0 3 1 4 0 . 3 9 8 6 0 . 0 0 6 3 0 . 0 7 3 9 0 . 0 3 1 2 0 . 0 8 3 9 0 . 0 1 7 8 0 . 8 6 9 8 0 . 1 5 7 3 0 . 0 4 9 6 0 . 0 1 7
M IT - 0 . 4 8 4 6 8 - 0 . 0 6 0 1 8 - 0 . 9 8 6 2 6 0 . 9 9 6 8 8 0 . 8 5 2 9 2 0 . 4 2 6 9 2 - 0 . 9 1 7 4 7 - 0 . 8 8 9 4 0 . 9 2 9 9 9 0 . 9 9 1 0 6 0 . 6 8 0 5 2 0 . 9 0 7 9 7 0 . 2 6 8 1 5 0 . 7 6 1 9 2 1
M I T 0 . 1 8 6 1 0 . 8 7 7 8 < . 0 0 0 1 < . 0 0 0 1 0 . 0 0 3 5 0 . 2 5 1 8 0 . 0 0 0 5 0 . 0 0 1 3 0 . 0 0 0 3 < . 0 0 0 1 0 . 0 4 3 6 0 . 0 0 0 7 0 . 4 8 5 4 0 . 0 1 7
Pr o b  >  | r |  u n d e r  H 0 :  R h o = 0
  
 
187 
 
Appendix ? D SAS Codes and Results for Principal Component Analysis 
(PCA) on NIR Data for Internal Validation of Models
 
SAS code for PCA on raw NIR spectral data for biomass dusts used for internal 
validation 
/******************************** 
 * Author: Jaskaran Dhiman      * 
 * Research work                * 
 * Date: 01/31/2014             * 
 ********************************/ 
 
ods html file='C:\Users\jzd0028\Desktop\Research\NIR new 
analysis\output_raw.html'; 
options nodate pageno=1; 
data work.jasnir; 
infile 'C:\Users\jzd0028\Desktop\Research\NIR new 
analysis\raw_data_SAS.csv' delimiter = ',' MISSOVER DSD lrecl=32767 
firstobs=1 n=1500; 
input Number$ Biomass$ MIT TORV TMML TOXY TRE TME  
w9994 w9984 w9974 w9964 w9954 w9944 w9934 w9924 w9914 w9904 w9894
 w9884 w9874  
w9864 w9854 w9844 w9834 w9824 w9814 w9804 w9794 w9784 w9774 w9764
 w9754 w9744 
w9734 w9724 w9714 w9704 w9694 w9684 w9674 w9664 w9654 w9644 w9634
 w9624 w9614 
w9604 w9594 w9584 w9574 w9564 w9554 w9544 w9534 w9524 w9514 w9504
 w9494 w9484 
w9474 w9464 w9454 w9444 w9434 w9424 w9414 w9404 w9394 w9384 w9374
 w9364 w9354 
w9344 w9334 w9324 w9314 w9304 w9294 w9284 w9274 w9264 w9254 w9244
 w9234 w9224 
w9214 w9204 w9194 w9184 w9174 w9164 w9154 w9144 w9134 w9124 w9114
 w9104 w9094 
w9084 w9074 w9064 w9054 w9044 w9034 w9024 w9014 w9004 w8994 w8984
 w8974 w8964 
w8954 w8944 w8934 w8924 w8914 w8904 w8894 w8884 w8874 w8864 w8854
 w8844 w8834 
w8824 w8814 w8804 w8794 w8784 w8774 w8764 w8754 w8744 w8734 w8724
 w8714 w8704 
w8694 w8684 w8674 w8664 w8654 w8644 w8634 w8624 w8614 w8604 w8594
 w8584 w8574 
w8564 w8554 w8544 w8534 w8524 w8514 w8504 w8494 w8484 w8474 w8464
 w8454 w8444 
w8434 w8424 w8414 w8404 w8394 w8384 w8374 w8364 w8354 w8344 w8334
 w8324 w8314 
  
 
188 
 
w8304 w8294 w8284 w8274 w8264 w8254 w8244 w8234 w8224 w8214 w8204
 w8194 w8184 
w8174 w8164 w8154 w8144 w8134 w8124 w8114 w8104 w8094 w8084 w8074
 w8064 w8054 
w8044 w8034 w8024 w8014 w8004 w7994 w7984 w7974 w7964 w7954 w7944
 w7934 w7924 
w7914 w7904 w7894 w7884 w7874 w7864 w7854 w7844 w7834 w7824 w7814
 w7804 w7794 
w7784 w7774 w7764 w7754 w7744 w7734 w7724 w7714 w7704 w7694 w7684
 w7674 w7664 
w7654 w7644 w7634 w7624 w7614 w7604 w7594 w7584 w7574 w7564 w7554
 w7544 w7534 
w7524 w7514 w7504 w7494 w7484 w7474 w7464 w7454 w7444 w7434 w7424
 w7414 w7404 
w7394 w7384 w7374 w7364 w7354 w7344 w7334 w7324 w7314 w7304 w7294
 w7284 w7274 
w7264 w7254 w7244 w7234 w7224 w7214 w7204 w7194 w7184 w7174 w7164
 w7154 w7144 
w7134 w7124 w7114 w7104 w7094 w7084 w7074 w7064 w7054 w7044 w7034
 w7024 w7014 
w7004 w6994 w6984 w6974 w6964 w6954 w6944 w6934 w6924 w6914 w6904
 w6894 w6884 
w6874 w6864 w6854 w6844 w6834 w6824 w6814 w6804 w6794 w6784 w6774
 w6764 w6754 
w6744 w6734 w6724 w6714 w6704 w6694 w6684 w6674 w6664 w6654 w6644
 w6634 w6624 
w6614 w6604 w6594 w6584 w6574 w6564 w6554 w6544 w6534 w6524 w6514
 w6504 w6494 
w6484 w6474 w6464 w6454 w6444 w6434 w6424 w6414 w6404 w6394 w6384
 w6374 w6364 
w6354 w6344 w6334 w6324 w6314 w6304 w6294 w6284 w6274 w6264 w6254
 w6244 w6234 
w6224 w6214 w6204 w6194 w6184 w6174 w6164 w6154 w6144 w6134 w6124
 w6114 w6104 
w6094 w6084 w6074 w6064 w6054 w6044 w6034 w6024 w6014 w6004 w5994
 w5984 w5974 
w5964 w5954 w5944 w5934 w5924 w5914 w5904 w5894 w5884 w5874 w5864
 w5854 w5844 
w5834 w5824 w5814 w5804 w5794 w5784 w5774 w5764 w5754 w5744 w5734
 w5724 w5714 
w5704 w5694 w5684 w5674 w5664 w5654 w5644 w5634 w5624 w5614 w5604
 w5594 w5584 
w5574 w5564 w5554 w5544 w5534 w5524 w5514 w5504 w5494 w5484 w5474
 w5464 w5454 
w5444 w5434 w5424 w5414 w5404 w5394 w5384 w5374 w5364 w5354 w5344
 w5334 w5324 
w5314 w5304 w5294 w5284 w5274 w5264 w5254 w5244 w5234 w5224 w5214
 w5204 w5194 
w5184 w5174 w5164 w5154 w5144 w5134 w5124 w5114 w5104 w5094 w5084
 w5074 w5064 
w5054 w5044 w5034 w5024 w5014 w5004 w4994 w4984 w4974 w4964 w4954
 w4944 w4934 
w4924 w4914 w4904 w4894 w4884 w4874 w4864 w4854 w4844 w4834 w4824
 w4814 w4804 
w4794 w4784 w4774 w4764 w4754 w4744 w4734 w4724 w4714 w4704 w4694
 w4684 w4674 
  
 
189 
 
w4664 w4654 w4644 w4634 w4624 w4614 w4604 w4594 w4584 w4574 w4564
 w4554 w4544 
w4534 w4524 w4514 w4504 w4494 w4484 w4474 w4464 w4454 w4444 w4434
 w4424 w4414 
w4404 w4394 w4384 w4374 w4364 w4354 w4344 w4334 w4324 w4314 w4304
 w4294 w4284 
w4274 w4264 w4254 w4244 w4234 w4224 w4214 w4204 w4194 w4184 w4174
 w4164 w4154 
w4144 w4134 w4124 w4114 w4104 w4094 w4084 w4074 w4064 w4054 w4044
 w4034 w4024 
w4014 w4004; 
datalines; 
; 
run; 
 
proc princomp data=work.jasnir out=prinvars std; 
var w9994 w9984 w9974 w9964 w9954 w9944 w9934 w9924 w9914 w9904
 w9894 w9884 w9874  
w9864 w9854 w9844 w9834 w9824 w9814 w9804 w9794 w9784 w9774 w9764
 w9754 w9744 
w9734 w9724 w9714 w9704 w9694 w9684 w9674 w9664 w9654 w9644 w9634
 w9624 w9614 
w9604 w9594 w9584 w9574 w9564 w9554 w9544 w9534 w9524 w9514 w9504
 w9494 w9484 
w9474 w9464 w9454 w9444 w9434 w9424 w9414 w9404 w9394 w9384 w9374
 w9364 w9354 
w9344 w9334 w9324 w9314 w9304 w9294 w9284 w9274 w9264 w9254 w9244
 w9234 w9224 
w9214 w9204 w9194 w9184 w9174 w9164 w9154 w9144 w9134 w9124 w9114
 w9104 w9094 
w9084 w9074 w9064 w9054 w9044 w9034 w9024 w9014 w9004 w8994 w8984
 w8974 w8964 
w8954 w8944 w8934 w8924 w8914 w8904 w8894 w8884 w8874 w8864 w8854
 w8844 w8834 
w8824 w8814 w8804 w8794 w8784 w8774 w8764 w8754 w8744 w8734 w8724
 w8714 w8704 
w8694 w8684 w8674 w8664 w8654 w8644 w8634 w8624 w8614 w8604 w8594
 w8584 w8574 
w8564 w8554 w8544 w8534 w8524 w8514 w8504 w8494 w8484 w8474 w8464
 w8454 w8444 
w8434 w8424 w8414 w8404 w8394 w8384 w8374 w8364 w8354 w8344 w8334
 w8324 w8314 
w8304 w8294 w8284 w8274 w8264 w8254 w8244 w8234 w8224 w8214 w8204
 w8194 w8184 
w8174 w8164 w8154 w8144 w8134 w8124 w8114 w8104 w8094 w8084 w8074
 w8064 w8054 
w8044 w8034 w8024 w8014 w8004 w7994 w7984 w7974 w7964 w7954 w7944
 w7934 w7924 
w7914 w7904 w7894 w7884 w7874 w7864 w7854 w7844 w7834 w7824 w7814
 w7804 w7794 
w7784 w7774 w7764 w7754 w7744 w7734 w7724 w7714 w7704 w7694 w7684
 w7674 w7664 
w7654 w7644 w7634 w7624 w7614 w7604 w7594 w7584 w7574 w7564 w7554
 w7544 w7534 
w7524 w7514 w7504 w7494 w7484 w7474 w7464 w7454 w7444 w7434 w7424
 w7414 w7404 
  
