Scholarly article on topic 'Effect of process variables on the density and durability of the pellets made from high moisture corn stover'

Effect of process variables on the density and durability of the pellets made from high moisture corn stover Academic research paper on "Agriculture, forestry, and fisheries"

CC BY-NC-ND
0
0
Share paper
Academic journal
Biosystems Engineering
OECD Field of science
Keywords
{}

Abstract of research paper on Agriculture, forestry, and fisheries, author of scientific article — Jaya Shankar Tumuluru

A flat die pellet mill was used to understand the effect of high levels of feedstock moisture content in the range of 28–38% (w.b.), with die rotational speeds of 40–60 Hz, and preheating temperatures of 30–110 °C on the pelleting characteristics of 4.8 mm screen size ground corn stover using an 8 mm pellet die. The physical properties of the pelletised biomass studied are: (a) pellet moisture content, (b) unit, bulk and tapped density, and (c) durability. Pelletisation experiments were conducted based on central composite design. Analysis of variance (ANOVA) indicated that feedstock moisture content influenced all of the physical properties at P < 0.001. Pellet moisture content decreased with increase in preheating temperature to about 110 °C and decreasing the feedstock moisture content to about 28% (w.b.). Response surface models developed for quality attributes with respect to process variables has adequately described the process with coefficient of determination (R 2) values of >0.88. The other pellet quality attributes such as unit, bulk, tapped density, were maximised at feedstock moisture content of 30–33% (w.b.), die speeds of >50 Hz and preheating temperature of >90 °C. In case of durability a medium moisture content of 33–34% (w.b.) and preheating temperatures of >70 °C and higher die speeds >50 Hz resulted in high durable pellets. It can be concluded from the present study that feedstock moisture content, followed by preheating, and die rotational speed are the interacting process variables influencing pellet moisture content, unit, bulk and tapped density and durability.

Academic research paper on topic "Effect of process variables on the density and durability of the pellets made from high moisture corn stover"

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/issn/153751 10

Research Paper

Effect of process variables on the density and durability of the pellets made from high moisture corn stover5

CrossMark

Jaya Shankar Tumuluru*

Biofuels and Renewable Energy Technologies Department, Energy Systems & Technologies Division, Idaho National Laboratory, 750 University Blvd., Idaho Falls, ID 83415-3570, USA

ARTICLE INFO

__A flat die pellet mill was used to understand the effect of high levels of feedstock moisture

Article history content in the range of 28—38% (w.b.), with die rotational speeds of40—60 Hz, and preheating

Received 14 May 2013 temperatures of 30—110 ° C on the pelleting characteristics of 4.8 mm screen size ground corn

Received in revised form stover using an 8 mm pellet die. The physical properties of the pelletised biomass studied are:

5 November 2013 (a) pellet moisture content, (b) unit, bulk and tapped density, and (c) durability. Pelletisation

Accepted 27 November 2013 experiments were conducted based on central composite design. Analysis of variance

Published online 14 February 2014 (ANOVA) indicated that feedstock moisture content influenced all of the physical properties

at P < 0.001. Pellet moisture content decreased with increase in preheating temperature to about 110 °C and decreasing the feedstock moisture content to about 28% (w.b.). Response surface models developed for quality attributes with respect to process variables has adequately described the process with coefficient of determination (R2) values of >0.88. The other pellet quality attributes such as unit, bulk, tapped density, were maximised at feedstock moisture content of 30—33% (w.b.), die speeds of >50 Hz and preheating temperature of >90 °C. In case of durability a medium moisture content of 33—34% (w.b.) and preheating temperatures of >70 °C and higher die speeds >50 Hz resulted in high durable pellets. It canbe concluded from the present study that feedstock moisture content, followed by preheating, and die rotational speed are the interacting process variables influencing pellet moisture content, unit, bulk and tapped density and durability.

© 2013 The Author. Published by Elsevier Ltd. on behalf of IAgre. All rights reserved.

1. Introduction

Our dependence on oil brings with it concerns for long-term environmental impacts, energy security, and rising costs of living. There is an opportunity to replace fossil fuels like petroleum and coal-based products with bio-based products and

biofuels from biomass. In general, agricultural and forestry biomass are widely used to generate electricity, supply heat and steam power, and heat homes. A recent joint study by the USA, Departments of Energy (DOE) and Department of Agriculture concluded that there is more than a billion tons of biomass in the USA, which can replace about 30% of its current petroleum consumption (Perlack et al., 2005).

