Characterization of Pecan Shells for Value-added Applications 
 
by 
 
Brad Lavoy Littlefield 
 
 
 
 
A thesis submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Master?s of Science 
 
Auburn, Alabama 
August 9, 2010 
 
 
 
 
Keywords: physical properties, flow properties, 
thermal decomposition 
 
Copyright 2010 by Brad Lavoy Littlefield 
 
 
 
 
 
 
 
 
Approved by 
 
 
Oladiran Fasina, Chair, Associate Professor of Biosystems Engineering  
Joey Shaw, Alumni Professor of Agronomy and Soils 
Sushil Adhikari, Assistant Professor of Biosystems Engineering 
Brian Via, Assistant Professor of Forestry and Wildlife Sciences
ii 
 
 
 
 
 
 
Abstract 
 There is a growing need in the United States to decrease dependence on fossil 
fuels because of energy security and environmental concerns associated with their use.  
Energy derived from biomass is especially important to the southeast due to availability 
of vegetation and optimum climate conditions.  Pecan shells, the by-product from the 
Pecan (Carya illinoinensis) shelling process, are a potential biomass feedstock.  In this 
study some of the physical characteristics of pecan shells (such as bulk, particle and tap 
densities, compressibility and flowability) that are important for its storage and process 
design were investigated.  It was found that particle size and moisture content 
significantly affected the bulk, tap and particle densities.  The porosity of pecan shells 
was significantly affected by particle size and moisture content as well.  Hausner ratio 
was affected by particle size but not by moisture content.  Mechanical compressibility 
was found to increase as particle size decreases and as moisture content increases.  The 
mechanical compressibility of the samples increased with pressure and decreased particle 
size or increased moisture content.  The flow behavior of pecan shells was not affected by 
particle size but moisture content indicated that lower moisture contents exhibited better 
flowability than pecan shells at higher moisture contents.  Rate of moisture sorption was 
determined using the Page model and equilibrium moisture content and equilibrium 
relative humidity (EMC-ERH) relationships for pecan shells were sigmoidal in shape and 
best predicted by the Henderson and Chung-Pfost equations.  The thermal decomposition 
iii 
 
characteristics of pecan shells were examined in nitrogen and air atmospheres at heating 
rates of 5, 10, 20, 30, and 40 ?C/min.  Four main stages of mass loss were observed 
during the thermal decomposition of pecan shells: moisture evaporation, hemicelluloses 
decomposition, cellulose decomposition and lignin degradation.  The moisture 
evaporation stage occurred within the temperature range of 30-150?C in both 
atmospheres.  The thermal decomposition of pecan shells demonstrated mass loss rate 
peaks attributed to hemicelluloses decomposition (275-330?C and 270-331?C) and 
cellulose decomposition (348-386?C and 315-339?C) for nitrogen and air thermal 
decomposition, respectively, increasing with increased heating rate.  The thermal 
decomposition of pecan shells was considered essentially complete at 600?C.  Volatilized 
gases during thermal decomposition of pecan shells were analyzed using a Fourier-
transform infrared spectrometer.  Gaseous products volatilized during thermochemical 
conversion processes were identified and quantified by concentration.  The major gases 
produced from nitrogen thermal decomposition of pecan shells were carbon dioxide 
(CO
2
), carbon monoxide (CO), ethanol (CH
3
CH
2
OH) and acetic acid (CH
3
COOH).  The 
major gases produced from air thermal decomposition of pecan shells were carbon 
monoxide, carbon dioxide and methyl isocyanate (C
2
H
3
NO).  A differential scanning 
calorimeter (DSC) was used to determine energy requirements at two temperature zones: 
moisture evaporation and thermal decomposition.  It was found that they energy required 
to drive off moisture was more than the energy required to raise the pecan shells to 
thermal decomposition temperatures.  The energy requirements at the two stages were not 
affected by heating rate.  It was also found that the energy required in both temperature 
zones was about 30% of the energy contained in raw pecan shells.  
iv 
 
 
 
 
 
 
Acknowledgements 
 First and foremost, I thank God for guiding me in the direction in which He knew 
was best for me.  I would like to express my sincere gratitude to all of those in the 
Biosystems Engineering Department that helped, in any part, toward the finishing of my 
research.  I was welcomed into the Department from my first day and I am very grateful 
to have been able to work in an environment that was always positive.  I would like to 
especially thank those that served on my committee, Dr. Joey Shaw, Dr. Sushil Adhikari 
and Dr. Brian Via.  I would also like to thank Dr. Oladiran Fasina for his guidance and 
providing the opportunity for me to earn my degree.  I appreciate all the patience and 
knowledge that you shared with me during these past two years.  I would also like to 
thank my wife, Laurel, without whom I would not have made it through these past few 
months.  All of the love, support, understanding and patience you have given me during 
this endeavor was more than I could ever ask.  Lastly, I would like to thank my parents 
(Brent and Gail) and the rest of my family and friends who have supported me 
throughout my academic career.   
v 
 
 
 
 
 
 
Table of Contents 
 
Abstract ........................................................................................................................... ii 
Acknowledgements ............................................................................................................ iv 
Table of Contents ................................................................................................................ v 
List of Figures .................................................................................................................... ix 
List of Tables ..................................................................................................................... xi 
Chapter 1. Introduction ....................................................................................................... 1 
Chapter 2. Literature Review .............................................................................................. 4 
2.1. Bioenergy .............................................................................................................. 4 
2.2. Pecan Shells .......................................................................................................... 6 
2.3. Physical Characteristics ........................................................................................ 8 
2.3.1. Particle Size ...................................................................................................8 
2.3.2. Bulk density ...................................................................................................9 
2.3.3. Particle density .............................................................................................10 
2.3.4. Porosity ........................................................................................................10 
2.3.5. Hausner ratio ................................................................................................11 
2.3.6. Compressibility ............................................................................................11 
2.3.7. Flowability ...................................................................................................12 
2.3.8. Moisture sorption isotherms ........................................................................17 
2.3.9. Rate of Moisture Sorption ............................................................................20 
2.4. Thermochemical Conversion of Biomass ........................................................... 21?
2.4.1. Chemical Composition of Biomass .............................................................21 
2.4.2. Thermochemical Conversion Processes ......................................................23 
2.4.2.1. Direct combustion ............................................................................... 24 
2.4.2.2. Gasification ......................................................................................... 24 
2.4.2.3. Pyrolysis .............................................................................................. 25 
2.4.2.4. Liquefaction ........................................................................................ 27 
2.4.3. Thermo-analytical techniques for biomass thermochemical                    
conversion ...................................................................................................27?
2.4.3.1. Thermogravimetric analysis ................................................................ 27 
2.4.3.2. TGA-FTIR .......................................................................................... 28 
2.5. Determination of Energy Requirement at Different Thermal                        
Decomposition Temperatures ............................................................................ 30?
2.6. Summary ............................................................................................................. 31 
2.7 References ............................................................................................................ 32 
vi 
 
 
Chapter 3. Effect of Moisture and Particle Size on Physical Properties                                     
of Pecan Shells ................................................................................................ 40?
3.1. Introduction ......................................................................................................... 40 
3.2. Materials and Methods ....................................................................................... 42 
3.2.1. Preparation ...................................................................................................42 
3.2.2. Particle Size Distribution .............................................................................42 
3.2.3. Mechanical Separation of Pecan Shells .......................................................43 
3.2.4. Compositional Analysis ...............................................................................45 
3.2.4.1. Energy Content ................................................................................... 45 
3.2.4.2. Ash Content ......................................................................................... 45 
3.2.4.3. Ultimate Analysis ................................................................................ 46 
3.2.4.4. Structural Carbohydrates ..................................................................... 46 
3.2.5. Poured Bulk Density ....................................................................................48 
3.2.6. Particle Density ............................................................................................49 
3.2.7. Tap Bulk Density .........................................................................................49 
3.2.8. Hausner Ratio...............................................................................................50 
3.2.9. Porosity ........................................................................................................50 
3.2.10. Mechanical Compressibility ......................................................................51 
3.2.11. Flowability .................................................................................................51 
3.2.12. Rate of Moisture Sorption ..........................................................................53 
3.2.13. Equilibrium Moisture Isotherm..................................................................55 
3.2.14. Data Analysis .............................................................................................56 
3.3. Results and Discussion ....................................................................................... 56?
3.3.1. Particle Size Distribution .............................................................................56 
3.3.2. Chemical Composition Analysis ..................................................................57 
3.3.2.1. Energy and Ash Content ..................................................................... 57 
3.3.2.2. Ultimate Analysis ................................................................................ 58 
3.3.2.3. Structural Carbohydrates ..................................................................... 59 
3.3.3. Particle Size Effect on Physical Properties of Pecan Shells ........................60?
3.3.3.1. Poured Bulk Density ........................................................................... 60 
3.3.3.2. Particle Density ................................................................................... 61 
3.3.3.3. Tap Bulk Density ................................................................................ 62 
3.3.3.4. Porosity ............................................................................................... 63 
3.3.3.5. Hausner Ratio ...................................................................................... 63 
3.3.3.6. Mechanical Compressibility................................................................ 64 
3.3.3.7. Flowability .......................................................................................... 65 
3.3.4. Moisture Content Effect on Physical Properties of Pecan Shells ................68?
3.3.4.1. Moisture Adjustment ........................................................................... 68 
3.3.4.2. Particle Size Distribution .................................................................... 68 
3.3.4.3. Poured Bulk Density ........................................................................... 70 
3.3.4.4. Particle Density ................................................................................... 71 
3.3.4.5. Tap Bulk Density ................................................................................ 72 
3.3.4.6. Porosity ............................................................................................... 73 
vii 
 
3.3.4.7. Hausner Ratio ...................................................................................... 74 
3.3.4.8. Mechanical Compressibility................................................................ 75 
3.3.4.9. Flowability .......................................................................................... 76 
3.3.5. Rate of Moisture Sorption ............................................................................79 
3.3.6. Equilibrium Moisture Isotherm ....................................................................83 
3.4. Conclusions ......................................................................................................... 87 
3.5. References ........................................................................................................... 89 
 
Chapter 4: TG-FTIR analysis of pecan shells during thermal decomposition.................. 93 
4.1. Introduction ......................................................................................................... 93 
4.2. Methods and Materials ....................................................................................... 95 
4.2.1. Sample Preparation ......................................................................................95 
4.2.2. Thermal Decomposition Study ....................................................................95 
4.2.3. Fourier Transform Infrared Spectroscopy ...................................................96 
4.2.4. Data Analysis ...............................................................................................96 
4.3. Results and Discussion ....................................................................................... 97?
4.3.1. Thermal Decomposition of Pecan Shells in Nitrogen Atmosphere .............97 
4.3.2. Thermal Decomposition of Pecan Shells in Air Atmosphere ....................100 
4.3.3. FTIR Analysis ............................................................................................104 
4.3.3.1. Thermal Decomposition in Nitrogen Atmosphere ............................ 104 
4.3.3.2. Thermal Decomposition in Air Atmosphere ..................................... 109 
4.4. Conclusions ....................................................................................................... 114 
4.5. References ......................................................................................................... 116 
 
Chapter 5. Determination of Energy Requirement at Different thermal degradation 
Temperatures................................................................................................. 119?
5.1. Introduction ....................................................................................................... 119 
5.2. Materials and Methods ..................................................................................... 120 
5.2.1. Sample Preparation ....................................................................................120 
5.2.2. Differential Scanning Calorimetry .............................................................120 
5.2.3. Data Analysis .............................................................................................120 
5.3. Results and Discussion ..................................................................................... 121 
5.4. Conclusions ....................................................................................................... 124 
5.5. References ......................................................................................................... 125 
 
Chapter 6. Conclusion ..................................................................................................... 126 
Chapter 7. Future Work .................................................................................................. 129?
Appendices ...................................................................................................................... 131?
Appendix A. Effect of Particle Size Physical Property Data ................................... 132?
Table A.1. Particle size distribution of pecan shells ............................................132 
viii 
 
Table A.2. Bulk density .......................................................................................133 
Table A.3. Tap density .........................................................................................133 
Table A.4. Particle density ...................................................................................133 
Table A.5 Porosity ...............................................................................................134 
Table A.6. Hausner Ratio.....................................................................................134 
Table A.7. Compressibility of pecan shells .........................................................134 
Table A.8. Flow function data .............................................................................135 
Table A.9. Cohesion and angle of internal friction ..............................................135 
Appendix B. Effect of moisture content on physical properties .............................. 136?
Table B.1. Particle size distribution .....................................................................136 
Table B.2. Bulk density .......................................................................................136 
Table B.3. Tap Density ........................................................................................137 
Table B.4. Particle Density ..................................................................................137 
Table B.5. Porosity ..............................................................................................138 
Table B.6. Hausner Ratio .....................................................................................138 
Table B.7. Compressibility of pecan shells .........................................................139 
Table B.8. Flow functions of pecan shells ...........................................................140 
Table B.9. Cohesion and angle of internal friction ..............................................141 
Appendix C. Rate of Moisture Sorption of Pecan Shells ........................................ 142?
Figure C.1. Moisture change in pecan shells exposed to air at 25?C                             
and relative humidity levels of 50, 65 and 80%. .......................................142?
Figure C.2. Moisture change in pecan shells exposed to air at 35?C                              
and relative humidity levels of 50, 65 and 80%. .......................................142 
Figure C.3. Moisture change in pecan shells exposed to air at 50%                        
relative humidity and temperatures of 15, 25 and 35?C. ...........................143 
Figure C.4. Moisture change in pecan shells exposed to air at 65%                      
relative humidity and temperatures of 15, 25 and 35?C. ...........................143 
Appendix D. Equipment Photos .............................................................................. 144?
Figure D.1. Annular shear cell .............................................................................144 
Figure D.2. Texture analyzer with compression cell and tight fitting piston ......144 
Figure D.3. Compression cell with tight fitting piston filled with fine,                    
medium and coarse pecan shell samples (left to right) .............................145 
Figure D.4. Thermogravimetric analyzer (right) connected to                               
Fourier-transform infrared spectrometer (left) by heated Teflon tube. .....145 
 
 
  
ix 
 
 
 
 
 
 
List of Figures 
 
Figure 2.1 ? Schematic illustrations of mass-flow and funnel-flow designed                    
storage silos ................................................................................................13 
Figure 2.2 ? Consolidation and shear of Jenike Shear cell ...............................................14 
Figure 2.3 ? Typical plot of Mohr circles, used to analyze Jenike shear cell                  
experiments ................................................................................................15 
Figure 2.4 ? Flow function graph .....................................................................................16 
Figure 2.5 ? Hemicellulose, cellulose and lignin in plant cell wall ..................................23 
Figure 2.6 ? Main processes, intermediate energy carriers and final energy                   
products from the thermochemical conversion of biomass .......................24 
Figure 3.1 ? Ripe pecan (left) and cracked pecans with shell (right) ...............................40 
Figure 3.2 ? Particle size distribution of pecan shells ......................................................44 
Figure 3.3 ? Kason vibrating screen separator..................................................................45 
Figure 3.4 ? Schematic diagram of rate of set up for moisture sorption study .................54 
Figure 3.5 ? Particle size distribution of pecan shells ......................................................57 
Figure 3.6 ? Effect of particle size and pressure on the mechanical                      
compressibility of pecan shells ..................................................................65 
Figure 3.7 ? Flow functions of pecan shells fractions ......................................................66 
Figure 3.8 ? Particle size distribution of pecan shells at different                                      
moisture contents .......................................................................................69 
Figure 3.9 ? Effect of moisture content on the poured bulk density of                                    
pecan shells ................................................................................................70 
Figure 3.10 ? Effect of moisture content on particle density of pecan shells ...................71 
Figure 3.11 ? Effect of moisture content on the tap bulk density of pecan shells ............73 
Figure 3.12 ? Effect of moisture content on porosity of pecan shells ..............................74 
Figure 3.13 ? Effect of moisture content and applied pressure on the                         
mechanical compressibility of pecan shells ...............................................76 
Figure 3.14 ? Flow functions of pecan shells at different moisture contents ...................77 
Figure 3.15 ? Cohesive strength of pecan shells at different moisture contents ...............79 
x 
 
Figure 3.16 ? Moisture change in pecan shells exposed to air at 15?C and                     
relative humidity levels of 50, 65 and 80% ...............................................80 
Figure 3.17 ? Moisture change in pecan shells exposed to air at 80% relative                  
humidity and temperatures of 15, 25 and 35?C. ........................................81 
Figure 3.18 ? Moisture sorption isotherms for pecan shells at 10, 20, 35 and                      
50?C ...........................................................................................................83 
Figure 4.1 ? Mass loss from thermal decomposition of pecan shells at different                   
heating rates with nitrogen as carrier gas ...................................................98 
Figure 4.2 ? Mass loss rate from the thermal decomposition of pecan shells at             
different heating rates with nitrogen as carrier gas ....................................99 
Figure 4.3 ? Mass loss from the thermal decomposition of pecan shells at                      
different heating rates with air as carrier gas ...........................................101 
Figure 4.4 ? Mass loss rate from the thermal decomposition of pecan shells                            
at different heating rates with air as carrier gas .......................................103 
Figure 4.5 ? Three-dimensional spectral plot of the gases produced from                                
the thermal degradation of pecan shells in nitrogen atmosphere                     
at a heating rate of 30?C/min ...................................................................104 
Figure 4.6 ? Concentration of the major gases volatilized during nitrogen thermal 
decomposition of pecan shells at heating rates of 5-40?C/min ................107 
Figure 4.7 ? Concentration of the minor gases volatilized during nitrogen thermal 
decomposition of pecan shells at heating rates of 5-40?C/min ................108 
Figure 4.8 ? Three-dimensional spectral plot of the gases produced from the                        
thermal decomposition of pecan shells in air atmosphere at a                    
heating rate of 30?C/min ..........................................................................110 
Figure 4.9 ? Concentration of the major gases volatilized during air thermal 
decomposition of pecan shells at heating rates of 5-40?C/min ................112 
Figure 4.10 - Concentration of the minor gases volatilized during air thermal 
decomposition of pecan shells at heating rates of 5-40?C/min ................113 
Figure 5.1 ? Heat flow (W/g) of pecan shells pyrolysis at different heating rates .........121 
 
 
 
  
xi 
 
 
 
 
 
 
List of Tables 
 
 
Table 2.1 ? Flowability predictions based on compressibility of bulk solids ...................12 
Table 2.2 ? Classification of powder flowability by flow index ......................................17 
Table 2.3 ? Chemical Constituents of Several Types of Biomass ....................................21 
Table 2.4 - Range of the main operating parameters for pyrolysis processes ..................26 
Table 3.1 ? Geometric mean diameter and geometric standard deviation of                       
pecan shells fractions .................................................................................57 
Table 3.2 ? Energy and ash content (dry basis) of pecan shells fractions ........................58 
Table 3.3 ? Carbon, hydrogen and nitrogen contents (dry basis) of pecan                            
shells fractions ...........................................................................................59 
Table 3.4 ? Cellulose, hemicelluloses and lignin contents (dry basis) of                                
pecan shells fractions .................................................................................60 
Table 3.5 ? Poured bulk density of pecan shells fractions ................................................61 
Table 3.6 ? Particle density of pecan shells fractions .......................................................62 
Table 3.7 ? Tap bulk density of pecan shells fractions .....................................................62 
Table 3.8 ? Porosity of pecan shells fractions ..................................................................63 
Table 3.9 ? Hausner ratio of pecan shells fractions ..........................................................64 
Table 3.10 ? Flow index of pecan shells fractions............................................................66 
Table 3.11 ? Cohesive strength (kPa) of pecan shells fractions .......................................67 
Table 3.12 ? Angle of internal friction (AIF,?) of pecan shell fractions ..........................68 
Table 3.13 ? Geometric mean diameters and standard deviation of pecan shells                     
at different moisture contents.....................................................................69 
Table 3.14 ? Hausner ratio of pecan shells at different moisture content levels ..............74 
Table 3.15 ? Flow index of pecan shells at various moisture contents.............................78 
Table 3.17 ? Angle of internal friction (?) of pecan shells at different                             
moisture contents .......................................................................................79 
Table 3.17 ? Estimated values of moisture sorption rate constant (k), constant n                   
and equilibrium moisture content (Me) obtained from non-linear 
regression analysis using Equation 3.20 for pecan shells ..........................82 
xii 
 
Table 3.18 ? Model coefficients and values for the standard error of estimate                     
(SEE), mean relative deviation (MRD) and coefficient of                 
determination (R2) for the temperature range of 10-50?C.........................86 
Table 4.1 ?Temperature (?C) peaks in mass loss rate for the main chemical                
constituents of pecan shells when thermally degraded in                             
nitrogen and air  .......................................................................................103 
Table 4.2 ? Effect of heating rate on the concentration (ml) of gases                       
produced during thermal decomposition of pecan shells in                   
nitrogen atmosphere .................................................................................109 
Table 4.3 ? Effect of heating rate on the concentration (ml) of gases                              
produced during thermal decomposition of pecan shells in                                 
air atmosphere ..........................................................................................111 
Table 5.1 ? Energy requirements for moisture evaporation and thermal                       
degradation of pecan shells at different heating rates ..............................123 
  
1 
 
CHAPTER 1. INTRODUCTION 
 There is a growing need in the United States to decrease dependence on fossil 
fuel.  National security, price stabilization, new government mandates placed on fossil 
fuel electricity producers and petroleum refineries and environmental concerns are 
primary reasons for the need to replace energy from fossil fuels with energy from 
renewable sources.  Examples of environmental concerns include carbon dioxide 
emissions from fossil fuel power plants and the environmental effect of drilling (e.g. BP 
oil spill in Gulf of Mexico) and mining disasters (e.g. Upper Big Branch Mine (June 
2010), West Virginia).  Energy from biomass is an example of alternative energy that is 
applicable to the Southeast and has the potential to relieve some of the stress placed on 
fossil fuel energy production.  Mean daily average temperatures of at least 10?C (50?F) 
and mean annual precipitation amounts of at least 114 cm (45 in.) create optimal 
conditions for producing biomass in the Southeast (Milbrandt, 2008).  In addition, forest 
lands occupy at least 60% of the total land area in the states that are in the southeastern 
part of the United States (NASS, 2010).  Energy by photovoltaic panels are not best 
suited due to government regulations which prohibit solar panels from existing on lands 
where minimum insolation values of six kilowatt-hours per square meter per day 
(Kwh/m
2
/day) can be achieved (NREL, 2009).  Most of the Southeastern U.S., including 
Alabama, have insolation values of 5 kWh/m
2
/day (NREL, 2009).  Wind power is 
available in the Southeast, however wind requirements of class 5 wind power (6.4 ? 8.0 
2 
 
m/s, 14.3 ? 17.9 mph) are only available on off-shore sites on the Atlantic coast or Gulf 
of Mexico (DOE, 2008).  New sites for hydropower have virtually been exhausted in 
Southeastern rivers and lakes (INL, 2006).  Thus, biomass is the only viable renewable 
energy resource for the Southeastern part of the U.S. 
Biomass is defined as organic materials that are plant or animal based (ASABE, 
2006).  A few examples of biomass are energy and agricultural crops, forestry and wood 
residues, food, feed, and fiber crop residue and agricultural wastes.  Pecan shells are 
waste products obtained from shelling Pecan (Carya illinoinensis).  They can therefore be 
categorized as a secondary biomass feedstock (ASABE, 2006).  Pecan shells have been 
utilized in a variety of applications, such as a source of activated carbon (Paraskeva et al., 
2008), mulch (Brison, 1974) and smokeless fuel briquettes (Blesa et al., 2001). 
Pecan shells are, however, a bulk solid.  Characterization of the physical 
properties of pecan shells is required for efficient design and selection of handling, 
processing and storage equipment and facilities (Woodcock and Mason, 1987).  This 
information is currently lacking for pecan shells. 
In addition, thermal behavior of pecan shells during thermal decomposition must 
be quantified to assess the possibility of utilizating pecan shells as a feedstock for 
thermochemical processes (e.g. gasification or pyrolysis).  These processes are used to 
convert biomass feedstocks such as pecan shells into value-added products.  
Therefore, the main objective of this study is to quantify the properties of pecan 
shells that are essential for process design and value added utilization.  To achieve this, 
the following tasks were carried out: 
3 
 
(1)  Quantify the effect of particle size and moisture content on the physical 
properties (particle size distribution, bulk density, tap bulk density, particle 
density, porosity, Hausner ratio, compressibility, flowability, moisture 
sorption rate and equilibrium moisture content) of pecan shells, 
(2)  Determine the rate of thermal decomposition and quantify the composition of 
gas volatilized at temperatures of 30 ? 800?C; and 
(3)  Determine the energy required to thermally decompose pecan shells.
4 
 
CHAPTER 2. LITERATURE REVIEW 
2.1. Bioenergy
 The last decade has proven to the United States and other countries around the 
world that are fossil fuel dependent that renewable and sustainable sources of fuel and 
energy are needed.  Rising transportation fuel costs, environmental implications of 
greenhouse gas emissions and energy security are the motivation for increased renewable 
energy research.  Bioenergy is one form of renewable energy that is derived from 
biomass or organic materials that are plant or animal based (ASABE, 2006).  Other 
examples of renewable energy are solar, hydro, geothermal and wind.  However, biomass 
is the only source of carbon in which the conversion to solid, liquid and gaseous energy 
sources can be produced (McKendry, 2002a).  Economic considerations, environmental 
issues, specifically greenhouse gas emissions, energy security and price stabilization are 
all well known advantages of renewable energy.  Recently, the United States government 
has introduced new policies and incentives for renewable energy research and 
development to try and alleviate some of the environmental and economic issues caused 
by the dependence on foreign sources of fossil fuels, such as renewable portfolio 
standards (RPS).  RPS is a regulation that requires the increased production of energy 
from renewable energy sources, such as wind, solar, biomass or geothermal.  The RPS 
mechanism generally places an obligation on eletricity production companies to produce 
a specified fraction of their electricity from renewable energy sources.  July 2008 marked 
5 
 
the highest crude-oil per barrel cost at $137.11 (www.eia.doe.gov).  Despite the oil prices 
falling again to roughly $70 per barrel, the need for renewable energy remains.  
Worldwide, the demand for energy and its resources is increasing exponentially with 
outgrowth of population and urbanization.  Since 1973 the energy resources have doubled 
in developed countries but the demand remains higher.  World energy demand is 
expected to increase approximately 2.5 times the present level by the year 2030 (Goyal et 
al, 2008; EIA, 2008).  The depletion of fossil fuels and the greenhouse gas emissions 
dilemma suggest that future energy supplies must come from renewable and sustainable 
sources. 
 Biomass is regarded the renewable energy source with the highest potential to 
contribute to the energy needs of modern society for both developed and developing 
economies world-wide (Bridgwater, 2003).  In contrast to other forms of renewable 
energy, biomass can be a substitute for all fossil-fuel based products, using a wide variety 
of technologies for conversion into solid, liquid and gaseous products.  In addition, 
biomass is used in traditional ways, such as the burning of biomass for heat and cooking 
(EIA, 2008).  In 2007 the United States produced approximately 102 quadrillion BTUs 
(10.7 x 10
13
 MJ) of energy and roughly 6.8 quadrillion BTUs (7.2 x 10
12
 MJ), or 7%, 
were from biomass (www.eia.doe.gov).  The ?Billion Ton Report? was introduced in 
2005 which outlined the ability of the United States to produce 1.36 billion tons of 
biomass for the production of bioenergy, which is sufficient enough to replace more than 
30% of the petroleum fuel consumption in the country (Perlack et al., 2005). 
Biomass stores energy during the process of photosynthesis.  This energy can be 
recovered by thermochemical conversion processes into more usable forms, such as 
6 
 
ethanol, bio-oils and producer gases (Mani et al., 2004).  Combustion, gasification and 
pyrolysis are three forms of thermochemical conversion processes in which biomass is 
converted to sources of energy.  Direct combustion is the oldest form of energy where 
biomass is burned in an air atmosphere to convert chemical energy stored in biomass into 
heat, mechanical power or electricity (Demirbas, 2004; McKendry, 2002b).  Combustion 
of biomass produces hot gases at temperatures at approximately 900-1000?C (McKendry, 
2002b)  Pyrolysis is the thermal degradation of biomass by heat in the absence of oxygen, 
which results in the production of charcoal (solid), bio-oil (liquid) and fuel gas products 
(Demirbas and Arin, 2002).  Pyrolysis is usually conducted at temperatures ranging from 
400-600?C (ASABE, 2006).  Converting biomass into a combustible gas mixture by 
partial oxidation of biomass is the conversion process known as gasification and occurs at 
temperatures of 800-900?C (McKendry, 2002b).  Thermochemical conversion processes 
of converting biomass to value-added energy products are discussed further in Section 
2.4.2. 
 