 
190 
 
w7394 w7384 w7374 w7364 w7354 w7344 w7334 w7324 w7314 w7304 w7294
 w7284 w7274 
w7264 w7254 w7244 w7234 w7224 w7214 w7204 w7194 w7184 w7174 w7164
 w7154 w7144 
w7134 w7124 w7114 w7104 w7094 w7084 w7074 w7064 w7054 w7044 w7034
 w7024 w7014 
w7004 w6994 w6984 w6974 w6964 w6954 w6944 w6934 w6924 w6914 w6904
 w6894 w6884 
w6874 w6864 w6854 w6844 w6834 w6824 w6814 w6804 w6794 w6784 w6774
 w6764 w6754 
w6744 w6734 w6724 w6714 w6704 w6694 w6684 w6674 w6664 w6654 w6644
 w6634 w6624 
w6614 w6604 w6594 w6584 w6574 w6564 w6554 w6544 w6534 w6524 w6514
 w6504 w6494 
w6484 w6474 w6464 w6454 w6444 w6434 w6424 w6414 w6404 w6394 w6384
 w6374 w6364 
w6354 w6344 w6334 w6324 w6314 w6304 w6294 w6284 w6274 w6264 w6254
 w6244 w6234 
w6224 w6214 w6204 w6194 w6184 w6174 w6164 w6154 w6144 w6134 w6124
 w6114 w6104 
w6094 w6084 w6074 w6064 w6054 w6044 w6034 w6024 w6014 w6004 w5994
 w5984 w5974 
w5964 w5954 w5944 w5934 w5924 w5914 w5904 w5894 w5884 w5874 w5864
 w5854 w5844 
w5834 w5824 w5814 w5804 w5794 w5784 w5774 w5764 w5754 w5744 w5734
 w5724 w5714 
w5704 w5694 w5684 w5674 w5664 w5654 w5644 w5634 w5624 w5614 w5604
 w5594 w5584 
w5574 w5564 w5554 w5544 w5534 w5524 w5514 w5504 w5494 w5484 w5474
 w5464 w5454 
w5444 w5434 w5424 w5414 w5404 w5394 w5384 w5374 w5364 w5354 w5344
 w5334 w5324 
w5314 w5304 w5294 w5284 w5274 w5264 w5254 w5244 w5234 w5224 w5214
 w5204 w5194 
w5184 w5174 w5164 w5154 w5144 w5134 w5124 w5114 w5104 w5094 w5084
 w5074 w5064 
w5054 w5044 w5034 w5024 w5014 w5004 w4994 w4984 w4974 w4964 w4954
 w4944 w4934 
w4924 w4914 w4904 w4894 w4884 w4874 w4864 w4854 w4844 w4834 w4824
 w4814 w4804 
w4794 w4784 w4774 w4764 w4754 w4744 w4734 w4724 w4714 w4704 w4694
 w4684 w4674 
w4664 w4654 w4644 w4634 w4624 w4614 w4604 w4594 w4584 w4574 w4564
 w4554 w4544 
w4534 w4524 w4514 w4504 w4494 w4484 w4474 w4464 w4454 w4444 w4434
 w4424 w4414 
w4404 w4394 w4384 w4374 w4364 w4354 w4344 w4334 w4324 w4314 w4304
 w4294 w4284 
w4274 w4264 w4254 w4244 w4234 w4224 w4214 w4204 w4194 w4184 w4174
 w4164 w4154 
w4144 w4134 w4124 w4114 w4104 w4094 w4084 w4074 w4064 w4054 w4044
 w4034 w4024 
w4014 w4004; 
run; 
 
 
proc reg data=prinvars outest=est press; 
  
 
191 
 
model MIT = prin1-prin10 / selection=stepwise vif p; 
model TORV= prin1-prin10 / selection=stepwise vif p; 
model TMML= prin1-prin10 / selection=stepwise vif p; 
model TOXY= prin1-prin10 / selection=stepwise vif p; 
model TRE= prin1-prin10 / selection=stepwise vif p; 
model TME= prin1-prin10 / selection=stepwise vif p; 
run; 
 
proc print data=prinvars; 
var prin1-prin10; 
run; 
ods html close; 
 
Table D.1 ANOVA results and model parameter estimates for MIT (raw data 
model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 10 6263.135 626.3135 322.8 <.0001  
Error 19 36.86515 1.94027    
Corrected Total 29 6300     
       
Root MSE 1.39294 R-Square 0.9941    
Dependent Mean 290 Adj R-Sq 0.9911    
Coeff Var 0.48032          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 290 0.25431 1140.32 <.0001 0 
Prin1 1 5.19486 0.25866 20.08 <.0001 1 
Prin2 1 -1.61427 0.25866 -6.24 <.0001 1 
Prin3 1 -8.44735 0.25866 -32.66 <.0001 1 
Prin4 1 5.2108 0.25866 20.15 <.0001 1 
Prin5 1 3.58786 0.25866 13.87 <.0001 1 
Prin6 1 1.8311 0.25866 7.08 <.0001 1 
Prin7 1 -4.84083 0.25866 -18.71 <.0001 1 
Prin8 1 -2.07599 0.25866 -8.03 <.0001 1 
Prin9 1 6.04782 0.25866 23.38 <.0001 1 
Prin10 1 -2.706 0.25866 -10.46 <.0001 1 
 
  
 
192 
 
Table D.2 ANOVA results and model parameter estimates for TORV (raw data 
model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 7 3415.039 487.8627 187.86 <.0001  
Error 22 57.13287 2.59695    
Corrected Total 29 3472.172     
       
Root MSE 1.61151 R-Square 0.9835    
Dependent Mean 283.8253 Adj R-Sq 0.9783    
Coeff Var 0.56778          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 283.8253 0.29422 964.67 <.0001 0 
Prin2 1 1.42119 0.29925 4.75 <.0001 1 
Prin3 1 -6.39864 0.29925 -21.38 <.0001 1 
Prin4 1 4.151 0.29925 13.87 <.0001 1 
Prin5 1 -4.34895 0.29925 -14.53 <.0001 1 
Prin8 1 -0.51608 0.29925 -1.72 0.0986 1 
Prin9 1 4.8256 0.29925 16.13 <.0001 1 
Prin10 1 -3.88595 0.29925 -12.99 <.0001 1 
 
 
 
 
 
 
 
 
  
 
193 
 
 
 
Table D.3 ANOVA results and model parameter estimates for TMML (raw data 
model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 6197.26 774.6576 67.75 <.0001  
Error 21 240.1086 11.43374    
Corrected Total 29 6437.369     
       
Root MSE 3.38138 R-Square 0.9627    
Dependent Mean 325.599 Adj R-Sq 0.9485    
Coeff Var 1.03851          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 325.599 0.61735 527.41 <.0001 0 
Prin2 1 -1.92959 0.62791 -3.07 0.0058 1 
Prin3 1 -5.95946 0.62791 -9.49 <.0001 1 
Prin4 1 6.82824 0.62791 10.87 <.0001 1 
Prin5 1 -5.83541 0.62791 -9.29 <.0001 1 
Prin6 1 2.47788 0.62791 3.95 0.0007 1 
Prin7 1 -7.32868 0.62791 -11.67 <.0001 1 
Prin9 1 5.5779 0.62791 8.88 <.0001 1 
Prin10 1 -1.67951 0.62791 -2.67 0.0142 1 
 
 
 
 
 
 
 
  
 
194 
 
 
 
Table D.4 ANOVA results and model parameter estimates for TOXY (raw data 
model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 6134.713 766.8392 7.35 0.0001  
Error 21 2190.255 104.2979    
Corrected Total 29 8324.969     
       
Root MSE 10.21263 R-Square 0.7369    
Dependent Mean 292.497 Adj R-Sq 0.6367    
Coeff Var 3.49153          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 292.497 1.86456 156.87 <.0001 0 
Prin1 1 3.27558 1.89644 1.73 0.0988 1 
Prin2 1 -3.8124 1.89644 -2.01 0.0574 1 
Prin3 1 -6.48977 1.89644 -3.42 0.0026 1 
Prin4 1 5.90992 1.89644 3.12 0.0052 1 
Prin7 1 -4.884 1.89644 -2.58 0.0176 1 
Prin8 1 -3.86455 1.89644 -2.04 0.0544 1 
Prin9 1 6.83182 1.89644 3.6 0.0017 1 
Prin10 1 -4.87563 1.89644 -2.57 0.0178 1 
 
 
 
 
 
 
 
  
 
195 
 
 
 
Table D.5 ANOVA results and model parameter estimates for TRE (raw data 
model). 
       
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 4107.185 513.3981 35.2 <.0001  
Error 21 306.2536 14.58351    
Corrected Total 29 4413.438     
       
Root MSE 3.81884 R-Square 0.9306    
Dependent Mean 239.0813 Adj R-Sq 0.9042    
Coeff Var 1.5973          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 239.0813 0.69722 342.91 <.0001 0 
Prin1 1 5.30679 0.70914 7.48 <.0001 1 
Prin2 1 -2.58712 0.70914 -3.65 0.0015 1 
Prin3 1 -6.34311 0.70914 -8.94 <.0001 1 
Prin5 1 7.30407 0.70914 10.3 <.0001 1 
Prin6 1 -2.03127 0.70914 -2.86 0.0093 1 
Prin7 1 -2.03972 0.70914 -2.88 0.009 1 
Prin8 1 -1.93073 0.70914 -2.72 0.0128 1 
Prin10 1 -1.08312 0.70914 -1.53 0.1416 1 
 
 
 
 
 
 
 
  
 
196 
 
 
 
Table D.6 ANOVA results and model parameter estimates for TME (raw data 
model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 7 6030.005 861.4292 28.49 <.0001  
Error 22 665.2383 30.2381    
Corrected Total 29 6695.243     
       
Root MSE 5.49892 R-Square 0.9006    
Dependent Mean 387.7237 Adj R-Sq 0.869    
Coeff Var 1.41826          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 387.7237 1.00396 386.19 <.0001 0 
Prin1 1 -5.66085 1.02112 -5.54 <.0001 1 
Prin2 1 -3.26223 1.02112 -3.19 0.0042 1 
Prin3 1 -5.15126 1.02112 -5.04 <.0001 1 
Prin4 1 4.5388 1.02112 4.44 0.0002 1 
Prin5 1 -8.74419 1.02112 -8.56 <.0001 1 
Prin6 1 5.56589 1.02112 5.45 <.0001 1 
Prin7 1 3.26613 1.02112 3.2 0.0041 1 
 
 
 
  
 
197 
 
Table D.7 Principal components (PC) obtained from raw NIR spectral data of 
biomass dusts used for PCA and internal validation of prediction models. 
 