5 This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-No Derivative Works License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited. * Tel.: +1 208 526 0529; fax: +1 208 526 2639. E-mail addresses: JayaShankar.Tumuluru@inl.gov, tjayashankar@yahoo.com. 1537-5110/$ — see front matter © 2013 The Author. Published by Elsevier Ltd. on behalf of IAgre. All rights reserved. http://dx.doi.org/10.1016Zj.biosystemseng.2013.11.012

Nomenclature

£ random error

ANOVA analysis of variance

BD bulk density (kg m~3)

bo intercept

bo, bi, bj coefficients of regression equations

D durability (%)

d.b. dry basis

DDGS distiller's dried grains with solubles

DOE department of energy

DS die speed (Hz)

FMC feedstock moisture content (%, w.b.)

HMPP high moisture pelletisation process

Hz hertz

n number of independent variables

P statistical significance

PHT preheating temperature

PMC pellet moisture content

R2 coefficient of determination

RSM response surface methodology

SD standard deviation

SEM scanning electron microscopy

TD tapped density (kg m~3)

UD unit density (kg m~3)

w.b. wet basis

xi feedstock moisture content

x2 die speed

x3 preheating temperature

y dependent variable

1.1. Density limitations

Although biomass has received a good deal of attention for energy generation, its use at the industrial level has attracted less attention. Some of the limiting factors include: (a) its lower mass and energy densities and dispersed feedstock, resulting in higher transportations costs and (b) its customary availability in less-convenient forms. One way to overcome the mass density and transportation limitations is to densify the biomass. Biomass densification employs equipment such as pellet mills, briquette presses, screw presses, or agglom-erators that are used for converting wood, plant, and agricultural residues into uniform commodity-type products for fuel applications (Tumuluru, Wright, Hess, & Kenney, 2011). These types of equipment help to produce densified biomass with improved handling characteristics while reducing storage losses and transportation costs. Among these densifica-tion processes, pelleting and briquetting have been applied for many years in several countries to produce densified biomass for fuel applications (Tumuluru et al., 2011).

1.2. Process variables effect on the binding behaviour of biomass

MacMahon (1984) stated that controlling densification process variables is very important to achieve desired quality attributes like moisture content, density, and durability. In their review on the densification of biomass, Tumuluru et al. (2011)

indicated that mechanical preprocessing, thermal preconditioning treatments, and densification equipment parameters can have a great influence on the quality and cost of densified products produced.

1.2.1. Feedstock moisture content

According to Tumuluru et al. (2011), biomass moisture content can effect densification in three ways: (a) by lowering the glass-transition temperature, (b) by promoting a solid-bridge formation, and (c) by increasing the contact area of particles by van der Waal's forces. Mani, Tabil, and Sokhansanj (2003) also observed that moisture in the biomass during the densi-fication process acts as a binder and increases the bonding via van der Waal's forces, thereby increasing the contact area of the particle. The impact of moisture content differs depending on the type of biomass being densified.

Demirbas, Sahin-Demirbas, and Hilal-Demirbas (2004) found that increasing the moisture content of spruce-wood sawdust from 7% to 15% significantly increased the strength of the pellets. In contrast, Mani et al. (2003) found that low moisture (i.e., 5-10%, (w.b.)) of corn stover resulted in denser, more stable, and more durable briquettes. Li and Liu (2000) found that an optimum moisture content of approximately 8% was recommended to produce high-density briquettes in a punch-and-die assembly from densified tree bark, sawmill waste, wood shavings, alfalfa hay, fresh alfalfa, and grass. They also recommended that a moisture content of 5-12% (w.b.) is necessary to produce good quality logs (in terms of good density and long-term storage properties) from hardwood, softwood, and bark. Tabil and Sokhansanj (1996) observed that during compression, the protoplasm present in fresh alfalfa containing about 19% (w.b.) moisture content acted as a binder and resulted in pellets with highest durability values. Sokhansanj, Mani, Bi, Zaini, and Tabil (2005) identified that the optimum moisture content for pelleting cellulosic materials is 8-12% (w.b.), whereas for starch and protein materials (i.e., mostly animal feeds), that content can reach up to 20% (w.b.).