2.2. Pecan Shells 
Pecan (Carya illinoinensis) is a tree nut crop grown primarily in the southern and eastern 
portions of the Unites States, Mexico, Israel and Australia.  The United States is 
responsible for more than 80% of the world?s production (Geisler, 2009).  Pecans are the 
only native tree nuts grown for commercial use in the U.S. and are considered the most 
important tree nut crop due to their export potential (Wood et al., 1994).  In 2008, the 
U.S. exported approximately 52.4 million kilograms of pecans valued at $238.5 million 
dollars (Geisler, 2009).  Top pecan production states include Georgia, New Mexico, 
7 
 
Texas, Arizona and Oklahoma.  Alabama produced approximately 2.7, 5.4 and 3.6 
million kilograms of pecans in the 2006-2008 crop years, respectively (NASS, 2009). 
 Pecans can be found growing in their original habitat and in a semi-domesticated 
form where wild trees have been cleared from competing trees and brush.  These pecans 
are called wild or native and improved, respectively (Harris et al., 1986).  Pecans also 
grow in orchards or dooryard plantings where they have been vegetatively propagated.  
The commodity value of improved varieties is higher than native or wild pecans because 
of the desired characteristics exploited by genetic manipulation.  Pecans are sold as an 
agriculture commodity on an in-shell or shelled (nutmeat only) basis.  Approximately 
54.5 million kilograms of pecan shells are produced annually from shelling operations in 
the United States (USDA-NASS, 2009). 
 Pecan shells are currently only utilized in a few ways.  For example, pecan shells 
are used for mulching due to their slow breakdown, relative permeability, tannic acid 
composition and they are relatively light-weight.  Pecan shells added to soils can increase 
acidity levels specifically for vegetation that thrive in acidic soil, such as azaleas, 
camellias and magnolias (Brison, 1974).  However, excessive applications of pecan shells 
to soil can kill grass, weeds and other vegetative plants that are sensitive to soil pH.  
Other known uses for pecan shells range from meat smoking in combination with 
hardwoods (i.e. hickory), charcoal briquettes, imitation fire logs, glue and soap abrasives.  
Pecan shells have also been utilized as a source of activated carbon for water filtration 
purposes (Ng et al, 2003; Bansode et al., 2003; Johns et al., 1999). 
 
8 
 
2.3. Physical Characteristics 
Pecan shells, similar to other agricultural crops and residues, are a bulk solid.  
Bulk solids consist of many random particles or granules of different sizes, densities and 
perhaps chemical compositions grouped together (Woodcock and Mason, 1987).  The 
design and selection of storage, handling and transportation facilities and equipment is 
dependent upon the bulk solid behavior.  Efficient and proper use of processing 
equipment and facilities could result in savings in handling costs, reduction in waste 
materials and improved utilization of labor, machines and space (Shamlou, 1988).  In 
order to define the bulk solid behavior of pecan shells, determination of the individual 
particles that form the bulk solid must be quantified.  Particulate size, shape, density, size 
distribution and surface area are some of pecan shells particle properties that are 
influential of its bulk solid behavior.  Bulk density, tap density, particle density, 
compressibility, porosity, flow properties and equilibrium moisture content and relative 
humidity are bulk properties needed to properly characterize pecan shells in bulk form.   
 
2.3.1. Particle Size 
 Particle size is crucial to the behavior of bulk material (Paraskeva et al., 2008).  
This is because it is directly used to calculate properties such as surface area and volume 
of individual particles that make up the bulk material.  Size is always represented as a 
distribution due to the shape and size variability throughout the bulk material.  There are 
several determination methods to quantify particle size, including mechanical sieving, 
digital imagery and laser-length scattering method.  Most researchers rely on the sieving 
procedure because of the relative ease of use, simple analysis and lower capital cost of 
9 
 
equipment.  The ASABE standard S319.4 (ASABE, 2008) is often used to estimate the 
geometric mean diameter and standard deviation of agricultural materials. 
 
2.3.2. Bulk density 
 Bulk density is defined as the ratio of bulk mass of a material to its bulk volume.  
There are three different types of bulk density: poured, tap and aerated.  Bulk density of 
agricultural materials is a key parameter that not only is important for the economic 
impact on transportation costs, but also affects the design and operation of conveyors, 
storage silos, processing and heat equipment (Woodcock and Mason, 1987).   
 Poured bulk density is the most common measurement technique for determining 
bulk density.  This is obtained by pouring a sample into a container of known volume.  
The ratio of the mass of the material that filled the container to the volume of the 
container is determined as poured bulk density (Barbosa-Canovas et al., 2005). 
 Tap bulk density, sometimes referred to as vibrational compressibility, is the ratio 
of mass of the material to the volume of the material after it has been dropped, or tapped, 
by its own weight from a height of 14 mm.  Tap density can also be determined using tap 
density testers that use graduated cylinders for volume measurements.  During 
measurements, the cylinders are systematically rotated with constant up and down 
tapping.  This results in reduction in volume of material.  The ASTM standard B527 is 
typically used for tap density determinations (ASTM, 2005).  Tap density is important for 
transportation requirements for biological materials. 
 Aerated bulk density is a measure of the material?s density when its particulates 
are at their most ?loosely packed? form.  The procedure to measure aerated bulk density is 
10 
 
very complicated and thus most researchers usually do not measure bulk density using 
this technique (Barbosa-Canovas et al., 2005). 
 
2.3.3. Particle density 
 There are two types of particle density; true and apparent.  True particle density is 
calculated by taking the mass of the particulate and dividing by its volume, excluding 
open and closed pores.  Apparent particle density is different only in that it includes the 
volume of closed pores on the surface of the particulate in the volume measurement 
(Woodcock and Mason, 1987).  True density is often measured with a helium pycnometer 
(Colley et al., 2006) or by the toluene displacement method (Deshpande et al., 1993).  
Apparent particle density is best obtained from size measurements of the particles. 
 
2.3.4. Porosity 
 Porosity is an indication of the voidage, or volume occupied by air spaces when 
particulate solids are placed in a container.  Poured bulk density and true particle density 
are used in the calculation of porosity, using the following equation: 
nullnull1null
null
null
null
null
     (2.1) 
Where ?
b
 is bulk density and ?
p
 is particle density.  An average porosity calculation of 0.4 
is normal for spheroid particles, whereas irregular shaped or very small particulates have 
higher porosity values (Woodcock and Mason, 1987). 
 
11 
 
2.3.5. Hausner ratio 
 The Hausner ratio is a function of the tap and poured bulk density (Equation 2.2). 
null
null
null 
null
null
null
null
     (2.2) 
Where, ?
t
 is tap bulk density and ?
b
 is bulk density.  This ratio is important because it 
deals with properties relevant to moving, rather than static, powder (Barbosa-Canovas et 
al., 2005).  Geldhart et al. (1984) proposed that Hausner ratios smaller than 1.25 are 
easily fluidized, while powders with Hausner ratios higher than 1.4 could present 
fluidization issues (Geldhart et al., 1984; Barbosa-Canovas et al., 2005) 
 
2.3.6. Compressibility 
 Compressibility is a physical characteristic that is of significant importance during 
storage, handling and transportation of bulk materials.  Unintentional compression can 
cause flow problems resulting from vibrations during transportation or by due to the 
compressive weight of the bulk material.  Mechanical compressibility is a measure of the 
compressibility of a bulk solid caused by its own weight during storage.  Mechanical 
compressibility (%) can be calculated by the relationships below: 
nullnull null  100null
null
null
null null
null
null
null
null null 100null1null 
null
nullnull
null
nullnull
null   (2.1) 
where, V
i
 is initial volume (cm
3
), V
f
 is compressed volume (cm
3
), ?
bi
 is initial bulk 
density (kg/m
3
) and ?
bf
 is final (compressed) bulk density (kg/m
3
).  Fayed and Skocir 
(1997) used mechanical compressibility values to classify the flowability of materials, as 
shown in Table 2.1 below: 
 
12 
 
Table 2.1 ? Flowability predictions based on compressibility of bulk solids 
(Fayed and Skocir, 1997) 
Percent Compressibility Bulk Solid Description Flow 
5 to 15 free-flowing granules excellent flow 
12 to 16 
free-flowing powdered 
granules 
good flow 
18 to 21 flowable powdered granules fair to passable flow 
23 to 28 very fluid powders poor flow 
28 to 35 fluid cohesive powders poor flow 
33 to 38 fluid cohesive powders very poor flow 
> 40 cohesive powders extremely poor flow 
 
2.3.7. Flowability 
 Flowability of bulk solids is another characteristic that is important in the design 
of large scale storage silos and processing equipment.  Flowability is a measure of the 
cohesiveness and adhesiveness of bulk solids and is influenced by other properties, such 
as bulk density, porosity and compressibility.  Particles within bulk solids have a strong 
tendency to ?stick? to one another due to attractive forces between particles.  Adhesion 
occurs when particles in storage silos or bins ?stick? to the walls or exterior surface of the 
storage container.  Flowability tests are designed to quantify the cohesiveness or 
adhesiveness of bulk solids. 
 Silos are typically the main storage facility for agricultural materials.  Usually, 
materials are gravity-discharged from the bottom of the silo.  Funnel and mass flow are 
two types of flow that can dictate the design of storage silos (Figure 2.1).  In mass flow, 
all the particles move downward when the outlet is opened resulting in uniform flow.  
During funnel flow, the material at the bottom of the silo and the material above the 
13 
 
outlet are the only particles that move while the rest of the particles remain still.  Simply 
stated, mass flow results in ?first in, first out? sequence, whereas the funnel flow particle 
sequence is ?first in, last out?.  Funnel flow can result in flow problems such as ratholing, 
increased segregation and possible degradation of material over time for the stagnant 
material. Therefore, funnel flow designed silos are limited to coarse, free flowing and 
non-degradable bulk solids. 
 
Figure 2.1 ? Schematic illustrations of mass-flow and funnel-flow designed storage 
silos 
 Mass flow designed silos can have issues as well, such as channeling and arches 
that prevent discharge, known as ?arching?.  Arches can form via two different forces; 
mechanical and cohesive.  Mechanical arches form from large particles interlocking with 
one another, whereas cohesive arches are caused by small particles consolidation to each 
other.  Major flow complications (e.g. ratholing or arching) can result in damage to the 
silo or total structural failure.  Flow problems can be averted if the materials flowability 
 
ch
fl
 
m
d
(?
A
th
to
F
 
cr
fa
co
al
aracteristic
ow problem
Flowa
easurement
eveloped in 
) of a mater
t a specific 
e material b
 fail the ma
igure 2.2 ? 
 
The y
iterion for d
ilure shear s
nsolidating
., 2004): 
s are known
s (Shamlou,
bility and co
s, typically w
1964 (Gane
ial while pre
normal force
egins to fail
terial.  
Consolidati
ield strength
etermining 
tress (?, Pa 
 stress, as sh
 and the stor
 1988). 
hesiveness 
ith the use 
san et al., 20
ssure (?) is 
 (V = A?), 
 or shear.  T
on and shea
 which a so
the flowabil
or kPa) vs. n
own in Figu
14 
age silos, bi
of bulk soli
of Jenike?s 
08).  Jenike
applied at v
the tangenti
he shear str
r of Jenike
lid develops
ity of bulk s
ormal stres
re 2.3 below
ns and hopp
ds is usually
tangential sh
?s shear test
arious level
al force (S =
ength is obta
 Shear cell
 as it flows i
olids.  Yield
s (?, Pa or k
 (Ganesan 
ers are desi
 obtained by
ear cell (Fig
 measures th
s to the top o
 A?) would 
ined from t
(Ganesan e
n a channel 
 locus (YL)
Pa) for a giv
et al., 2008;
gned to prev
 shear stren
ure 2.2) 
e shear stre
f the materi
increase unt
he force req
 
t al., 2008)
is an import
 is a plot of 
en 
 Fitzpatrick 
ent 
gth 
ss 
al.  
il 
uired 
ant 
et 
15 
 
 
Figure 2.3 ? Typical plot of Mohr circles, used to analyze Jenike shear cell 
experiments (Teunou et al., 1999). 
  Unconfined yield strength (UYS) is the compressive strength of a bulk solid and 
measured in units of pressure (Pa or kPa) (Schulze, 2006).   UYS is a very important 
factor concerned with arching of bulk solids in silos (Jenike, 1964).  The point of 
intersection of the Mohr circle (passing though the origin) and the normal stress (x-axis) 
determines UYS (Ganesan et al., 2008).  Major consolidating stress (MCS) can also be 
determined from the Mohr circle, the point at which the circle passes though the steady 
state point and is tangent to the Yield Locus (YL), also measured in units of pressure (Pa 
or kPa).  The effective angle of internal friction (AIF, ?
e
) is a measure of the inter-particle 
friction as a bulk solid starts to slide itself at the onset of flow (Jenike, 1964), measured 
in degrees.  Increases in pressure generally increase AIF, but not always (Ganesan et al., 
2008).  For any shearing to take place, the AIF of that material must be overcome.  UYS 
and MCS are calculated by Equations 2.3 and 2.4 shown below: 
 
 
 
 
W
 
(E
F
2
fl
qu
fl
ab
th
(G
here,          
Flow 
quation 2.3
igure 2.4.  T
004; Teunou
ow function
antity used
owability by
ove propert
emselves, b
anesan et a
                   
function gra
) vs. major c
his type of g
 et al., 1999
 (FF).  The i
 to classify m
 flow index
ies are influ
ut also by te
l., 2008). 
Figure 2.4
and c is coh
phs are obta
onsolidatin
raph is very
).  The slop
nverse of th
aterials flo
 developed b
enced not o
mperature, 
 ? Flow fun
16 
esion, defin
ined by plot
g stress (Equ
 good at ch
e of the line
e FF, called
wability (Je
y Jenike (1
nly by the p
time of stora
ction graph
 
ed as ? at ? =
ting the unc
ation 2.4) o
aracterizing 
ar trendline 
 flow index 
nike, 1964).
964) is show
hysical natu
ge, moistur
 (Fitzpatric
  
 
 0. 
onfined yiel
f a material
flow (Fitzpa
(or line of b
(ff) is a dim
  Classificat
n in Table 
re of the par
e and chemi
 
k et al., 200
 
 
d stress 
, shown in 
trick et al., 
est fit) is cal
ensionless 
ion of powd
2.2.  All of t
ticles 
cal composi
4) 
(2.3) 
(2.4) 
led 
er 
he 
tions 
17 
 
 
Table 2.2 ? Classification of powder flowability by flow index (Jenike, 1964) 
 Flowability 
 Very cohesive Cohesive Easy flowing Free flowing 
flow index (ff) ff < 2 2 < ff < 4 4 < ff < 10 Ff > 10 
 
Although the use of Jenike?s shear cell is well established, its limitation is that the 
tests cannot be automated.  This has led to the development of rotational shear testers, 
such as a ShearScan (Sci-Tec Inc., Worthington, Ohio). 
 
2.3.8. Moisture sorption isotherms 
 Biological materials are hygroscopic in nature.  Therefore, these materials have 
the ability to exchange moisture with the atmosphere (Colley et al., 2006).  Knowledge of 
the equilibrium moisture content (EMC) and equilibrium relative humidity (ERH) 
relationship is crucial to the design of post harvest operations, such as storage, drying, 
aeration, handling, packaging and processing (e.g. shelling) of biological materials 
(Durakova et al., 2005; Pagano and Mascheroni, 2005; Erbas et al., 2005; Igathinathane 
et al., 2007).  The equilibrium moisture content is the moisture content in which a 
hygroscopic material, such as pecan shells, is in equilibrium with the environment.  The 
relative humidity of that environment is considered the equilibrium relative humidity.  By 
plotting EMC vs. ERH at a particular temperature, moisture isotherm can be determined.   
 Moisture isotherm curves have been used to determine storage stability of 
biological materials when exposed to varying environmental conditions during storage 
and transportation (Erbas et al., 2005).  They also provide a method for evaluation of 
18 
 
physical, chemical and microbiological parameters suitable and deteriorative reactions for 
the determination of stability of dry materials (Erbas et al., 2005; Leiras and Iglesias, 
1991).  EMC data can also determine the upper and lower limits of drying conditions for 
biomass (Labuza, 1984; Singh 2004).  Singh (2004) determined the EMC of biomass 
briquettes at relative humidity ranges of 40-85% and reported that limited amounts of 
moisture are beneficial as the steam generated causes steam gasification reaction leading 
to better gas quality.  However, high moisture in biomass briquettes can result in 
swelling, disintegration and prevention of application for thermochemical conversion 
(Singh, 2004) 
 In the drying of biological materials, the difference between the product?s actual 
moisture content and the EMC is often used as a measure for the driving force (Banaszek 
and Sibenmorgen, 1993).  Thermodynamic analysis of water sorption can also be 
determined from EMC-ERH data.  Specific thermodynamic properties of biological 
materials that can be obtained from EMC-ERH relationships include the isosteric heat of 
sorption, free energy and entropy.  Isosteric heat of sorption measures the binding energy 
or the force between water vapor molecules and the polar (active) sites of adsorbent 
surfaces, which is beneficial in estimating the heat requirement during drying and the 
state of absorbed water in the solid materials (Erbas et al., 2005; Labuza, 1984).  The 
level of moisture at which the heat of sorption approaches the heat of vaporization of 
water is taken to be indicative of the amount of bound water in the material of interest.  
However, at moisture contents higher than this level, water is free in the void spaces of 
the system and readily available for microorganisms (Fasina and Sokhansanj, 1993). 
19 
 
 There are more than 270 documented equations (theoretical, semi-empirical and 
empirical) have been documented to determine the EMC-ERH relationship of biological 
materials.  Theoretical, or kinetic based, models are based on a monolayer or muiltilayer 
sorption where the material?s physical properties are used as constants.  Examples of 
theoretical isotherm models are the Langmuir, Brunauer-Emmet-Teller (BET) and 
Guggenheim-Anderson-deBoer (GAB).  Empirical models require the use of non-linear 
regression from curve-fitting software to determine the model constants (Barbosa-
Canovas et al., 2007).  Traditional empirical isotherm models are the Halsey, Henderson, 
Chung-Pfost, Chen-Clayton and  Iglesias-Chirife.  However, no single equation 
accurately describes the relationship for all biological materials over a broad range of 
relative humidities and temperatures (Soysal and Oztekin, 1999).  Models adopted as 
standard equations by the American Society of Agricultural and Biological Engineers 
(ASABE) are the modified Halsey, Modified Henderson, Modified Chung-Pfost, 
Modified Oswin and GAB (ASABE Standard D245.5, 2001).  The Halsey equation was 
developed for food materials with high protein and oil contents whereas the Henderson 
and Chung-Pfost equations have been reported to be best suited for starchy foods (Fasina 
and Sokhansanj, 1993; Soysal and Oztekin, 2001).  These equations have been utilized to 
model moisture sorption isotherms of various agricultural and biological materials such 
as amaranth grains (Pagano and Macheroni, 2005), medicinal and aromatic plants (Soysal 
and Oztekin, 2001), peanut hulls, kernels and whole pods (Correa et al., 2001), rapeseed 
(Sun and Byrne, 1998), rough rice, brown rice, corn kernels, corn cobs, soybeans and red 
beans (Chen, 2000), switchgrass pellets (Colley et al, 2006), pea seeds (Chen, 2003), 
20 
 
alfalfa pellets (Fasina and Sokhansaj, 1993), corn stover (Igathinathane, 2007), semolina 
and farina (Erbas et al., 2005) and poultry litter pellets (McMullen et al., 2005). 
 
2.3.9. Rate of Moisture Sorption 
 In the previous section, it was mentioned that the hygroscopic nature of biological 
materials enables it to absorb or desorb moisture until equilibrium is reached with its 
environment.  In this section, the rate at which moisture is absorbed or desorbed by 
biological materials is discussed.  The Page empirical model, developed in 1949, has 
been widely used to describe the drying characteristics of various vegetables and fruits 
such as eggplant, red pepper, purslane and tomatoes (Doymaz and Pala, 2002; Doymaz, 
2004; Kashaninejad and Tabil, 2004; Doymaz, 2007).  Other mathematical models used 
to predict moisture sorption rate include: the Peleg model (Dadgar et al., 2004), 
exponential model (Colley et al, 2006; McMullen et al, 2005), two-term exponential 
model (Dadgar et al., 2004), Fick?s diffusion model (Chhinnan, 1984; Dadgar et al., 
2004).  The Page model (Equation 2.5) was found to be best model for describing drying 
characteristics of rice kernel, tomatoes, field peas and apricots (Banaszek and 
Siebenmorgen, 1993; Doymaz, 2007; Dadgar et al., 2004; Akpinar and Bicer, 2004). 
nullnullnull
null
null
null
nullnull
null
nullexpnullnullnullnull
null
null     (2.5) 
where, t = time (min) 
          M = instantaneous moisture content (%, w.b.) 
         M
i
 = initial moisture content (%, w.b.) 
         M
e
 = equilibrium moisture content (%, w.b.) 
21 
 
           k = constant 
           n = constant 
 
2.4. Thermochemical Conversion of Biomass 
2.4.1. Chemical Composition of Biomass 
 Most biomass materials contain the same constituents; cellulose, hemicelluloses, 
lignin, ash and a small amount of other extractives.  However, proportions of cellulose, 
hemicelluloses and lignin can vary drastically between materials.  Table 2.3 below gives 
cellulose, hemicelluloses and lignin percentages (by mass) of several types of biological 
materials. 
 
Table 2.3 ? Chemical Constituents of Several Types of Biomass 
Biological Material Cellulose (%) Hemicelluloses (%) Lignin (%) 
Birch Wood
[a]
42.6 31.8 21.0 
Fir
[b]
 48.9 11.9 31.2 
White Pine
[b]
53.3 12.3 26.7 
Black Birch
[b]
 54.0 29.0 9.4 
Poplar
[b]
 60.7 19.1 14.8 
Rice Straw
[b]
 41.1 29.5 5.1 
Corn Straw
[b]
 37.1 27.6 3.8 
Olive Waste
[a]
 44.8 18.5 28.0 
Wheat Straw
[a]
43.6 27.3 21.7 
Almond shell
[c]
 40.5 19.7 27.2 
Hazelnut shell
[c]
27.8 14.8 19.2 
English walnut shell
[c]
 60.2 13.2 18.6 
Macadamia nut shell
[c]
25.8 11.7 47.6 
[a]
 Zanzi et al., 2003; 
[b]
 Liu et al., 2008; 
[c]
 Wartelle and Marshall, 2001 
 
22 
 
 Cellulose, the most abundant polymer on earth, is a linear organic glucose 
polymer (C
6
H
10
O
5
)
n
, consisting of chains of (1,4)-D-glucopyranose units linked in the ?-
1,4-configuration with a molecular weight range of 300,000 to 500,000 Da (McKendry, 
2002a).  The crystalline portion of cellulose is insoluble and forms the skeletal structure 
of the cell (Yaman, 2004). 
Hemicelluloses are shorter, complex polysaccharides containing several glucose 
chains that are also located with cellulose in the cell wall.  Generally, hemicelluloses 
have the chemical formula (C
5
H
8
O
4
)
n
.  Shorter chains of xylose, arabinose and galactose, 
5-carbon monosaccharides, are connected in the hemicelluloses chains creating an 
amorphous branched polymer.  Xylan is the most abundant hemicellulose polysaccharide 
in biological materials.  The xylans exist in softwoods and hardwoods, composing of 10-
30% of the dry weight of the species, respectively (Yaman, 2004).  In contrast to 
cellulose, hemicelluloses are heterogenous branched polysaccharide that binds tightly 
(non-covalently) to the surface of the cellulose microfibril (McKendry, 2002a).  Unlike 
cellulose, hemicelluloses are soluble in dilute alkaline solutions (Yaman, 2004). 
 Lignin is the second most abundant polymer on earth and the largest structure in 
the cell wall.  Lignin is comprised of three carbon chains attached to six-carbon rings, 
called phenyl-propanes (Yaman, 2004).  Lignin has several important functions in the cell 
wall, ranging from structure rigidity to water movement because of its hydrophobic 
nature.  Woody plants species are typically composed of tightly bound fibers, giving a 
hard external surface from higher concentrations of lignin.  Herbaceous plants are 
composed with more loosely bound fibers, indicating a lower proportion of lignin 
23 
 
(McKendry, 2002a).  As shown in Figure 2.5, hemicelluloses, cellulose and lignin are the 
three main components that form a complex matrix in the plant cell wall. 
 
Figure 2.5 ? Hemicellulose, cellulose and lignin in plant cell wall (Sierra et al., 2008) 
 
2.4.2. Thermochemical Conversion Processes 
 Biomass can be converted to fuel, energy and value-added products by 
thermochemical processes, such as direct combustion, gasification, pyrolysis and 
liquefaction.  The main processes, intermediate and final products that can be obtained 
from these thermochemical conversion processes can be found in Figure 2.6 below: 
 
 
F
2.
to
(A
te
ty
5
co
p
 
2.
d
igure 2.6 ? 
4.2.1. Direc
Direct
 convert che
SABE, 200
mperatures 
pe of bioma
0% (wet bas
ntent and th
arameters fo
4.2.2. Gasif
Gasifi
eficient envi
Main proce
from 
t combustio
 combustion
mical energ
6; Goyal et 
around 800-
ss with the 
is) (Goyal e
ermal beha
r the design
ication 
cation occur
ronment.  T
sses, interm
the thermo
n 
 is defined 
y stored in 
al., 2008).  
1000?C.  It 
condition th
t al., 2008). 
vior of agric
 and operati
s at tempera
he carbon-ri
24 
ediate ene
chemical co
as the burnin
biomass into
Combustion
should be n
at the moistu
 Volatile rel
ultural resid
on of combu
tures in the
ch biomass 
rgy carriers
nversion of
g of bioma
 heat, mech
 of biomass
oted that com
re content o
ease, ash m
ues during c
stion system
 range of 75
is converted
 and final e
 biomass 
ss in the pre
anical powe
 produces ho
bustion is 
f the mater
elting point,
ombustion 
s (Werther
0-850?C in 
 to a low ca
nergy prod
sence of air 
r or electric
t gases at 
feasible for 
ial is less tha
 moisture 
are importan
 et al., 2000)
an oxygen-
lorific-value
 
ucts 
so as 
ity 
any 
n 
t 
. 
 
25 
 
gaseous fuel known as producer gas (ASABE, 2006).  The producer gas is a mixture of 
carbon monoxide, hydrogen, carbon dioxide, methane and nitrogen which can be burned 
for heat or power, or further refined to form liquid fuels and chemicals (ASABE, 2006).  
The gas is regarded as more versatile than the original solid biomass and can be burned to 
produce heat or steam or used in gas turbines to produce electricity (Demirbas, 2001).  
Commercial gasifiers are available in a range of sizes and types utilizing a variety of 
materials, including wood, charcoal, coconut shells and rice husks (Demirbas, 2001). 
 
2.4.2.3. Pyrolysis 
 Pyrolysis is carried out in the absence of oxygen at temperatures ranging from 
400-600?C (ASABE, 2006).  This process produces solid (bio-char), liquid (bio-oil) and 
gas compounds (methane, carbon dioxide and carbon monoxide).  According to many 
researchers, pyrolysis is regarded as the most suitable technique for the thermochemical 
conversion of biomass to liquid fuels (bio-oil), charcoal, acetic acid, acetone and 
methanol (Bahng et al., 2009; Demirbas, 2001; Goyal et al., 2008; McKendry, 2002b).  
The pyrolysis process can be further subdivided into three classes:  conventional (slow) 
pyrolysis (carbonization), fast pyrolysis and flash pyrolysis.  The range of the main 
operating parameters for pyrolysis processes are given in Table 2.4. 
Conventional, or slow, pyrolysis occurs under very slow heating rates (0.1-
1?C/min).  The slow heating rates enable higher char proportions instead of liquid and 
gaseous products (Demirbas and Arin, 2002).  Slow pyrolysis has been used for centuries 
for the production of charcoal from wood.  Das et al. (2008) conducted slow pyrolysis on 
26 
 
poultry litter and pine wood.  They report product yields of 35-58% wt% liquid, 30-45% 
wt% solid char and 12-21% wt% non-condensable gases. 
Table 2.4 - Range of the main operating parameters for pyrolysis processes 
(Demirbas and Arin, 2002) 
 Conventional 
pyrolysis 
Fast 
pyrolysis 
Flash         
pyrolysis 
Pyrolysis temperature (?C) 275 - 675 575 - 975 775 - 1025 
Heating rate (?C/min) 0.1 ? 1 10 ? 200 > 1000 
Particle size (mm) 5 ? 50 < 1 < 0.2 
Solid residence time (s) 450-550 0.5 ? 10 < 0.5 
 
A higher yield of liquid product is achieved in the fast pyrolysis procedure.  Rapid 
heating rates ranging from 10-200?C/min are used to obtain a high-grade bio-oil.  The 
fast heating rate allows the conversion of thermally unstable biomass compounds to 
liquid before they have a chance to form undesirable char (Bahng et al., 2009; Goyal et 
al., 2008).  According to Bridgwater (2002), the fast pyrolysis procedure produces 60-
75% wt% of liquid fuel, 15-25% wt% of char and 10-20% wt% of noncondensable gases, 
depending on the feedstock used.  Because the fast pyrolysis process takes from a few 
seconds to a few minutes, the chemical reaction kinetics, heat and mass transfer processes 
and phase transition phenomena affect product distribution (Bahng et al., 2009; 
Bridgwater, 2002).   
Flash pyrolysis has very short reaction times due to extremely fast heating rates (> 
1000?C/min) (Demirbas and Arin, 2002).  Particle size is important in the flash pyrolysis 
procedure because of the high heating rates.  Many researchers have used flash pyrolysis 
to produce higher yields of liquid and gaseous products (Goyal et al., 2008).  Sun et al. 
(2010) report product yields of 2-10% wt% char, 10-38% wt% liquid and 60-85% wt% 
27 
 
gaseous products for the flash pyrolysis of sawdust.  Gercel (2002) suggest that the 
particle size should be approximately 105-250 ?m.  Reactor design for flash pyrolysis is 
also imperative to achieve high efficiency of conversion.   
 