 
 
 
 
 
 
 
 
 
 
B i o m a ss Pr i n 1 Pr i n 2 Pr i n 3 Pr i n 4 Pr i n 5 Pr i n 6 Pr i n 7 Pr i n 8 Pr i n 9 Pr i n 1 0
- 1 . 1 3 5 1 - 0 . 3 2 2 8 7 0 . 2 3 0 6 4 - 1 . 1 0 2 6 4 0 . 9 2 6 0 3 0 . 4 5 8 0 7 - 1 . 3 1 8 0 3 1 . 0 3 6 3 3 - 1 . 7 6 5 2 8 - 0 . 6 1 4 4 5
- 0 . 2 5 2 3 8 - 0 . 0 5 0 8 2 - 0 . 0 2 1 0 8 - 1 . 3 0 7 9 3 0 . 6 2 8 3 8 0 . 3 2 9 9 - 0 . 2 9 6 8 4 2 . 0 1 6 1 2 - 1 . 1 2 9 8 1 1 . 1 0 6 9 8
- 1 . 1 5 3 8 4 - 0 . 5 5 5 0 6 0 . 2 4 5 6 7 - 1 . 3 4 3 1 2 0 . 8 8 0 0 7 0 . 3 7 1 1 8 - 1 . 1 4 4 3 5 1 . 3 9 1 4 1 - 1 . 3 9 2 9 5 - 0 . 4 4 6 1 4
- 0 . 7 0 1 2 1 0 . 2 6 0 8 - 0 . 6 0 2 8 2 - 1 . 2 0 9 9 2 0 . 0 0 3 4 5 0 . 5 9 1 3 5 0 . 8 1 3 9 3 - 1 . 7 0 2 5 4 - 0 . 3 8 9 1 8 1 . 4 9 3 2 5
0 . 1 4 7 6 3 0 . 4 4 8 3 1 - 0 . 8 0 9 7 3 - 1 . 3 0 5 1 1 - 0 . 4 8 2 6 5 0 . 9 4 8 9 2 0 . 5 7 2 8 5 - 0 . 1 3 2 8 4 0 . 4 5 1 9 4 2 . 8 4 9 3 3
- 0 . 5 1 7 9 6 0 . 0 1 1 2 6 - 0 . 3 9 7 4 8 - 1 . 6 8 1 1 9 - 0 . 3 4 7 4 2 0 . 2 4 5 9 8 0 . 1 8 2 1 7 - 2 . 7 0 3 5 9 - 0 . 8 7 6 5 1 0 . 2 4 7 0 9
1 . 2 4 4 9 3 - 1 . 3 9 2 4 3 0 . 6 2 0 4 0 . 8 0 3 8 1 0 . 1 9 6 1 3 1 . 1 6 0 5 7 1 . 2 0 1 5 7 0 . 4 2 1 1 3 - 0 . 8 6 6 8 8 - 0 . 5 9 1 8 5
1 . 3 4 6 8 2 - 1 . 1 4 5 4 6 0 . 4 1 9 0 3 0 . 8 0 1 7 2 0 . 3 7 4 0 3 1 . 1 9 7 8 4 1 . 8 6 8 2 2 1 . 5 4 3 7 1 0 . 4 4 0 0 3 1 . 3 9 4 8 8
0 . 3 9 2 6 4 - 1 . 6 6 8 2 0 . 7 5 5 3 8 0 . 6 5 2 5 8 0 . 5 5 0 5 2 0 . 9 6 4 2 8 0 . 9 5 9 9 9 - 0 . 4 6 6 2 3 - 0 . 8 0 1 8 - 1 . 2 7 2 1 1
0 . 5 7 2 9 1 0 . 1 3 7 7 2 - 1 . 1 2 1 5 2 - 0 . 6 9 3 1 3 - 0 . 6 6 3 5 3 - 2 . 1 0 3 4 5 1 . 3 5 7 7 4 0 . 7 8 3 2 5 - 0 . 0 7 7 6 7 - 0 . 4 7 5 1 9
- 0 . 3 3 9 0 7 - 0 . 3 5 1 9 2 - 0 . 7 9 3 0 5 - 0 . 7 8 9 8 - 0 . 1 0 6 8 7 - 2 . 0 1 6 4 3 1 . 1 8 4 4 1 0 . 8 5 0 4 6 0 . 6 3 6 6 7 - 0 . 7 7 9 1
0 . 1 8 7 0 2 - 0 . 1 7 7 5 2 - 0 . 8 7 1 4 2 - 0 . 7 4 7 1 - 0 . 4 6 5 5 4 - 2 . 0 6 8 2 3 1 . 1 9 4 6 0 . 5 2 9 9 9 0 . 0 1 6 0 3 - 0 . 9 8 0 2 1
- 1 . 1 2 3 8 7 0 . 6 1 5 0 8 1 . 4 1 1 2 7 0 . 6 0 4 4 6 - 2 . 0 9 7 7 2 0 . 1 4 6 0 1 - 0 . 0 9 2 1 1 0 . 2 6 2 0 4 - 0 . 3 5 4 2 5 - 0 . 3 6 4 4 9
- 1 . 1 5 1 7 7 0 . 6 5 9 5 8 1 . 3 6 6 2 7 0 . 7 0 1 9 7 - 2 . 1 0 2 9 9 0 . 2 0 5 4 3 0 . 1 9 5 2 2 0 . 4 1 1 8 6 - 0 . 1 6 4 0 8 - 0 . 1 0 5 2 5
- 1 . 4 3 4 5 0 . 6 2 5 6 7 1 . 4 1 8 3 0 . 6 3 8 1 4 - 1 . 9 5 3 3 7 0 . 0 3 5 6 4 0 . 5 8 1 7 1 0 . 2 7 8 8 7 0 . 3 3 5 7 2 0 . 3 9 2 3 2
0 . 4 8 0 1 4 1 . 0 5 1 2 7 - 1 . 5 3 2 2 5 0 . 2 7 3 4 9 - 0 . 3 7 3 8 1 1 . 6 7 6 2 9 - 0 . 2 2 2 5 2 - 0 . 0 6 7 3 3 0 . 2 0 9 1 5 - 1 . 7 5 0 8 6
0 . 3 1 8 1 7 0 . 8 8 8 6 3 - 1 . 5 3 9 1 4 - 0 . 1 5 8 4 7 - 0 . 4 1 1 6 9 1 . 5 5 8 7 - 0 . 2 5 9 8 7 0 . 2 5 7 5 9 1 . 2 0 0 3 8 - 1 . 1 1 8 1 3
- 0 . 8 3 4 0 9 0 . 4 3 1 3 2 - 1 . 3 4 7 9 3 - 0 . 0 9 8 0 2 0 . 2 6 4 0 7 1 . 5 5 0 8 5 - 0 . 1 5 7 4 6 0 . 4 6 4 9 6 2 . 1 2 0 1 6 - 1 . 0 6 6 6 3
0 . 4 9 0 1 7 1 . 6 1 2 7 8 1 . 4 5 7 5 4 - 0 . 1 9 6 5 8 1 . 5 3 5 5 3 - 0 . 3 0 9 9 8 0 . 1 2 7 9 - 0 . 1 4 4 2 0 . 7 2 1 9 3 - 0 . 3 7 4 4
1 . 8 9 9 7 4 2 . 2 4 8 9 1 1 . 3 5 4 3 1 - 0 . 3 3 4 5 9 0 . 8 1 3 4 4 - 0 . 3 5 2 1 1 0 . 0 8 6 0 3 - 0 . 4 2 1 1 1 - 0 . 3 6 3 7 8 - 0 . 0 9 5 2
0 . 0 5 6 5 6 1 . 4 6 1 3 1 . 7 3 5 8 1 - 0 . 1 7 3 0 7 1 . 8 9 4 1 8 - 0 . 4 3 1 3 4 0 . 1 4 3 3 1 - 0 . 3 4 8 5 8 0 . 9 7 9 2 9 - 0 . 2 6 1 1 6
0 . 3 6 9 6 3 - 1 . 2 3 3 5 9 0 . 4 3 6 5 2 - 0 . 0 4 1 4 3 - 0 . 0 0 8 5 2 - 0 . 4 7 5 7 2 - 1 . 2 6 5 5 3 0 . 2 8 6 3 2 2 . 1 5 2 2 1 1 . 4 0 1 0 4
1 . 2 2 0 7 8 - 1 . 0 7 4 2 8 0 . 3 9 3 6 2 - 0 . 5 2 8 7 5 - 0 . 9 6 1 0 3 - 0 . 6 4 4 7 - 2 . 5 6 8 6 2 - 0 . 3 9 0 0 1 0 . 4 3 9 1 1 - 0 . 6 2 0 1 5
1 . 4 4 2 3 5 - 0 . 9 3 3 4 5 0 . 2 9 2 7 6 - 0 . 1 8 3 0 1 - 0 . 8 2 6 2 6 - 0 . 4 8 6 9 5 - 1 . 9 1 6 3 3 - 0 . 1 0 8 5 6 0 . 8 9 4 2 1 0 . 4 6 6 8 5
- 0 . 3 2 2 2 7 0 . 5 1 2 3 1 - 1 . 1 0 5 1 3 1 . 8 1 8 6 5 0 . 5 3 3 3 1 - 0 . 8 2 6 3 3 - 0 . 5 6 8 5 3 - 0 . 1 9 8 - 0 . 7 0 7 0 2 0 . 7 6 5 6 2
- 0 . 2 4 2 9 2 0 . 6 6 2 4 1 - 1 . 1 4 9 7 9 2 . 1 2 0 3 3 0 . 5 0 4 6 6 - 0 . 6 2 6 7 8 - 1 . 0 1 9 9 1 - 0 . 2 8 2 4 9 - 1 . 1 5 7 6 5 0 . 4 6 1 4
- 0 . 7 9 1 7 7 0 . 3 9 5 0 1 - 1 . 0 8 4 6 9 2 . 1 1 8 2 7 0 . 9 3 9 8 7 - 0 . 6 6 2 2 9 - 0 . 4 2 1 4 8 0 . 2 9 8 6 3 - 0 . 1 8 7 9 5 1 . 0 9 4 3 5
1 . 2 4 9 2 7 - 0 . 4 4 3 2 1 - 0 . 2 3 7 6 8 0 . 2 2 9 0 8 - 0 . 6 0 9 8 5 - 0 . 1 7 1 3 7 0 . 2 0 9 5 - 1 . 7 5 1 7 1 - 1 . 3 3 5 5 7 - 0 . 4 4 8 2 9
- 2 . 2 8 1 9 1 - 1 . 9 8 9 9 6 0 . 5 8 1 3 3 0 . 4 1 4 4 2 1 . 5 5 2 3 1 - 0 . 4 3 2 4 5 0 . 5 3 1 8 8 - 1 . 6 1 5 0 3 1 . 4 6 8 7 5 - 0 . 5 4 2 6 2
0 . 8 6 3 9 - 0 . 6 8 3 5 6 - 0 . 1 0 5 1 1 0 . 7 1 6 9 6 - 0 . 1 8 4 7 4 0 . 1 6 7 1 1 0 . 0 4 0 5 6 - 0 . 5 0 0 4 6 - 0 . 4 9 5 2 2 0 . 2 3 3 1
P o u l tr y  L i tt e r
S u g a r c a n e  b a g a s s e
S w e e tg u m
S w i tc h g r a s s
B e r mu d a  g r a s s
Co r n  Co b s
Co r n  s tov e r
E u c a l y p tu s
P e c a n  s h e l l
P i n e
  
 
198 
 
Table D.8 Result of Tukey test on principal component values of different biomass 
dust samples for PC1. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
A 1.0109 3 Sugarcane bagasse 
A       
A 0.9948 3 Corn stover 
A       
A 0.8155 3 Poultry Litter 
A       
A 0.1403 3 Eucalyptus 
A       
A -0.0119 3 Pine 
A       
A -0.0562 3 Switchgrass 
A       
A -0.3572 3 Corn Cobs 
A       
A -0.4523 3 Sweetgum 
A       
A -0.8471 3 Bermuda grass 
A       
A -1.2367 3 Pecan shell 
 
 
 
 
 
 
  
 
199 
 
 
 