1.2.2. Preheating

Preconditioning biomass by preheating it prior to densifica-tion can affect both the chemical composition and the mechanical preprocessing attributes, thereby changing the way the feedstock responds during densification and improving the overall quality of the pellets (Bhattacharya, Sett, & Shrestha, 1989; Tumuluru, Wright, Kenney, & Hess, 2010). Preheating can also increase the throughput of the pellet mill and reduce the energy requirement per kilogram of the biomass pellets being formed and also helps to produce different high-quality densified products for different end-use applications (Aqa & Bhattacharya, 1992; Bhattacharya, 1993). Preheating in the presence of moisture plays an important role as it softens some of the natural binders like starch, lignin, and protein in the biomass.

1.2.3. Die speed

Die speed significantly affects the amount of material that can be pelleted and the energy required for compression. It will also influence product properties like bulk density and durability by changing the retention time of material in die. In their

studies on the pelleting of distiller's dried grains with solubles (DDGS), Tumuluru, Tabil, Opoku, Mosqueda, and Fadeyi (2010) reported that larger die diameters such as 7.2 mm produce less-durable DDGS pellets compared to smaller dies (6.4 mm), both with and without the addition of steam. Previous studies on the effects of screw speed during the extrusion of feed pellets indicated that higher screw speeds reduces viscosity, die pressure, and product temperature (Rolfe, Huff, & Hsief, 2001). These conditions may also result in less flash-off of product moisture when the extrudate leaves the die, thereby increasing the density of the product. Our literature search indicated that studies related to the effect of die speed on the quality of fuel pellets made from lignocellulosic biomass are not available.

1.2.4. Scanning electron microscopy

Pelleting and briquetting are known as high-pressure compaction processes. Both of these processes result in the formation of solid bridges due to mechanical interlocking of particles and alteration of chemical components in biomasses like protein, starch, and lignin (Sokhansanj et al., 2005; Tumuluru et al., 2011). Solid bridges can also be developed during pelleting and briquetting due to diffusion of particles at high temperatures and pressure. Pordesimo, Hames, Sokhansanj, and Edens (2005) reported that lignin and protein content of corn stover are in the ranges of 14.3—26.0% (d.b.) and 2.4—9.1% (d.b.), respectively. During pelleting, the pressure and temperature developed in the die can help to plasticise protein and starch, and soften the lignin. In their studies on understanding the natural binding in biomass briquettes using scanning electron microscopy (SEM) at 600 x magnification, Kaliyan and Morey (2010) indicated that natural binders are activated by high moisture content of biomass or by applying elevated temperature or steam during pellet production processes. The major chemical component which affects the binding in lignocellulosic biomass is lignin, so it is important to ensure that lignin plays a major role in the binding mechanism of lignocellulosic biomass. Kaliyan and Morey (2009a) showed that there is an inverse relationship between the glass-transition temperature of lignin and the moisture content of the biomass. They concluded that increasing the biomass moisture to about 20% (w.b.) reduces the glass-transition temperature of lignin from 120—140 °C to 70—90 °C. This reduction in glass transition temperature can help to bind the biomass particles to form densified products in high pressure and temperature compaction systems like pellet mill and briquettes press. In the present study scanning electron microscopy (SEM) was used to understand the state of lignin in the raw and pelletised corn stover.

1.3. Response surface methodology

Response surface methodology (RSM) is a collection of statistical and mathematical techniques that are widely used to design, develop, and formulate new products (Khuri & Cornell, 1987). The advantages of using RSM are (a) it helps in analysing the effects of independent variables and (b) it generates a mathematical model that describes the processes under study (Bas & Boyaci, 2007). Many researchers have used RSM

for understanding both process and product properties (Braga, Gomes, & Kalil, 2012; Ovissipour, Kenari, Motamedzadegan, & Nazari, 2012; Santilian-Moreno, Martínez-Bustos, Castaño-Tostado, & Amaya-Llano, 2011; Shankar & Bandyopadhyay, 2004; 2005; Shankar, Shahab, Bandyopadhyay, & Bawa, 2010; Tavakkoli, Hamidi-Esfahani, & Azizi, 2012; Tumuluru, Sokhansanj, Bandyopadhyay, & Bawa, 2013).