2.4.2.4. Liquefaction 
Liquefaction is the thermochemical conversion of biomass (in liquid phase) at low 
temperatures (250-350?C) and high hydrogen pressures (100-200 bar) into stable liquid 
hydrocarbons (McKendry, 2002b).  Liquefaction and pyrolysis are similar in that they 
both convert biomass into liquid products; however, liquefaction requires the use of a 
catalyst, such as sodium carbonate, to decompose macro-molecules into lighter molecules 
that repolymerize into oily compounds (Balat, 2008).  The complexity of the reactors and 
fuel-feeding systems and overall higher capital than for pyrolysis processes has caused 
interest in liquefaction to remain low (McKendry, 2002b; Goyal et al., 2008). 
 
2.4.3. Thermo-analytical techniques for biomass thermochemical conversion 
 Thermo-analytical techniques include thermo gravimetric analysis (TGA), 
differential scanning calorimetry (DSC) and differential thermal analysis (DTA).  The 
thermal behavior of biomass and the kinetic parameters of the thermal reactions are 
examined with the use of these techniques (Murugan et al., 2008; Bahng et al., 2009). 
2.4.3.1. Thermogravimetric analysis 
 Thermogravimetric analysis instruments are equipped with a balance that is used 
to monitor weight of the material as it is being heated.  Reina et al. (1998) used TGA to 
28 
 
determine degradation, oxidation and reduction reactions, evaporation, sublimation and 
other heat-changes that occur in biomass.  Mothe and Miranda (2009) used DTA-TGA 
methods to determine the decomposition of sugarcane bagasse and coconut fibers.  Both 
TG curves of sugarcane and coconut fibers showed moisture evaporation temperatures of 
100?C and 50?C, respectively.  Sugarcane demonstrated mass loss at temperatures of 
250-380?C, indicating hemicelluloses and cellulose decomposition.  Mass loss peaks at 
temperatures of 320?C and 330?C are attributed to the decomposition of hemicelluloses 
and celluloses in sugarcane bagasse, respectively.  Lignin degradation was determined to 
continue to 800?C and pyrolysis was determined to be essentially complete at 600?C.  
Coconut fibers demonstrated mass loss in the temperature range of 250-370?C with peaks 
at 300?C and 350?C indicating hemicelluloses and cellulose decomposition, respectively.  
Similar to sugarcane, coconut fiber lignin showed degradation until 600?C.  The 
pyrolysis of coconut fiber was considered essentially complete at 800?C with 20% 
residual char (of original weight) remaining versus 15% residual char of sugarcane 
(Mothe and Miranda, 2009). 
 
2.4.3.2. TGA-FTIR 
 Fourier transform infrared spectroscopy in combination with a thermogravimetric 
analyzer, TG-FTIR, is a well-known analytical tool for compositional analysis of evolved 
gaseous products from biomass pyrolysis (Bahng et al., 2009; Bassilakis et al., 2001).  
TGA combined with FTIR is a useful tool in dynamic analysis because it monitors 
continuously both the mass of the non-volatile material and the time dependent evolution 
of gases (Souza et al., 2009).  Quantitative data are routinely obtained on about 20 
29 
 
commonly evolved volatile species, including water, carbon monoxide, carbon dioxide, 
carbonyl sulfide (COS), sulfur dioxide (SO
2
), methane (CH
4
), ethylene (C
2
H
4
), hydrogen 
cyanide (HCN), ammonia (NH
3
), acetic acid, acetaldehyde, formic acid, formaldehyde, 
methanol, phenol, acetone and levoglucosan (Bassilakis et al., 2001; Bahng et al., 2009)  
One approach for measuring concentrations of these gases is based on evaluation of the 
concentration from the measured spectra using molecular spectroscopic databases, like 
HITRAN or QASOFT (den Blanken, 2008; Lee and Fasina, 2009).  This approach is 
novel for FTIR spectroscopy with an intermediate resolution (Bahng et al., 2009; den 
Blanken, 2008).  The combination of TGA and FTIR coupled together is extremely 
advantageous in the study of the pyrolysis because of the ?real-time? gas analysis.  In 
other words, as soon as the volatilization of biomass occurs during pyrolysis, the gases 
volatilized are captured, sent through a pre-heated Teflon tube and analyzed in a FTIR to 
quantify and identify volatile compounds.  There has been much research conducted on 
cellulose and lignin pyrolysis coupled with FTIR (Wang et al., 2007; Li et al., 2001; 
Bassilakis et al., 2001; Liu et al., 2008).  There has also been extensive research with 
other biomass types, such as dry distiller?s grains with solubles (DDGS) (Guintoli et al., 
2008), chicken manure (Guintoli et al., 2008), wood lignin (Liu et al., 2008), switchgrass 
(Lee and Fasina, 2009), coconut shells (Souza et al., 2009; Mothe and Miranda, 2009), 
corn stalk (Souza et al., 2009), peanut shells (Souza et al., 2009), tobacco (Bassilakis et 
al., 2001), wheat straw (Bassilakis et al., 2001) and sugarcane bagasse (Souza et al., 
2009; Mothe and Miranda, 2009; Bilba and Ouensanga, 1996). 
 
30 
 
2.5. Determination of Energy Requirement at Different Thermal Decomposition 
Temperatures 
 Differential scanning calorimetry (DSC) is another analytical tool used to 
determine heat capacities, phase transition and thermal behavior of biomass, such as the 
energy used for heating during pyrolysis.  Temperature ranges of thermal decomposition 
can be determined using TGA, however, the use of a DSC enables the determination of 
specific kinetic parameters during thermoconversion of biomass.  DSC instruments 
calculate specific heat by measuring the difference in amount of heat required to increase 
the temperature of the sample (in the sample cell) verses the reference cell, measured as a 
function of temperature (Murugan et al., 2008).  Enthalpy of pyrolysis is defined as the 
energy required to raise biomass from room temperature to the reaction temperature and 
convert the solid biomass into gaseous, liquid and char products (Daugaard and Brown, 
2003).  This value is important to the efficient design of pyrolysis reactors.  The design of 
the system may vary with respect to heat transfer or sizing of the reactor to handle 
multiple biomass feedstocks (Daugarrd and Brown, 2003).  Several researchers have used 
DSC in determining kinetic parameters necessary to model the kinetics of pyrolysis of 
biomass, such as lignin (Murugan et al., 2008), wheat straw (Stenseng et al., 2001), corn 
stalk (Cai and Chen, 2008), beech wood and artichoke thistle (Gomez et al., 2009), cotton 
stalk, peanut shells and pine (He et al., 2006) and palm oil mill by-products (Garcia-
Nunez et al., 2008).  Daugaard and Brown (2003) determined the enthalpy of pyrolysis of 
oak, oat hulls, pine and corn stover to be 2.0, 1.5, 0.8, 1.6 and 1.4 MJ/kg (dry basis), 
respectively.  Moisture content is another important parameter in the consideration of 
biomass for thermochemical conversion by pyrolysis.  It is feasible to thermally 
31 
 
decompose any type of biological material with a moisture content less than 50% (wet 
basis) (McKendry, 2002b).  However, pre-drying of biological materials before 
thermochemical conversion can take place has an economic impact on the overall net 
energy conversion.  Much is known about the energy required to evaporate pure water, 
however, less is known about the energy required to remove water from materials with 
strong water affinity, such as biological materials (Park et al, 2007).  Park et al. (2007) 
determined the energy required to remove moisture from softwood fibers to be 2.36 and 
2.24 KJ/g at moisture contents of 16.8% and 21.5%  (wet basis), respectively. 
 
2.6. Summary 
 The use of pecan shells as a feedstock for the production of renewable fuels 
through several thermochemical conversion processes is very promising.  Understanding 
the effect of particle size and moisture content on the physical properties of pecan shells 
is needed for the proper design of equipment and facilities needed to store, transport and 
handle the material.  The thermal behavior of pecan shells is also important to the 
efficient design of pyrolysis reactors or gasifiers.  This knowledge will eventually result 
in waste reduction and improved utilization of labor, machines and storage space.  
Furthermore, the development of a large scale biorefinery for pecan shells will be crucial 
to its value-added utilization in the bioenergy economy.
32 
 
2.7 References 
Akpinar, E.K., Bicer, Y. 2004. Modelling of the drying of eggplants in thin-layers. 
International Journal of Food Science and Technology 39: 1-9. 
 
Altuna?, E., Yildiz, M. 2007. Effect of moisture content of some physical and mechanical 
properties of faba bean (Vicia faba L.) grains. Journal of Food Engineering. 78: 178-183. 
 
ASABE Standard. 2000. D245.5 Moisture relationship of plant based agricultural 
products. St. Joseph, MI: ASABE. 
 
ASABE Standard. 2006. S593 Terminology and definitions for biomass production, 
harvesting and collection, storage, processing, conversion and utilization. St. Joseph, MI: 
ASABE. 
 
ASABE Standard. 2008. S319.4 Method of determining and expressing fineness of feed 
materials by sieving. St. Joseph, MI: ASABE. 
 
ASTM Standards. 2005. B 527. Standard test method for determation of tap density of 
metallic powders and compounds. West  Conshohocken, PA: ASTM. 
Aydin, C. 2002. Physical properties of hazel nuts. Biosystems Engineering 82(3): 297-
303. 
 
Bahng, M.K., Mukarakate, C., Robichaud, D., Nimlos, M. 2009. Current technologies for 
the analysis of biomass thermochemical processing: a review. Analytica Chimica Acta 
651: 117-138. 
 
Balat, M. 2008. Mechanisms of Thermochemical Conversion Processes. Part 3: Reactions 
of Liquefaction. Energy Sources, Part A, 30: 649-659 
 
Balasubramanian, D. 2001. Physical properties of raw cashew nut. Journal of 
Agricultural Engineering Research. 78(3): 291-297. 
 
Banaszek, M.M., Siebenmorgen, T.J. 1993. Individual rice kernel drying curves. 
Transactions of the ASAE 36(2): 521-528. 
 
Bansode, R.R., Losso, J.N., Marshall, W.E., Rao, R.M., Portier, R.J. 2003. Adsorption of 
volatile organic compounds by pecan-shells and almond-shells based granular activated 
carbons. Bioresource Technology 90: 175-184. 
 
Barbosa-Canovas, G.V., Ortega-Rivas, E., Juliano, P., Yan, H. 2005. Food Powders: 
Physical Properties, Processing, and Functionality. New York, NY: Kluwer 
Academic/Plenum Pub. 
 
33 
 
Bassilakis, R., Carangelo, R.M., Wojitowitcz, M.A. 2001. TG-FTIR analysis of biomass 
pyrolysis. Fuel 80: 1765-1786. 
 
Bilba, K., Ouensanga, A. 1996. Fourier transform infrared spectroscopic study of thermal 
degradation of sugar cane bagasse. Journal of Analytical and Applied Pyrolysis 38: 61-
73. 
 
Blahovec, J., Yanniotis, S.  2009.  Modified classification of sorption isotherms. Journal 
of Food Engineering 91: 72-77. 
 
Bridgwater, A.V. 2003. Renewable fuels and chemicals by thermal processing of 
biomass. Chemical Engineering Journal 91: 87-102. 
 
Brison, F. 1974. Pecan Culture. Austin, TX: Capital Printing. 247-254. 
 
Cai, J., Chen, S. 2008. Determination of drying kinetics for biomass by 
thermogravimetric analysis under nonisothermal conditions. Drying Technology 26: 
1464-1468. 
 
Chen, C.  2003.  Moisture sorption isotherms of pea seeds. Journal of Food Engineering 
58: 45-51. 
 
Chhinnan, M.S. 1984. Evaluation of selected mathematical models for describing thin 
layer drying of in shells pecans. Transactions of the ASAE 27: 610-615. 
 
Chum, H.L. 1991. Polymers from biobased materials. New Jersey, USA: Noyes Data 
Corporation. 
 
Colley, Z., Fasina, O.O., Bransby, D., Lee, Y.  2006.  Moisture effect of the physical 
characteristics of switchgrass pellets. Transactions of the ASABE.  (49)6: 1845-1851. 
 
Correra, P.C., Pinto, F.A.C., Queiroz, D.M., Afonso Jr., P.C. 2001. Equilibrium moisture 
content for peanut hulls, kernels and whole pods. An ASAE Meeting Presentation, Paper 
No. 01-6083, American Society of Agricultural Engineers, Annual International Meeting, 
July 30-August 1, Sacremento, CA. 
 
Dadgar S., Tabil, L.G., Crerar, B.  2004.  Moisture adsorption of field peas as a funtion of 
temperature and relative humidity during storage, An ASABE/CSAE Annual Meeting 
Presentation, Paper No. 046089, Sponsored by ASAE/CSAE, August 1-4, Ottawa, 
Ontario, Canada. 
 
 
Das, K.C., Garcia-Perez, M., Bibens, B., Melear, N. 2008. Slow pyrolysis of poultry litter 
and pine woody biomass:  Impact of chars and bio-oils on microbial growth. Journal of 
Environmental Science and Health Part A 43: 714-724. 
34 
 
 
Daugaard, D., Brown, R. 2003. Enthalpy for pyrolysis for several types of biomass. 
Energy and Fuels 17: 934-939. 
 
Demirba?, Ahyan. 1998. Kinetics for non-isothermal flash pyrolysis of hazelnut shell. 
Bioresource Technology 66: 247-252 
 
Demirbas, Ayhan. 2001. Biomass resource facilities and biomass conversion processing 
for fuels and chemicals. Energy Conversion and Management 42: 1357-1378. 
 
Demirbas, A., Gonenc, A. 2002. An overview of biomass pyrolysis. Energy Sources 
24:471-482. 
 
Demirba?, Ahyan. 2004. Combustion characteristics of different biomass fuels. Progress 
in Energy and Combustion Science 30: 219-230. 
 
den Blanken, I.R. 2008. Equivalent width method for quantitative FTIR analysis of 
Biomass Pyrolysis using HITRAN database. Master's Thesis, University of Technology 
of Eindhoven, TUE, Eindhoven, Netherlands. 
 
Deshpande, S.D., Bal, S., Ojha, T.P. 1993. Physical properties of soybean. Journal of 
Agricultural Engineering Research. 56(2): 89-98. 
 
DOE. 2008. 20% Wind Energy by 2030, Increasing Wind Energy's Contribution to the 
U.S. Electrical Supply. U.S. Department of Energy: Washington D.C. 
 
Doymaz, I. 2004. Effect of the pre-treatments using potassium metabisulphide and 
alkaline ethyl oleate on the drying kinetics of apricots. Biosystems Engineering 89: 281-
287. 
 
Doymaz, I. 2007. Air drying characteristics of tomatoes. Journal of Food Engineering 
78: 1291-1297. 
 
Doymaz, I., Pala, M. 2002. Hot-air drying characteristics of red pepper. Journal of Food 
Engineering 55: 331-335. 
 
Durakova, A.G., Menkov, N.D. 2005. Moisture sorption chacteristics of chickpea flour. 
Journal of Food Engineering 68: 535-539. 
 
EIA. 2008. Annual Energy Outlook 2008. Washington, DC: DOE Energy Information 
Administration. Available at: http://www.eia.doe.gov/piaf/aeo/index.html. Accessed 12 
April 2009. 
 
Erba?, M., Ertugay, M., Certel, M.  2005. Moisture adsorption behaviour of semilina and 
farina. Journal of Food Engineering 69: 191-198. 
35 
 
 
Fasina, O.O., Sokhansanj, S.  1993. Equilibrium moisture relations and heat of sorption 
of alfalfa pellets. Journal of Agricultural Enginnering Research 56: 51-63. 
 
Fayed, M.E., Skocir, T.S. 1997. Mechanical Conveyors: Selection and Operation. 
Lancaster, PA: Technomic Publishing Company, Inc. 
 
Fitzpatrick, J.J., Barringer, S.A., Iqbal, T. Flow property measurement of food powders 
and sensitivity of Jenike's hopper design methodology to the measured values. Journal of 
Food Engineering 61: 399-405. 
 
Ganesan, V., Muthukumarappan, K., Rosentrater, K.A. 2008. Flow properties of DDGS 
with varying soluble and moisture contents using jenike shear testing. Powder 
Technology 187: 130-137. 
 
Garcia-Nunez, J.A., Garcia-Perez, M., Das, K.C. 2008. Determination of kinetic 
parameters of thermal degradation of palm oil mill by-products using thermogravimetric 
analysis and differential scanning calorimetry. Transactions of the ASABE 51(2): 547-
557. 
 
Geisler, Malinda. 2009. Pecans. Agriculture Marketing Resource Center, Iowa State 
University. Available at: http://www.agmrc.org. Accessed 3 May 2010. 
 
Geldhart, D., Harnby, N., Wong, A. 1984. Fluidization of cohesive powders. Powder 
Technology 37(1): 25-37. 
 
Gercel, H.F. 2002. Production and characterization of pyrolysis liquids from sunflower 
pressed bagasse. Bioresource Technology 85: 113-117. 
 
Gomez, C., Velo, E., Barontini, F., Cozzani, V. 2009. Influence of secondary reactions on 
the heat of pyrolysis of biomass. Industrial Engineering and Chemical Research 48: 
10222-10233. 
 
Goyal, H.B., Seal, D., Saxena, R.C. 2008. Bio-fuels from thermochemical conversion of 
renewable resources: a review. Renewable and Sustainable Energy Reviews 12: 504-517. 
 
Guintoli, H., de Jong, W., Arvelakis, S., Spliethoff, H., Verkooijen, A.H.M. 2009. 
Quantitative and kinetic TG-FTIR study of biomass residue pyrolysis: dry distiller's 
grains with solubles (DDGS) and chicken manure. Journal of Analytical and Applied 
Pyrolysis 85: 301-312. 
 
Gupta, R.K., Das, S.K. 1997. Physical properties of sunflower seed. Journal of 
Agricultrual Engineering and Research 66(1): 1-8. 
 
36 
 
Harris, M.K., Cutler, B.L., Ring, D.R. 1986. Pecan nut loss from pollination to harvest. 
Journal of Economic Entomology 79(6): 1653-1657. 
 
He, F., Yi, W., Bai, X. 2006. Investigation on caloric requirement of biomass pyrolysis 
using TG-DSC analyzer. Energy Conversion and Management 47: 2461-2469 
 
Igathinathane, C., Womac, A.R., Sokhansanj, S., Pordesimo, L.O.  2007.  Moisture 
sorption thermodynamic properties of corn stover fractions. Transactions of ASABE 
50(6): 2151-2160. 
 
Jenike, A.W. 1964. Storage and flow of solids. Bulletin No. 123. University of Utah, Salt 
Lake City, Utah: Engineering and Experiment Station. 
Jenkins, B.M. 1989. Physical Properties of Biomass. In Biomass Handbook, 860-891. 
Amsterdam, Holland: Gordon and Breach Science. 
 
Johns, M., Marshall, W.E., Toles, C.A. 1999. The effect of activation method on the 
properties of pecan shells-activated carbons. Journal of Chemical Technology and 
Biotechnology 74: 1037-1044. 
 
Joshi, D.C., Das, S.K., Mukherjee, R.K. 1993. Physical properties of pumpkin seeds. 
Journal of Agricultural Engineering Research 54(3): 219-229. 
 
Kashaninejad, M., Tabil, L.G. 2004. Drying characteristics of purslane (Portulaca 
oleraceae L.). Drying Technology 22: 2183-2200. 
 
Kotwaliwale, N., Brusewitz, G.H., Weckler, P.R.  2004.  Physical characteristics of pecan 
components:  effect of cultivar and relative humidity. Transactions of the ASAE 47(1): 
227-231. 
 
Labuza, T.P. 1984. Moisture sorption: practical aspects of isotherm measurement and 
use. Minneapolis, MN: American Association of Cereal Chemist. 
 
Lee, S-B, Fasina, O. 2009. TG-FTIR analysis of switchgrass pyrolysis. Journal of 
Analytical and Applied Pyrolysis 86: 39-43. 
 
Leiras, M.C., Iglesias, H.A. 1991. Water vapor sorption isotherms of two cake mizes and 
their components. International Journal of Food Science 26: 91-95. 
 
Li, S., Lyons-Hart, J., Banyasz, J., Shafer, K. 2001. Real-time evolved gas analysis by 
FTIR method: an experimental study of cellulose pyrolysis. Fuel 80: 1809-1817. 
 
Liu, Q., Wang, S., Zheng, Y., Luo, Z., Cen, K. 2008. Mechanism study of wood lignin 
pyrolysis by using TG-FTIR analysis. Journal of Analytical and Applied Pyrolysis 82: 
170-177. 
 
37 
 
Luo, Z., Wang, S., Liao, Y., Zhou, J., Gu, Y., Cen, K. 2004. Research on biomass fast 
pyrolysis for liquid fuel. Biomass and Bioenergy 26: 455-462. 
 
Market Data.2009. Commodity Futures - Brent Crude Futures. Available at: 
http://www.bloomberg.com. Accessed 4 April 2009. 
 
McKendry, P. 2002a. Energy production from biomass (part 1): overview of biomass. 
Bioresource Technology 83: 37-46. 
 
McKendry, Peter. 2002b. Energy production from biomass (part 2): conversion 
technologies. Bioresource Technology 83: 47-54. 
 
McMullen, J., Fasina, O.O., Wood, C.W., Feng, Y.  2004.  Storage and handling 
characteristics of pellets from poultry litter. Applied Engineering in Agriculture 21: 645-
651. 
 
Milbrandt, A. 2008. A geographic perspective on the current biomass resource 
availability in the United States. NREL TP-560-39181. DOE, Washington, D.C. 
 
Mothe C.G., Miranda, I.C. 2009. Characterization of sugarcane and coconut fibers by 
thermal analysis and FTIR. Journal of Thermochemical and Analytical Calorimetry 
97:661-665. 
 
Murugan, P, Mahinpey, N., Johnson, K.E., Wilson, M. 2008.  Kinetics of the Pyrolysis of 
Lignin Using Thermogravimetric and Differential Scanning Calorimetry Methods. 
Energy and Fuels 22: 2720-2724. 
 
Ng., C., Marshall, W.E., Ramu, R., Bansode, R., Losso, J. 2003. Activated carbon from 
pecan shells: process description and economic analysis. Industrial Crops and Products 
17: 209-217. 
 
Nimkar, P.M., Chattopadhyay, P.K. 2001. Some physical properties of green gram. 
Journal of Agricultural Engineering Research 80: 183-189. 
Ozarslan, C. 2002. Physical properties of cotton seed. Biosystems Engineering 82(2): 
169-174.  
 
Pagano, A.M., Mascheroni, R.H.  2004.  Sorption isotherms for amaranth grains. Journal 
of Food Engineering 67: 441-450. 
 
Paraskeva, P., Kalderis, D., Diamadopoulos, E. 2008. Production of activated carbon 
from agricultural by-products. Journal of Chemical Technology and Biotechnology 83(5): 
581-592. 
 
38 
 
Park, S., Venditti, R., Jameel, H., Pawlak, J. 2007. Studies of the heat of vaporization of 
water associated with cellulose fibers characterized by thermal analysis. Cellulose 14: 
195-204. 
 
Perlack, R.D., Wright, L.L., Turhollow, A.F., Grahm, R.L., Stokes, B.J. Erback, D.C. 
2005. Biomass feedstock for a bioenergy and bioproducts indrustry: the technical 
feasibility of a billion ton annual supply.  Available at: 
http://www.feedstockreview.ornl.gov/pdf/billion_ton_vision.pdf. Accessed 1 May 2010. 
 
Reina, J., Velo, E. Puighaner, L. 1998. Thermogravimetic study of the pyrolysis of waste 
wood. Thermochimica Acta 320: 161-167. 
 
Renne, D., George, R., Wilcox, S., Stoffel, T., Myers, D., Heimiller, D. 2008. Solar 
Resource Assessment. NREL TP-581-42301. DOE, Washington, D.C. 
 
Schulze, D. 2006. Storage of powders and bulk solids in solos. Dietmar Schulze. 
Available at: http://www.dietmar-schulze.de/storagepr.html. Accessed 13 April 2010. 
 
Sentorun-Shalaby, C., Ucak-Astarlioglu, M.G., Artok, L., Sarici, C. 2006. Preparation 
and characterization of activated carbons by one-step steam pyrolysis/activation from 
apricot stones. Microporous and Mesoporous Materials 88: 126-134. 
 
Shamlou, P.A. 1988. Handling of Bulk Solids: Theory and Practice. London, UK: 
Butterworth-Heinemann. 
 
Sierra, R., Smith, A., Granda, C., Holtzapple, M. 2008. Producing fuels and chemicals 
from lignocellulosic biomass. Chemical Engineering Progress, August 1, 2008: 10-18. 
 
Singh, R.N.  2004.  Equilibrium moisture content of biomass briquettes. Biomass and 
Bioenergy 26: 251-253. 
 
Souza, B.S., Moreira, A.P., Teixeira, A.M. 2009. TG-FTIR coupling to monitor the 
pyrolysis products from agricultural residues. Journal of Thermal Analysis and 
Calorimetry 97: 637-642. 
 
Soysal, Y., ?stekin, S.  1999.  Equilibrium moisture content equations for some 
medicinal and aromatic plants. Journal of Agricultural Engineering Research 74: 371-
324. 
 
Soysal, Y., Oztekin, S. 2001. Comparison of seven equilibrium moisture content 
equations for some medicinal and aromatic plants. Journal of Agricultural Engineering 
Research 78: 57-63. 
 
39 
 
Stenseng, M., Jensen, A., Dam-Johnson, K. 2001. Investigation of biomass pyrolysis by 
thermogravimetric analysis and differential scanning calorimetry. Journal of Analytical 
and Applied Pyrolysis 58-59: 765-780. 
 
Sun, D., Byrne, C. 1998. Selection of EMC/ERH isotherm equations for rapeseed.  
Journal of Agricultural Engineering Research 69: 307-315. 
 
Sun, S., Tian, H., Zhao, Y., Sun, R., Zhou, H. 2010. Experimental and numerical study of 
biomass flash pyrolysis in an entrained flow reactor. Bioresource Technology 101: 3678-
3684. 
 
Teunou, E., Fitzpatrick, J.J. 1999. Effect of relative humidity and temperature on food 
powder flowability. Journal of Food Engineering 42: 109-116. 
 
USDA-NASS. 2008. Noncitrus Fruits and Nuts Summary. Washington, DC.: USDA 
National Agricultural Statistics Service. 
 
Wang, S., Liu, S., Luo, Z., Wen, L., Cen, K. 2007. Mechanism study on cellulose 
pyrolysis using thermogravimetric analysis coupled with infrared spectroscopy. Frontiers 
of Energy and Power Engineering in China 1(4): 413-419. 
 
Wartelle, L.H., Marshall, W.E. 2001. Nutshells as granular activated carbons: physical, 
chemical and adsorptive properties. Journal of Chemical Technology and Biotechnology 
76(5): 451-455. 
 
Werther, J., Saenger, M., Hartge, E.-U., Ogada, T., Siagi, Z. 2000. Combusition of 
agricultural residues. Progress in Energy and Combustion Science 26: 1-27. 
 
Wood, B., Payne, J., Grauke, L. 1994. An overview of the evolution of the U.S. pecan 
industry. In Pecan Technology, 1-11. New York, NY: Chapman and Hall. 
 
Woodcock, C.R., Mason, J.S. 1987. Bulk Solids Handling: An Introduction to the 
Practice and Technology. Glasgow, Scotland: Blackie and Son, Ltd. 
 
Worley, R. 1994. Pecan Production. In Pecan Technology, 12-38. New York, NY: 
Chapman and Hall. 
 
Yaman, Serder. 2004. Pyrolysis of biomass to produce fuels and chemical feedstocks. 
Energy Conversion and Management 45: 651-671. 
 
Zanzi, R., Sjostrom, K., Bjornbom, E. 2002. Rapid pyrolysis of agricultural residues at 
high temperature. Biomass and Bioenergy 23: 357-366. 
40 
 
CHAPTER 3. EFFECT OF MOISTURE AND PARTICLE SIZE ON PHYSICAL 
PROPERTIES OF PECAN SHELLS 
3.1. Introduction 
 Pecan (Carya illinionensis) is a tree nut crop grown in most of the southern states 
of the country.  Pecan is the only native tree nut grown for commercial use in the U.S.  
The country produces over 80% of the world?s supply (USDA-ERS, 2003; Geisler, 
2009). 
 