Table D.9 Result of Tukey test on principal component values of different biomass 
dust samples for PC2. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A   1.7743 3 Poultry Litter 
  A         
B A   0.7904 3 Pine 
B           
B C   0.6334 3 Pecan shell 
B C         
B C   0.5232 3 Sweetgum 
B C         
B C   0.2401 3 Corn Cobs 
B C         
B C D -0.1306 3 Eucalyptus 
  C D       
  C D -0.3096 3 Bermuda grass 
    D       
  E D -1.0389 3 Switchgrass 
  E D       
  E D -1.0804 3 Sugarcane bagasse 
  E         
  E   -1.4020 3 Corn stover 
 
 
 
 
  
 
200 
 
Table D.10 Result of Tukey test on principal component values of different 
biomass dust samples for PC3. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A 1.5159 3 Poultry Litter 
  A       
  A 1.3986 3 Pecan shell 
          
  B 0.5983 3 Corn stover 
  B       
  B 0.3743 3 Sugarcane bagasse 
  B       
  B 0.1517 3 Bermuda grass 
  B       
  B 0.0795 3 Switchgrass 
          
  C -0.6033 3 Corn Cobs 
  C       
D C -0.9287 3 Eucalyptus 
D C       
D C -1.1132 3 Sweetgum 
D         
D   -1.4731 3 Pine 
 
 
 
 
 
 
  
 
201 
 
Table D.11 Result of Tukey test on principal component values of different 
biomass dust samples for PC4. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A 2.0191 3 Sweetgum 
          
  B 0.7527 3 Corn stover 
  B       
  B 0.6482 3 Pecan shell 
  B       
C B 0.4535 3 Switchgrass 
C         
C D 0.0057 3 Pine 
  D       
  D -0.2347 3 Poultry Litter 
  D       
E D -0.2511 3 Sugarcane bagasse 
E         
E   -0.7433 3 Eucalyptus 
          
  F -1.2512 3 Bermuda grass 
  F       
  F -1.3987 3 Corn Cobs 
 
 
 
 
 
 
  
 
202 
 
Table D.12 Result of Tukey test on principal component values of different 
biomass dust samples for PC5. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A   1.4144 3 Poultry Litter 
  A         
B A   0.8115 3 Bermuda grass 
B A         
B A C 0.6593 3 Sweetgum 
B A C       
B A C 0.3736 3 Corn stover 
B A C       
B A C 0.2526 3 Switchgrass 
B   C       
B   C -0.1738 3 Pine 
B   C       
B   C -0.2755 3 Corn Cobs 
B   C       
B   C -0.4120 3 Eucalyptus 
    C       
    C -0.5986 3 Sugarcane bagasse 
            
  D   -2.0514 3 Pecan shell 
 
 
 
 
 
 
  
 
203 
 
Table D.13 Result of Tukey test on principal component values of different 
biomass dust samples for PC6. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A 1.5953 3 Pine 
          
  B 1.1076 3 Corn stover 
          
  C 0.5954 3 Corn Cobs 
  C       
  C 0.3864 3 Bermuda grass 
  C       
D C 0.1290 3 Pecan shell 
D         
D E -0.1456 3 Switchgrass 
  E       
F E -0.3645 3 Poultry Litter 
F E       
F E -0.5358 3 Sugarcane bagasse 
F         
F   -0.7051 3 Sweetgum 
          
  G -2.0627 3 Eucalyptus 
 
 
 
 
 
 
  
 
204 
 
Table D.14 Result of Tukey test on principal component values of different 
biomass dust samples for PC9. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A 1.1766 3 Pine 
  A       
  A 1.1618 3 Sugarcane bagasse 
  A       
B A 0.4458 3 Poultry Litter 
B A       
B A 0.1917 3 Eucalyptus 
B A       
B A -0.0609 3 Pecan shell 
B A       
B A -0.1207 3 Switchgrass 
B A       
B A -0.2713 3 Corn Cobs 
B A       
B A -0.4096 3 Corn stover 
B A       
B A -0.6842 3 Sweetgum 
B         
B   -1.4293 3 Bermuda grass 
 
 
 
 
 
 
  
 
205 
 
Table D.15 Result of Tukey test on principal component values of different 
biomass dust samples for PC10. 
Means with the same letter are not significantly 
different. 
Tukey Grouping Mean N Biomass 
  A 1.5299 3 Corn Cobs 
  A       
B A 0.7738 3 Sweetgum 
B A       
B A 0.4159 3 Sugarcane bagasse 
B A       
B A 0.0155 3 Bermuda grass 
B A       
B A -0.0258 3 Pecan shell 
B A       
B A -0.1564 3 Corn stover 
B A       
B A -0.2436 3 Poultry Litter 
B A       
B A -0.2526 3 Switchgrass 
B A       
B A -0.7448 3 Eucalyptus 
B         
B   -1.3119 3 Pine 
 
 
 
 
 
 
  
 
206 
 
SAS code for PCA on first derivative NIR spectral data and internal validation of 
prediction models 
 
/******************************** 
 * Author: Jaskaran Dhiman      * 
 * Research work                * 
 * Date: 01/31/2014             * 
 ********************************/ 
 
ods html file='C:\Users\jzd0028\Desktop\Research\NIR new 
analysis\output_firstderivative.html'; 
options nodate pageno=1; 
data work.jasnir; 
infile 'C:\Users\jzd0028\Desktop\Research\NIR new 
analysis\firstderivative_data_SAS.csv' delimiter = ',' MISSOVER DSD 
lrecl=32767 firstobs=1 n=1500; 
input Number$ Biomass$ MIT TORV TMML TOXY TRE TME  
w9994 w9984 w9974 w9964 w9954 w9944 w9934 w9924 w9914 w9904 w9894
 w9884 w9874  
w9864 w9854 w9844 w9834 w9824 w9814 w9804 w9794 w9784 w9774 w9764
 w9754 w9744 
w9734 w9724 w9714 w9704 w9694 w9684 w9674 w9664 w9654 w9644 w9634
 w9624 w9614 
w9604 w9594 w9584 w9574 w9564 w9554 w9544 w9534 w9524 w9514 w9504
 w9494 w9484 
w9474 w9464 w9454 w9444 w9434 w9424 w9414 w9404 w9394 w9384 w9374
 w9364 w9354 
w9344 w9334 w9324 w9314 w9304 w9294 w9284 w9274 w9264 w9254 w9244
 w9234 w9224 
w9214 w9204 w9194 w9184 w9174 w9164 w9154 w9144 w9134 w9124 w9114
 w9104 w9094 
w9084 w9074 w9064 w9054 w9044 w9034 w9024 w9014 w9004 w8994 w8984
 w8974 w8964 
w8954 w8944 w8934 w8924 w8914 w8904 w8894 w8884 w8874 w8864 w8854
 w8844 w8834 
w8824 w8814 w8804 w8794 w8784 w8774 w8764 w8754 w8744 w8734 w8724
 w8714 w8704 
w8694 w8684 w8674 w8664 w8654 w8644 w8634 w8624 w8614 w8604 w8594
 w8584 w8574 
w8564 w8554 w8544 w8534 w8524 w8514 w8504 w8494 w8484 w8474 w8464
 w8454 w8444 
w8434 w8424 w8414 w8404 w8394 w8384 w8374 w8364 w8354 w8344 w8334
 w8324 w8314 
w8304 w8294 w8284 w8274 w8264 w8254 w8244 w8234 w8224 w8214 w8204
 w8194 w8184 
w8174 w8164 w8154 w8144 w8134 w8124 w8114 w8104 w8094 w8084 w8074
 w8064 w8054 
w8044 w8034 w8024 w8014 w8004 w7994 w7984 w7974 w7964 w7954 w7944
 w7934 w7924 
w7914 w7904 w7894 w7884 w7874 w7864 w7854 w7844 w7834 w7824 w7814
 w7804 w7794 
w7784 w7774 w7764 w7754 w7744 w7734 w7724 w7714 w7704 w7694 w7684
 w7674 w7664 
  
 
207 
 
w7654 w7644 w7634 w7624 w7614 w7604 w7594 w7584 w7574 w7564 w7554
 w7544 w7534 
w7524 w7514 w7504 w7494 w7484 w7474 w7464 w7454 w7444 w7434 w7424
 w7414 w7404 
w7394 w7384 w7374 w7364 w7354 w7344 w7334 w7324 w7314 w7304 w7294
 w7284 w7274 
w7264 w7254 w7244 w7234 w7224 w7214 w7204 w7194 w7184 w7174 w7164
 w7154 w7144 
w7134 w7124 w7114 w7104 w7094 w7084 w7074 w7064 w7054 w7044 w7034
 w7024 w7014 
w7004 w6994 w6984 w6974 w6964 w6954 w6944 w6934 w6924 w6914 w6904
 w6894 w6884 
w6874 w6864 w6854 w6844 w6834 w6824 w6814 w6804 w6794 w6784 w6774
 w6764 w6754 
w6744 w6734 w6724 w6714 w6704 w6694 w6684 w6674 w6664 w6654 w6644
 w6634 w6624 
w6614 w6604 w6594 w6584 w6574 w6564 w6554 w6544 w6534 w6524 w6514
 w6504 w6494 
w6484 w6474 w6464 w6454 w6444 w6434 w6424 w6414 w6404 w6394 w6384
 w6374 w6364 
w6354 w6344 w6334 w6324 w6314 w6304 w6294 w6284 w6274 w6264 w6254
 w6244 w6234 
w6224 w6214 w6204 w6194 w6184 w6174 w6164 w6154 w6144 w6134 w6124
 w6114 w6104 
w6094 w6084 w6074 w6064 w6054 w6044 w6034 w6024 w6014 w6004 w5994
 w5984 w5974 
w5964 w5954 w5944 w5934 w5924 w5914 w5904 w5894 w5884 w5874 w5864
 w5854 w5844 
w5834 w5824 w5814 w5804 w5794 w5784 w5774 w5764 w5754 w5744 w5734
 w5724 w5714 
w5704 w5694 w5684 w5674 w5664 w5654 w5644 w5634 w5624 w5614 w5604
 w5594 w5584 
w5574 w5564 w5554 w5544 w5534 w5524 w5514 w5504 w5494 w5484 w5474
 w5464 w5454 
w5444 w5434 w5424 w5414 w5404 w5394 w5384 w5374 w5364 w5354 w5344
 w5334 w5324 
w5314 w5304 w5294 w5284 w5274 w5264 w5254 w5244 w5234 w5224 w5214
 w5204 w5194 
w5184 w5174 w5164 w5154 w5144 w5134 w5124 w5114 w5104 w5094 w5084
 w5074 w5064 
w5054 w5044 w5034 w5024 w5014 w5004 w4994 w4984 w4974 w4964 w4954
 w4944 w4934 
w4924 w4914 w4904 w4894 w4884 w4874 w4864 w4854 w4844 w4834 w4824
 w4814 w4804 
w4794 w4784 w4774 w4764 w4754 w4744 w4734 w4724 w4714 w4704 w4694
 w4684 w4674 
w4664 w4654 w4644 w4634 w4624 w4614 w4604 w4594 w4584 w4574 w4564
 w4554 w4544 
w4534 w4524 w4514 w4504 w4494 w4484 w4474 w4464 w4454 w4444 w4434
 w4424 w4414 
w4404 w4394 w4384 w4374 w4364 w4354 w4344 w4334 w4324 w4314 w4304
 w4294 w4284 
w4274 w4264 w4254 w4244 w4234 w4224 w4214 w4204 w4194 w4184 w4174
 w4164 w4154 
w4144 w4134 w4124 w4114 w4104 w4094 w4084 w4074 w4064 w4054 w4044
 w4034 w4024 
w4014; 
  