1.4. Objective

Many researchers have worked on the densification of herbaceous and woody biomass using pellet mills and screw/ piston presses (Li & Liu, 2000; Mani, Tabil, & Sokhansanj, 2006; Ndiema, Manga, & Ruttoh, 2002; Tabil & Sokhansanj, 1996; Tumuluru, Tabil, et al., 2010). All of these researchers have carried out pelleting studies by adjusting the moisture content and other process variables like pressure and retention time. Our literature review indicated that most of the research on pelleting and briquetting of different biomass was carried out at feedstock moisture content in the range of 7-23% (w.b.) (Brackley & Parrent, 2011; Demirbas et al., 2004; Kaliyan & Morey, 2009b; Larsson, Thyrel, Geladi, & Lestander, 2008; Li & Liu, 2000; Mani et al., 2003; Nielsen, Gardner, Poulsen, & Felby, 2009; Serrano, Monedero, Laupuerta, & Portero, 2011). Recently Zafari and Kianmehr (2013) pelleted compost at high moisture contents in the range of 35-45% (w.b.). The other process variables these authors have varied are particle size, die length, and piston speed.

Studies on pelletisation of high moisture feedstocks for producing densified biomass fuels for energy applications are not available. Other pelleting process variables like preheating and die rotational speed will affect the quality of pellets produced. The objectives of the present research are to: (a) understand the effect of feedstock moisture content (28-38% (w.b.)), die rotational speed (40-60 Hz), and preheating temperatures of (30-110 °C) on pellet physical properties like unit density, bulk density, and durability; and (b) develop response surface models and plots for the pelleting process variables with respect to physical properties.

2. Material and methods

2.1. Feedstock

Corn stover, maize leaves and stalks, was harvested from farms in Iowa, Boone, in autumn, 2011 and was baled for transportation. At the Idaho National Laboratory the baled corn stover was further size reduced to bigger particle sizes using to 50.8 mm screen size using Vermeer HG200 grinder (Vermeer Corporation-Agriculture, Pella, Iowa, USA). These bigger size corn stover particles were further size reduced to smaller grinds using a hammer mill (Bliss Eliminator Hammer mill, model E-4424-TF, manufactured by Bliss Industries, Ponca City, OK, USA) fitted with 4.8 mm screen size. The ground corn stover was stored in airtight containers and was measured for properties such as moisture, bulk, and tapped densities. The initial physical properties such as moisture content, and bulk and tapped density of the corn stover are given in Table 1. Further, the moisture content of the corn

Table 1 - Physical properties of ground corn stover.

S. No Physical properties Raw material

1 Moisture content (% (w.b.)) 8.39 (0.14)

2 Bulk density (kg m~3) 111 (3.51)

3 Tapped density (kg m~3) 139 (3.61)

Note: Numbers in the parenthesis are the standard deviation.

stover was adjusted to the desired moisture levels based on the experimental design for the pelleting studies.

2.2. Pellet mill

The flat-die pellet mill is the simplest of all pellet mill designs. In a flat-die pellet mill, the die is placed horizontally to the ground and the rollers are placed on it. Biomass is then fed by gravity to the mill. During the process, material is compressed between the rollers and the die then further extruded through the die holes. There are two types of designs commonly available: (a) flat-die pellet mills, which have a rotating die with a stationary roller shaft; and (b) a stationary die, which employs a rotating roller shaft. Normally, flat die pellet mills are preferred for research and development applications as they are simple in design and weigh less than other mills. Ring-die pellet mills, on the other hand, are used for large-scale production.

In the present study, a laboratory-scale EC0-10 (Colorado Mill Equipment, Canon City, CO, USA) flat-die pellet mill with a rotating die and stationary roller shaft was used for the present pelletisation studies (Fig. 1). The rotational speed of the die at a maximum of 60 Hz was 1750 rpm. The mill was equipped with a 7.46 kW, 460-V, 3-phase motor, and is built to withstand most pelletisation applications. The rated output of this pellet mill is 20-50 kg h_1, but this value depends on the