Figure 3.1 ? Ripe pecan (left) and cracked pecans with shell (right) (UGCE, 2008) 
 
Pecan shells are obtained as by-product of the pecan shelling process.  It is 
therefore an agricultural by-product that have current uses as value-added products, such 
as mulch, imitation fire logs, glue and soap abrasives and as a source of activated carbon 
for water filtration and commercial absorbents (Bansode et al., 2003; Ng et al., 2003; Guo 
and Rockstraw, 2007).   
41 
 
Pecan shells are however, a bulk solid.  Bulk solids consist of many particles of 
different sizes randomly grouped together to form a bulk (Woodcock and Mason, 1987).  
The physical characteristics of a bulk solid are important to the efficient design and 
selection of equipment for its handling and storage (Woodcock and Mason, 1987).  
Several published studies have reported that moisture content and particle size have 
significant influences on the physical properties of bulk biological materials, such as raw 
cashew nut (Balasubramanian, 2001), switchgrass pellets (Colley et al., 2006), poultry 
litter (Berhart et al., 2008) and barley straws, corn stover and switchgrass (Mani et al., 
2004). 
Therefore, the objectives of this study were to: 
(a)  Quantify the effect of particle size and moisture content on poured bulk 
density, particle density, tap bulk density, Hausner ratio and porosity of pecan 
shells; 
(b)  Quantify the effect of particle size, moisture content and applied pressure on 
the compressibility of pecan shells; 
(c)  Measure the flowability, cohesion and internal angle of friction of pecan 
shells at the different particle sizes and different moisture levels; and 
(d)  Determine the equilibrium moisture isotherm and rate of sorption of moisture 
by pecan shells. 
 
42 
 
3.2. Materials and Methods 
3.2.1. Preparation 
 The pecan shells used in this study were obtained from the Louisville Pecan 
Company (Louisville, AL).  Before use, the shells were stored in the Biosystems 
Engineering process engineering laboratory that has an average temperature of 
approximately 25?C.  Physical property experiments were performed at moisture contents 
of 4.42, 8.41, 14.73, 19.36 and 24.70% (w.b.).  In order to reach the desired moisture 
content, the samples were either: a) placed in an oven set at 45-50?C (for moisture 
content reduction) or, b) sprayed with a known quantity of distilled water (to increase 
moisture content of samples).  In both cases, the samples were allowed to equilibrate for 
at least 24 hours in an air-tight container before further use.  After the 24 hour moisture 
equilibration period, sample moisture content was verified by duplicate moisture content 
determinations using a moisture analyzer (MB45, Ohaus Corporation, Pine Brook NJ).  
Initial moisture content of the pecan shell sample was 14.59% (w.b.).  All moisture 
contents are reported in wet basis unless otherwise noted. 
 
3.2.2. Particle Size Distribution 
 Particle size distribution was determined by mechanical sieving according to 
ASABE Standard S319.4 (ASABE, 2008).  A Ro-Tap sieve shaker (W.S. Tyler Model 
RX-29, Mentor Ohio) was used for the sieving method.  Approximately 100 grams of 
pecan shells were placed in the first (top) sieve and shaken in the Ro-Tap sieve shaker for 
fifteen minutes.  After the shaking period, particle size distribution curves were obtained 
43 
 
from weights of samples retained on each sieve.  Geometric mean diameter (d
gw
) and 
geometric standard deviation (S
gw
) of the samples were also calculated according to the 
ASABE Standard S319.4 (ASABE, 2008) (Equations 3.1 ? 3.3). 
null
nullnull
nullnullnullnull
nullnull
null
?nullnull
null
nullnullnullnull
null
nullnullnull
null
null
nullnullnull
?nullnull
null
null
null
nullnullnull
null                                         (3.1) 
null
nullnullnull
nullnull
? null
null
nullnullnullnullnull
null
nullnullnull
nullnullnullnullnull
nullnull
null
null
null
nullnullnull
?nullnull
null
null
null
nullnullnull
null
null/null
                        (3.2) 
null
nullnull
null 
null
null
null
nullnull
nullnullnullnull
nullnull
 null
nullnullnull
null nullnullnullnull
nullnull
 null
nullnullnull
null
nullnull
null                       (3.3) 
where, d
gw
  =  geometric mean diameter (or median size) of particles by mass (mm) 
S
log
  =  geometric standard deviation of log-normal distribution by mass in ten-
based logarithm (dimensionless) 
 
S
gw
  = geometric standard deviation of particle diameter by mass (mm) 
W
i
   = mass on i
th
 sieve (g) 
n     = number of sieves plus one pan 
nullnull
null
   = nullnull
null
nullnull
nullnullnull
null
null/null
 
d
i
    = nominal sieve aperture size of the ith sieve (mm) 
d
i+1
 = nominal sieve aperture size os the next larger than ith sieve (mm) 
 
3.2.3. Mechanical Separation of Pecan Shells 
 Initial particle size analysis of the pecan shell sample demonstrated three separate 
peaks in the particle size distribution curve at 0.11, 1.4 and 2.4 mm aperture openings 
(Figure 3.2). 
 
44 
 
 
 
Figure 3.2 ? Particle size distribution of pecan shells 
The separation of the shells into the three distinct particle sizes observed in the initial 
sieve analysis was accomplished by the use of 30 inch Kason vibrating screen separator 
(Figure 3.3) (Kason Corporation, Millburn New Jersey).  The following screens were 
used in the separation: #10 (1.885 mm aperture size) screen and #14 (1.295 mm aperture 
size).  Pecan shells retained on the #10 screen were classified as ?coarse?, those retained 
on #14 but passed through #10 were classified as ?medium? and those that passed though 
#14 screen were classified as ?fine?.  These three particle size fractions (coarse, medium 
and fine) and the raw sample (making a total of four samples) were used for further 
analysis. 
 
0
5
10
15
20
25
30
012345
Distribution (%
)
Sieve Aperture Size (mm)
45 
 
 
Figure 3.3 ? Kason vibrating screen separator 
3.2.4. Compositional Analysis 
3.2.4.1. Energy Content 
 Heating value of pecan shells was obtained with an IKA C200 calorimeter (IKA 
Works Inc., Wilmington, NC).  To conduct measurements, approximately 0.5 g of sample 
was weighed and pressed into a pellet using IKA C21 pelleting press (IKA Works, 
Wilmington, NC).  After pressing, the pellet was reweighed and placed in the bomb 
according to manufacturer guidelines.  Energy content of pecan shells was determined in 
triplicate for the four samples. 
3.2.4.2. Ash Content 
 Ash content of samples was determined using the National Renewable Energy 
Laboratory?s (NREL) laboratory analytical procedure for the determination of ash in 
biomass (NREL, 2005).  To conduct a test, approximately one gram of each sample was 
placed in a crucible and then transferred into a muffle furnace (Type-1500, Thermolyn 
Sybron Corp., Milwaukee, WI).  Ashing of samples was carried out at 575? ? 25?C for 
three hours.  After the three hour period, the furnace was allowed to cool to 105?C.  The 
46 
 
samples were then removed from the furnace and allowed to cool in a dessicator before 
they were re-weighed.  Percent ash was calculated using Equation 3.4 and experiments 
were conducted in triplicate. 
 
% nullnullnull null 
nullnullnullnullnullnull
nullnullnullnullnullnullnullnull nullnullnullnull nullnullnull
nullnullnullnullnullnullnull
nullnullnullnullnullnullnullnull
nullnullnullnull nullnullnull nullnullnullnullnullnull
nullnullnullnullnullnull
 null 100                            (3.4) 
3.2.4.3. Ultimate Analysis 
 Carbon, hydrogen and nitrogen were determined (in duplicate) for the samples 
using an elemental analyzer (Model 2400 Series II Perkin-Elmer, Shelton, CT).  The 
samples were ground in a wiley mill to pass through a standard #40 screen before they 
were analyzed.  Approximately 300 mg of sample was used.  Helium was used as a 
carrier gas. 
 
3.2.4.4. Structural Carbohydrates 
 Van Soest analysis was used to determine the hemicelluloses, cellulose and lignin 
fractions of the sample.  In this process, the samples were separated progressively into 
neutral-detergent fiber (NDF), acid-detergent fiber (ADF) and acid-detergent-lignin 
(ADL).  This method of analysis has been used for other biological materials, such as 
hazelnut shells, wood, rice straw and corn stover (Haykiri-Acma, 2006; Liu et al., 2008).  
NDF represents the insoluble matrix of the pecan shells cell wall, substances covalently 
linked and intimately associated through hydrogen bonding, crystallinity, or other 
intramolecular association that they are resistant to solutions within the range of 
physiological concentrations.  ADF is considered as a rough estimate of the insoluble cell 
47 
 
wall, or cellulose, lignin, ash and acid insoluble ash (silica).  ADL was determined using 
the Klason lignin methodology, which does not recover all of the true lignin, but by 
hydrolysis of cellulose in the cell wall, the residue not hydrolyzed is an effective 
approximation of lignin content in forages (Van Soest and Robertson, 1985).  To 
determine NDF, ADF and ADL, the pecan shells were ground in a hammer mill to pass 
though a 1 mm screen.  Approximately 1.0 g of pecan shells were weighed (Model 
AM50, DeltaRange, Mettler-Toledo, Heightstown, NJ), recorded and placed in a 600 ml 
beaker.  Fifty milliliters of room-temperature solution (2% w/v amylase and deionized 
water) was added to the beaker containing the sample and placed on a hot plate and 
evenly boiled for 40 minutes.  After 40 minutes, another 50 ml of enzyme solution was 
added to the beaker and replaced on the hot plate and boiled for another 20 minutes.  
After the hour of boiling, the enzyme solution was removed from the beaker, washed 
with deionized water in a filtering crucible (50 ml, 40-60 ?m) and the solution of water 
and enzyme solution was removed via vacuum pump.  The crucibles were dried overnight 
in a forced air oven set at 105?C (Precision Scientific, Chicago, IL).  Then, the crucibles 
were removed, placed in a dessicator to cool, reweighed and recorded.  NDF was 
determined for each particle size.  For ADF determinations, approximately 1.0 dried and 
ground pecan shells were placed in a beaker and 100 ml acid-detergent solution was 
added.  The beaker, containing solution and sample, was placed on a hot plate and boiled 
for 1 hour.  Then the beaker was removed from the hot plate and the sample and solution 
mixture was rinsed with deionized water, filtered through a crucible and liquids removed 
via vacuum pump.  The crucibles were dried overnight in an oven, allowed to cool, and 
weights recorded.  ADL was determined using the residual of the ADF procedure.  
48 
 
Seventy-two percent (w/w 12M sulfuric acid and deionized water) was added to the 
crucibles (approx. 20 ml) and stirred using a glass stirring rod.  The solution was stirred 
occasionally for 3 hours after which the crucibles were placed in a vacuum manifold and 
residue was washed with deionized water, rinsing off crucible sides, etc. to make sure 
that no fiber was left unfiltered.  The crucibles were dried in a forced oven set at 105?C 
overnight.  The crucibles were removed from the oven, allowed to cool in a dessicator 
and weights recorded.  ADL was considered lignin composition whereas cellulose and 
hemicelluloses were calculated using the following Equations (3.5-3.6): 
% nullnullnullnullnullnullnullnullnull null %nullnullnull null%nullnullnull    (3.5) 
% nullnullnullnullnullnullnullnullnullnullnullnullnullnull null %nullnullnull null%nullnullnull             (3.6) 
NDF, ADF and ADL were conducted in duplicates. 
3.2.5. Poured Bulk Density 
 Poured bulk density was determined by a bulk density measurement apparatus 
(Burrows Co., Evanston, IL) and according to ASABE Standard S269.4 (2002).  This 
procedure involves pouring the bulk solid into a container of known volume (947 mm
3
) 
via a funnel placed above the container.  The material was then leveled off the top of the 
container and weighed.  The bulk density (?
b
) of the pecan shells were calculated by 
dividing the mass of sample in the container (m
c
) by the volume of the container (V
c
) 
(Equation 3.7).  This procedure was done in triplicate and the average value was reported. 
null
null
 null 
null
null
null
null
     (3.7) 
 
 
49 
 
3.2.6. Particle Density 
 Particle density of pecan shell samples was measured by gas comparison 
pycnometry (Model AccuPyc II 1340, Micrometrics Instrument Corp., Norcross, GA) 
that involves the use of known quantity of inert gas under pressure (helium) that is 
allowed to flow from a previously known reference volume into a cell containing a 
sample of pecan shells.  By applying the ideal gas law to the pressure change from the 
reference cell to the sample cell, the pycnometer calculates the volume of the material in 
the sample cell.  True particle density (?
p
) was taken as the ratio of the mass of material 
in the sample cell (m
p
) to the volume (V
p
) measured by the pycnometer (Equation 3.8).  
Sample mass was obtained with a digital balance accurate to 0.001 grams (Model 
AR3130, Ohaus Corp., Pinebrook, NJ).  This procedure was carried out in triplicate. 
null
null
 null 
null
null
null
null
      (3.8) 
3.2.7. Tap Bulk Density 
 Tap bulk density of pecan shell samples were measured using an automated tap 
density tester (Model TD-12, Pharma Alliance Group Inc., Valencia, CA) that is based on 
ASTM Standard B 527 (ASTM, 2005).  A 250 ml graduated cylinder was filled (to 250 
ml) with a sample and weighed using a balance accurate to 0.01 g (Model PE3600 
DeltaRange, Mettler-Toledo, Heightstown, NJ).  The cylinder was then placed in the tap 
density tester.  The tester taps the graduated cylinder in two steps.  For the first step, the 
cylinder was tapped 500 times at a rate of 300 taps per minute.  The volume of the 
material in the cylinder was measured after the first tapping.  This was followed by 
tapping of the cylinder for 750 times at the rate of 300 taps per minute.  After this step, 
50 
 
the new volume was measured.  The tap bulk density (?
t
) of the pecan shells was 
calculated as the mass of sample in the cylinder (m
c
) divided by the final tap volume of 
the sample (V
t
) (Equation 3.9).  This procedure was carried out in triplicate. 
null
null
 null 
null
null
null
null
     (3.9) 
 
3.2.8. Hausner Ratio 
 Hausner ratio is a function of the tap and poured bulk density.  The Hausner ratio 
is often used to indicate how easily a powder is fluidized (Geldhart et al., 1984).  It is 
mathematically defined by Equation 3.10 and was calculated as the ratio of tap bulk 
density (?
t
) and poured bulk density (?
b
).   
null
null
 null 
null
null
null
null
     (3.10) 
Powders that are easily fluidized have Hausner ratios less than 1.25, whereas powders 
with fluidization problems have Hausner ratios greater than 1.40 (Geldhart et al., 1984). 
 
3.2.9. Porosity 
 Porosity is the percentage of a bulk solid that is occupied by air spaces 
(Woodcock and Mason, 1987).  Porosity is mathematically defined by Equation 3.11 and 
was calculated using the bulk and particle densities that were obtained in the previous 
sections. 
null null null1null 
null
null
null
null
null null100                                                    (3.11) 
 
51 
 
3.2.10. Mechanical Compressibility 
 A texture analyzer (Model TA-HD, Stable MicroSystems, Surrey, UK) was used 
to measure the mechanical compressibility of pecan shell samples.  The measurement 
system consisted of a compression cell (internal diameter of 49.55 mm and internal 
height of 101.83 mm) and a tight-fitting piston (diameter of 49.00 mm) that was attached 
to the crosshead of the texture analyzer.  For each test, a known mass of pecan shells 
(Model PM2000, DeltaRange, Mettler-Toledo, Heightstown, NJ, accurate to 0.01 g) was 
loaded into the compression cell.  The piston compressed the sample (at a rate of 1 mm/s) 
inside the cell until a consolidating pressure of 1.5, 3, 6, 9, 12, or 15 kPa was reached.  
The software provided by the manufacturer of the analyzer recorded the force (or 
pressure) on the sample and the distance traveled (by compression of the sample) by the 
piston.  This procedure was carried out in triplicate.  Mechanical compressibility (Cm, %) 
was calculated as follows: 
nullnull null  100null
null
null
null null
null
null
null
null null 100null1null 
null
nullnull
null
nullnull
null                                (3.12) 
where, V
i
 is initial volume (cm
3
), V
f
 is compressed volume (cm
3
), ?
bi
 is initial bulk 
density (kg/m
3
) and ?
bf
 is final (compressed) bulk density (kg/m
3
). 
 
3.2.11. Flowability 
 An automated shear tester (ShearScan TS-12, Sci-Tec Inc., Worthington, Ohio) 
was used to quantify the flow behavior of pecan shells.  The shear cell used in this 
procedure was of the annular ?fixed? cell type.  The cell (height of 27 mm) consisted of a 
base ring attached to an inner (diameter of 109.6 mm) and outer ring (diameter of 55.0 
52 
 
mm), and to a cell lid.  To run a sample, the space between the inner and outer rings was 
filled with a pecan shell sample.  The cell lid was then placed on top of the sample 
between the inner and outer rings. 
 Using settings on the software provided by the manufacturer, a compression load 
was applied to the sample (in the cell) at a constant rate of 7.5 mm/min until the 
consolidating stress, or preset normal load, was reached.  This was followed by twisting 
of load cell at 2.5 mm/min.  Shear stress was determined by the shear cell load and 
increased until it reached the predetermined or programmed steady state point.  When this 
point was reached, the load cell gradually lifted and decreased the load to a preset point.  
Shear action began once again until the sample failed due to shear stress.  Consolidating 
stresses of 3, 6, 9, 12, 15 and 21 kPa were used for all samples.  Each consolidating stress 
was run in triplicate.  Maximum shear stress was recorded for six different loads (25, 35, 
45, 55, 65 and 75% of the consolidating stress).  Yield locus was obtained by plotting the 
maximum shear stress versus normal stress.   
 The software provided by the manufacturer was used to calculate ultimate yield 
stress (UYS) and major consolidating stress (MCS) (see Equations 3.13 and 3.14, 
respectively).  The flow index was obtained as the inverse of the slope of the plot of UYS 
vs. MCS. 
 
nullnullnull null 
nullnullnullnullnull nullnullnullnullnull
nullnullnullnull
    (3.13) 
nullnullnull null null
nullnull nullnull
null
 nullnullnull
null
nullnull null
null 
nullnullnull
null
  null 
nullnullnull
null
 null
nullnull1null sinnullnullnull null
null
nullnullnullnull
null   (3.14) 
nullnullnullnull tannullnullnullnull    (3.15)  
53 
 
Where: 
? = shear stress (kPa) 
? = normal stress (kPa) 
? = angle of internal friction (?) 
c = cohesion (kPa) 
nullnull nullnull 
null
nullnullnullnull
 , and c is cohesion (defined as ? at ? = 0).  
 
3.2.12. Rate of Moisture Sorption 
 Agricultural materials are hygroscopic in nature.  Therefore, they have the ability 
to exchange moisture with the atmosphere (Singh, 2004; Igathinanthane et al., 2007).  
This experiment was designed to determine the rate at which pecan shells absorb or 
desorb moisture from an environment.  Moisture sorption property of pecan shell samples 
was carried out in a chamber (0.9 m x 1.8 m x 0.9 m) that was supplied with conditioned 
air from a conditioner (Model #9221-2110, Parameter Generation and Control, Inc., 
Black Mountain, NC).  The combination of temperatures and relative humidity levels 
used in this study were 15, 25 and 35?C (? 1?C) and 50, 65 and 80?C (? 3%) 
respectively.  A schematic diagram of the test system is shown in Figure 3.4 below: 
 
 
 
m
O
p
h
th
M
ex
th
ea
B
p
Figure 3
To co
esh basket h
H) (accurate
ersonal com
ours, depend
e chamber)
anchester, U
periment w
an 0.01 g w
ch experim
rook NJ).  E
erformed in 
 
.4 ? Schema
nduct a test,
ung from a 
 to 0.01g)).
puter at 10 m
ing on the t
.  A commun
K) was use
as considere
ithin a span 
ent was veri
xperimenta
duplicate. 
tic diagram
 a thin layer
digital balan
  Sample ma
inute interv
emperature 
ication soft
ed to interfac
d complete 
of one hour
fied using a
tion at each 
54 
 of rate of 
 of pecan sh
ce (Model 
ss was mon
als until eq
and relative 
ware (Wind
e the balanc
when the m
.  The moist
 moisture an
temperature
set up for m
ells (100-10
PM4600, M
itored and c
uilibrium w
humidity of
mill RS232,
e to the per
ass of samp
ure content 
alyzer (MB
 and relative
oisture sor
5 g) was pla
ettler-Toled
ontinuously
as reached (
 the conditio
 Windmill S
sonal compu
le did not ch
of a sample 
45, Ohaus C
 humidity c
ption study
ced in a wir
o, Columbu
 recorded on
generally 48
ned air ente
oftware Ltd
ter.  An 
ange by mo
after the end
orporation, 
ombination 
 
 
e 
s, 
 a 
-60 
ring 
., 
re 
 of 
Pine 
was 
55 
 
 
3.2.13. Equilibrium Moisture Isotherm 
 A water activity instrument (HygroLab 2 ? H3, Rotronic Instrument Corp., 
Huntington, NY) was used to measure the equilibrium relative humidity of 
preconditioned pecan shell samples.  To conduct a test, the conditioned sample was 
placed in the sample holder of the water activity instrument.  The water activity probe 
was then placed on the top of the sample holder to form a sealed measurement system.  A 
small fan, located in the probe, circulated air within the sample container and a thin film 
capacitance sensor measured relative humidity from 0 to 100% (with an accuracy of ? 
1.5%) while a platinum RTD (resistance temperature detector) temperature probe (with 
an accuracy of ? 0.3?C) was used to measure temperature.  The measurement probes and 
containers were placed into a temperature controlled-chamber (Model AA-5460A, Espec. 
Corp., Hudsonville, MI) set at 10, 20, 35, or 50?C.  The pecan shells were adjusted to 
moisture content levels of 4.12, 7.06, 11.28, 14.56, 19.43, and 25.58% (wet basis), using 
the moisture content adjustment methods discussed previously.  Relative humidity and 
dry-bulb temperature output from the water activity meter were continuously recorded on 
a personal computer until equilibrium was reached (usually less than 3 hours).  Each 
combination of sample moisture content and chamber temperature were performed in 
triplicate.  The equilibrium moisture content (EMC) and equilibrium relative humidity 
(ERH) were taken as the average of the three moisture contents and relative humidity 
combinations for each sample.  
 
56 
 
 
3.2.14. Data Analysis 
 Statistical analysis was performed on all data sets using SAS statistical software 
package (Version 9.1, SAS institute Inc., Cary, NC, 2009-2010).  Microsoft Excel 
(Microsoft Office XP Professional, 2007) software was used for data analysis and data 
plotting. 
 
 
3.3. Results and Discussion 
3.3.1. Particle Size Distribution  
 Particle size analysis of the three fractions of pecan shells (fine, medium and 
coarse) was carried out by sieve analysis.  Figure 3.5 shows that the raw sample was 
successfully separated into three sizes.  The geometric mean diameter (d
gw
) and 
geometric standard deviation (S
gw
) of the three fractions and the raw pecan shell samples 
are given in Table 3.1 (Equations 3.1-3.3, ASABE 2008).  Mechanical separation of raw 
pecan shells reduced the variability in particle size (based on the coefficient of variation 
values in Table 3.1).  
 
57 
 
 
Figure 3.5 ? Particle size distribution of pecan shells 
 
Table 3.1 ? Geometric mean diameter and geometric standard deviation of 
pecan shells fractions 
  d
gw  
(mm) S
gw  
(mm) CV (%) 
fine 0.212 0.244 115.09 
medium 1.241 0.442 35.62 
coarse 2.194 0.696 31.72 
raw 1.200 2.262 188.50 
 
3.3.2. Chemical Composition Analysis 
3.3.2.1. Energy and Ash Content 
 The energy and ash contents of the four particle size fractions of pecan shells are 
given in Table 3.2.  In general, energy content decreased as particle size increased, but 
the decrease was not significant except for the fine fraction.  The heating values are 
comparable to values reported for other biomass feedstocks, such as switchgrass (19.22 
0
5
10
15
20
25
30
35
40
012345
Distribution (%
)
Sieve Aperature Size (mm)
fine
medium
coarse
raw
58 
 
MJ/kg), peanut hull (19.83 MJ/kg), almond shell (20.2 MJ/kg) and wood pellets (19.0 
MJ/kg) (Fasina, 2007; Blesa et al., 2001; Biagini et al., 2006). 
Ash content of the pecan shell fractions ranged from 3.02% in the fine fraction to 
2.32% in the medium fraction.  In general, the ash content of the pecan shell fractions are 
similar to that of other reported agricultural materials, such as almond shell (2.4%) and 
wood pellets (2.3%) (Di Blasi, 1999; Blesa et al., 2001).  Table 3.2 gives the energy 
content (MJ/kg) and ash content (%) of the four particle size fractions of pecan shells.  
Statistical analysis of energy and ash content was performed using analysis of variance 
(ANOVA) and Tukey?s multiple comparison tests (at ?=0.05) in SAS Version 9.3 (2009). 
 
 
Table 3.2 ? Energy and ash content (dry basis) of pecan shells fractions 
Sample HHV (MJ/kg) St.dev. Ash Content (%) St.dev. 
fine 20.50
[a]
 0.09 3.02
[a]
 0.06 
medium 19.78
[b,c]
 0.03 2.32
[c]
 0.09 
coarse 19.50
[c]
 0.11 2.53
[b]
 0.03 
raw 20.06
[b]
 0.03 2.49
[b]
 0.01 
        samples (n=3) in each column with same letter are not significantly different (P<0.05) 
 
3.3.2.2. Ultimate Analysis 
 Carbon, hydrogen and nitrogen contents of the pecan shells fractions are shown in 
Table 3.3.  The fine fraction of pecan shells had significantly (P < 0.05) less carbon than 
the medium, coarse and raw fractions.  As reported in the previous section, the fine 
fraction had the highest energy content (Table 3.2).  There is no scientific explanation for 
the fine fraction to have the lowest carbon level even though it has the highest energy 
59 
 
content.  However, it is postulated that the fine fraction contains soil particles from 
harvesting processes or pith particles (from the inside of the shell, between nut-halves).  
The pecan shell pith contains higher concentrations of phenolic compounds and tannins 
(compared to polysaccharide carbon chains) that have been extracted for the production 
of resins (McGraw, 1993; Hamud and Chung, 1989).  Hydrogen percentages were not 
significantly affected by particle size in the four fractions.  In terms of nitrogen 
(percentage by mass, dry basis), a trend of decreased nitrogen as particle size increased 
was observed.  The carbon, hydrogen and nitrogen percentages are similar to other 
agricultural materials, such as cherry bay branch, cotton and maize straw and hazelnut 
shell (Hu et al., 2007; Haykiri-Acma, 2006). 
Table 3.3 ? Carbon, hydrogen and nitrogen contents (dry basis) of pecan shells 
fractions 
Component Fine Medium Coarse Raw 
Carbon (%) 42.38
[b]
 46.84
[a]
 46.14
[a]
 46.97
[a]
 
(0.778) (0.18) (0.35) (0.39) 
Hydrogen (%) 5.55
[a]
 5.41
[a]
 5.67
[a]
 5.42
[a]
 
(0.01) (0.15) (0.08) (0.01) 
Nitrogen (%) 0.60
 [a]
 0.44
[a,b]
 0.23
[b]
 0.57
[a]
 
?? (0.06) (0.02) (0.10) (0.11) 
Note:  Standard deviations are notated in parenthesis below each measurement. 
samples (n=3) in each row with same letter are not significantly different (P<0.05) 
 
3.3.2.3. Structural Carbohydrates 
 In general, the cellulose composition in pecan shells increased as particle size 
fractions increased.  As expected from the ultimate analysis previously discussed, the fine 
fraction had significantly (P < 0.05) less cellulose and hemicelluloses than the medium, 
60 
 
coarse and raw particle size fractions (hence less carbon).  There is no clear trend of an 
effect of particle size on hemicelluloses and lignin percentages.  In general, the lignin 
composition of the samples was similar reported values of fir (31.2%), white pine 
(26.7%), olive waste (28.0%) and almond shell (27.2%) in Table 2.3.  The structural 
carbohydrate analysis of particle size fractions of pecan shells is given in Table 3.4 
below. 
Table 3.4 ? Cellulose, hemicelluloses and lignin contents (dry basis) of pecan 
shells fractions 
Sample Cellulose (%)
A
Hemicelluloses (%)
B
Lignin (%)
C
 Extractives (%)
D
 
fine 9.53
[a]
 7.38
[a]
 26.53
[b]
 
56.56 
medium 24.17
[b]
 18.43
[c]
 25.16
[a]
 
32.24 
coarse 36.03
[c]
 16.33
[b]
 30.59
[d]
 
17.05 
raw 35.56
[c]
 16.73
[b]
 29.78
[c]
 
17.93 
samples (n=3) in each column with the same letter are not significantly different (P<0.05) 
A
 = Acid-detergent fiber (ADF) ? acid-detergent lignin (ADL) (Eq. 3.5) 
B
 = Neutral-detergent fiber (NDF) ? ADF (Eq. 3.6) 
C
 = ADL 
D
 = Extractives include lipids, proteins or non-structured carbohydrates (i.e. starch) 
 
3.3.3. Particle Size Effect on Physical Properties of Pecan Shells 
3.3.3.1. Poured Bulk Density 
 The poured bulk density of pecan shells ranged from 417.39 to 469.13 kg/m
3
 and 
decreased with an increase in particle size.  Therefore, the volume required to store or 
transport (with identical mass) pecan shells is significantly increased as particle size 
increases.  Mani et al. (2006) found that bulk and particle densities of wheat straw, barley 
straw, corn stover and swichgrass decreased with increased particle size.  It should be 
mentioned that although the poured bulk density measurements were similar to other 
61 
 
biological materials (soybean ? Deshpande, et al., 1993; faba bean ? Altuna? and and 
Yildiz, 2007; cashew nut ? Balasubramanian, 2001; alfalfa cubes and pellets ? Fasina and 
Sokhansanj, 1992), it is noticeably higher than some other bioenergy feedstocks, such as 
switchgrass and corn stover (<200 kg/m
3
) (Mani et al., 2006; Fasina, 2006; Bernhart and 
Fasina, 2009).  The average poured bulk density (?
b
) of pecan shell samples are shown in 
Table 3.5. 
Table 3.5 ? Poured bulk density of pecan shells fractions 
Sample Mean ?
b
 (kg/m
3
) St. dev. 
fine 459.93
[a]
 7.02 
medium 434.86
[b]
 6.44 
coarse 417.39
[c]
 5.23 
raw 469.13
[a]
 6.09 
samples (n=3) with same letter are not significantly different (P<0.05) 
 
3.3.3.2. Particle Density 
 The particle density of pecan shells showed a significant decrease (P < 0.05) as 
particle size increased with particle densities ranging in value from 1498.51 to 
1439.11kg/m
3
.  Therefore, the individual particles became less dense as the sample 
increased in size.  It is postulated that the process of grinding or breaking of the particles 
removed some of the open pores within the particles.  The particle density (?
p
) of pecan 
shell samples is shown in Table 3.6 below. 
 