 
208 
 
datalines; 
; 
run; 
 
proc princomp data=work.jasnir out=prinvars std; 
var w9994 w9984 w9974 w9964 w9954 w9944 w9934 w9924 w9914 w9904
 w9894 w9884 w9874  
w9864 w9854 w9844 w9834 w9824 w9814 w9804 w9794 w9784 w9774 w9764
 w9754 w9744 
w9734 w9724 w9714 w9704 w9694 w9684 w9674 w9664 w9654 w9644 w9634
 w9624 w9614 
w9604 w9594 w9584 w9574 w9564 w9554 w9544 w9534 w9524 w9514 w9504
 w9494 w9484 
w9474 w9464 w9454 w9444 w9434 w9424 w9414 w9404 w9394 w9384 w9374
 w9364 w9354 
w9344 w9334 w9324 w9314 w9304 w9294 w9284 w9274 w9264 w9254 w9244
 w9234 w9224 
w9214 w9204 w9194 w9184 w9174 w9164 w9154 w9144 w9134 w9124 w9114
 w9104 w9094 
w9084 w9074 w9064 w9054 w9044 w9034 w9024 w9014 w9004 w8994 w8984
 w8974 w8964 
w8954 w8944 w8934 w8924 w8914 w8904 w8894 w8884 w8874 w8864 w8854
 w8844 w8834 
w8824 w8814 w8804 w8794 w8784 w8774 w8764 w8754 w8744 w8734 w8724
 w8714 w8704 
w8694 w8684 w8674 w8664 w8654 w8644 w8634 w8624 w8614 w8604 w8594
 w8584 w8574 
w8564 w8554 w8544 w8534 w8524 w8514 w8504 w8494 w8484 w8474 w8464
 w8454 w8444 
w8434 w8424 w8414 w8404 w8394 w8384 w8374 w8364 w8354 w8344 w8334
 w8324 w8314 
w8304 w8294 w8284 w8274 w8264 w8254 w8244 w8234 w8224 w8214 w8204
 w8194 w8184 
w8174 w8164 w8154 w8144 w8134 w8124 w8114 w8104 w8094 w8084 w8074
 w8064 w8054 
w8044 w8034 w8024 w8014 w8004 w7994 w7984 w7974 w7964 w7954 w7944
 w7934 w7924 
w7914 w7904 w7894 w7884 w7874 w7864 w7854 w7844 w7834 w7824 w7814
 w7804 w7794 
w7784 w7774 w7764 w7754 w7744 w7734 w7724 w7714 w7704 w7694 w7684
 w7674 w7664 
w7654 w7644 w7634 w7624 w7614 w7604 w7594 w7584 w7574 w7564 w7554
 w7544 w7534 
w7524 w7514 w7504 w7494 w7484 w7474 w7464 w7454 w7444 w7434 w7424
 w7414 w7404 
w7394 w7384 w7374 w7364 w7354 w7344 w7334 w7324 w7314 w7304 w7294
 w7284 w7274 
w7264 w7254 w7244 w7234 w7224 w7214 w7204 w7194 w7184 w7174 w7164
 w7154 w7144 
w7134 w7124 w7114 w7104 w7094 w7084 w7074 w7064 w7054 w7044 w7034
 w7024 w7014 
w7004 w6994 w6984 w6974 w6964 w6954 w6944 w6934 w6924 w6914 w6904
 w6894 w6884 
w6874 w6864 w6854 w6844 w6834 w6824 w6814 w6804 w6794 w6784 w6774
 w6764 w6754 
w6744 w6734 w6724 w6714 w6704 w6694 w6684 w6674 w6664 w6654 w6644
 w6634 w6624 
  
 
209 
 
w6614 w6604 w6594 w6584 w6574 w6564 w6554 w6544 w6534 w6524 w6514
 w6504 w6494 
w6484 w6474 w6464 w6454 w6444 w6434 w6424 w6414 w6404 w6394 w6384
 w6374 w6364 
w6354 w6344 w6334 w6324 w6314 w6304 w6294 w6284 w6274 w6264 w6254
 w6244 w6234 
w6224 w6214 w6204 w6194 w6184 w6174 w6164 w6154 w6144 w6134 w6124
 w6114 w6104 
w6094 w6084 w6074 w6064 w6054 w6044 w6034 w6024 w6014 w6004 w5994
 w5984 w5974 
w5964 w5954 w5944 w5934 w5924 w5914 w5904 w5894 w5884 w5874 w5864
 w5854 w5844 
w5834 w5824 w5814 w5804 w5794 w5784 w5774 w5764 w5754 w5744 w5734
 w5724 w5714 
w5704 w5694 w5684 w5674 w5664 w5654 w5644 w5634 w5624 w5614 w5604
 w5594 w5584 
w5574 w5564 w5554 w5544 w5534 w5524 w5514 w5504 w5494 w5484 w5474
 w5464 w5454 
w5444 w5434 w5424 w5414 w5404 w5394 w5384 w5374 w5364 w5354 w5344
 w5334 w5324 
w5314 w5304 w5294 w5284 w5274 w5264 w5254 w5244 w5234 w5224 w5214
 w5204 w5194 
w5184 w5174 w5164 w5154 w5144 w5134 w5124 w5114 w5104 w5094 w5084
 w5074 w5064 
w5054 w5044 w5034 w5024 w5014 w5004 w4994 w4984 w4974 w4964 w4954
 w4944 w4934 
w4924 w4914 w4904 w4894 w4884 w4874 w4864 w4854 w4844 w4834 w4824
 w4814 w4804 
w4794 w4784 w4774 w4764 w4754 w4744 w4734 w4724 w4714 w4704 w4694
 w4684 w4674 
w4664 w4654 w4644 w4634 w4624 w4614 w4604 w4594 w4584 w4574 w4564
 w4554 w4544 
w4534 w4524 w4514 w4504 w4494 w4484 w4474 w4464 w4454 w4444 w4434
 w4424 w4414 
w4404 w4394 w4384 w4374 w4364 w4354 w4344 w4334 w4324 w4314 w4304
 w4294 w4284 
w4274 w4264 w4254 w4244 w4234 w4224 w4214 w4204 w4194 w4184 w4174
 w4164 w4154 
w4144 w4134 w4124 w4114 w4104 w4094 w4084 w4074 w4064 w4054 w4044
 w4034 w4024 
w4014; 
run; 
 
 
proc reg data=prinvars outest=est press; 
model MIT = prin1-prin10 / selection=stepwise vif p; 
model TORV= prin1-prin10 / selection=stepwise vif p; 
model TMML= prin1-prin10 / selection=stepwise vif p; 
model TOXY= prin1-prin10 / selection=stepwise vif p; 
model TRE= prin1-prin10 / selection=stepwise vif p; 
model TME= prin1-prin10 / selection=stepwise vif p; 
run; 
 
proc print data=prinvars; 
var prin1-prin10; 
run; 
ods html close; 
  
 
210 
 
Table D.16 ANOVA results and model parameter estimates for MIT (first derivative 
data model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 6146.751 768.3439 105.29 <.0001  
Error 21 153.249 7.29757    
Corrected 
Total 29 6300     
       
Root MSE 2.7014 R-Square 0.9757    
Dependent 
Mean 290 Adj R-Sq 0.9664    
Coeff Var 0.93152          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 290 0.49321 587.99 <.0001 0 
Prin1 1 2.27234 0.50164 4.53 0.0002 1 
Prin2 1 9.93925 0.50164 19.81 <.0001 1 
Prin3 1 1.09205 0.50164 2.18 0.041 1 
Prin4 1 6.38861 0.50164 12.74 <.0001 1 
Prin5 1 2.32867 0.50164 4.64 0.0001 1 
Prin6 1 -6.66013 0.50164 -13.28 <.0001 1 
Prin7 1 -3.39066 0.50164 -6.76 <.0001 1 
Prin9 1 2.17281 0.50164 4.33 0.0003 1 
 
 
 
 
 
  
 
211 
 
Table D.17 ANOVA results and model parameter estimates for TORV (first 
derivative data model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 6 3348.447 558.0746 103.74 <.0001  
Error 23 123.7242 5.37931    
Corrected 
Total 29 3472.172     
       
Root MSE 2.31933 R-Square 0.9644    
Dependent 
Mean 283.8253 Adj R-Sq 0.9551    
Coeff Var 0.81717          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 283.8253 0.42345 670.27 <.0001 0 
Prin1 1 -3.85968 0.43069 -8.96 <.0001 1 
Prin2 1 4.62802 0.43069 10.75 <.0001 1 
Prin3 1 6.09119 0.43069 14.14 <.0001 1 
Prin4 1 1.06455 0.43069 2.47 0.0213 1 
Prin5 1 2.01164 0.43069 4.67 0.0001 1 
Prin6 1 -6.0717 0.43069 -14.1 <.0001 1 
 
 
 
 
 
 
 
  
 
212 
 
Table D.18 ANOVA results and model parameter estimates for TMML (first 
derivative data model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 6072.09 759.0113 43.64 <.0001  
Error 21 365.279 17.39424    
Corrected 
Total 29 6437.369     
       
Root MSE 4.17064 R-Square 0.9433    
Dependent 
Mean 325.599 Adj R-Sq 0.9216    
Coeff Var 1.28091          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 325.599 0.76145 427.6 <.0001 0 
Prin1 1 -1.70486 0.77447 -2.2 0.039 1 
Prin2 1 4.58655 0.77447 5.92 <.0001 1 
Prin3 1 8.90202 0.77447 11.49 <.0001 1 
Prin4 1 3.79045 0.77447 4.89 <.0001 1 
Prin5 1 3.87898 0.77447 5.01 <.0001 1 
Prin6 1 -4.58461 0.77447 -5.92 <.0001 1 
Prin7 1 -3.88601 0.77447 -5.02 <.0001 1 
Prin9 1 6.3765 0.77447 8.23 <.0001 1 
 
 
 
 
 
 
  
 
213 
 
Table D.19 ANOVA results and model parameter estimates for TOXY (first 
derivative data model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 6643.051 830.3814 10.37 <.0001  
Error 21 1681.918 80.09131    
Corrected 
Total 29 8324.969     
       
Root MSE 8.94938 R-Square 0.798    
Dependent 
Mean 292.497 Adj R-Sq 0.721    
Coeff Var 3.05965          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 292.497 1.63392 179.01 <.0001 0 
Prin2 1 6.20974 1.66186 3.74 0.0012 1 
Prin3 1 6.56672 1.66186 3.95 0.0007 1 
Prin4 1 4.34758 1.66186 2.62 0.0161 1 
Prin5 1 4.32033 1.66186 2.6 0.0167 1 
Prin6 1 -7.3818 1.66186 -4.44 0.0002 1 
Prin7 1 -5.20823 1.66186 -3.13 0.005 1 
Prin8 1 4.10755 1.66186 2.47 0.0221 1 
Prin9 1 3.36644 1.66186 2.03 0.0557 1 
 
 
 
 
 
  
 
214 
 
Table D.20 ANOVA results and model parameter estimates for TRE (first derivative 
data model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 8 4073.756 509.2195 31.48 <.0001  
Error 21 339.682 16.17533    
Corrected 
Total 29 4413.438     
       
Root MSE 4.02186 R-Square 0.923    
Dependent 
Mean 239.0813 Adj R-Sq 0.8937    
Coeff Var 1.68221          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 239.0813 0.73429 325.6 <.0001 0 
Prin1 1 4.96446 0.74684 6.65 <.0001 1 
Prin2 1 8.69835 0.74684 11.65 <.0001 1 
Prin3 1 -3.48872 0.74684 -4.67 0.0001 1 
Prin4 1 3.29752 0.74684 4.42 0.0002 1 
Prin5 1 -2.49227 0.74684 -3.34 0.0031 1 
Prin6 1 -2.69359 0.74684 -3.61 0.0017 1 
Prin7 1 -1.2945 0.74684 -1.73 0.0977 1 
Prin8 1 1.40704 0.74684 1.88 0.0735 1 
 
 
 
 
 
 
  
 
215 
 
Table D.21 ANOVA results and model parameter estimates for TME (first derivative 
data model). 
             