type of biomass material processed. This machine is intended for research and development applications, to test the pelleting characteristics of both raw and pretreated biomass. The pellet mill is equipped with both a hopper and a screw feeder to uniformly feed biomass during pelletisation tests. The hopper is provided with a flexible rectangular heater (Silicon Rubber Heater, Etched foil elements, 152 mm x 508 mm, 240VAC, Type J thermocouple, Branom Instrument, WA, USA) and a feeder with a flexible tape heater (Briskheat Xtreameflex (BriskHeat, Columbus, OH, USA), grounded heavy-insulated heating tapes, 2438.4 mm long, 240VAC). In addition, J-type thermocouples and controllers (Model 96A-FDAA-00RG, Wat-low, Burnaby, BC, Canada) were provided to heat the biomass both in the hopper and the feeder to obtain a constant temperature. The die of the pellet mill was provided with a variable frequency drive (Altivar 71, Schneider Electric, Palatine, IL, USA) variable-frequency motor driver for 7.46 kW, 480VAC/ 3-phase to control the rotational speed of the die. A variable frequency drive (Lenze AC Tech, SMVector-frequency inverter (Lenze Americas Corporation, Uxbridge, MA, USA)) was also provided with a cooler (CME ECO-HC6 (Colorado Mill Equipment, Canon City, CO, USA) incline pellet cooling conveyor with a standard 1830 mm long x 152 mm wide, 457 mm-711 mm in-feed height, and 813 mm-1067 mm discharge height). The maximum rotational speed of the cooler at 60 Hz was 33.3 rpm. The incline cooling conveyor was provided with a perforated steel bottom pan to facilitate ambient air draw. The cooler was equipped with Hi Flite Rubber Paddles (Colorado Mill Equipment, Canon City, CO, USA) mounted on a C2050H (Colorado Mill Equipment, Canon City, CO, USA) conveyor chain. Two 102 mm diameter vacuum ports were provided for ambient air input. The pellet cooler moves the pellets up a conveyor where ambient air is flowing counter to the direction of the pellets. The cooled pellets are collected in a drum or container provided at the end of cooler.

Fig. 1 - Flat-die pellet mill with other accessories.

The data on pellet-mill motor current and die rotational speeds was logged by a computer using Labview software (National Instruments Corporation, Austin, TX, USA).

2.3. Experimental design

The process variables that were studied consisted of feedstock moisture content, die rotational speed, and preheating. All of the pelleting experiments were conducted at one particle size (i.e., biomass ground using Bliss Eliminator Hammer mill, fitted with a 4.8 mm screen) using an 8 mm pellet mill die. Table 2 indicates the levels of the process variables used in the present study. A central composite experimental design was used to generate the experimental data.

2.4. High moisture pelleting process (HMPP)

Pelleting tests were carried out using a flat-die pellet mill based on a central composite design. Figure 2 shows the process-flow diagram for making pellets using high-moisture corn stover feedstock. About 3 kg of corn stover, with calculated amounts of water, was mixed in a ribbon blender (Model: RB 500, Colorado Mill Equipment, Canon City, CO, USA), to adjust the moisture content to desired levels based on the experimental design. The moisture-adjusted corn stover was stored overnight in a cold storage unit set at about 4 °C. The moisture-adjusted corn stover was further loaded into the feeder hopper of the pellet mill, where it was preheated for about 4-5 min at different temperatures based on experimental design. The preheated biomass was fed continuously using the feeder, and care was taken that there were no flow irregularities inside the pellet mill. The pellets produced were further cooled using the horizontal cooler at different cooling rates to reduce the pellet moisture content. The moisture content of the pellets after cooling was also measured to determine how much moisture is lost due to preheating, pelleting, and cooling. As the moisture in the pellets after the cooling step was found to be high, pellets were further dried in a mechanical oven at 70 °C for about 2-4 h to reduce the moisture content to about 7-9% (w.b.) for safe storage without any degradation. These dried pellets were immediately further analysed for physical properties such as unit, bulk, and tapped density, and durability. Details of the standard procedures used for measuring the physical properties of pellets are given in Table 3.