 
 
62 
 
Table 3.6 ? Particle density of pecan shells fractions 
Sample Mean ?
p
 (kg/m
3
) St. dev. 
fine 1498.51
[a]
 2.40 
medium 1456.98
[b]
 4.24 
coarse 1439.11
[c]
 2.75 
raw 1432.41
[c]
 2.88 
samples (n=3) with same letter are not significantly different (P<0.05) 
 
3.3.3.3. Tap Bulk Density 
 The tap bulk density of pecan shells significantly decreased (P < 0.05) from 
601.67 kg/m
3
 to 486.00 kg/m
3
 with an increase in particle size.  It is postulated that the 
smaller size of the particles in the fine fraction made it easier to fill the voids during 
tapping.  The larger sized samples have larger sized particles that cannot easily fill the 
void space during tapping.  Similar trends have been reported for biological materials 
such as wheat straw, barley straw, corn stover and switchgrass (Mani et al., 2004; Lam et 
al., 2008).  Statistical analysis showed that there was a significant effect (P < 0.05) of 
particle size on the tap bulk density of pecan shells. Table 3.7 shows the average tap bulk 
density (?
t
) of pecan shell samples. 
Table 3.7 ? Tap bulk density of pecan shells fractions 
Sample Mean ?
t
 (kg/m
3
) St. dev. 
fine 601.67
[a]
 9.87 
medium 546.33
[c]
 5.03 
coarse 486.00
[d]
 2.65 
raw 579.33
[b]
 2.52 
samples (n=3) with same letter are not significantly different (P<0.05) 
 
63 
 
3.3.3.4. Porosity 
 Porosity significantly increased (P < 0.05) from 0.672 to 0.710 as particle size 
increased from the fine to coarse fraction, respectively.  These values are typical of 
extremely irregular-shaped particles that have cohesive tendencies (Woodcock and 
Mason, 1987).  It is suspected that the hardness of the pecan shells and shelling process 
operations caused the irregularities in the particle shapes of pecan shells, hence the high 
porosity values.  It should be noted that the porosity values of spherical particles is about 
0.4.  Increasing porosity values with increasing particle size has been reported for 
switchgrass and corn stover (Lam et al., 2008).  High porosity values are a sign of 
logistical and economic problems that can be encountered during the storage and 
transportation of pecan shells, unless some form of densification is utilized (Fasina, 
2007).  Table 3.8 below gives the average porosity values of the pecan shell fractions. 
 
Table 3.8 ? Porosity of pecan shells fractions 
Sample Mean St. dev. 
fine 0.693
[b]
 0.004 
medium 0.702
[a,b]
 0.004 
coarse 0.710
[a]
 0.003 
raw 0.672
[c]
 0.005 
samples (n=3) with same letter are not significantly different (P<0.05) 
 
3.3.3.5. Hausner Ratio 
 Hausner ratio (H
R
) values significantly increased (P < 0.05) with increase in 
particle size (Table 3.9).  The values of Hausner ratio show that the coarse and raw 
64 
 
samples can be easily fluidized since their values are lower than 1.25.  However, the fine 
and medium samples may present fluidization problems (Barbosa-Canovas et al., 2005). 
Table 3.9 ? Hausner ratio of pecan shells fractions 
Sample Mean H
R
 St. dev. 
fine 1.27
[a]
 0.01 
medium 1.28
[a]
 0.02 
coarse 1.15
[b]
 0.01 
raw 1.15
[b]
 0.00 
samples (n=3) with same letter are not significantly different (P<0.05) 
 
3.3.3.6. Mechanical Compressibility 
 As expected, the compressibility of the samples decreased with increase in 
particle size.  The percent compressibility of pecan shells ranged from 1.33% at applied 
pressure of 1.5 kPa to 10.5% at an applied pressure of 15 kPa for the coarse and fine 
samples, respectively.  A trend of increased compressibility as particle size decreased and 
applied pressure increased is shown in Figure 3.6.  The highest percentages of 
compressibility were obtained from the fine sample, which is indicative of its 
deformation property.  The fine sample?s consistency was similar to fine powders, 
whereas the larger particle size samples had larger particulates that were not able to 
deform as easily as the smaller particle size sample.  The inability of the larger sized 
particles to fill the inter-particle void space (as was observed for tap bulk density) was 
also the reason for reduced compressibility values.  High compressibility values can be 
associated with reduced flowing ability (Tabil and Sokhansanj, 1997; Barbosa-Canovas, 
2005).  According to Fayed and Skocir (1997), percent compressibility values ranging 
from 5 to 15% have excellent flow.  Based on this and the values presented in Figure 3.6., 
65 
 
pecan shells can be classified as free-flowing granules with excellent flow.  Equation 
3.21 was developed to show the relationship between mechanical compressibility (%), 
mean diameter particle size (mm) and pressure (kPa) of pecan shells. 
 
C
M
 = 4.432 ? 4.307*PS + 0.941*PS
2
 + 4.917*log(?) , R
2
 = 0.942  (3.21) 
Where, C
M
 is percent compressibility, PS is mean geometric particle size (mm) and ? is 
applied pressure (kPa). 
 
Figure 3.6 ? Effect of particle size and pressure on the mechanical compressibility of 
pecan shells 
3.3.3.7. Flowability 
 The flow function (FF) plots (Figure 3.7) or plots of the ultimate stress (UYS) vs. 
the major consolidating stress (MCS) indicate the flowability of the fine pecan shell 
sample is higher than that of the medium and coarse samples.  This is confirmed from the 
values of the flow index (obtained from the inverse of the slope of the linear fit of the 
flow function graph; Table 3.10). 
0
2
4
6
8
10
12
0 2 4 6 8 10 12 14 16
Compr
e
ssibility (%
)
Consolidating Pressure (kPa)
fine
medium
coarse
fitted
66 
 
 
 
Figure 3.7 ? Flow functions of pecan shells fractions 
 
Table 3.10 ? Flow index of pecan shells fractions 
Sample Flow Function (FF) Flow Index (FI) Flow Characteristic 
fine 0.052 19.120 Free Flowing 
medium 0.215 4.651 Easy Flowing 
coarse 0.162 6.180 Easy Flowing 
  
According to Jenike (1964) the fine sample can be classified as a free flowing 
material.  The medium and coarse samples can be classified as easy flowing materials.  
Therefore, the fine sample has the best flow characterization of the pecan shells samples. 
The flow characterizations are expected from the results from the mechanical 
compressibility determinations in the previous section of this thesis.  The flow properties 
of the pecan shell samples indicate that in storage silos or transportation equipment, 
R? = 0.86
R? = 0.83
R? = 0.67
0
2
4
6
8
10
12
14
0 10203040506070
Unconfined yield str
e
ss 
(k
Pa)
Major consolidating stress (kPa)
medium
coarse
fine
67 
 
pecan shells separated into these particle sizes will most likely not need the use of flow 
aids or specialized discharge openings.  Similar trends of flow behavior and particle size 
were reported for thirteen food powders by Fitzpatrick et al. (2004).  
Cohesive strength and angle of internal friction (AIF) were also quantified.  
Cohesive strength increased as pressure increased for fine, medium and coarse particle 
size fractions (Table 3.11).  The mean cohesive strength of the medium size fraction was 
significantly higher than those of fine and coarse fractions.  In addition, analysis of 
variance indicated pressure had a significant effect (P < 0.05) on the cohesion of pecan 
shells at 9, 15 and 21 kPa.   
Table 3.11 ? Cohesive strength (kPa) of pecan shells fractions 
Consolidating Pressure 
(kPa) Fine Medium Coarse 
3 0.26 0.40 0.44 
6 0.49 0.58 0.67 
9
?
 0.77 1.49 1.40 
12 0.72 0.90 0.97 
15
?
 0.78 2.16 0.66 
21
?
 0.40 2.85 2.33 
Mean Cohesion (kPa) 0.57
[a]
 1.40
[b]
 1.08
[a]
 
*samples (n=3) in each column with same letter are not significantly (P<0.05) different 
?
 Comparisons of cohesion at that pressure are statistically different 
 
Statistical testing also showed that the angle of internal friction (AIF) of the 
medium size fraction was significantly lower than those of the fine and coarse fractions.  
Pressure did not have a significant effect of the values of angle of internal friction for all 
particle size fractions.  Average values for angle of internal friction for all pressures and 
particle size samples are shown in Table 3.12. 
68 
 
 
 
Table 3.12 ? Angle of internal friction (AIF,?) of pecan shell fractions 
Consolidating 
Pressure (kPa) Fine Medium Coarse 
3 43.49 34.97 35.63 
6 44.63 37.85 42.14 
9 43.74 40.90 42.74 
12 44.59 39.47 44.98 
15 45.81 35.27 43.14 
21 46.98 38.08 44.61 
Mean AIF (?) 44.87
[a]
 37.76
[b]
 42.21
[a]
 
* samples (n=3) with same letter are not significantly different (P<0.05) 
 
3.3.4. Moisture content effect on physical properties of pecan shells 
3.3.4.1. Moisture Adjustment 
The study of the effect of moisture content on the physical properties of pecan 
shells was carried out with samples conditioned to five moisture content levels (4.42, 
8.41, 14.73, 19.36 and 24.70%, wet basis).  The initial moisture content of the shells was 
14.59% (w.b.).  Only the medium size sample (geometric mean diameter 1.241 mm) was 
used for the moisture content effect study. 
 
3.3.4.2. Particle Size Distribution 
 The particle size distribution of the medium size fraction at different moisture 
contents is shown in Figure 3.8.  The values of geometric mean diameter (d
gw
) and 
geometric standard deviation (S
gw
) (calculated according to ASAE S319.4) can be found 
in Table 3.13.  Analysis of variance (ANOVA) and Tukey?s multiple comparison test 
69 
 
indicated that moisture content had a significant effect (P < 0.05) on the geometric mean 
diameter of pecan shells.  The increase in geometric mean diameter is a result of the 
additional moisture that caused the pecan shell particles to swell and increase in size. 
 
 
Figure 3.8 ? Particle size distribution of pecan shells at different moisture contents 
 
Table 3.13 ? Geometric mean diameters and standard deviation of pecan shells 
at different moisture contents 
Moisture Content (%, w.b.) d
gw
 S
gw
 CV (%) 
4.21 1.219
[c]
 0.459 37.65 
8.96 1.243
[b,c]
 0.464 37.33 
14.59 1.284
[b]
 0.421 32.79 
18.36 1.356
[a]
 0.384 28.32 
24.56 1.404
[a]
 0.381 27.14 
samples (n=3) with same letter are not significantly different (P<0.05) 
 
0
5
10
15
20
25
30
35
40
0 0.5 1 1.5 2
Distribution (%
)
Aperature opening size (mm)
MC = 4.21%
MC = 8.96%
MC = 14.59%
MC = 18.36%
MC = 24.56%
70 
 
3.3.4.3. Poured Bulk Density 
 Poured bulk density of pecan shells reduced from 460 kg/m
3
 to 396 kg/m
3
 as 
moisture content increased from 4.42% to 24.70% (Figure 3.9).  This shows that the bulk 
material expanded volumetrically at a faster rate than the increased mass due to addition 
of moisture.  Practical application of this is that when designing the storage, handling and 
transportation of shells, the volume necessary to store or move the shells (with identical 
mass) will increase as moisture content increases.  Similar results have been documented 
for other biological materials (Balasubramanian, 2001; Nimkar and Chattopadhyay, 
2001; Fasina and Sokhansanj, 1992; Deshpande et al., 1993; Mozammel et al., 2006; 
Altunata? and Yildiz, 2005; Bernhart and Fasina, 2009; McMullen et al., 2005; Colley et 
al., 2006).  The following equation describes the relationship between poured bulk 
density (kg/m
3
) and moisture content (%, w.b.). 
?
b
 =463.873? 0.104*MC
2
 , R
2
 = 0.942    (3.22) 
 
Figure 3.9 ? Effect of moisture content on the poured bulk density of pecan shells 
375
390
405
420
435
450
465
480
0 5 10 15 20 25 30
Bulk Density (kg/m
3
)
Moisture Content (%, w.b.)
Experimental data
Fitted
Note that error bars are included but too small to be visible 
71 
 
 
3.3.4.4. Particle Density 
 The particle density of pecan shells also decreased from 1429 kg/m
3
 to 1375 
kg/m
3
 as moisture content increased from 4.42% to 24.7%, respectively (Figure 3.10).  
The decline in particle density is due to the volumetric expansion of the particles, 
occurring at a faster rate than the increase in mass of particles due to addition of 
moisture.  Similar trends have been reported for other biological materials (Joshi et al., 
1993; Deshpande et al., 1993; Colley and Fasina, 2006).  Equation 3.23 explains the 
relationship between particle density (kg/m
3
) and moisture content (%, w.b.) for pecan 
shells. 
?
p
 =1444.706 ? 2.694*MC , R
2
 = 0.963   (3.23) 
 
Figure 3.10 ? Effect of moisture content on particle density of pecan shells 
1360
1375
1390
1405
1420
1435
1450
0 5 10 15 20 25 30
Density (kg/m
3
)
Moisture content (%, w.b.)
Experimental Data
Fitted
Note that error bars are included but too small to be visible 
72 
 
3.3.4.5. Tap Bulk Density 
 The tap bulk density of the pecan shells decreased with increased moisture 
content (Figure 3.11).  A maximum value of 537 kg/m
3
 was obtained at the moisture 
content level of 4.42%.  The decrease in tap density as moisture increases is indicative of 
the same moisture content ? density relationship previously discussed in this thesis that 
showed that bulk and particle volume increased at a rate faster than the mass of the bulk.  
The overall higher values of tap bulk density versus poured bulk density is an implication 
that during the transport of pecan shells, the material will compact as a result of 
vibrations that could result in difficulties in unloading pecan shells from transportation 
vehicles.  Similar trends have been reported for other biological materials (Lam et al., 
2008; Bernhart and Fasina, 2009).   
 
The following equation was developed to show the relationship of the effect of 
tap bulk density (kg/m
3
) and moisture content (%, w.b.) for pecan shells. 
?
t
 = 541.580 ? 0.108*MC
2
 , R
2
 = 0.942    (3.24) 
 
73 
 
 
Figure 3.11 ? Effect of moisture content on the tap bulk density of pecan shells 
3.3.4.6. Porosity 
 Porosity increased from 0.677 to 0.712 as moisture content of pecan shells 
increased.  Generally, as moisture content increased, an increase in porosity was 
observed.  Figure 3.12 and Equation 3.25 demonstrate the linear relationship between 
porosity and moisture content.  Similar trends have been obtained for other biological 
materials (Balasubramanian, 2001; Altuna? and Yildiz, 2007; McMullen et al., 2005; 
Suthar and Das, 1996; Gupta and Das, 1997; Aviara et al., 1999).  The average porosity 
for spherical particles is 0.40.  Non-spherical or irregular particles have higher porosity 
percentages (Woodcock and Mason, 1987).  Therefore, the porosity values obtained in 
this study is indicative of the irregular shape of pecan shell particles. 
?  = 0.675 + 0.000055*MC
2
 , R
2
 = 0.858   (3.25)  
450
465
480
495
510
525
540
555
0 5 10 15 20 25 30
Density (kg/m
3
)
Moisture Content (%, w.b.)
Experimental Data
Fitted
Error bars indicate standard deviation of sample (n=3) 
74 
 
 
Figure 3.12 ? Effect of moisture content on porosity of pecan shells 
 
3.3.4.7. Hausner Ratio 
 In this study, statistical analysis of the Hausner ratio of pecan shells showed no 
significant effect (P > 0.05) of moisture content on Hausner ratio.  The average Hausner 
ratio of pecan shells at different moisture contents was 1.144, less than the critical value 
of 1.25, therefore suggesting that pecan shells (with geometric mean diameter of 1.241 
mm) can be easily fluidized at moisture content (%, w.b.) levels ranging from 4.42% to 
24.70% (Barbosa-Canovas et al., 2005). 
 
Table 3.14 ? Hausner ratio of pecan shells at different moisture content levels 
M.C. (%, w.b.) Mean H
R
 St. dev. 
4.42 1.142
[a]
 0.005 
8.41 1.141
[a]
 0.006 
14.73 1.133
[a]
 0.006 
0.66
0.67
0.68
0.69
0.70
0.71
0.72
0 5 10 15 20 25 30
Por
o
sity
Moisture Content (%, w.b.)
Experimental Data
Fitted
Error bars indicate standard deviation of sample (n=3) 
75 
 
19.36 1.154
[a]
 0.003 
24.70 1.150
[a]
 0.006 
*samples (n=3) with same letter are not significantly different (P<0.05) 
 
3.3.4.8. Mechanical Compressibility 
 The percent compressibility of pecan shells ranged from 0.78% to 10.02% within 
the range of moisture contents and applied pressures used in this study (Figure 3.13).  As 
was observed for other biological materials (Bernhart and Fasina, 2009; Barbosa-
Canovas et al., 2005), compressibility of pecan shells increased with moisture content 
and pressure.  Increased compressibility with moisture content is due to moisture added 
that formed liquid bridges between particles thereby making them more cohesive 
(Teunou and Fitzpatrick, 1999).  In addition, biological materials become softer when 
moisture is adsorbed, thus deformation of the material is greater at higher moisture 
contents (Bernhart and Fasina, 2009).  Based on Fayed and Skocir?s (1997) 
classifications, the compressibility values show that pecan shells have excellent flow, 
since the compressibility values are less than 15%.  The following equation was 
developed to show the relationship of mechanical compressibility, applied pressure and 
moisture content: 
           C
M
 = 0.783 ? 0.284*MC + 0.016*MC
2
 +4.961 log(?) , R
2
 = 0.927               (3.26) 
Where C
M
 is compressibility (%), MC is moisture content (%, w.b.) and ? is applied 
pressure (kPa). 
 
76 
 
 
Figure 3.13 ? Effect of moisture content and applied pressure on the mechanical 
compressibility of pecan shells 
 
3.3.4.9. Flowability 
 The flow function plots of the ultimate yield stress (UYS) and major 
consolidating stress (MCS) (Figure 3.14) are fitted with linear best-fit lines that indicate 
that as moisture content increases, flowability decreases.  Slopes of these lines that are 
lying towards the bottom of the graph (x-axis) represent easy flow, and more difficult 
flow is represented as the linear fit lines move upwards in an counter-clockwise direction 
(see Figure 2.4).  Flowability of the pecan shells is further characterized by flow index 
(FI) values obtained from the inverse of the slope of the linear fit of the flow function 
plot (Table 3.15).  Similar to the effect of moisture on mechanical compressibility 
discussed in the previous section, higher moisture content levels in pecan shells will 
compress more than lower moisture content shells.  Therefore, at higher moisture 
0
2
4
6
8
10
12
0 3 6 9 12 15
Compr
e
ssibility (%
)
Pressure  (kPa)
MC = 24.70%
MC = 19.36%
MC = 14.73%
MC = 8.41%
MC = 4.42%
Fitted
Error bars indicate standard deviation of sample (n=3) 
77 
 
contents, the material is compressing more and the force required to shear the material is 
greater than the force required to shear lower moisture content samples that compress less 
at the same pressure.  The general trend for reduction in flowability due to increase in 
moisture content observed in this study is similar to other biological materials (Ganesan 
et al, 2008; Zou and Brusewitz, 2002; Teunou and Fitzpatrick, 1999) 
 
 
Figure 3.14 ? Flow functions of pecan shells at different moisture contents 
 
 
Utilizing Jenike?s classification of flowability by flow index, the reduction in flowability 
is confirmed as pecan shells at the lowest moisture content (4.11%, w.b.) is characterized 
R? = 0.68
R? = 0.80
R? = 0.62
R? = 0.71
R? = 0.89
0
5
10
15
20
25
30
20406080
Unconfined yield str
e
ss 
(k
Pa)
Major consolidating stress (kPa)
4.11%
8.53%
14.57%
18.67%
24.23%
78 
 
as free flowing and the 24.23% (w.b.) moisture content sample is classified as a cohesive 
material.  The implication of the shift in flow behavior due to moisture content indicates 
that pecan shells at moisture contents above 24% (w.b.) will require flow aids during 
discharge from storage equipment. 
 
Table 3.15 ? Flow index of pecan shells at various moisture contents 
M.C. (%, w.b.) Flow Function (FF) Flow Index (FI) Flow Characteristic 
4.11 0.063 15.773 Free Flowing 
8.53 0.150 6.676 Easy Flow 
14.57 0.225 4.448 Easy Flow 
18.67 0.230 4.352 Easy Flow 
24.23 0.319 3.138 Cohesive 
 
 Figure 3.15 shows the cohesion at all pressures for all samples.  As expected, 
cohesion increased with pressure.  The cohesive strength of the 4.11% (w.b.) sample was 
significantly (P<0.05) lower than that of the other samples whereas the angle of internal 
friction of the highest moisture content sample (24.23%, w.b.) was significantly higher 
than those of the other samples.  Similar to cohesive strength, angle of internal friction 
also increased with increased pressure.  Mean angle of internal friction at all pressures 
and moisture content levels are given in Table 3.16. 
 
 
 
 
 
79 
 
 
 
Figure 3.15 ? Cohesive strength of pecan shells at different moisture contents 
 
Table 3.17 ? Angle of internal friction (?) of pecan shells at different moisture 
contents 
Consolidating 
Pressure (kPa) 
Moisture Content (%, w.b.) 
4.11 8.53 14.53 18.67 24.23 
3 37.03 40.52 32.97 36.91 35.62 
6 32.97 36.46 37.85 33.39 39.10 
9 42.52 38.42 34.71 40.34 48.78 
12 42.81 36.33 42.35 39.87 47.76 
15 - 39.98 35.27 46.31 47.89 
21 - 35.51 38.08 45.28 54.28 
Mean (kPa) 38.83
[a]
37.87
[a]
 36.87
[a]
 40.35
[a]
 45.57
[b]
 
*samples (n=3) with same letter are not significantly different (P<0.05) 
- denotes values that could not be determined 
 
3.3.5. Rate of Moisture Sorption 
 The effect of air temperature and relative humidity on the rate of moisture 
sorption by pecan shells is shown in Figures 3.16 and 3.17.  Overall, the rate of moisture 
sorption was affected by the environment that the samples were exposed to. Figure 3.16 
R? = 0.80
R? = 0.95
R? = 0.95
R? = 0.72
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0 5 10 15 20
Cohesion (kPa)
Unconfined yield stress (kPa)
8.53%
14.57%
18.67%
24.23%
80 
 
demonstrates the effect of relative humidity for pecan shell samples exposed to a 
temperature of 15?C.  At 50% relative humidity, the moisture content at the end of the 
experiment (13.47%, w.b.) decreased from the initial moisture content of 15.02% (w.b.).  
An increase in relative humidity to 80% (compared to 50% RH) resulted in an increase in 
final moisture content (16.68%, w.b.).  Figure 3.17 demonstrates the effect of temperature 
on the rate of moisture sorption of pecan shells in the same relative humidity environment 
of 80%.  The rate of moisture sorption increases with increased temperature (at 80% 
relative humidity).  At 80% relative humidity, final moisture contents of 16.68, 16.92 and 
17.62% (w.b.) were observed for temperatures of 15, 25 and 35?C, respectively.  The 
implication of the data depicted in the two graphs is that air relative humidity and 
temperature will affect the moisture content of pecan shells during storage and transport 
and that these changes can occur in a period of two days or less. 
 
Figure 3.16 ? Moisture change in pecan shells exposed to air at 15?C and relative 
humidity levels of 50, 65 and 80%.  Initial moisture content was 15.02% (w.b.) 
 
12
13
14
15
16
17
18
0 8 16 24 32 40 48 56
Moistur
e
 content (%
, 
w
.
b.)
Time (hours)
50% RH 65% RH 80% RH fitted
Error bars indicate standard deviation (n=2)
81 
 
 
Figure 3.17 ? Moisture change in pecan shells exposed to air at 80% relative 
humidity and temperatures of 15, 25 and 35?C. 
The Page model (Equation 3.27) has been used to model the drying characteristics of 
biological materials and fruits (Dadgar, 2004; Doymaz, 2007;Kashinaninejad and Tabil, 
2004; Akpinar and Bicer, 2004).  Table 3.17 gives the estimates of rate of moisture 
sorption constants k and n and the equilibrium moisture content (M
e
).  A non-linear 
regression procedure in SAS (Version 9.3, 2008) was used to estimate the values of the 
constants.   
nullnull null
null
null
null 
null null
null
nullexpnullnullnullnull
null
null                                                (3.27) 
where, t = time (min) 
 M = instantaneous moisture content (%, w.b.) 
 M
i
 = initial moisture content (%, w.b.) 
 M
e
 = equilibrium moisture content (%, w.b.) 
 k = constant 
 n = constant 
14
15
16
17
18
19
0 8 16 24 32 40 48 56
Moistur
e
 content (%
, 
w
.
b.)
Time (hours)
35?C 25?C 15?C fitted
Error bars indicate standard deviation (n=2)
82 
 
Table 3.17 ? Estimated values of constant (k), constant (n) and equilibrium 
moisture content (M
e
) obtained from non-linear regression analysis using 
Equation 3.20 for pecan shells 
Relative Humidity (%) Temperature (?C) k  n M
e
 
50 
15 0.1007 0.9349 13.47 
25 0.0337 1.3566 11.46 
35 0.2321 0.8449 9.86 
65 
15 0.0066 1.5464 15.05 
25 0.5131 0.5646 14.28 
35 0.2067 0.7429 14.20 
80 
15 0.3639 0.6389 16.68 
25 0.3008 0.7209 16.92 
35 0.5003 0.6223 17.62 
 
Statistical analysis showed that estimated values of k, n and M
e
 were significantly (P < 
0.05) affected by relative humidity and temperature.  In general, k was lower for samples 
that desorbed moisture and higher for those that absorbed moisture.  Overall, samples that 
did not absorb or desorb moisture had low k values and n values higher than 1.  High 
coefficient of determination (R
2
) values indicates that the model used to determine the 
constants were appropriate with values between 0.892 and 0.991.  Standard errors of the 
model ranged from 0.0005 and 0.0052, giving further confidence that the model is 
acceptable in the determination of moisture sorption rate, constant n and equilibrium 
moisture content of pecan shells.  Other biological materials have shown the similar 
relationships of k, n and M
e
 on relative humidity and temperature (Dadgar et al, 2004; 
Doymaz, 2007; Doymaz and Pala, 2002). 
 
 
83 
 
3.3.6. Equilibrium Moisture Isotherm 
 Typically, equilibrium moisture relationships for biological materials are 
represented in terms of dry basis, or grams of water per grams of dry material.  Since 
most moisture content measurements have been reported in wet basis, equation 3.28 was 
utilized to convert wet basis (MC
wb
, %) to dry basis (MC
db
, %) (ASAE Standard D245.6, 
2007). 
nullnull
nullnull 
null 
nullnullnullnull nullnull
nullnull
nullnullnullnull nullnull
nullnull
    (3.28) 
 
The equilibrium moisture content (EMC) ? equilibrium relative humidity (ERH) curves 
are illustrated in Figure 3.18.  The curves are type II (sigmoidal) in shape that is typically 
obtained for biological materials (Labuza, 1984; Erbas at al., 2005; Chen, 2003; 
Igathinathane et al., 2007).   
 