Analysis of Variance  
Source DF Sum of Mean F Value Pr > F  Squares Square  
Model 7 5992.439 856.0627 26.8 <.0001  
Error 22 702.8042 31.94565    
Corrected 
Total 29 6695.243     
       
Root MSE 5.65205 R-Square 0.895    
Dependent 
Mean 387.7237 Adj R-Sq 0.8616    
Coeff Var 1.45775          
              
Parameter Estimates 
Variable DF Parameter Standard t Value Pr > |t| Variance 
Estimate Error Inflation 
Intercept 1 387.7237 1.03192 375.73 <.0001 0 
Prin1 1 -3.41304 1.04956 -3.25 0.0037 1 
Prin2 1 2.49212 1.04956 2.37 0.0267 1 
Prin3 1 11.28486 1.04956 10.75 <.0001 1 
Prin4 1 -1.96302 1.04956 -1.87 0.0748 1 
Prin5 1 6.35524 1.04956 6.06 <.0001 1 
Prin6 1 3.65403 1.04956 3.48 0.0021 1 
Prin9 1 -1.95801 1.04956 -1.87 0.0755 1 
 
  
 
216 
 
Table D.22 Principal components (PC) obtained from first derivative NIR spectral 
data of biomass dusts used for internal validation of prediction models. 
 
 
 
 
 
B i o m a ss Pr i n 1 Pr i n 2 Pr i n 3 Pr i n 4 Pr i n 5 Pr i n 6 Pr i n 7 Pr i n 8 Pr i n 9 Pr i n 1 0
0 . 2 3 7 0 9 0 . 0 7 0 5 5 - 1 . 1 2 7 4 8 - 0 . 6 2 2 4 8 0 . 0 9 1 2 4 0 . 9 1 4 3 2 1 . 0 0 7 5 4 0 . 5 0 7 6 4 1 . 8 3 2 1 9 - 0 . 4 7 3 6 4
0 . 3 5 8 7 6 0 . 0 9 0 2 7 - 1 . 0 7 5 5 7 - 0 . 9 4 6 3 9 0 . 2 9 1 0 8 1 . 6 3 5 7 3 1 . 3 2 5 9 8 0 . 7 5 9 8 2 1 . 6 3 5 7 8 0 . 6 9 4 8 1
0 . 4 6 0 5 8 0 . 0 7 4 9 6 - 1 . 0 5 6 6 8 - 0 . 9 2 7 1 6 0 . 0 7 2 5 2 1 . 3 0 6 3 3 1 . 3 4 0 4 3 0 . 8 8 8 7 9 1 . 0 7 8 1 2 - 0 . 7 5 6 6 4
- 0 . 4 6 1 0 4 0 . 5 0 9 5 1 - 0 . 5 3 4 1 4 - 1 . 4 4 0 1 4 0 . 1 3 7 2 1 1 . 3 2 5 0 3 - 1 . 3 3 0 6 9 - 1 . 1 4 2 3 7 - 1 . 0 2 4 8 1 - 0 . 4 5 9 1 9
- 0 . 2 9 4 5 3 0 . 5 4 0 3 5 - 0 . 4 5 1 8 4 - 1 . 2 9 6 0 1 0 . 4 8 6 6 5 1 . 1 9 1 9 - 1 . 4 6 6 8 - 0 . 8 0 0 8 1 - 0 . 3 0 3 4 3 0 . 0 9 4 1 9
- 0 . 1 4 3 8 0 . 2 8 2 1 8 - 0 . 5 1 2 1 9 - 1 . 7 9 6 9 5 - 0 . 0 5 9 3 5 0 . 9 8 1 8 5 - 1 . 4 7 4 1 1 - 1 . 4 5 6 3 3 - 1 . 0 8 6 2 6 0 . 3 6 7 5 4
1 . 7 1 7 6 - 0 . 2 8 6 5 7 0 . 3 7 0 3 1 . 0 6 4 3 7 1 . 2 0 3 2 8 0 . 8 4 4 3 9 2 . 0 3 5 7 3 - 0 . 8 9 8 9 8 - 0 . 7 1 3 5 9 0 . 8 6 7 3 2
1 . 5 6 4 2 8 - 0 . 1 4 5 7 0 . 2 9 6 9 8 1 . 0 3 1 6 4 1 . 1 1 1 7 4 0 . 2 5 2 6 7 0 . 9 2 2 0 1 - 0 . 6 0 4 7 9 - 2 . 0 1 1 0 8 - 2 . 2 3 9 5 4
1 . 6 3 1 7 3 - 0 . 2 8 7 4 4 0 . 3 5 9 7 0 . 6 1 0 6 7 0 . 6 7 0 8 5 0 . 4 2 3 7 7 0 . 1 0 6 6 3 - 1 . 0 8 2 0 1 - 1 . 1 5 5 2 1 . 9 3 8 4 8
- 0 . 0 8 6 8 0 . 6 5 2 1 7 0 . 5 5 3 9 5 - 1 . 1 6 4 8 2 - 1 . 4 0 6 6 3 - 1 . 0 7 4 9 5 0 . 8 0 2 4 7 0 . 5 1 0 8 7 - 0 . 5 9 5 3 4 - 0 . 7 8 4 6
0 . 0 6 5 6 2 0 . 5 6 4 0 8 0 . 4 8 6 1 9 - 1 . 3 1 2 4 9 - 1 . 8 4 7 1 6 - 1 . 0 9 3 0 2 1 . 5 8 1 5 6 0 . 1 5 2 0 9 - 1 . 2 4 8 0 6 - 0 . 3 8 3 6 7
0 . 0 6 1 8 2 0 . 5 1 5 5 4 0 . 6 0 9 6 1 - 1 . 2 5 4 6 1 - 1 . 6 4 0 2 3 - 1 . 1 0 5 2 1 . 1 2 6 4 7 0 . 2 7 3 9 8 - 0 . 7 5 7 6 7 1 . 7 6 3 5 9
- 1 . 3 2 7 5 7 - 2 . 0 5 2 8 1 . 4 3 9 6 3 - 0 . 1 2 8 9 4 0 . 3 4 6 8 6 0 . 2 8 6 6 7 0 . 1 2 0 0 2 0 . 5 4 8 4 1 - 0 . 1 7 0 1 1 0 . 9 6 0 0 5
- 1 . 4 1 7 6 4 - 2 . 0 2 7 5 9 1 . 4 5 4 9 3 - 0 . 1 4 1 3 3 0 . 4 2 5 5 0 . 4 0 2 6 6 - 0 . 3 5 1 1 7 0 . 3 7 3 5 8 0 . 2 0 3 1 1 0 . 6 8 4 9 7
- 1 . 4 6 7 7 - 2 . 0 8 0 2 6 1 . 4 7 3 1 - 0 . 4 2 8 6 7 0 . 3 4 7 7 0 . 1 6 2 6 6 0 . 8 8 8 2 - 0 . 9 6 5 6 9 0 . 4 9 1 7 6 - 1 . 8 0 2 6 1
- 1 . 0 1 8 8 2 1 . 3 0 7 3 3 0 . 0 2 7 8 6 0 . 2 0 9 0 3 1 . 9 2 5 1 9 - 1 . 2 9 9 5 4 - 0 . 3 2 1 3 4 0 . 1 1 2 8 5 - 0 . 2 0 0 4 9 0 . 4 5 2 4 4
- 0 . 8 7 6 1 3 1 . 2 7 9 2 9 - 0 . 0 2 0 3 3 - 0 . 1 5 4 8 8 2 . 0 9 1 2 7 - 1 . 4 1 0 4 6 0 . 5 1 4 5 3 0 . 4 0 9 5 5 0 . 5 7 9 2 7 - 0 . 1 3 7 7
- 0 . 8 7 8 5 2 1 . 3 4 8 2 9 - 0 . 0 1 4 1 - 0 . 2 7 9 5 5 1 . 9 0 3 6 1 - 1 . 4 9 9 4 5 0 . 1 4 8 5 - 0 . 1 2 0 0 4 0 . 3 3 3 6 9 - 0 . 0 6 0 9 5
- 0 . 5 3 3 9 1 - 1 . 0 2 7 7 2 - 2 . 1 4 8 9 1 0 . 8 2 8 2 6 - 0 . 5 4 7 6 3 - 1 . 0 4 3 9 1 - 0 . 0 0 8 2 9 - 0 . 3 0 8 6 7 - 0 . 2 7 7 8 4 0 . 6 8 9 7 7
- 0 . 5 5 4 8 4 - 1 . 2 1 5 7 4 - 2 . 2 8 8 8 3 0 . 9 6 9 4 1 - 0 . 5 2 7 9 8 - 0 . 5 9 8 2 3 - 0 . 4 1 1 6 9 0 . 0 9 4 6 2 - 0 . 4 2 5 9 3 0 . 4 0 1 1 9
- 0 . 5 0 3 8 7 - 1 . 1 6 3 2 4 - 2 . 2 4 7 9 0 . 9 0 5 4 2 - 0 . 3 4 9 9 8 - 0 . 9 2 8 9 2 - 0 . 0 3 6 4 6 0 . 2 0 4 2 4 - 0 . 7 7 8 5 1 - 0 . 7 4 9 3 2
1 . 3 1 5 2 6 - 0 . 2 4 6 8 0 . 4 9 1 4 5 0 . 0 1 6 3 - 0 . 5 2 1 8 2 - 0 . 9 2 4 5 2 - 1 . 3 6 0 9 9 - 0 . 9 7 9 6 2 1 . 5 4 4 6 8 - 0 . 7 9 4 2 4
1 . 4 8 4 4 - 0 . 4 1 2 1 1 0 . 4 4 5 2 5 - 0 . 0 4 8 3 4 - 0 . 5 1 7 0 8 - 1 . 0 5 5 2 1 - 0 . 9 9 6 1 - 0 . 6 1 3 1 1 1 . 1 7 3 5 3 - 1 . 4 9 5 6 9
1 . 4 2 4 3 7 - 0 . 3 4 0 6 9 0 . 5 1 9 5 5 0 . 1 9 1 5 4 - 0 . 3 6 6 8 6 - 1 . 2 6 6 2 8 - 0 . 6 2 4 7 1 - 1 . 1 7 4 7 2 . 1 1 9 6 3 1 . 3 8 7 4 1
- 0 . 9 4 3 8 1 . 2 5 1 9 3 0 . 5 0 9 2 4 1 . 5 4 2 0 8 - 1 . 1 9 7 9 5 0 . 9 2 2 8 9 - 0 . 1 7 9 9 5 - 0 . 6 3 0 5 1 0 . 5 0 8 5 2 - 0 . 5 6 5 4 7
- 1 . 0 8 2 4 7 1 . 2 8 1 9 6 0 . 5 0 2 2 6 1 . 9 3 1 6 2 - 1 . 1 4 3 4 7 1 . 0 3 5 1 1 0 . 2 3 5 2 - 0 . 5 2 2 1 2 0 . 3 2 5 6 4 0 . 8 6 3 5
- 1 . 0 3 1 1 1 1 . 3 7 2 5 9 0 . 5 2 6 2 3 1 . 7 8 0 5 6 - 1 . 0 4 9 8 2 0 . 8 2 5 5 9 - 0 . 0 9 1 8 2 - 0 . 2 4 7 5 7 0 . 2 2 5 6 1 - 0 . 5 5 5 1 2
0 . 7 1 6 5 6 0 . 0 7 0 4 2 0 . 4 7 8 9 7 0 . 1 3 7 3 3 0 . 1 0 4 0 4 0 . 2 8 5 4 9 - 1 . 0 9 5 0 1 1 . 9 7 3 3 1 - 0 . 1 7 4 1 1 . 0 9 0 6 4
0 . 8 0 8 3 3 0 . 0 0 7 2 2 0 . 4 4 4 7 1 0 . 0 1 6 5 2 - 0 . 2 0 4 - 0 . 0 8 1 3 9 - 1 . 1 6 7 9 1 . 9 6 4 3 9 - 0 . 7 9 3 9 8 - 0 . 3 4 4 2 3
0 . 7 7 6 1 6 0 . 0 6 8 0 1 0 . 4 8 8 0 8 0 . 7 0 8 0 2 0 . 1 7 1 2 0 . 5 8 4 0 2 - 1 . 2 3 8 2 4 2 . 7 7 3 1 9 - 0 . 3 3 5 1 4 - 0 . 6 5 3 2 9
P o u l tr y  L i tt e r
S u g a r c a n e  b a g a s s e
S w e e tg u m
S w i tc h g r a s s
B e r mu d a  g r a s s
Co r n  Co b s
Co r n  s tov e r
E u c a l y p tu s
P e c a n  s h e l l
P i n e
  