2.5. Response surface methodology (RSM)

Response surface models (see Eq. (1)) were developed for pellet quality attributes like unit, bulk, and tapped densities, and

a Die rotational speed at 60 Hz is 1750 rpm.

durability with respect to the process variables, feedstock moisture content, die rotational speed, and preheating:

n n n n

y = b0 + J2 b'x' + J2 bx2 + J2 J2 byX'Xj+ e (i)

i=1 i=1 i=1 j=1

where xt and Xj are the coded independent variables, b0, b;, bj are coefficients, n is the number of independent variables, is a random error and y is the dependent variable (Khuri & Cornell, 1987; Tumuluru, Tabil, et al., 2010). Analysis of variance (ANOVA) analysis is carried out to find the statistical significance of the independent variables with respect to the dependent variables under study. The RSM models developed for the pellet-quality attributes were used for drawing the surface plots to understand the interactive effect of the process variables on pellet-quality attributes. STATISTICA software, version 9.1 (StatSoft Inc., 2010) was used for the experimental data analysis.

3. Results

3.1. Raw materials

Feedstock moisture content of the ground corn stover was about 8.39% (w.b). The corn stover was further conditioned to different target moisture contents (28, 33, and 38% (w.b.)). The final moisture values measured after conditioning deviated ± 0.5% (w.b.) from the target values. The bulk and tapped densities of the corn stover at these moistures were in the range of 128-157 kg m~3 and 153-189 kg m~3, and the measured standard deviation values were <3 kg m~3.

3.2. Pellet properties

Figure 3 shows the ground corn stover used and the subsequent pellets produced at different process conditions. The moisture content of the pellets produced at low and high preheating temperatures and feedstock moisture content at a cooling rate of 10 Hz were in the range of 15-30% (see Fig. 4). The pellet moisture content in all the experimental conditions was not provided in the present study, as different cooling rates were employed for each experiment. The standard deviation values of the measured pellet moisture content were in the range of 0.04-1.20% (w.b.). The experimental data indicated that final moisture content of the pellet was dependent on the feedstock moisture content and preheating temperatures. Lower feedstock moisture of 28% (w.b.) and higher preheating temperatures of 110 °C employed for 4-5 min reduced the pellet moisture content to about 14-16% (w.b.), whereas higher feedstock moisture content of 38% (w.b.) and higher preheating temperatures of 110 °C resulted in 28-30% (w.b.) moisture in the pellets after cooling. The decrease in feedstock moisture content after pelleting can be attributed to preheating temperature, frictional heat developed in the die (typically the pellet die temperature reached about 100-110 °C), and cooling, which might have caused the pellets to loose most of its surface moisture due to flash-off after extrusion through a constricted die hole and further cooling. Shankar et al. (2004) in their studies on extrusion

Table 2 - Levels of the pelletisation process variables used in the present study.

S. No Process variables Coded levels

-1 0 +1

1 Moisture content (%) 28 33 38

2 Die rotational speed (Hz)a 40 50 60

3 Preheating (°C) 30 70 110

Fig. 2 - Pelletisation process flow diagram for high moisture corn stover.

cooking of food indicated that density of pellet decreased with increasing moisture (>30% (w.b.)) content and preheating temperature (>100 °C). They reasoned that the decrease in the pellet density is due flash off of moisture as the extrudate leaves the die, resulting in expanded products.

The other physical properties of pellets, unit, bulk, and tapped density, and durability measured after drying the partially dried pellets to <9% (w.b.) in laboratory oven at 70 °C for 2-4 h are given in Table 4. The average pellet length was in the range of 25-33 mm with standard deviation values in the range of 3-8 mm (not shown in Table 4). The pellet length values are larger than regular commercial pellets, which are typically in the range of 11-17 mm (Tumuluru, Sokhansanj, et al., 2010). In the present study the pellet mill was not equipped with a cutter, and thus it produced longer pellets. The unit, bulk and tapped density and durability of the pellets measured were in the range of 813-1180 kg m~3, 399-626 kg m~3 and 445-681 kg m~3 and 76-96%. Higher unit, bulk, and tapped densities resulted from feedstock moisture content values <33% (w.b.), medium to high preheating temperatures of 70-110 °C, and die rotational speeds of >50 Hz. Lower bulk and tapped density values, compared to commercial pellets, were observed in the present study, and can be attributed to the longer pellet length, which may have resulted in less compaction, resulting more void spaces in the container. Higher durability values of about >95% can be achieved at medium moisture contents of about 33% (w.b.), at 40-60 Hz die speeds and 30-110 °C preheating temperatures. From the data in Table 4, it is clear that not all the experimental conditions used resulted in good pellets with high durability and density values. Thus, the data was further analysed using response surface methodology (RSM) to understand the trends in the process variables to maximise density and durability values.