 
Figure 3.18 ? Moisture sorption isotherms for pecan shells at 10, 20, 35 and 50?C 
0
5
10
15
20
25
30
35
0.0 0.2 0.4 0.6 0.8 1.0
Equilibrium moistur
e
 (%
, 
d.b.)
Relative humidity (decimal)
10?C
20?C
35?C
50?C
84 
 
 
The EMC and ERH data were analyzed using the following four equations (Chen, 2008; 
Ajibola et al., 2003; Colley et al., 2006; Fasina and Sokhansanj, 1993): 
 
Modified Halsey equation:  nullnullnull null nullnullnullnullnull 
nullnullnullnullnullnullnullnullnullnull
nullnull
nullnull
null
null                 (3.29) 
Modified Henderson equation:   nullnullnull null 1nullnullnullnullnullnullnull nullnullnullnullnullnullnullnullnullnull
nullnull
nullnullnull               (3.30) 
Modified Chung-Pfost equation: nullnullnull null nullnullnullnullnull
null
nullnullnull
 exp nullnullnull nullnullnull
nullnull
nullnull            (3.31) 
Modified Oswin equation:         nullnullnull null nullnull
nullnullnullnull
nullnull
nullnull
null
null
null 1null
nullnull
                                         (3.32) 
 
Where, ERH is equilibrium relative humidity (in decimal), MC
db
 is equilibrium moisture 
content (in dry basis), t is temperature (?C) and a,b and c are model constants.  The 
experimental data collected were fitted to the four isotherm equations using the NLIN 
procedure (SAS, 2008), that involves minimizing the sum of squares deviations in a 
series of iterations.  The coefficient of determination (R
2
), standard error of estimate 
(SEE) and mean relative deviation (MRD) were utilized to determine the goodness-of-fit 
for each isotherm prediction equation (3.22 ? 3.24).  The standard error of the estimate is 
defined as:  
nullnullnull null 
null?nullnullnull
nullnull
nullnullnull
nullnull
null
null
null
null
nullnullnull
nullnull
                                               (3.33) 
Where, MC
db 
is the experimental data, nullnull
nullnull
null
 is the estimated value, m is the number of 
data points and df is degrees of freedom on the regression model.   Mean relative 
85 
 
deviation (MRD) gives insight of the mean divergence of the measured data from the 
predicted data, and is defined as: 
nullnullnull null 
null
null
 ? null
null| nullnull
nullnull
nullnullnull
nullnull
null
|null
nullnull
nullnull
null
null
nullnullnull
                 (3.34) 
The SEE value represents the fitting ability of a model in relation to the number of data 
points, however used alone is not able to provide direct visualization of the goodness-of-
fit.  Therefore, SEE is used in combination with MRD and R
2
 in this study.  In general, 
the best model has the smallest SEE and MRD values and highest R
2
 value (Sun and 
Byrne, 1998).  Multiple authors have reported the use of multiple statistical parameters to 
select the best moisture sorption isotherm equation (Chen, 2003; Fasina and Sokhansanj, 
1993; Sun and Bryne; 1998; Colley et al., 2006).  The model coefficients and statistics of 
fitting are located in Table 3.18.  All of the models have R
2
 values of 0.999.  According 
to Pagano and Mascheroni (2005), isotherm equations that have MRD values less than 
0.05 are considered to be a good fit.  The Henderson and Chung-Pfost equations have 
SEE values and MRD values less than 0.05, and both could be used as isotherm equations 
for pecan shells.  The Chung-Pfost and Henderson equations have also been reported as 
the best isotherm equation for other biological materials (switchgrass pellets? Colley et 
al., 2006; cowpea ? Ajibola et. al., 2003)   
 
 
 
 
86 
 
Table 3.18 ? Model coefficients and values for the standard error of estimate 
(SEE), mean relative deviation (MRD) and coefficient of determination (R
2
) for 
the temperature range of 10-50?C 
Coefficients Statistics of fitting 
Equation a b c SEE MRD R
2
 
Halsey 3.4093 -0.0068 1.5359 0.0773 0.0718 0.999 
Stdev ? 0.5490 ? 0.0050 ? 0.1968 
Henderson 1.020E-04 142.500 1.5065 0.0307 0.0265 0.999 
Stdev ? 3.200E-05 ? 50.006 ? 0.0578 
Chung-Pfost 528.3000 136.7000 0.1313 0.0291 0.0245 0.999 
Stdev ? 150.9000 ? 46.8436 ? 0.0048 
Oswin 12.2736 -0.0444 2.2105 0.0539 0.0508 0.999 
Stdev ? 0.8161 ? 0.0236 ? 0.1611 
 
 
  
87 
 
3.4. Conclusions 
 Compositional analysis of pecan shells indicated that energy content and ash 
content were significantly affected by particle size.  Elemental analysis of pecan shells at 
different particle sizes indicated that there was significantly less carbon percentage in the 
fine fraction whereas the composition of nitrogen decreased as particle size increased.  
Hydrogen levels were statistically similar at a 95% confidence level.  The structural 
carbohydrate analysis indicated that the cellulose, hemicelluloses and lignin percentages 
were significantly affected by particle size.  Cellulose levels increased as particle size 
increased, however, there was no noticeable trend for the effect of particle size in 
hemicelluloses and lignin levels. 
It can be concluded from this study that particle size and moisture content 
significantly affect the physical characteristics of pecan shells.  Increasing particle size 
and moisture content resulted in decreases in the poured bulk, particle and tap bulk 
densities.  Porosity of pecan shells indicated that the particles were irregular and non-
spherical in shape and increased with particle size and moisture content.  The Hausner 
ratio of the particle size fractions indicated that the coarse and raw samples can be easily 
fluidized as well as the samples at all moisture contents.  The compressibility of pecan 
shells increased as particle size decreased, as moisture content and applied pressure 
increased.  Flow characteristics of pecan shells were determined as free-flowing for the 
fine fraction and easy flowing for the medium and coarse particle size fractions.  
Cohesion was significantly affected by pressure and particle size.  Cohesion of pecan 
shells increased as particle size increased.  Angle of internal friction was also 
88 
 
significantly affected by particle size, where a 17% decrease was observed as particle size 
decreased by 10%. 
 Temperature and relative humidity had a significant effect on the moisture 
sorption rate (k), constant (n) and equilibrium moisture content (M
e
).  The Henderson and 
Chung-Pfost isotherm equations were determined to be the most appropriate fit to the 
experimental data whereas the Halsey equation was determined to be the least 
appropriate. 
 
 
89 
 
3.5. References 
Ajibola, O.O., Aviara, N.A., Ajetumobi, O.E. 2003. Sorption equilibrium and 
thermodynamic properties of cowpea (Vigna unguiculata). Journal of Food Engineering 
58: 317-324. 
 
Akpinar, E.K., Bicer, Y. 2004. Modelling of the drying of eggplants in thin-layers. 
International Journal of Food Science and Technology 39: 1-9. 
 
Altuna?, E., Yildiz, M. 2007. Effect of moisture content of some physical and mechanical 
properties of faba bean (Vicia faba L.) grains. Journal of Food Engineering 78: 178-183. 
 
ASABE Standard. 2006. S593 Terminology and definitions for biomass production, 
harvesting and collection, storage, processing, conversion and utilization. St. Joseph, MI: 
ASABE. 
 
ASABE Standard. 2007.  S269.4. Definition of Methods and Determining Density, 
Durability, and Moisture Content. St. Joseph, MI: ASABE 
 
ASABE Standard. 2008. S319.4 Method of determining and expressing fineness of feed 
materials by sieving. St. Joseph, MI: ASABE. 
 
ASTM Standard. 2005. B 527. Standard test method for determation of tap density of 
metallic powders and compounds. West  Conshohocken, PA: ASTM. 
 
Aviara, N.A., Gwandzang, M.I., Haque, M.A. 1999. Physical properties of Guna Seeds. 
Journal of Agricultural Engineering Research 73: 105-111. 
 
Balasubramanian, D. 2001. Physical properties of raw cashew nut. Journal of 
Agricultural Engineering Research 78(3): 291-297. 
 
Barbosa-Canovas, G.V., Ortega-Rivas, E., Juliano, P., Yan, H. 2005. Food Powders: 
Physical Properties, Processing, and Functionality. New York, NY: Kluwer 
Academic/Plenum Pub. 
 
Bernhart, M., Fasina, O.O. 2009. Moisture effect of the storage, handling and flow 
properties of poultry litter. Waste Management 29: 1392-1398. 
 
Biagini, E., Barontini, F., Tognotti, L. 2006. Devolatilization of biomass fuels and 
biomass components studied by TG-FTIR technique. Industrial and Engineering 
Chemistry Research 45: 4486-4493. 
 
Chen, C. 2000. Factors which effect equilibrium relative humidity of agricultural 
products. Transactions of the ASAE (43)2: 673-683. 
90 
 
Chen, C. 2003. Moisture sorption isotherms of pea seeds. Journal of Food Engineering 
58: 45-51. 
 
Colley, Z., Fasina, O.O., Bransby, D., Lee, Y. 2006. Moisture effect of the physical 
characteristics of switchgrass pellets. Transactions of the ASABE (49)6: 1845-1851. 
 
Dadgar S., Tabil, L.G., Crerar, B. 2004. Moisture adsorption of field peas as a funtion of 
temperature and relative humidity during storage, An ASABE/CSAE Annual Meeting 
Presentation, Paper No. 046089, Sponsored by ASAE/CSAE, August 1-4, Ottawa, 
Ontario, Canada. 
 
Deshpande, S.D., Bal, S., Ojha, T.P. 1993. Physical properties of soybean. Journal of 
Agricultural Engineering Research 56(2): 89-98. 
 
Di Blasi, C., Signorelli, G., Di Russo, C., Rea, G. 1999. Product distribution from 
pyrolysis of wood and agricultural residues. Industrial Engineering and Chemistry 
Research 38: 2216-2224. 
 
Doymaz, I. 2007. Air-drying characteristics of tomatoes. Journal of Food Engineering 
78: 1291-1297. 
 
Doymaz, I., Pala, M. 2002. Hot-air drying characteristics of red pepper. Journal of Food 
Engineering 55: 331-335. 
 
Erba?, M., Ertugay, M., Certel, M. 2005. Moisture adsorption behavior of semilina and 
farina. Journal of Food Engineering 69: 191-198. 
 
Fasina, O.O., Ranatunga, T., Vaughan, B. 2007. Physical and chemical characteristics of 
biofuel feedstocks indigenous to southeastern united states, An ASABE Annual Meeting 
Presentation, Paper No. 076247, June 17-20, Minneapolis Convention Center, 
Minneapolis, MN. 
 
Fasina, O.O., Sokhansanj, S. 1992. Hygroscopic moisture absorption by alfalfa cubes and 
pellets. Transactions of the ASAE 35: 1615-1619. 
 
Fasina, O.O., Sokhansanj, S. 1993. Equilibrium moisture relations and heat of sorption of 
alfalfa pellets. Journal of Agricultural Engineering Research 56: 51-63. 
 
Fayed, M.E., Skocir, T.S. 1997. Mechanical Conveyors: Selection and Operation. 
Lancaster, PA: Technomic Publishing Company, Inc. 
 
Fitzpatrick, J.J., Barringer, S.A., Iqbal, T. Flow property measurement of food powders 
and sensitivity of Jenike's hopper design methodology to the measured values. Journal of 
Food Engineering 61: 399-405. 
 
91 
 
Ganesan, V., Muthukumarappan, K., Rosentrater, K.A. 2008. Flow properties of DDGS 
with varying soluble and moisture contents using jenike shear testing. Powder 
Technology 187: 130-137. 
 
Geisler, Malinda. 2009. Pecans. Agriculture Marketing Resource Center, Iowa State 
University. Available at: http://www.agmrc.org. Accessed 3 May 2010. 
 
Gupta, R.K., Das, S.K. 1997. Physical properties of sunflower seeds. Journal of 
Agricultural Engineering Research 66: 1-8. 
 
Haykiri-Acma, H. 2006. The role of particle size in the non-isothermal pyrolysis of 
hazelnut shells. Journal of Analytical and Applied Pyrolysis 75: 211-216. 
 
Igathinathane, C., Womac, A.R., Sokhansanj, S., Pordesimo, L.O. 2007. Moisture 
sorption thermodynamic properties of corn stover fractions. Transactions of ASABE 
50(6): 2151-2160. 
 
Jenike, A.W. 1964. Storage and flow of solids. Bulletin No. 123. University of Utah, Salt 
Lake City, Utah: Engineering and Experiment Station. 
 
Joshi, D.C., Das, S.K., Mukherjee, R.K. 1993. Physical properties of pumpkin seeds. 
Journal of Agricultural Engineering Research 54(3): 219-229. 
 
Kashaninejad, M., Tabil, L.G. 2004. Drying characteristics of purslane (Portulaca 
oleraceae L.). Drying Technology 22: 2183-2200. 
 
Labuza, T.P. 1984. Moisture sorption: practical aspects of isotherm measurement and 
use. Minneapolis, MN: American Association of Cereal Chemist. 
 
Lam, P.S., Sokhansanj, S., Bi, X., Lim, C.J., Naimi, L.J., Hoque, M., Mani, S., Womac, 
A.R., Ye, X.P., Narayan, S. 2008. Bulk density of wet and dry wheat straw and 
switchgrass particles. Applied Engineering in Agriculture 24(3): 351-358. 
 
Mani, S., Tabil, L.G., Sokhansanj, S. 2004. Grinding performance and physical properties 
of wheat and barley straws, corn stover and switchgrass. Biomass and Bioenergy 27: 339-
352. 
 
Mani, S., Tabil, L.G., Sokhansanj, S. 2006. Effects of compressive force, particle size and 
moisture content on mechanical properties of biomass pellets from grasses. Biomass and 
Bioenergy 30: 648-654. 
 
McMullen, J., Fasina, O.O., Wood, C.W., Feng, Y. 2004. Storage and handling 
characteristics of pellets from poultry litter. Applied Engineering in Agriculture 21: 645-
651. 
92 
 
Mozammel, H., Shahab, S., Bi, T., Sudhagar, M., Ladan, J., Lim, J. 2006. Interaction of 
particle size, moisture content, and compression pressure on the bulk density of wood 
chip and straw, An SCBE/SCGAB Annual Meeting Presentation, Paper No. 06-100, 
Sponsored by CSBE/SCGAB, July 16-19, Edmonton, Alberta, Canada.  
 
Nimkar, P.M., Chattopadhyay, P.K. 2001. Some physical properties of green gram. 
Journal of Agricultural Engineering Research 80: 183-189. 
 
NREL. 2005. Laboratory Analytical Procedure - Determination of Ash in Biomass. 
United States Department of Energy, Washington, D.C. 
 
Pagano, A.M., Mascheroni, R.H. 2004. Sorption isotherms for amaranth grains. Journal 
of Food Engineering 67: 441-450. 
 
Sun, D., Byrne, C.  1998.  Selection of EMC/ERH isotherm equations for rapeseed.  
Journal of Agricultural Engineering Research 69: 307-315. 
 
Suthar, S.H., Das, S.K. 1996. Some physical properties of Karingda [Citrullus lanatus 
(Thumb) Mansf] seeds. Journal of Agricultural Engineering Research 65: 15-22. 
 
Tabil, L.G., Sokhansanj, S. 1997. Bulk properties of alfalfa grind in relation to its 
compaction characteristics. Applied Engineering in Agriculture 13(4): 499-505. 
 
Teunou, E., Fitzpatrick, J.J. 1999. Effect of relative humidity and temperature on food 
powder flowability. Journal of Food Engineering 42: 109-116. 
 
USDA-ERS. 2003. Fruit and Tree Nuts Outlook: 28 May 2003. USDA - Economic 
Research Service, Washington DC. 
 
USDA-NASS. 2008. Noncitrus Fruits and Nuts Summary. Washington, DC.: USDA 
National Agricultural Statistics Service. 
 
Van Soest, P.J., Robertson, J.B. 1985. Analysis of forages and fibrous foods. A 
Laboratory Manual for Animal Science 613. Cornell University, Ithaca, NY. 
 
Woodcock, C.R., Mason, J.S. 1987. Bulk Solids Handling: An Introduction to the 
Practice and Technology. Glasgow, Scotland: Blackie and Son, Ltd. 
 
Zou, Y., Brusewitz, G.H. 2002. Flowability of uncompacted marigold powder as affected 
by moisture content. Journal of Food Engineering 55: 165-171. 
 
 
 
93 
 
CHAPTER 4: TG-FTIR ANALYSIS OF PECAN SHELLS DURING THERMAL 
DECOMPOSITION 
4.1. Introduction 
 Biomass can be converted to energy via biological and thermochemical 
conversion platforms.  Biological conversion processes typically use fermentation to 
produce ethanol.  High conversion efficiency is only possible if the biomass feedstock 
contain high concentrations of starch or sugar compounds (that are easily accessible) and 
low concentrations of lignin (e.g. maize or sugar cane).  Pecan shells do not contain 
easily accessible cellulose chains because of a significant amount of lignin concentration 
(see Chapter 3, Section 3.3.2).  Thermochemical conversion processes are the preferred 
conversion process for biomass feedstocks that have high amounts of lignin, such as 
wood chips or pecan shells.  Combustion, gasification and pyrolysis are the main 
thermochemical conversion processes that can be used to convert biomass to value-added 
products, such as heat, syngas, bio-oil or bio-char.  Gasification and pyrolysis are 
considered the most promising because of high energy efficiencies and the ability to 
handle most biomass materials (Yang et al., 2004; Yaman, 2004).  Thermogravimetric 
analysis (TGA) is the most common thermo-analytical method for quantifying thermal 
decomposition of biological materials (Vamvuka et al., 2003; Vuthaluru et al., 2003).  
The fast, repeatable and accurate data collection from TGA enables in-depth analysis of 
mass loss and determination of kinetic parameters (Yang et al., 2004).  
94 
 
In addition to TGA, other thermo-analytical methods used to study the thermal 
decomposition of biological materials include: differential scanning calorimetry (DSC) 
and differential thermal analysis (DTA).  TGA, DSC and DTA are all able to provide 
valuable information on the rate of decomposition.  However, they are unable to provide 
information about the composition and rate of the gas compounds volatilized during 
thermal decomposition.  Thus, it is advantageous to use TGA in conjunction with other 
analytical tools which can identify and quantify the evolved gas products, or ?real-time 
analysis?, during thermal decomposition.  TGA is currently the only thermo-analytical 
method that can be coupled to other gas analytical instruments, such as a gas 
chromatogram (GC), mass spectrometer (MS) and Fourier transform infrared (FTIR) 
spectrometer.  These gas analytical instruments are used for ?real-time? analysis of the 
gases produced during thermochemical analysis (Butterman et al., 2009; Arias et al., 
2006; Yang et al., 2007; Bahng et al., 2009). 
  
Therefore, the objectives of this study were to: 
(1)  Quantify the rate of thermal decomposition of pecan shells and 
(2)  Quantify the composition of the gases evolved during the thermal decomposition 
process. 
95 
 
4.2. Methods and Materials 
4.2.1. Sample Preparation 
 The pecan shells used in this study were obtained from the Louisville Pecan 
Company, Louisville, AL.  The samples were stored in the Biosystems Process 
Engineering Laboratory (average temperature of 25?C).  Before use, the pecan shells 
were ground in a Wiley mill to pass through a standard 40-mesh screen.  Energy content, 
ash content and ultimate analysis of pecan shells were previously discussed in Section 
3.3.2 and values given in Tables 3.2 and 3.3, respectively.   
4.2.2. Thermal Decomposition Study 
 The thermal decomposition for pecan shells was carried out in a 
thermogravimetric analyzer (TGA) (Pyris 1, Perkin-Elmer, Shelton, CT).  A sample mass 
of approximately 5 mg was used for thermogravimetric analysis.  Samples were heated at 
rates of 5?C/min, 10?C/min, 20?C/min, 30?C/min and 40?C/min within the temperature 
range of 30?C to 800?C.  Nitrogen and air were used as carrier gases, each at a flow rate 
of 20 ml/min.  Prior to use, the TGA was calibrated for temperature and mass according 
to the manufacturer?s procedure.  Temperature calibration was performed using the Curie 
point of four magnetic standards ? alumel (T
Curie
 of 154.2?C), nickel (T
Curie
 of 355.3?C), 
perkalloy (T
Curie
 of 596.0?C) and iron (T
Curie
 of 780.0?C).  Calibration for mass was 
performed using a 100 mg calibration mass provided by the manufacturer. 
 
96 
 
4.2.3. Fourier Transform Infrared Spectroscopy 
 A Fourier Transform infrared (FTIR) spectrometer (Model 100, Perkin-Elmer, 
Shelton, CT) was used to identify and quantify the gases volatilized from the TGA.  A 
transfer line (heated to 220?C) was used to connect the FTIR to the TGA.  The transfer 
line was heated to eliminate condensation of the volatile gases or adsorption of semi 
volatile products produced during pyrolysis (Yang et al., 2004).  The software provided 
by the FTIR manufacturer was used to obtain spectra of the gas flowing through the 
measurement cell at 20 second intervals within the wavelength number range of 4000 to 
700 cm
-1
.  Quantitative analysis of the spectra was performed in two steps: (a) 
identification of peaks by matching spectra against a library software package (QASOFT, 
Infrared Analysis Inc., Anaheim, CA) and (b) using the software to quantify the 
concentration of the identified gases in ppm.  Thermogravimetric analysis and FTIR 
experiments were run simultaneously in triplicate. 
 
4.2.4. Data Analysis 
 Concentration of volatile gases was determined using library software (QASOFT, 
Infrared Analysis Inc., Anaheim, CA) and plotted as a function of temperature using 
Microsoft Excel (Microsoft Office XP Professional, 2007).  Analysis of variance 
(ANOVA) and Studentized t-tests were performed using the Proc anova and Proc ttest 
functions in SAS statistical software package (Version 9.3, SAS Institute Inc., Cary, NC, 
2008-2009), respectively. 
 
97 
 
4.3. Results and Discussion 
4.3.1. Thermal Decomposition of Pecan Shells in Nitrogen Atmosphere 
Observed thermal behavior (TG curve) of pecan shells during thermal 
decomposition is shown in Figure 4.1.  For each heating rate, there was an initial mass 
loss (approximately 14% of total weight)  recorded between 30?C and 150?C due to the 
release of moisture (Fang et al., 2006; de Jong et al., 2007; Baker et al., 2005).  The 
thermal decomposition of pecan shells was therefore determined to begin at 180?C, 
similar to other reported biological materials (wood lignin ? Liu et al., 2008; sugar cane 
and coconut fibers ? Mothe and Miranda; 2009; coconut shells, sugarcane bagasse, corn 
stalks and peanut shells ? Souza et al., 2009).  The figure also showed that significant 
loss in sample mass (40% original weight) occurred between the temperature range of 
180?C and 400?C, and thermal decomposition was essentially complete at 600?C.  There 
was a residual weight of 30% after the thermal decomposition procedure.  This residue is 
typically referred to as bio-char.  In general, there was an increase in mass loss as heating 
rate decreased within the temperature range of 180?C to 400?C.  Similar results were 
reported for textile wastes (Miranda et al., 2007), seaweed (Wang et al., 2007) and rice 
straw, camphor branch, rice husk, cherry bay branch, cotton and maize straw (Hu et al., 
2007).  For temperatures higher than 600?C, mass losses of 5.72% and 6.16% were 
obtained at heating rates of 5?C/min and 40?C/min, respectively.  Therefore, the amount 
of gases volatilized at temperatures higher than 600?C may not justify the amount of 
energy required to maintain the high temperatures (Biagini et al., 2006).  
 
98 
 
 
Figure 4.1 ? Mass loss from thermal decomposition of pecan shells at different 
heating rates with nitrogen as carrier gas 
Figure 4.2 shows the mass loss rate (d?/dt) curves (derivative thermogram ? DTG curves) 
within the temperature range of 150?C to 800?C.  Mass loss rate at temperatures lower 
than 150?C was not shown in order to isolate peaks caused by the thermal decomposition 
of the hemicelluloses and cellulose compounds in the pecan shell samples.  The mass loss 
fraction (?) was calculated using the equation below: 
nullnull 
nullnull null
null
null
null
null null
null
                                                             (4.1)      
Where: m is instantaneous mass (g), m
o
 is initial mass (at 30?C, g) and m
f
 is final mass 
(at 800?C, g).  The DTG curve provides a clear representation of the general trend of 
increasing mass loss with heating rate.  Thermal decomposition of biomass frequently 
presents overlapping peaks; one large peak with a shoulder peak on the left side, which 
represents the thermal decomposition of hemicelluloses and cellulose.  The shoulder peak 
on the left side is attributed to the peak of hemicelluloses thermal decomposition while 
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800
Mass Change 
(%
)
Temperature (?C)
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
99 
 
the larger peak is regarded as the peak decomposition for cellulose.  The hemicelluloses 
shoulder and the cellulose peak phenomena has been shown in other biological materials, 
such as dry distiller?s grains with solubles (DDGS) - Giuntoli et al., 2009; sugarcane 
bagasse ? Mothe and Miranda, 2009; palm oil waste ? Yang et al., 2004; coconut shells - 
Souza et al., 2009; seaweed ? Wang et al., 2007; and coconut and cashew nut shells - 
Tsamba et al., 2006.  The flat, tailing section of the DTG curves at higher temperatures 
indicates the thermal degradation of lignin.   
 
 
Figure 4.2 ? Mass loss rate from the thermal decomposition of pecan shells at 
different heating rates with nitrogen as carrier gas 
The cellulose peaks increased as heating rates increased, occurring at temperatures of 
348, 360, 371, 385 and 386?C at heating rates of 5, 10, 20, 30 and 40?C, respectively.   
Hemicelluloses decomposition peaks were at 275, 285, 305, 315 and 330?C at heating 
rates of 5, 10, 20, 30 and 40?C, respectively.  Hemicelluloses decompose at lower 
temperatures than cellulose due to the amorphous structure and composition of shorter, 
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
100 200 300 400 500 600 700 800
Mass loss rate 
(min
-1
)
Temperature (?C)
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
100 
 
branched 5 carbon chains.  In contrast, cellulose consists of a long polymer of glucose 
without branches, hence the higher thermal stability of cellulose when compared to 
hemicelluloses (Yang et al., 2007).  The temperature ranges observed during 
decomposition of hemicelluloses and cellulose in pecan shells is consistent with other 
agricultural waste materials (cotton cocoon shells ? ?a?lar and Demirba?; hazelnut shells 
? Demirba?, 1998; grape residues, olive and rice husk ? Di Blasi et al., 1999; coconut and 
cashew shells ? Tsamba et al., 2006; olive kernel, cotton waste and forest residue ? 
Vamvuka et al., 2003).  There was no identifiable mass loss peak for lignin 
decomposition.  Previous researchers have determined that lignin decomposition occurs 
throughout the thermal decomposition temperature range (180-800?C).  This is because 
of the large molecular composition of lignin that consists of many aromatic rings with 
various branches and that the activities of the chemical bonds are covered in an extremely 
wide range (Yang et al., 2007; Gronli et al., 1999; Vamvuka et al., 2003).   
4.3.2. Thermal Decomposition of Pecan Shells in Air Atmosphere 
 The reduction in mass of pecan shells during thermal decomposition in air is 
shown in Figure 4.3.  The trend of reducing mass loss as heating rate increased that was 
observed during thermal decomposition in nitrogen was also obtained when pecan shells 
was thermal decomposed in air.  However, the oxygen content in the air caused the pecan 
shells to experience total oxidation, meaning total decomposition of all the 
hemicelluloses, cellulose and lignin compounds.  In contrast to the thermal 
decomposition of pecan shells in nitrogen where there was a 30% (of original weight) 
residual ?bio-char? left over after the process, the air thermal decomposition yielded ash 
as residual material consisting of less than 2% (of original weight).  The rate of 
101 
 
decomposition in air atmosphere is much faster than that of the nitrogen atmosphere.  The 
oxygen content in the air promoted the combustion of the char residue generated from 
thermal decomposition at lower temperatures (Liu et al., 2002).  The morphology of the 
char formed by the thermochemical conversion of pecan shells in an air atmosphere also 
aids in the total oxidation phenomena observed (Gani and Naruse, 2007). 
 
 
Figure 4.3 ? Mass loss from the thermal decomposition of pecan shells at different 
heating rates with air as carrier gas 
 The mass loss curves for pecan shell thermal decomposition in air were similar to 
those obtained when thermally decomposed in nitrogen.  Moisture release occurred 
between the temperature ranges of 30?C to 150?C (12.58% - 16.01% ).  Thermal 
decomposition of pecan shells during combustion was determined to start at 180?C and 
considered essentially finished at 600?C.  Figure 4.4 shows the mass loss rate (d?/dt) of 
pecan shells (DTG curve), calculated using the mass loss fraction (Equation 4.1).  Similar 
to the thermal decomposition in nitrogen results, the hemicelluloses-cellulose shoulder 
0
10
20
30
40
50
60
70
80
90
100
0 200 400 600 800
Mass Change 
(%
)
Temperature (?C)
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
102 
 
peaks were observed.  Peak thermal degradation for hemicelluloses occurred at 270?C, 
292?C, 309?C, 318?C and 331?C for heating rates of 5, 10, 20, 30 and 40?C/min, 
respectively.  Temperatures at which cellulose decomposition peaked were 315?C, 
327?C, 334?C, 336?C and 339?C for heating rates of 5, 10, 20, 30 and 40?C/min 
respectively.  Unlike the thermal decomposition in nitrogen, an additional peak in the 
temperature range of 430?C-600?C was observed.  This can be attributed to the 
combustion of the char residue formed from decomposition at lower temperatures.  Mass 
loss during the range of thermal decomposition of the char increased from 16.53% to 
29.31% (of total mass), increasing as heating rate increased from 5-40?C/min.   
 It should be noted that the mass loss rate is higher during air decomposition 
(Figure 4.4) compared to the rate obtained during nitrogen decomposition (Figure 4.2).  
The total oxidation of the pecan shells in an air atmosphere (e.g. combustion) allows for 
total decomposition of the cellulose, hemicelluloses and lignin compounds.  Sufficient 
amounts of oxygen in the air atmosphere allow for the decomposition of the char residue 
whereas under the nitrogen atmosphere, the nitrogen environment does not promote the 
further degradation of the lignin-rich bio-char (Liu et al., 2002).  The aromatic rings and 
complex structure of lignin must have sufficient amounts of oxygen in the air to cleave 
the C-O bonds to experience total decomposition of biological materials (Demirbas, 
2002).  Peak temperatures of mass loss rate of hemicelluloses and cellulose during 
thermal decomposition in air and nitrogen are given in Table 4.1.  Although the peak 
temperatures of hemicelluloses decomposition were similar in nitrogen and air 
atmospheres, cellulose peak decomposition temperatures in air were lower than that of 
the nitrogen environment.  This suggests that the oxygen content in the air atmosphere 
103 
 
has a more significant effect on the activation energy of pecan shells during cellulose 
decomposition than hemicelluloses decomposition (Fang et al., 2006). 
 