 
217 
 
Appendix E ?Heating and Ignition Properties, NIR Spectra and SAS Code 
for Principal Component Analysis for Biomass Dusts Used for External 
Validation
 
 
Table E.1 Minimum hot surface ignition temperature (MIT) values for biomass 
dusts used for external validation of prediction models. 
Sample  MIT (?C) Mean MIT (?C) Standard Deviation (?C) 
Eucalyptus 
280.00 
280.00 0.00 280.00 
280.00 
Pine 
320.00 
320.00 0.00 320.00 
320.00 
Sweetgum 
300.00 
300.00 0.00 300.00 
300.00 
Switchgrass 
290.00 
290.00 0.00 290.00 
290.00 
 
 
 
 
 
 
  
 
218 
 
Table E.2 Temperature of onset of rapid volatilization (TORV) values for biomass 
dusts used for external validation of prediction models. 
Sample  TORV (?C) Mean TORV (?C) Standard Deviation (?C) 
Eucalyptus 
292.45 
292.36 0.91 291.41 
293.22 
Pine 
300.00 
301.23 1.10 301.59 
302.11 
Sweetgum 
301.16 
293.58 6.70 291.11 
288.46 
Switchgrass 
284.01 
284.00 8.24 275.75 
292.23 
 
 
 
 
Table E.3 Temperature of maximum rate of mass loss (TMML) values for biomass 
dusts used for external validation of prediction models. 
Sample  TMML (?C) Mean TMML (?C) Standard Deviation (?C) 
Eucalyptus 
328.09 
326.66 2.34 327.93 
323.96 
Pine 
353.12 
351.83 3.50 347.87 
354.50 
Sweetgum 
325.87 
324.97 2.93 321.70 
327.35 
Switchgrass 
324.31 
319.04 7.65 310.26 
322.55 
 
 
  
 
219 
 
Table E.4 Oxidation temperature (TOXY) values for biomass dusts used for 
external validation of prediction models. 
Sample  TOXY (?C) Mean TOXY (?C) Standard Deviation (?C) 
Eucalyptus 
293.95 
292.15 2.72 289.03 
293.48 
Pine 
335.79 
325.54 8.91 319.65 
321.17 
Sweetgum 
299.48 
300.43 2.19 298.87 
302.93 
Switchgrass 
289.37 
291.74 2.12 293.47 
292.38 
 
 
 
 
Table E.5 Temperature of rapid exothermic reaction (TRE) values for Biomass 
Dusts Used for External Validation of Prediction Models. 
Sample  TRE (?C) Mean TRE (?C) Standard Deviation (?C) 
Eucalyptus 
237.84 
238.76 0.81 239.36 
239.07 
Pine 
239.32 
244.32 4.45 245.79 
247.84 
Sweetgum 
242.22 
245.33 2.70 247.00 
246.77 
Switchgrass 
231.90 
237.20 5.18 242.25 
237.46 
 
 
  
 
220 
 
 
Table E.6 Maximum temperature reached during exothermic reaction (TME) values 
for Biomass Dusts Used for External Validation of Prediction Models. 
Sample  TME (?C) Mean TME (?C) Standard Deviation (?C) 
Eucalyptus 
399.86 
398.70 1.03 397.88 
398.35 
Pine 
403.66 
398.21 6.61 390.86 
400.10 
Sweetgum 
392.14 
390.94 1.08 390.07 
390.60 
Switchgrass 
397.87 
392.08 8.98 381.74 
396.63 
 
 
 
Figure E.1 NIR spectra showing average absorbance vs. wavenumber plot for 
biomass dusts used for external validation of prediction models. 
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
40005000600070008000900010000
Av
erag
e A
bsor
ban
ce
 (A)
Wavenumber (cm-1)
Eucalyptus
Pine
Sweetgum
Switchgrass
  
 
221 
 
SAS code: Obtaining principal components (PC) from raw NIR spectral data of 
biomass dusts used for external validation of prediction models.  
/******************************** 
 * Author: Jaskaran Dhiman      * 
 * Research work                * 
 * Date: 04/21/2014             * 
 ********************************/ 
 
ods html file='C:\Users\jzd0028\Desktop\extvalraw.html'; 
options nodate pageno=1; 
data work.jasnir; 
infile 'C:\Users\jzd0028\Desktop\extvalraw.csv' delimiter = ',' 
MISSOVER DSD lrecl=32767 firstobs=1 n=1500; 
input Biomass$ 
w9994 w9984 w9974 w9964 w9954 w9944 w9934 w9924 w9914 w9904 w9894
 w9884 w9874  
w9864 w9854 w9844 w9834 w9824 w9814 w9804 w9794 w9784 w9774 w9764
 w9754 w9744 
w9734 w9724 w9714 w9704 w9694 w9684 w9674 w9664 w9654 w9644 w9634
 w9624 w9614 
w9604 w9594 w9584 w9574 w9564 w9554 w9544 w9534 w9524 w9514 w9504
 w9494 w9484 
w9474 w9464 w9454 w9444 w9434 w9424 w9414 w9404 w9394 w9384 w9374
 w9364 w9354 
w9344 w9334 w9324 w9314 w9304 w9294 w9284 w9274 w9264 w9254 w9244
 w9234 w9224 
w9214 w9204 w9194 w9184 w9174 w9164 w9154 w9144 w9134 w9124 w9114
 w9104 w9094 
w9084 w9074 w9064 w9054 w9044 w9034 w9024 w9014 w9004 w8994 w8984
 w8974 w8964 
w8954 w8944 w8934 w8924 w8914 w8904 w8894 w8884 w8874 w8864 w8854
 w8844 w8834 
w8824 w8814 w8804 w8794 w8784 w8774 w8764 w8754 w8744 w8734 w8724
 w8714 w8704 
w8694 w8684 w8674 w8664 w8654 w8644 w8634 w8624 w8614 w8604 w8594
 w8584 w8574 
w8564 w8554 w8544 w8534 w8524 w8514 w8504 w8494 w8484 w8474 w8464
 w8454 w8444 
w8434 w8424 w8414 w8404 w8394 w8384 w8374 w8364 w8354 w8344 w8334
 w8324 w8314 
w8304 w8294 w8284 w8274 w8264 w8254 w8244 w8234 w8224 w8214 w8204
 w8194 w8184 
w8174 w8164 w8154 w8144 w8134 w8124 w8114 w8104 w8094 w8084 w8074
 w8064 w8054 
w8044 w8034 w8024 w8014 w8004 w7994 w7984 w7974 w7964 w7954 w7944
 w7934 w7924 
w7914 w7904 w7894 w7884 w7874 w7864 w7854 w7844 w7834 w7824 w7814
 w7804 w7794 
w7784 w7774 w7764 w7754 w7744 w7734 w7724 w7714 w7704 w7694 w7684
 w7674 w7664 
w7654 w7644 w7634 w7624 w7614 w7604 w7594 w7584 w7574 w7564 w7554
 w7544 w7534 
w7524 w7514 w7504 w7494 w7484 w7474 w7464 w7454 w7444 w7434 w7424
 w7414 w7404 
  
 
222 
 
w7394 w7384 w7374 w7364 w7354 w7344 w7334 w7324 w7314 w7304 w7294
 w7284 w7274 
w7264 w7254 w7244 w7234 w7224 w7214 w7204 w7194 w7184 w7174 w7164
 w7154 w7144 
w7134 w7124 w7114 w7104 w7094 w7084 w7074 w7064 w7054 w7044 w7034
 w7024 w7014 
w7004 w6994 w6984 w6974 w6964 w6954 w6944 w6934 w6924 w6914 w6904
 w6894 w6884 
w6874 w6864 w6854 w6844 w6834 w6824 w6814 w6804 w6794 w6784 w6774
 w6764 w6754 
w6744 w6734 w6724 w6714 w6704 w6694 w6684 w6674 w6664 w6654 w6644
 w6634 w6624 
w6614 w6604 w6594 w6584 w6574 w6564 w6554 w6544 w6534 w6524 w6514
 w6504 w6494 
w6484 w6474 w6464 w6454 w6444 w6434 w6424 w6414 w6404 w6394 w6384
 w6374 w6364 
w6354 w6344 w6334 w6324 w6314 w6304 w6294 w6284 w6274 w6264 w6254
 w6244 w6234 
w6224 w6214 w6204 w6194 w6184 w6174 w6164 w6154 w6144 w6134 w6124
 w6114 w6104 
w6094 w6084 w6074 w6064 w6054 w6044 w6034 w6024 w6014 w6004 w5994
 w5984 w5974 
w5964 w5954 w5944 w5934 w5924 w5914 w5904 w5894 w5884 w5874 w5864
 w5854 w5844 
w5834 w5824 w5814 w5804 w5794 w5784 w5774 w5764 w5754 w5744 w5734
 w5724 w5714 
w5704 w5694 w5684 w5674 w5664 w5654 w5644 w5634 w5624 w5614 w5604
 w5594 w5584 
w5574 w5564 w5554 w5544 w5534 w5524 w5514 w5504 w5494 w5484 w5474
 w5464 w5454 
w5444 w5434 w5424 w5414 w5404 w5394 w5384 w5374 w5364 w5354 w5344
 w5334 w5324 
w5314 w5304 w5294 w5284 w5274 w5264 w5254 w5244 w5234 w5224 w5214
 w5204 w5194 
w5184 w5174 w5164 w5154 w5144 w5134 w5124 w5114 w5104 w5094 w5084
 w5074 w5064 
w5054 w5044 w5034 w5024 w5014 w5004 w4994 w4984 w4974 w4964 w4954
 w4944 w4934 
w4924 w4914 w4904 w4894 w4884 w4874 w4864 w4854 w4844 w4834 w4824
 w4814 w4804 
w4794 w4784 w4774 w4764 w4754 w4744 w4734 w4724 w4714 w4704 w4694
 w4684 w4674 
w4664 w4654 w4644 w4634 w4624 w4614 w4604 w4594 w4584 w4574 w4564
 w4554 w4544 
w4534 w4524 w4514 w4504 w4494 w4484 w4474 w4464 w4454 w4444 w4434
 w4424 w4414 
w4404 w4394 w4384 w4374 w4364 w4354 w4344 w4334 w4324 w4314 w4304
 w4294 w4284 
w4274 w4264 w4254 w4244 w4234 w4224 w4214 w4204 w4194 w4184 w4174
 w4164 w4154 
w4144 w4134 w4124 w4114 w4104 w4094 w4084 w4074 w4064 w4054 w4044
 w4034 w4024 
w4014 w4004; 
datalines; 
; 
run; 
 
  
 
223 
 
proc princomp data=work.jasnir out=prinvars std; 
var w9994 w9984 w9974 w9964 w9954 w9944 w9934 w9924 w9914 w9904
 w9894 w9884 w9874  
w9864 w9854 w9844 w9834 w9824 w9814 w9804 w9794 w9784 w9774 w9764
 w9754 w9744 
w9734 w9724 w9714 w9704 w9694 w9684 w9674 w9664 w9654 w9644 w9634
 w9624 w9614 
w9604 w9594 w9584 w9574 w9564 w9554 w9544 w9534 w9524 w9514 w9504
 w9494 w9484 
w9474 w9464 w9454 w9444 w9434 w9424 w9414 w9404 w9394 w9384 w9374
 w9364 w9354 
w9344 w9334 w9324 w9314 w9304 w9294 w9284 w9274 w9264 w9254 w9244
 w9234 w9224 
w9214 w9204 w9194 w9184 w9174 w9164 w9154 w9144 w9134 w9124 w9114
 w9104 w9094 
w9084 w9074 w9064 w9054 w9044 w9034 w9024 w9014 w9004 w8994 w8984
 w8974 w8964 
w8954 w8944 w8934 w8924 w8914 w8904 w8894 w8884 w8874 w8864 w8854
 w8844 w8834 
w8824 w8814 w8804 w8794 w8784 w8774 w8764 w8754 w8744 w8734 w8724
 w8714 w8704 
w8694 w8684 w8674 w8664 w8654 w8644 w8634 w8624 w8614 w8604 w8594
 w8584 w8574 
w8564 w8554 w8544 w8534 w8524 w8514 w8504 w8494 w8484 w8474 w8464
 w8454 w8444 
w8434 w8424 w8414 w8404 w8394 w8384 w8374 w8364 w8354 w8344 w8334
 w8324 w8314 
w8304 w8294 w8284 w8274 w8264 w8254 w8244 w8234 w8224 w8214 w8204
 w8194 w8184 
w8174 w8164 w8154 w8144 w8134 w8124 w8114 w8104 w8094 w8084 w8074
 w8064 w8054 
w8044 w8034 w8024 w8014 w8004 w7994 w7984 w7974 w7964 w7954 w7944
 w7934 w7924 
w7914 w7904 w7894 w7884 w7874 w7864 w7854 w7844 w7834 w7824 w7814
 w7804 w7794 
w7784 w7774 w7764 w7754 w7744 w7734 w7724 w7714 w7704 w7694 w7684
 w7674 w7664 
w7654 w7644 w7634 w7624 w7614 w7604 w7594 w7584 w7574 w7564 w7554
 w7544 w7534 
w7524 w7514 w7504 w7494 w7484 w7474 w7464 w7454 w7444 w7434 w7424
 w7414 w7404 
w7394 w7384 w7374 w7364 w7354 w7344 w7334 w7324 w7314 w7304 w7294
 w7284 w7274 
w7264 w7254 w7244 w7234 w7224 w7214 w7204 w7194 w7184 w7174 w7164
 w7154 w7144 
w7134 w7124 w7114 w7104 w7094 w7084 w7074 w7064 w7054 w7044 w7034
 w7024 w7014 
w7004 w6994 w6984 w6974 w6964 w6954 w6944 w6934 w6924 w6914 w6904
 w6894 w6884 
w6874 w6864 w6854 w6844 w6834 w6824 w6814 w6804 w6794 w6784 w6774
 w6764 w6754 
w6744 w6734 w6724 w6714 w6704 w6694 w6684 w6674 w6664 w6654 w6644
 w6634 w6624 
w6614 w6604 w6594 w6584 w6574 w6564 w6554 w6544 w6534 w6524 w6514
 w6504 w6494 
w6484 w6474 w6464 w6454 w6444 w6434 w6424 w6414 w6404 w6394 w6384
 w6374 w6364 
  
 
224 
 
w6354 w6344 w6334 w6324 w6314 w6304 w6294 w6284 w6274 w6264 w6254
 w6244 w6234 
w6224 w6214 w6204 w6194 w6184 w6174 w6164 w6154 w6144 w6134 w6124
 w6114 w6104 
w6094 w6084 w6074 w6064 w6054 w6044 w6034 w6024 w6014 w6004 w5994
 w5984 w5974 
w5964 w5954 w5944 w5934 w5924 w5914 w5904 w5894 w5884 w5874 w5864
 w5854 w5844 
w5834 w5824 w5814 w5804 w5794 w5784 w5774 w5764 w5754 w5744 w5734
 w5724 w5714 
w5704 w5694 w5684 w5674 w5664 w5654 w5644 w5634 w5624 w5614 w5604
 w5594 w5584 
w5574 w5564 w5554 w5544 w5534 w5524 w5514 w5504 w5494 w5484 w5474
 w5464 w5454 
w5444 w5434 w5424 w5414 w5404 w5394 w5384 w5374 w5364 w5354 w5344
 w5334 w5324 
w5314 w5304 w5294 w5284 w5274 w5264 w5254 w5244 w5234 w5224 w5214
 w5204 w5194 
w5184 w5174 w5164 w5154 w5144 w5134 w5124 w5114 w5104 w5094 w5084
 w5074 w5064 
w5054 w5044 w5034 w5024 w5014 w5004 w4994 w4984 w4974 w4964 w4954
 w4944 w4934 
w4924 w4914 w4904 w4894 w4884 w4874 w4864 w4854 w4844 w4834 w4824
 w4814 w4804 
w4794 w4784 w4774 w4764 w4754 w4744 w4734 w4724 w4714 w4704 w4694
 w4684 w4674 
w4664 w4654 w4644 w4634 w4624 w4614 w4604 w4594 w4584 w4574 w4564
 w4554 w4544 
w4534 w4524 w4514 w4504 w4494 w4484 w4474 w4464 w4454 w4444 w4434
 w4424 w4414 
w4404 w4394 w4384 w4374 w4364 w4354 w4344 w4334 w4324 w4314 w4304
 w4294 w4284 
w4274 w4264 w4254 w4244 w4234 w4224 w4214 w4204 w4194 w4184 w4174
 w4164 w4154 
w4144 w4134 w4124 w4114 w4104 w4094 w4084 w4074 w4064 w4054 w4044
 w4034 w4024 
w4014 w4004; 
run; 
 
 
proc print data=prinvars; 
var prin1-prin10; 
run; 
ods html close; 
 
 
 
  
 
225 
 
Table E.7 Principal components (PC) obtained from raw NIR spectral data of 
biomass dusts used for external validation of prediction models. 
 
S a m p l e Pr i n 1 Pr i n 2 Pr i n 3 Pr i n 4 Pr i n 5 Pr i n 6 Pr i n 7 Pr i n 8 Pr i n 9 Pr i n 1 0
0 . 4 3 5 - 0 . 5 6 9 - 1 . 0 8 7 - 0 . 9 2 5 0 . 2 8 5 0 . 9 4 2 - 1 . 2 9 3 - 0 . 3 3 5 1 . 6 9 0 1 . 1 2 6
0 . 3 5 3 - 0 . 9 7 8 - 0 . 9 9 4 - 0 . 7 4 3 1 . 9 7 6 0 . 4 5 7 1 . 0 7 8 - 0 . 2 7 7 - 1 . 1 0 0 - 0 . 9 1 8
0 . 3 8 2 - 1 . 4 5 9 - 1 . 0 9 2 - 0 . 3 2 7 - 1 . 9 3 2 - 1 . 2 4 5 - 0 . 1 6 0 0 . 7 3 6 - 0 . 5 4 5 - 0 . 2 4 5
0 . 5 3 1 - 0 . 2 5 0 - 0 . 2 0 9 2 . 1 4 6 - 0 . 4 5 8 0 . 7 5 4 0 . 6 5 2 - 1 . 1 5 2 - 0 . 8 0 0 1 . 3 7 3
0 . 6 1 7 0 . 3 5 8 - 0 . 0 8 2 1 . 7 1 1 0 . 7 4 0 0 . 1 2 7 - 0 . 6 8 2 1 . 4 1 3 1 . 1 0 7 - 1 . 0 4 1
0 . 7 7 7 2 . 4 7 1 - 1 . 0 7 4 - 0 . 3 2 2 - 0 . 2 6 2 - 0 . 8 5 2 0 . 1 3 2 - 0 . 3 1 2 - 0 . 3 6 0 - 0 . 4 5 9
0 . 6 3 8 0 . 5 9 2 1 . 2 5 9 - 1 . 0 7 5 - 0 . 8 1 5 1 . 9 0 3 - 0 . 1 0 3 1 . 0 5 4 - 0 . 9 8 5 0 . 1 6 5
0 . 6 3 0 - 0 . 2 8 0 1 . 5 0 3 - 0 . 2 7 7 0 . 9 4 7 - 1 . 6 6 8 0 . 6 5 4 0 . 8 7 4 0 . 2 3 5 1 . 4 1 1
0 . 5 7 6 - 0 . 4 1 4 1 . 5 4 3 - 0 . 2 5 3 - 0 . 4 5 0 - 0 . 3 2 3 - 0 . 2 4 1 - 2 . 0 3 0 0 . 7 3 6 - 1 . 4 4 3
- 1 . 6 9 6 - 0 . 4 5 5 0 . 3 9 2 0 . 6 2 5 - 0 . 0 5 8 0 . 2 2 6 - 0 . 5 5 3 0 . 6 3 5 - 0 . 4 8 0 - 0 . 9 3 4
- 1 . 5 9 5 0 . 5 4 7 0 . 0 6 5 - 0 . 2 4 1 0 . 6 0 0 - 0 . 6 2 5 - 1 . 5 1 7 - 0 . 6 6 8 - 0 . 9 1 5 0 . 8 0 8
- 1 . 6 4 9 0 . 4 3 7 - 0 . 2 2 3 - 0 . 3 1 9 - 0 . 5 7 4 0 . 3 0 5 2 . 0 3 2 0 . 0 6 0 1 . 4 1 6 0 . 1 5 7
Pr i n c i p a l  Co m p o n e n t s
E u c a l y p t u s
Pi n e
S w e e t g u m
S w i t c h g r a ss