Figure 4.4 ? Mass loss rate from the thermal decomposition of pecan shells at 
different heating rates with air as carrier gas 
 
Table 4.1 ?Temperature (?C) peaks in mass loss rate for the main chemical 
constituents of pecan shells when thermally degraded in nitrogen and air 
Thermal 
Decomposition 
Atmosphere 
Chemical 
Constituent 
Heating Rates (?C/min) 
5 10 20 30 40 
Nitrogen hemicelluloses 275 285 305 315 330 
 cellulose 348 360 371 385 386 
Air hemicelluloses 270 292 309 318 331 
 cellulose 315 327 334 336 339 
 
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
100 200 300 400 500 600 700 800
Mass Loss 
Rate (min
-1
)
Temperature (?C)
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
104 
 
 4.3.3. FTIR Analysis 
4.3.3.1. Thermal Decomposition in Nitrogen Atmosphere 
 A typical 3-dimensional plot of the spectral obtained from the gas volatilized 
during thermal degradation of pecan shells in nitrogen atmosphere is shown in Figure 4.5.   
 
Figure 4.5 ? Three-dimensional spectral plot of the gases produced from the thermal 
degradation of pecan shells in nitrogen at a heating rate of 30?C/min 
Generally the thermal decomposition of biomass results in gas compounds evolving from 
volatilization of hemicelluloses, cellulose and lignin, the main organic constituents of 
biomass (Bassilakis et al., 2001; Baker et al., 2005; Yang et al., 2004).  The following 15 
gas compounds were identified from the FTIR spectra: acetic acid (CH
3
COOH), 
ammonia (NH
3
), carbon dioxide (CO
2
), carbon monoxide (CO), carbonyl sulfide (COS), 
ethanol (CH
2
CH
3
OH), ethylene (C
2
H
4
), formaldehyde (CH
2
O), hydrogen cyanide (HCN), 
isocyanic acid (HCNO), formic acid (HCOOH), methane (CH
4
), methanol (CH
3
OH), 
methyl isocyanate (C
2
H
3
NO) and nitrous oxide (N
2
O) (Bassilakis et al., 2001; Yang et 
al., 2004; Gronli et al., 1999).  Of the gas compounds analyzed, four were determined to 
volatilize in higher concentrations.  These four compounds were carbon dioxide, carbon 
105 
 
monoxide, ethanol and acetic acid and are referred to as major gases.  Carbon dioxide 
demonstrated a wide concentration peak ranging from 360-425?C that is representative of 
the degradation of hemicelluloses.  The concentration of CO
2
 at this thermal degradation 
range is attributed to the cracking and cutting of C-C and C-O bonds connected with the 
main branch of hemicelluloses, thus leading to the high thermal decomposition reactivity 
of hemicelluloses.  The degradation of lignin was also responsible for the volatilization of 
CO
2
, occurring at temperatures 400-600?C (Yang et al., 2007).  Lower concentrations of 
isocyanic acid, carbonyl sulfide, methane and methyl isocyanate were also found 
(referred as minor gases).  A general trend of increased concentration of these gases 
increased with heating rate.  Acetic acid, ethanol and carbon dioxide peak concentrations 
were found to correspond with the hemicelluloses-cellulose shoulder-peak phenomena 
observed in the mass loss rate peaks during the pyrolysis procedure previously discussed.  
The major and minor gaseous products volatilized during nitrogen thermal decomposition 
are shown in Figures 4.6 and 4.7. 
The shoulder peak phenomenon was not observed for carbon monoxide.  Carbon 
monoxide volatilization resulted in a peak, at approximately 400?C, due to the cracking 
of carbonyl (C-O-C) and carboxyl (C=O) bonds.  At higher temperatures (> 475?C) 
lignin degradation resulted in concentration peaks of carbon monoxide.  However, carbon 
monoxide concentrations increased in the temperature region beyond hemicelluloses and 
cellulose decomposition, known as the char combustion region (Baker et al., 2005).  The 
increased concentrations of carbon monoxide and carbon dioxide in the char combustion 
region has been attributed to thermal cracking of residue in the sample, to oxidation of 
106 
 
the carbonized char and to high temperature reaction that is given by: (Yang et al., 2007, 
Baker et al., 2005; Lee and Fasina, 2009):   
CO
2
 + C = 2CO         (4.2) 
Total volume of the gaseous products volatilized during the nitrogen thermal 
decomposition of pecan shells was obtained by the following equation: 
nullnull 
nullnull
nullnull
nullnull
nullnull
               (4.3) 
Where, I is the value obtained by integrating the concentration-time curve (using 
Simpson?s Rule, ppm) and F is the flow rate of nitrogen carrier gas through the infrared 
spectrometer (20 ml/min).  Table 4.2 shows the total volume of each gaseous compound 
analyzed during the course of the thermal decomposition experiment.  Analysis of 
variance of total volume of carbon monoxide, carbon dioxide and ethanol were 
determined to significantly (P < 0.05) decrease with increased heating rate.  There was 
not a significant trend identified in the other gases analyzed.  The total volumes of all 
gases quantified were also found to be significantly (P < 0.05) decreasing as heating rate 
decreased.  This is due to the shorter residence times in higher heating rates resulted in 
higher residue during the thermal decomposition process (Lee and Fasina, 2009).  There 
was not sufficient time for the residue in higher heating rates for the oxidation of carbon 
containing gases to carbon dioxide during thermal decomposition, thus the reduction in 
carbon concentration as heating rate increased. 
107 
 
0
50
100
150
200
250
300
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Acetic Acid
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
50
100
150
200
250
300
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Carbon Dioxide
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
50
100
150
200
250
300
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Carbon Monoxide
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
50
100
150
200
250
300
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Ethanol
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
Figure 4.6 ? Concentration of the major gases volatilized during nitrogen thermal decomposition of pecan shells at heating 
rates of 5-40?C/min 
108 
 
0
5
10
15
20
25
30
35
40
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Methyl isocyanate
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
5
10
15
20
25
30
35
40
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Methane
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
5
10
15
20
25
30
35
40
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Isocyanic acid
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
5
10
15
20
25
30
35
40
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)Carbonyl sulfide
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
Figure 4.7 ? Concentration of the minor gases volatilized during nitrogen thermal decomposition of pecan shells at heating 
rates of 5-40?C/min 
109 
 
Table 4.2 ? Effect of heating rate on the concentration (ml) of gases produced during 
thermal decomposition of pecan shells in nitrogen atmosphere 
Heating rate (?C/min) 
Gas Type 
5 10 20 30 40 
Carbon dioxide 0.0129 0.0146 0.0118 0.0093 0.0069 
Carbon monoxide 0.0349 0.0331 0.0328 0.0314 0.0294 
Isocyanic acid 0.0062 0.0049 0.0017 0.0020 0.0039 
Carbonyl sulfide 0.0015 0.0012 0.0014 0.0012 0.0010 
Acetic acid 0.0041 0.0035 0.0047 0.0034 0.0046 
Nitrous oxide 0.0004 0.0001 0.0002 0.0001 0.0001 
Methane 0.0034 0.0026 0.0012 0.0012 0.0010 
Ethylene 0.0002 0.0003 0.0001 0.0002 0.0001 
Ammonia 0.0004 0.0006 0.0004 0.0002 0.0002 
Hydrogen cyanide 0.0001 0.0003 0.0001 0.0004 0.0001 
Formic acid 0.0001 0.0007 0.0003 0.0001 0.0001 
Formaldehyde 0.0003 0.0001 0.0001 0.0001 0.0002 
Methyl isocyanate 0.0006 0.0007 0.0014 0.0007 0.0001 
Ethanol 0.0173 0.0149 0.0148 0.0098 0.0054 
Methanol 0.0010 0.0012 0.0007 0.0005 0.0003 
Total 
0.0834 0.0788 0.0716 0.0605 0.0533 
 
4.3.3.2. Thermal Decomposition in Air Atmosphere 
 Figure 4.8 shows a 3-dimensional plot of the spectra obtained from the gas volatilized 
during thermal decomposition of pecan shells in air.  The major gases identified from the spectra 
were carbon dioxide, carbon monoxide and methyl isocyanate and are shown in Figure 4.9 
below: 
110 
 
 
Figure 4.8 ? Three-dimensional spectral plot of the gases produced from the thermal 
decomposition of pecan shells in air atmosphere at a heating rate of 30?C/min 
The concentration-temperature curves of carbon dioxide and carbon monoxide gases 
demonstrated the shoulder peak trend previously shown in the mass loss rate curves in air 
thermal decomposition (Figure 4.4).  It should be mentioned that the concentration of carbon 
dioxide was higher in temperatures beyond the hemicelluloses-cellulose decomposition region.  
Similar to the nitrogen decomposition, carbon monoxide and carbon dioxide were volatilized due 
to the abscission of C-C and C-O complexes in these organic compounds.  The additional CO
2
 
peak observed at temperatures beyond hemicelluloses and cellulose decomposition range can be 
attributed to the thermal degradation of lignin present in the char residue.  The released CO
2
 was 
caused by the cracked C-O and COOH bonds in the lignin-rich char (Chouchene et al., 2010).  
Thus, carbon from the lignin compound reacts with the oxygen to form CO
2
 and CO, as 
previously shown in Equation 4.2.  Similar results were also reported by Chouchene et al. (2010) 
with olive solid waste.  Similar to the FTIR results in a nitrogen atmosphere, at any temperature, 
the concentration of the main gases volatilized increased as heating rate increased during air 
111 
 
thermal decomposition. Lower concentrations of isocyanic acid, carbonyl sulfide and acetic acid 
were found and shown in Figure 4.10. 
Equation 4.3 was utilized to calculate the total volumes of the gas products volatilized 
from the combustion of pecan shells are located in Table 4.3.  Statistical testing of total volume 
of gas products evolved from the combustion of pecan shells indicated that total volume was 
significantly (P < 0.05) affected by heating rate.  As the heating rate increased, total volume of 
carbon monoxide, isocyanic acid, carbonyl sulfide and acetic acid decreased significantly (P < 
0.05).  It is postulated that the residence times in all heating rates was sufficient for complete 
oxidation of carbon containing gases into carbon dioxide and carbon monoxide. 
Table 4.3 ? Effect of heating rate on the concentration (ml) of gases produced during 
thermal decomposition of pecan shells in air atmosphere 
Heating rate (?C/min)
Gas Type 5 10 20 30 40 
Carbon dioxide 0.0924 0.0994 0.0903 0.0826 0.0889 
Carbon monoxide 0.0197 0.0164 0.0110 0.0053 0.0031 
Isocyanic acid 0.0128 0.0098 0.0083 0.0070 0.0012 
Carbonyl sulfide 0.0041 0.0028 0.0018 0.0007 0.0004 
Acetic acid 0.0032 0.0034 0.0043 0.0025 0.0007 
Nitrous oxide 0.0002 0.0003 0.0003 0.0004 0.0005 
Methane 0.0004 0.0003 0.0002 0.0001 0.0003 
Ethylene 0.0005 0.0003 0.0009 0.0001 0.0001 
Ammonia 0.0006 0.0009 0.0003 0.0004 0.0003 
Hydrogen cyanide 0.0022 0.0021 0.0023 0.0016 0.0011 
Formic acid 0.0010 0.0001 0.0003 0.0001 0.0001 
Formaldehyde 0.0002 0.0004 0.0007 0.0002 0.0002 
Methyl isocyanate 0.0385 0.0379 0.0274 0.0346 0.0348 
Ethanol 0.0057 0.0059 0.0052 0.0049 0.0051 
Methanol 0.0003 0.0002 0.0002 0.0003 0.0001 
Total 0.1818 0.1801 0.1535 0.1409 0.1370 
112 
 
0
200
400
600
800
1000
1200
1400
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)
Carbon Dioxide
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
100
200
300
400
500
600
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature ?C)
Carbon Monoxide
10?C/min
20?C/min
30?C/min
40?C/min
5?C/min
 
 
 
 
 
 
 
 
Figure 4.9 ? Concentration of the major gases volatilized during air thermal decomposition of pecan shells at heating rates of 
5-40?C/min 
0
100
200
300
400
500
600
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)
Methyl isocyanate
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
113 
 
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)
Acetic Acid
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)
Carbonyl sulfide
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
0
50
100
150
200
250
0 100 200 300 400 500 600 700 800
Concentration (ppm)
Temperature (?C)
Isocyanic acid
40?C/min
30?C/min
20?C/min
10?C/min
5?C/min
 
 
 
 
 
 
 
 
Figure 4.10 - Concentration of the minor gases volatilized during air thermal decomposition of pecan shells at heating rates of 
5-40?C/min
114 
 
4.4. Conclusions 
 The thermal decomposition of pecan shells in nitrogen and air atmospheres were 
observed at five different heating rates.  There was evidence of the thermal degradation 
of the major chemical constituents of pecan shells (i.e. hemicelluloses, cellulose and 
lignin).  The four main stages of mass loss were observed to be moisture evaporation, 
hemicelluloses and cellulose decomposition and lignin degradation.  Moisture was 
released in the temperature range of 30 to 150?C in both atmospheres.  Mass loss rate 
increased as heating rate increased in both nitrogen and air atmospheres. 
During thermal decomposition under nitrogen, hemicelluloses and cellulose 
decomposition peaks were observed.  Mass loss rate peaks at temperatures of 275, 285, 
305, 315 and 330?C at heating rates of 5, 10, 20, 30 and 40?C/min were determined to 
attribute to hemicelluloses decomposition.  Cellulose nitrogen decomposition 
demonstrated mass loss rate peaks at temperatures of 348, 360, 371, 385 and 386?C for 
heating rates of 5-40?C/min, respectively.  The nitrogen decomposition of 
hemicelluloses, cellulose and some of the lignin resulted in approximately 40% mass loss 
of total weight of the pecan shell sample.  Similar to other researchers, a thermal 
degradation mass loss rate peak was not observed for lignin.  However, similar to other 
researchers, a flat, tailing section of mass loss rate at higher temperatures indicated 
thermal degradation of lignin.  The nitrogen thermal degradation of pecan shells was 
determined to be essentially complete at 600?C, with a residual content of 30% of 
original weight. 
The thermal decomposition of pecan shells in an air atmosphere was similar to the 
thermal degradation observed under nitrogen atmosphere in that the mass loss rate 
115 
 
increased with increased heating rate and the hemicelluloses-cellulose decomposition 
shoulder-peaks were observed.  However, mass loss rate in the air atmosphere was higher 
compared to nitrogen thermal decomposition due to the oxidation of the char residue 
generated during lower temperature decomposition.  Mass loss rate during hemicelluloses 
decomposition exhibited peaks at temperatures of 270, 292, 309, 318, 331?C for heating 
rates of 5, 10, 20, 30 and 40 ?C min
-1
, respectively.  Cellulose decomposition mass loss 
rate peaks also increased with increased heating rate.  Cellulose mass loss rate peaks (in 
air atmosphere) were observed at temperatures of 315, 327, 334, 336, 333?C for heating 
rates of 5 ? 40?C/min, respectively.  Combustion of the char residue demonstrated peaks 
at temperatures of 450, 456, 467, 469 and 517?C at heating rates of 5, 10, 20, 30 and 
40?/min, respectively. 
Heating rate significantly (P < 0.05) affected the total volume of gaseous products 
volatilized from the thermal decomposition of pecan shells in nitrogen and air 
atmospheres.  The major gases evolved from nitrogen thermal decomposition of pecan 
shells were carbon dioxide, carbon monoxide, ethanol and acetic acid.  The major gases 
volatilized from air thermal decomposition of pecan shells were carbon dioxide, carbon 
monoxide and methyl isocyanate. 
 
 
 
 
 
  
116 
 
4.5. References 
Arias, M.E., Polvillo, O., Rodriguez, J., Hernandez, M., Gonzalez-Perez, J.A., Gonzalez-
Vila, F.J. 2006. Thermal transformations of pine wood components under pyrolysis-gas 
chromatography/mass spectrometry conditions. Journal of Analytical and Applied 
Pyrolysis 77: 63-67. 
 
ASABE Standard. 2006. S593 Terminology and definitions for biomass production, 
harvesting and collection, storage, processing, conversion and utilization. St. Joseph, MI: 
ASABE. 
 
Bahng, M.K., Mukarakate, C., Robichaud, D., Nimlos, M. 2009. Current technologies for 
the analysis of biomass thermochemical processing: a review. Analytica Chimica Acta 
651: 117-138. 
 
Baker, R.R., Coburn, S., Liu, C., Tetteh, J. 2005. Pyrolysis of saccharide tobacco 
ingredients: a TGA-FTIR investigation. Journal of Analytical and Applied Pyrolysis 74: 
171-180. 
 
Bassilakis, R., Carangelo, R.M., Wojitowitcz, M.A. 2001. TG-FTIR analysis of biomass 
pyrolysis. Fuel 80: 1765-1786. 
 
Biagini, E., Barontini, F., Tognotti, L. 2006. Devolatilization of biomass fuels and 
biomass components studied by TG-FTIR technique. Industrial and Engineering 
Chemistry Research 45: 4486-4493. 
 
Butterman, H., Castildi, M.J. 2009. CO2 as a carhon neutral fuel source via enhanced 
biomass gasification. Environmental Science and Technology 43(23): 9030-9037. 
 
?a?lar, A., Demirba?, A. 2002. Conversion of cotton cocoon shells to hydrogen rich 
gaseous products by pyrolysis. Energy Conversion and Management 43: 489-497. 
 
Chouchene, A., Jeguirim, M., Khiari, B., Zagrouba, F., Trouve, G. 2010. Thermal 
degradation of olive solid waste: Influence of particle size and oxygen concentration. 
Resources, Conservation and Recycling 54: 271-277. 
 
de Jong, W., Di Nola, G., Venneker, B.C.H., Spliethoff, H., Wojtowicz, M.A. 2007. TG-
FTIR pyrolysis of coal and secondary biomass fuels: determination of pyrolysis kinetic 
parameters for main species and NOx precursors. Fuel 86: 2367-2376. 
 
Demirba?, Ahyan. 1998. Kinetics for non-isothermal flash pyrolysis of hazelnut shells. 
Bioresource Technology 66: 247-252. 
 
117 
 
Di Blasi, C., Signorelli, G., Di Russo, C., Rea, G. 1999. Product distribution from 
pyrolysis of wood and agricultural residues. Industrial Engineering and Chemistry 
Research 38: 2216-2224. 
 
Fang, M.X., Shen, D.K., Li, Y.X., Yu, C.J., Luo, Z.Y., Cen, K.F. 2006. Kinetic study on 
pyrolysis and combustion of wood under different oxygen concentrations by using TG-
FTIR analysis. Journal of Analytical and Applied Pyrolysis 77: 22-27. 
 
Gani, A., Naruse, I. 2007. Effect of cellulose and lignin content on pyrolysis and 
combustion characteristics for several type of biomass. Renewable Energy 32: 649-661. 
 
Gronli, M., Antal Jr., M., Varhegyi, G. 1999. A round-robin study of cellulose pyrolysis 
kinetics by thermogravimetry. Industrial Engineering and Chemical Research 38: 2238-
2244. 
 
Guintoli, H., de Jong, W., Arvelakis, S., Spliethoff, H., Verkooijen, A.H.M. 2009. 
Quantitative and kinetic TG-FTIR study of biomass residue pyrolysis: dry distiller's 
grains with solubles (DDGS) and chicken manure. Journal of Analytical and Applied 
Pyrolysis 85: 301-312. 
 
Hu, S., Jess, A., Xu, M. 2007. Kinetic study of chinese biomass slow pyrolysis: 
comparison of different kinetic models. Fuel 86: 2778-2788. 
 
Lee, S-B, Fasina, O. 2009. TG-FTIR analysis of switchgrass pyrolysis. Journal of 
Analytical and Applied Pyrolysis 86: 39-43. 
 
Liu, Q., Wang, S., Zheng, Y., Luo, Z., Cen, K. 2008. Mechanism study of wood lignin 
pyrolysis by using TG-FTIR analysis. Journal of Analytical and Applied Pyrolysis 82: 
170-177. 
 
Miranda, R., Sosa_Blanco, C., Bustos-Martinez, D., Vasile, C. 2007. Pyrolysis of textile 
wastes I. Kinetics and yields. Journal of Analytical and Applied Pyrolysis 80: 489-495. 
Mothe C.G., Miranda, I.C. 2009. Characterization of sugarcane and coconut fibers by 
thermal analysis and FTIR. Journal of Thermochemical and Analytical Calorimetry 
97:661-665. 
 
Rao, T.R., Sharma, A. 1998. Pyrolysis rates of biomass materials. Energy (23) 11: 973-
978. 
 
Souza, B.S., Moreira, A.P., Teixeira, A.M. 2009. TG-FTIR coupling to monitor the 
pyrolysis products from agricultural residues. Journal of Thermal Analysis and 
Calorimetry 97: 637-642. 
 
Tsamba, A., Yang, W., Blasiak, W. 2006. Pyrolysis characteristics and global kinetics of 
coconut and cashew nut shells. Fuel Processing Technology 87: 523-530. 
118 
 
Wang, S., Jian, X.M., Wang, N., Yu, Y.J., Li, Z., He, P.M. 2007. Research on pyrolysis 
characteristics of seaweed. Energy and Fuels 21: 3723-3729. 
 
Vamvuka, D., Kakaras, E., Kastanaki, E., Grammelis, P. 2003. Pyrolysis characteristics 
and kinetics of biomass residuals mixtures with lignite. Fuel 82: 1949-1960. 
 
Vuthaluru, H.B. 2003. Thermal behavior of coal/biomass blends during co-pyrolysis. 
Fuel Processing Technology 85: 141-155. 
 
Yang, H., Yan, R., Chin, T., Lian, D.T., Chen, H., Zheng, C. 2004. Thermogravimetric 
analysis - Fourier transform infrared analysis of palm oil waste pyrolysis. Energy and 
Fuels 18: 1814-1821. 
 
Yang, H., Yan, R., Chen, H., Lee, D.H., Sheng, C. 2007. Characteristics of hemicellulose, 
cellulose and lignin pyrolysis. Fuel 86: 1781-1788. 
 
119 
 
CHAPTER 5. DETERMINATION OF ENERGY REQUIREMENT AT 
DIFFERENT THERMAL DEGRADATION TEMPERATURES 
 
5.1. Introduction 
 Thermal degradation of biomass in the absence of an oxidizing agent is the first 
step in the thermochemical conversion of biomass into char, bio-oil and synthesis gas 
(Garcia-Nunez et al., 2008; Chen et al., 2003; Yaman, 2004; Guo and Lua, 2001).  The 
economics of thermochemical conversion of biomass into value-added products depends 
on the amount of energy consumed and produced during the process (He et al, 2006).  
Enthalpy is therefore defined as the energy required to raise biomass from room 
temperature to the reaction temperature and convert the solid biomass into gas, liquid and 
solid (char) products (Daugaard and Brown, 2003).  Enthalpy needed for thermal 
degradation can vary with different types of biomass. Enthalpy will therefore play an 
important role in reactor design, specification of operation parameters, energy balance 
analysis and the analysis of biomass pyrolysis, gasification and combustion (He et al., 
2006; Daugaard and Brown, 2003). 
 Therefore, the objective of this study was to quantify the energy required to 
thermally degrade pecan shells. 
120 
 
5.2. Materials and Methods 
5.2.1. Sample Preparation 
 The pecan shells used in this study were obtained from the Louisville Pecan 
Company, Louisville, AL.  The samples were stored in the Biosystems Engineering 
Process Engineering Laboratory with an average temperature of 25?C.  Before use, the 
pecan shells were ground in a Wiley mill to pass through a standard 40-mesh screen.  
Energy content, ash content and ultimate analysis of pecan shells were previously 
discussed in Section 3.3.2 and values given in Tables 3.2 and 3.3, respectively. 
 
5.2.2. Differential Scanning Calorimetry 
 The differential scanning calorimeter (DSC) used in this study was a DSC Q200 
(Perkin-Elmer, Shelton, CT) that was operated under nitrogen atmosphere at a flow rate 
of 20 ml/min.  Samples  (approximately 5 mg) were equilibrated at 30?C for 2 minutes, 
then heated to 550?C at five different heating rates (5, 10, 20, 30 and 40?C/min) and then 
held at 550?C for 2 minutes.  Temperature (?C), time (min), heat flow (W/g) and sample 
purge flow (ml/min) were all recorded by the software provided by the manufacturer of 
the DSC.  The equipment was calibrated before use according to manufacturer?s 
specifications. 
 
5.2.3. Data Analysis 
 Statistical analysis was performed using SAS statistical package (Version 9.2, 
SAS Institute Inc., Cary, NC, 2008-2009) at the 95% confidence interval.  The 
121 
 
experimental data was plotted using Microsoft Excel (Microsoft Office XP Professional, 
2007). 
 
5.3. Results and Discussion 
 Moisture evaporation typically occurs between ambient (30?C) and 150?C.  
Figure 5.1 below shows the heat flow (W/g) of pecan shells pyrolysis from 30?C to 
550?C at heating rates of 5, 10, 20, 30 and 40?C/min.  The negative values of heat flow 
indicate endothermic reactions.  The large endothermic peaks at temperatures around 
100?C are due to moisture evaporation from the pecan shells samples (He et al., 2006; 
Gomez et al., 2009).  
 
Figure 5.1 ? Heat flow (W/g) of pecan shells pyrolysis at different heating rates 
 
   Since the thermal decomposition of pecan shells begin at approximately 180?C 
(as reported in Section 4.3.1. of this thesis), the energy requirement to thermally degrade 
-45
-40
-35
-30
-25
-20
-15
-10
-5
0
0 100 200 300 400 500 600
Heat Flow
 
(W/g)
Temperature (?C)
5?C/min
10?C/min
20?C/min
30?C/min
40?C/min
122 
 
pecan shells was divided into two zones: a) the energy required to evaporate moisture in 
the sample and b) the energy required to heat the now moisture-free pecan shells.  Based 
on the DSC curves obtained at the different heating rates, the temperature that was used 
to divide the two zones was 175?C.  Therefore, the moisture evaporation temperature 
zone was calculated from 30-175?C and the thermal decomposition energy was 
calculated from 175-550?C.  The energy required in these two temperature zones were 
calculated based on sigmoidal tangent baseline calculation where two tangent lines were 
fit within the selected limits (30-175?C and 175-550?C) and a sigmoidal curve joined the 
tangent lines together and the area under the curve was calculated via the integration 
equation below (He et al., 2006): 
null
null
null,null
null null
nullnull
null
null
null
nullnull
nullnull
nullnull
null
null
null
null
null
null
null,null
null
null
nullnull    (5.1) 
Where, Q is caloric requirement of pecan shells pyrolysis (J), m
s,o
 is initial mass of 
sample (g), ms is mass of sample during experiment (g), c
p
 is specific heat of sample (J/g 
?C), dT/dt is the ratio of the sample temperature and time of DSC run (?C/s) and null
null
null
 is 
heat flow caused by the reaction heat of pecan shells thermal decomposition. 
 Table 5.1 shows the energy required to evaporate moisture and the thermally 
degrade pecan shells.  The energy required to remove moisture from the sample was 
significantly higher (P < 0.05) than that required for thermal degradation of pecan shells.  
Similar results were reported by Park et al. (2007) for wood cellulose fibers and 
Daugaard and Brown (2003) for oak, oat hulls, corn stover and pine. 
 
123 
 
Table 5.1 ? Energy requirements for moisture evaporation and thermal 
degradation of pecan shells at different heating rates 
Heating rates(?C/min) 
Temperature Zones 5 10 20 30 40 
0 - 175?C 3.16 3.41 3.24 3.47 2.94 Energy 
(MJ/kg) 175 - 550?C 2.93 3.22 2.62 2.72 2.78 
% of Energy in 
Raw* Pecan Shells 30.38 33.04 29.19 30.86 28.53  
% of Energy in  
Dry* Pecan Shells 14.61 16.05 13.06 13.56 13.86  
*Raw pecan shells energy calculation based on moisture content level of 15.02% (w.b.) 
*Dry pecan shells energy calculation based on moisture-free or 0% moisture content. 
 
 Statistical analysis indicated that the energy required to remove moisture (30-
175?C) and pyrolysis temperatures (175-550?C) were not significantly (P > 0.05) affected 
by heating rate.  It should be mentioned that the energy required to drive off moisture and 
raise the temperature of the pecan shells (moisture content of 15.02%, w.b.) to the 
thermal decomposition temperature range was approximately 6 MJ/kg.  As previously 
reported in Section 3.3.2.1. of this thesis, the energy content of raw pecan shells was 
20.06 MJ/kg.  Therefore, the energy calculated in the two temperature zones of pyrolysis 
required approximately 30% of the energy available in pecan shells.  However, if the 
pecan shells are bone-dry (or 0% moisture) the energy to thermally degrade the pecan 
shells is approximately 14%.  This is an important logistical and economic parameter for 
the potential use of pecan shells for value-added applications.  Ideal biomass feedstocks 
for the conversion to value-added products have to contain more energy than the 
conversion process requires.  If not, the conversion process will not be economically 
feasible. 
 
124 
 
5.4. Conclusions 
 This study found that the energy requirement to remove moisture during thermal 
degradation of pecan shells under nitrogen is significantly greater than the energy 
required to thermally degrade pecan shells.  The energy required in the moisture 
evaporation and the thermal degradation stages were not significantly affected by heating 
rate.  The energy required to drive off moisture and raise the temperature of raw pecan 
shells to thermal degradation temperatures was approximately 30% of the energy 
available in the sample.  However, if the pecan shells are dried to 0% moisture, the 
thermal degradation of the material uses approximately 14% of the energy available. 
125 
 
5.5. References 
Cai, J., Liu, R. 2007. Research on water evaporation in the process of biomass pyrolysis. 
Energy and Fuels 21: 3695-3697. 
 
Cai, J., Chen, S. 2008. Determination of drying kinetics for biomass by 
thermogravimetric analysis under nonisothermal conditions. Drying Technology 26: 
1464-1468. 
 
Cozzani, V., Petarca, L., Tognotti, L. 1995. Devolatilization and pyrolysis of refuse 
derived fuels: characterization and kinetic modelling by a thermogravimetric and 
calorimetric approach. Fuel 74(6): 903-912. 
 
Daugaard, D., Brown, R. 2003. Enthalpy for pyrolysis for several types of biomass. 
Energy and Fuels 17: 934-939. 
 
Garcia-Nunez, J.A., Garcia-Perez, M., Das, K.C. 2008. Determination of kinetic 
parameters of thermal degradation of palm oil mill by-products using thermogravimetric 
analysis and differential scanning calorimetry. Transactions of the ASABE 51(2): 547-
557. 
 
Gomez, C., Velo, E., Barontini, F., Cozzani, V. 2009. Influence of secondary reactions on 
the heat of pyrolysis of biomass. Industrial Engineering and Chemical Research 48: 
10222-10233. 
 
He, F., Yi, W., Bai, X. 2006. Investigation on caloric requirement of biomass pyrolysis 
using TG-DSC analyzer. Energy Conversion and Management 47: 2461-2469 
 
Murugan, P, Mahinpey, N., Johnson, K.E., Wilson, M. 2008.  Kinetics of the Pyrolysis of 
Lignin Using Thermogravimetric and Differential Scanning Calorimetry Methods. 
Energy and Fuels 22: 2720-2724. 
 
Park, S., Venditti, R., Jameel, H., Pawlak, J. 2007. Studies of the heat of vaporization of 
water associated with cellulose fibers characterized by thermal analysis. Cellulose 14: 
195-204. 
 
Stenseng, M., Jensen, A., Dam-Johnson, K. 2001. Investigation of biomass pyrolysis by 
thermogravimetric analysis and differential scanning calorimetry. Journal of Analytical 
and Applied Pyrolysis 58-59: 765-780. 
 
Yang, H., Yan, R., Chen, H., Lee, D.H., Sheng, C. 2007. Characteristics of hemicellulose, 
cellulose and lignin pyrolysis. Fuel 86: 1781-1788. 
 
126 
 
CHAPTER 6. CONCLUSION 
 It can be concluded from this study that particle size and moisture content 
significantly affected the physical properties of pecan shells. Increases in particle size and 
moisture content resulted in decreases in bulk, tap and density.  The porosity of pecan 
shells separated by particle size indicated that the particles were irregular and non-
spherical in shape.  Compressibility of pecan shells increased with increased pressure and 
decreased particle size or increased moisture content.  Flow characterizations of particle 
sizes fractions ranged from free-flowing to easy-flowing as particle size increased and 
from easy flowing at a moisture content of 4.11% (w.b.) to cohesive at a moisture content 
of 24.70% (w.b.).  Cohesion of particle sizes increased as particle size and moisture 
content increased.  Average angle of internal friction for the particle sizes were 
determined to be 36.87, 42.56 and 44.87? for the fine, medium and coarse particle size 
fractions (for all pressures), respectively. 
 The rate of moisture sorption was determined.  The Page model gave a food fit to 
the experimental data at three relative humidity levels of 50, 65 and 80% and 
temperatures of 15, 25 and 35?C.  The Henderson and Chung-Pfost equations were the 
best fit to the EMC-ERH data.
 Thermal decomposition of pecan shells was carried out at 5 different heating rates 
(5, 10, 20, 30 and 40?C/min).  Four stages of thermal degradation were observed.  The 
stages include: moisture evaporation, hemicelluloses decomposition, cellulose 
127 
 
decomposition and lignin degradation.  During nitrogen thermal decomposition, peak 
hemicelluloses decompositions were found to occur at 275-330?C, whereas cellulose 
decomposition peaks were determined to occur at temperatures ranging from 348-386?C.  
No identifiable peak was determined for lignin degradation in the nitrogen atmosphere.  
Thermal decomposition under air demonstrated hemicelluloses decomposition peaks at 
temperatures of 270-331?C.  Cellulose decomposition in an air atmosphere occurred at 
temperatures of 315-339?C.  Oxidation of the char residue demonstrated peaks at 
temperatures of 450-517?C.  Mass loss rate and thermal decomposition peaks of 
hemicellululoses, cellulose and char increased with increases in heating rate in both 
atmospheres. 
 Fifteen gases were identified to evolve during thermal decomposition of pecan 
shells.  The major gases evolved from the nitrogen decomposition process were carbon 
dioxide, carbon monoxide, acetic acid and ethanol while the major gases volatilized 
during the air decomposition process were carbon dioxide, carbon monoxide and methyl 
isocyanate.  For each gas analyzed, concentration increased at all temperatures as heating 
rate increased.  The total volume of gaseous products volatilized during thermal 
decomposition in air and nitrogen atmospheres decreased as heating rate increased. 
 The energy requirement for the thermal decomposition of pecan shells was 
determined in two stages; moisture evaporation and the thermal degradation.  It was 
determined that the energy required to remove moisture from pecan shells was greater 
128 
 
than the energy required to heat pecan shells to the thermal decomposition temperature 
range.  Heating rate did not affect the energy requirements at the two temperature zones 
examined.  The energy consumed to drive off moisture and raise the raw pecan shells to 
thermal degradation temperatures was approximately 30% of the energy content of the 
pecan shell sample.
129 
 
CHAPTER 7. FUTURE WORK 
 This study provides useful information regarding particle size and moisture 
content and their effect on physical properties, compressibility and flowability of pecan 
shells.  This will aide in the efficient design and selection of proper equipment for the 
storage, handling and transportation of pecan shells.  The thermal decomposition 
characteristics were also examined, providing useful data to develop thermochemical 
conversion techniques specific for pecan shells.  However, more research can be done to 
gain further understanding of the material. 
 In order for pecan shells to be considered as a biomass feedstock for value-added 
applications, a form of densification must be determined to increase bulk density.  
Pelleting has been developed for other biomass materials, such as switchgrass and wood 
chips.  Research into optimal moisture content and particle size for the pelletization of 
pecan shells should be considered to make transportation of the raw material as 
economically advantageous as possible.
Additional tests on the flowability of pecan shells should be completed.  As 
observed in the present data, pecan shells at a moisture content of 19 % (w.b.) were 
characterized as an easy flowing material, whereas shells at a moisture content of 24% 
were characterized as a cohesive material.  The point at which the material changes from 
easy flow to cohesive must be examined in order to determine the specific moisture 
content at which the pecan shells gain cohesive properties.  In addition, the effect of 
130 
 
moisture content on the pecan shells was carried out on a single particle size.  Further 
physical property measurements must be determined for the other particle sizes used in 
the effect of particle size. 
 The data in Chapter 4 could be added to additional tests run in order to determine 
the kinetic parameters of pecan shells thermal decomposition.  For example, activation 
energy of pecan shells (determined by a DSC) is necessary to fit mass loss data to 
mathematical kinetic models.  The knowledge of kinetic parameters of the thermal 
decomposition of pecan shells will allow better design and higher efficiencies of energy 
products produced.  A further study of the feasibility of pecan shells to be used in the 
formation of bio-oil should also be considered. 
 Therefore, future research objectives could include: 
 (a)  Determination of optimal conditions for the pelleting of pecan shells, 
(b)  Characterization of the effect of moisture content on the other particle sizes 
determined in the sieve analysis of the bulk material, and 
(c)  Application of the thermal decomposition characteristics to determine kinetic 
parameters for pecan shells thermal decomposition. 
 
 
 
 
 
 
131 
 
APPENDICES 
 
132 
 
Appendix A. Effect of Particle Size Physical Property Data 
 
Table A.1. Particle size distribution of pecan shells 
Raw (D
gw
 = 1.200 mm)  Medium (D
gw
 = 1.241 mm) 
U.S. 
Sieve No. 
Sieve 
Aperture 
Size (mm) 
Distribution 
(%) 
U.S. 
Sieve No.
Sieve 
Aperture 
Size (mm) 
Distribution 
(%) 
4 4.75 6.40 10 2 0.89 
5 4.00 6.47 12 1.7 16.50 
6 3.36 10.27 14 1.4 37.04 
8 2.36 26.53 16 1.18 25.01 
10 2.00 12.92 18 1 14.16 
14 1.40 16.98 20 0.85 3.97 
25 0.71 10.59 60 0.25 0.99 
140 0.11 6.70 140 0.106 0.17 
pan 0 3.14 pan 0.09 1.27 
Dust (D
gw
 = 0.212 mm) Coarse (D
gw
 = 2.194 mm) 
U.S. 
Sieve No. 
Sieve 
Aperture 
Size (mm) 
Distribution 
(%) 
U.S. 
Sieve No.
Sieve 
Aperture 
Size (mm) 
Distribution 
(%) 
20 0.85 9.41 4 4.75 13.88 
25 0.71 13.47 5 4 10.46 
35 0.50 16.55 6 3.36 15.78 
60 0.25 17.89 8 2.36 35.33 
80 0.18 5.72 10 2 17.03 
120 0.13 4.81 12 1.7 6.09 
230 0.06 15.53 18 1 0.93 
325 0.05 5.32 25 0.71 0.08 
pan 0.04 11.27 50 0.3 0.04 
pan 0.25 0.37 
 
 
 
133 
 
Table A.2. Bulk density 
d
gw
 (mm) ?
b
 (kg/m
3
) Mean (kg/m
3
) St.dev. 
fine 464.439 459.935 7.022 
451.844 
463.522     
medium 427.562 434.860 6.441 
439.748 
  437.269     
coarse 422.704 417.394 5.228 
412.252 
  417.228     
 
Table A.3. Tap density 
d
gw
 (mm) ?
t
 (kg/m
3
) Mean (kg/m
3
) St.dev. 
fine 613.000 601.667 9.866 
595.000 
  597.000     
medium 551.000 546.333 5.033 
541.000 
  547.000     
coarse 485.000 486.000 2.646 
484.000 
  489.000     
 
Table A.4. Particle density 
d
gw
 (mm) ?
p
 (kg/m
3
) Mean (kg/m
3
) St.dev. 
fine 1497.654 1498.512 2.398 
1496.662 
  1501.221     
medium 1452.304 1456.984 4.243 
1460.580 
  1458.068     
coarse 1440.521 1439.110 2.745 
1435.946 
  1440.862     
 
134 
 
Table A.5 Porosity 
d
gw
 (mm) ? Mean St.dev. 
fine 0.690 0.693 0.004
0.698 
  0.691 
medium 0.706 0.702 0.004 
0.699 
  0.700 
coarse 0.707 0.710 0.003 
0.713 
  0.710     
Table A.6. Hausner Ratio 
d
gw
 (mm) H
R
 Mean St.dev. 
fine 1.269 1.270 0.006
1.277 
  1.265     
medium 1.300 1.281 0.017 
1.276 
  1.266     
coarse 1.136 1.145 0.012 
1.142 
  1.159     
Table A.7. Compressibility of pecan shells 
d
gw
 
(mm) 
P 
(kPa) 
C  
(%) 
d
gw
 
(mm) 
P 
(kPa) 
C 
(%) 
d
gw
 
(mm) 
P 
(kPa) 
C 
(%) 
0.212 1.5 3.740 1.241 1.5 2.180 2.194 1.5 1.538 
0.212 1.5 3.480 1.241 1.5 1.846 2.194 1.5 1.388 
0.212 1.5 3.466 1.241 1.5 1.253 2.194 1.5 1.331 
0.212 3 5.028 1.241 3 2.665 2.194 3 1.815 
0.212 3 4.982 1.241 3 3.302 2.194 3 1.982 
0.212 3 4.977 1.241 3 3.242 2.194 3 2.399 
0.212 6 7.015 1.241 6 3.906 2.194 6 3.377 
0.212 6 7.604 1.241 6 4.679 2.194 6 3.228 
0.212 6 7.429 1.241 6 4.252 2.194 6 4.028 
0.212 9 8.374 1.241 9 4.645 2.194 9 4.020 
0.212 9 8.604 1.241 9 4.645 2.194 9 3.573 
0.212 9 8.429 1.241 9 5.164 2.194 9 3.741 
0.212 12 9.658 1.241 12 5.428 2.194 12 4.655 
0.212 12 9.723 1.241 12 5.838 2.194 12 4.337 
0.212 12 9.337 1.241 12 5.625 2.194 12 4.214 
0.212 15 10.375 1.241 15 6.631 2.194 15 5.359 
0.212 15 10.392 1.241 15 6.156 2.194 15 5.102 
0.212 15 10.452 1.241 15 6.670 2.194 15 4.572 
135 
 
Table A.8. Flow function data 
d
gw
 (mm) 0.212 1.241 2.194 
P (kPa) 
M.C.S. 
(kPa) 
U.Y.S. 
(kpa) 
M.C.S. 
(kPa) 
U.Y.S. 
(kpa) 
M.C.S. 
(kPa) 
U.Y.S. 
(kpa) 
3 8.31 1.215 5.05 1.839 5.15 2.661 
3 
7.29 0.977 4.32 0.870 8.74 1.869 
3 
8.66 1.459 6.24 1.703 6.99 2.724 
6 17.48 1.246 9.45 2.764 15.36 2.422 
6 
15.75 3.589 9.88 2.606 17.23 3.744 
6 
15.62 2.407 13.82 1.562 16.04 2.803 
9 24.85 3.243 27.28 2.843 24.89 6.029 
9 
24.27 2.311 23.04 5.888 27.56 4.958 
9 
26.29 5.173 13.55 4.148 21.97 8.564 
12 33.34 2.501 32.18 1.829 32.98 3.363 
12 
32.17 5.742 28.53 4.587 24.22 2.254 
12 
33.33 2.039 21.55 5.249 33.71 8.169 
15 40.46 3.070 29.14 4.514 52.69 2.824 
15 
44.96 3.335 30.09 12.410 45.92 1.146 
15 
 34.21 7.829 46.61 6.675 
21   31.80 11.486 64.63 12.144 
21 
  41.72 10.626 45.92 13.994 
21 
    57.67 13.255 61.39 6.929 
Table A.9. Cohesion and angle of internal friction 
d
gw
 (mm) 0.212 1.241 2.194 
P (kPa) c (kPa) ? (?) c (kPa) ? (?) c (kPa) ? (?) 
3 0.26 43.66 0.54 29.16 0.43 28.54 
3 
0.22 41.52 0.24 32.23 0.48 35.63 
3 0.30 45.29 0.42 37.51 0.41 31.00 
6 0.25 46.28 0.76 32.39 0.52 43.52 
6 
0.72 43.54 0.65 36.98 0.93 37.17 
6 0.51 44.07 0.33 44.19 0.57 45.73 
9 0.65 46.31 1.81 38.14 1.34 42.07 
9 
0.50 43.21 1.26 43.66 0.83 46.87 
9 1.16 41.69 1.39 22.34 2.03 39.27 
12 0.54 43.29 0.35 48.12 0.61 50.12 
12 
1.21 44.29 0.98 43.73 0.50 42.15 
12 0.41 46.18 1.36 35.21 1.79 42.67 
15 0.63 46.66 1.16 35.60 0.48 52.45 
15 
0.70 44.46 1.93 37.51 0.34 43.31 
15 
      1.16 51.67 
21  3.23 31.29 2.58 43.96 
21  2.71 35.95 3.02 43.31 
21 
    2.61 47.01     
136 
 
Appendix B. Effect of moisture content on physical properties 
Table B.1. Particle size distribution 
 
  Distribution (%) 
US 
Sieve 
No. 
Aperture 
size (mm) 
MC = 
4.21% 
MC = 
8.96% 
MC = 
14.59% 
MC = 
18.36% 
MC = 
24.56% 
10 2.00 0.68 0.75 0.98 2.91 5.07 
12 1.70 13.10 16.52 20.19 26.94 30.90
14 1.40 36.83 36.56 36.87 36.73 35.28 
16 1.18 27.66 26.78 25.08 22.20 20.52
18 1.00 14.36 14.27 13.18 8.88 6.87 
20 0.85 5.71 3.28 2.45 1.41 0.67
60 0.25 0.22 0.63 0.31 0.52 0.27 
140 0.11 0.42 0.11 0.30 0.14 0.16
pan  1.01 1.09 0.63 0.28 0.26 
 
Table B.2. Bulk density 
M.C. (%, wb) ?
b
 (kg/m
3
) Mean (kg/m
3
) St.dev. 
4.42 460.480 460.837 1.468 
459.581 
  462.450     
8.41 462.078 458.978 4.160 
460.607 
  454.250     
14.73 434.790 436.089 1.172 
436.407 
  437.069     
19.36 437.778 434.684 2.719 
432.674 
  433.601     
24.70 398.376 396.369 1.826 
395.924 
  394.807     
 
 
 
 
 
137 
 
Table B.3. Tap Density 
M.C. (%, wb) ?
t
 (kg/m
3
) Mean (kg/m
3
) St.dev. 
4.42 537.000 537.000 1.000 
536.000 
  538.000     
8.41 526.000 536.333 8.963 
541.000 
  542.000     
14.73 510.000 514.333 4.041 
518.000 
  515.000     
19.36 506.000 508.667 2.517 
511.000 
  509.000     
24.70 471.000 472.667 1.528 
474.000 
  473.000     
 
Table B.4. Particle Density 
M.C. (%, wb) ?
p
 (kg/m3) Mean (kg/m
3
) St.dev. 
4.42 1430.647 1429.833 0.993 
1428.727 
  1430.125     
8.41 1421.400 1421.747 0.375 
1421.694 
  1422.146     
14.73 1412.281 1412.409 0.390 
1412.098 
  1412.847     
19.36 1390.704 1390.944 0.265 
1391.229 
  1390.901     
24.70 1376.527 1375.651 0.858 
1375.614 
  1374.813     
 
 
 
138 
 
Table B.5. Porosity 
M.C. (%, wb) ? Mean St.dev. 
4.42 0.678 0.678 0.001
0.678 
  0.677     
8.41 0.675 0.677 0.003 
0.676 
  0.681     
14.73 0.692 0.691 0.001 
0.691 
  0.691     
19.36 0.685 0.687 0.002 
0.689 
  0.688     
24.70 0.711 0.712 0.001 
0.712 
  0.713     
 
Table B.6. Hausner Ratio 
M.C. (%, wb) H
R
 Mean St.dev. 
4.42 1.116 1.137 0.018
1.147 
  1.147     
8.41 1.147 1.141 0.006 
1.141 
  1.136     
14.73 1.136 1.133 0.006 
1.136 
  1.126     
19.36 1.152 1.154 0.003 
1.152 
  1.157     
24.70 1.147 1.150 0.006 
1.147 
  1.157     
 
 
 
 
139 
 
Table B.7. Compressibility of pecan shells 
M.C.        
(%, w.b.) 
P 
(kPa) 
C 
(%) 
M.C.       
(%, w.b.) 
P 
(kPa) C (%) 
M.C.        
(%, w.b.) P (kPa) C (%) 
4.42 1.5 1.957 8.41 1.5 0.769 14.73 1.5 2.180 
4.42 1.5 1.398 8.41 1.5 1.057 14.73 1.5 0.846 
4.42 1.5 0.977 8.41 1.5 0.541 14.73 1.5 1.253 
4.42 3 2.197 8.41 3 1.383 14.73 3 2.665 
4.42 3 2.277 8.41 3 2.097 14.73 3 3.302 
4.42 3 2.504 8.41 3 1.845 14.73 3 3.242 
4.42 6 3.728 8.41 6 2.880 14.73 6 3.906 
4.42 6 3.234 8.41 6 3.111 14.73 6 3.679 
4.42 6 3.037 8.41 6 2.807 14.73 6 4.252 
4.42 9 4.697 8.41 9 4.824 14.73 9 4.645 
4.42 9 3.714 8.41 9 4.917 14.73 9 4.645 
4.42 9 4.765 8.41 9 4.622 14.73 9 4.164 
4.42 12 4.908 8.41 12 4.963 14.73 12 5.428 
4.42 12 4.674 8.41 12 5.125 14.73 12 5.838 
4.42 12 5.046 8.41 12 5.175 14.73 12 5.625 
4.42 15 4.627 8.41 15 6.571 14.73 15 6.631 
4.42 15 4.759 8.41 15 5.678 14.73 15 6.156 
4.42 15 4.930 8.41 15 5.048 14.73 15 6.670 
M.C.        
(%, w.b.) 
P 
(kPa) 
C 
(%) 
M.C.       
(%, w.b.) 
P 
(kPa) C (%) 
19.36 1.5 1.510 24.7 1.5 3.559 
19.36 1.5 1.869 24.7 1.5 4.163 
19.36 1.5 1.917 24.7 1.5 3.491 
19.36 3 3.642 24.7 3 5.444 
19.36 3 2.839 24.7 3 5.826 
19.36 3 2.869 24.7 3 5.164 
19.36 6 4.354 24.7 6 7.252 
19.36 6 4.728 24.7 6 7.459 
19.36 6 3.832 24.7 6 7.227 
19.36 9 5.328 24.7 9 8.317 
19.36 9 5.409 24.7 9 7.165 
19.36 9 5.025 24.7 9 8.535 
19.36 12 5.778 24.7 12 9.494 
19.36 12 6.478 24.7 12 8.946 
19.36 12 6.256 24.7 12 10.316 
19.36 15 6.264 24.7 15 10.792 
19.36 15 6.996 24.7 15 10.337 
19.36 15 6.244 24.7 15 10.928 
 
 
 
 
140 
 
Table B.8. Flow functions of pecan shells 
M.C. (%, w.b.) 4.11 8.53 14.57 
P (kPa) 
U.Y.S. 
(kPa) 
M.C.S. 
(kPa) 
U.Y.S. 
(kPa) 
M.C.S. 
(kPa) 
U.Y.S. 
(kPa) 
M.C.S. 
(kPa) 
3 0.52 5.25 1.23 4.23 1.84 5.05 
3 0.68 4.74 1.20 7.20 0.87 4.32 
3 
0.74 7.42 0.37 6.37 1.70 6.24 
6 1.55 15.82 2.23 10.23 2.76 9.45 
6 1.90 12.06 2.87 12.87 2.61 9.88 
6 
1.70 8.22 2.91 11.91 1.56 13.82 
9 1.65 19.09 3.18 20.18 7.44 27.28 
9 1.54 19.54 4.56 23.56 5.89 23.04 
9 
2.79 24.32 2.20 19.20 4.15 13.55 
12 3.15 32.49 3.14 33.14 1.83 32.18 
12 2.51 27.80 3.34 30.34 4.59 28.53 
12 
2.13 36.74 3.67 27.71 5.25 21.55 
15 6.51 35.17 4.51 29.14 
15 4.81 32.81 12.41 30.09 
15 
    6.32 30.32 7.83 34.21 
21 8.24 40.24 11.48 31.80 
21 8.21 45.30 10.62 41.72 
21 
    7.67 56.67 13.26 57.67 
M.C. (%, w.b.) 18.67 24.23 
P (kPa) 
U.Y.S. 
(kPa) 
M.C.S. 
(kPa) 
U.Y.S. 
(kPa) 
M.C.S. 
(kPa) 
3 1.09 5.42 1.341 6.79
3 1.12 5.34 1.089 5.22 
3 0.53 5.77 3.549 8.14
6 2.38 10.05 8.116 12.48 
6 2.57 10.08 6.33 9.59 
6 2.67 14.21 3.07 11.67 
9 2.42 17.61 6.754 19.44 
9 1.63 20.88 9.762 22.78 
9 4.63 27.86 4.242 28.11 
12 6.24 28.38 16.655 45.03 
12 6.69 30.30 10.04 33.21 
12 7.34 24.85 
15 
8.91 33.09 16.885 54.19
15 
9.73 36.88 17.948 60.4
15 
3.88 38.87     
21 
18.34 50.69 25.668 63.29
21 
8.66 57.33 20.162 66.7
21 
13.29 61.85 13.187 39
141 
 
Table B.9. Cohesion and angle of internal friction 
M.C. (%, w.b.) 4.11 8.53 14.57 
P (kPa) ? (kPa) c (kPa) ? (kPa) c (kPa) ? (kPa) c (kPa) 
3 40.12 0.12 37.89 0.3 29.16 0.54
3 28.77 0.2 39.97 0.28 32.23 0.24 
3 42.19 0.16 43.71 0.08 37.51 0.42 
6 54.21 0.25 35.84 0.57 32.39 0.76 
6 34.39 0.5 34.23 0.76 36.98 0.65 
6 10.32 0.71 39.31 0.69 44.19 0.33 
9 42.94 0.36 40.64 0.73 38.14 1.81 
9 45.84 0.11 34.27 1.15 43.66 1.26 
9 38.78 0.19 40.34 0.51 22.34 1.39 
12 46.29 0.23 39.52 0.74 48.12 0.35 
12 40.94 0.8 41.78 0.3 43.73 0.98 
12 41.21 0.71 27.7 1.11 35.21 1.36 
15     31.35 2.67 35.6 1.16 
15 46.03 0.77 32.7 3.39 
15 
    42.56 0.73 37.51 1.93 
21 33.36 2.76 31.29 3.23 
21 34.08 3.24 35.95 2.71 
21 
    39.08 2.54 47.01 2.61 
M.C. (%, w.b.) 18.67 24.23 
P (kPa) ? (kPa) c (kPa) ? (kPa) c (kPa) 
3 
37.16 0.27 34.86 0.35 
3 
32.03 0.31 30.66 0.31 
3 
41.53 0.12 41.35 0.35 
6 
31.28 0.67 54.24 0.18 
6 
31.5 0.72 25.52 1.05 
6 
37.39 0.66 37.54 0.51 
9 
39.47 0.57 43.21 2.11 
9 
41.18 0.37 49.1 1.82 
9 
40.37 1.07 54.04 2.08 
12 
41.64 1.4 44.83 1.8 
12 
41.96 1.49 44.42 2.21 
12 
36 1.87     
15 
49.6 1.64 55.71 2.46 
15 
43.12 2.11 49.38 4.75 
15 
46.22 0.78     
21 
36.32 4.64 49.38 4.75 
21 
50.61 1.55 54.16 3.26 
21 
48.92 2.49 59.3 1.81 
142 
 
Appendix C. Rate of Moisture Sorption of Pecan Shells 
 
Figure C.1. Moisture change in pecan shells exposed to air at 25?C and relative 
humidity levels of 50, 65 and 80%. Initial moisture content was 15.02% (w.b.) 
 
 
Figure C.2. Moisture change in pecan shells exposed to air at 35?C and relative 
humidity levels of 50, 65 and 80%. 
 
14
15
16
17
0 8 16 24 32 40 48 56
Moistur
e
 content (%
, 
w
.
b.)
Time (hours)
50% RH
65% RH
80% RH
fitted
8
10
12
14
16
18
0 8 16 24 32 40 48 56
Moistur
e
 content (%
, 
w
.
b.)
Time (hours)
80% RH
65% RH
50% RH
Predicted
143 
 
 
Figure C.3. Moisture change in pecan shells exposed to air at 50% relative humidity 
and temperatures of 15, 25 and 35?C. 
 
 
Figure C.4. Moisture change in pecan shells exposed to air at 65% relative humidity 
and temperatures of 15, 25 and 35?C. 
 
 
 
9
10
11
12
13
14
15
16
0 8 16 24 32 40 48 56
Moistur
e
 content (%
, 
w
.
b.)
Time (hours)
15?C
25?C
35?C
Predicted
12
13
14
15
16
0 8 16 24 32 40 48 56
Moistur
e
 content (%
, 
w
.
b.)
Time (hours)
15?C
25?C
35?C
Predicted
144 
 
Appendix D. Equipment Photos 
 
Figure D.1. Annular shear cell 
 
Figure D.2. Texture analyzer with compression cell and tight fitting piston 
 
145 
 
Figure D.3. Compression cell with tight fitting piston filled with fine, medium and 
coarse pecan shell samples (left to right) 
 
 
Figure D.4. Thermogravimetric analyzer (right) connected to Fourier-transform 
infrared spectrometer (left) by heated Teflon tube.