The results from the analysis of variance (ANOVA) indicated the significance of the linear, quadratic, and interaction

terms of the process variables on the quality attributes. In the case of unit, bulk, and tapped densities, the linear and quadratic terms of feedstock moisture content (xa & xi2) were found to be significant at P < 0.01 and P < 0.05. The durability of the pellets was more influenced by linear terms of feedstock moisture content (x1) and preheating (x3) at P < 0.01 and P < 0.05, quadratic term of feedstock moisture content (x12) at P < 0.05, and interactive term of feedstock moisture content and preheating (x1x3) at P < 0.01.

3.3. Response surface models

The regression equations developed for these properties in original terms are given in Table 5. All of the equations developed were statistically significant at P < 0.01. The coefficient of determination values (R2) were >0.88 for all of the equations, indicating that they have adequately described the effect of process variables on quality attributes.

3.4. Response surface plots

Response surface plots were drawn for pellet-quality attributes, such as unit, bulk and tapped densities, and durability using the response surface model developed for the experimental data. These plots were drawn for each of the two individual process variables by keeping the other one at the centre point of the central composite design. These plots were customised to attach different colours to the levels of the surface for better understanding of the trends of the process variables with respect to desired output.

3.5. Unit, bulk, and tapped density

The surface plots drawn between feedstock moisture content and preheating indicated that higher preheating temperatures

Table 3 - Standards used for measurement of pellets physical properties.

Physical property

Standard

Method description

Number of measurements

Moisture content (%) ASABE (2007)

Unit density (kgm-3)

Bulk and tapped density (kg m~3)

Tumuluru, Tabil, et al. (2010) and Tumuluru, Sokhansanj, et al. (2010) ASABE (2007)

Durability (%)

ASABE (2007)

Scanning electron microscopy

Ray, Hoover, Nagle, Chen, and Gresham (2013)

The moisture content of raw and pelleted com stover was measured by drying about 50 g samples in a heated convection oven set at 105 °C for 24 h following ASABE Standards (2007). The samples were weighed before and after drying, and the moisture content was expressed in wet basis. The reported values are an average of three measurements. The two ends of single pellet were smoothed with sand paper. Pellet length and diameter were measured using a vernier. The mass of the pellet was measured on a balance with 0.001 g precision. Unit density is measured by dividing mass of an individual pellet by its volume. Bulk density of pellets produced was determined based on ASABE Standard S269.4 (ASABE Standards, 2007). A cylindrical container 143 (152) mm in diameter and 146 (122) mm in height was used. The values given in parenthesis indicate the recommended values based on ASABE standard. The samples are poured slowly into the bucket until the bucket was overflowing. The drop height during filling was 500 mm from the bottom of the container. The excess material was removed by striking a straight edge across the top. The weight of the material with the bucket was recorded. For dense-filled bulk density, the loosely filled container was tapped on the laboratory bench five times. Tapping and filling was repeated until the container was overflowing. The filled container was weighed to 0.01-g precision. Loose-filled and dense-filled bulk density was calculated by dividing the mass over the bucket volume (Shankar et al., 2007, 2008). The reported values are an average of four measurements. Durability is defined as the ability of densified biomass to remain intact when handled. Durability of pellets was determined by tumbling the test sample at 50 rpm for 10 min, in a dust-tight enclosure. A four-compartment pellet-durability tester (Seedburo Equipment Co., Des Plaines, IL 60018, USA) was used in the present study. A sample of pellets to be tested was sieved on the appropriate sieve to remove fines. About 500 g of sample of sieved pellets are placed in each compartment of the tumbler. After tumbling for 10 min, the sample is removed, sieved using a 6.2 mm screen, and the percent of whole pellets is calculated. Fines are removed from the tumbled samples by screening on a wire sieve having openings just smaller than the nominal pellet diameter. Pellet durability is defined as:

Durability=Masstofpelrbfetu:mbis:gg * 100

Scanning electron microscopy (SEM) was performed with a JEOL JCM-5000 NeoScope™ in a high vacuum mode (JEOL; Arvada, CO, USA). The raw and corn stover pellets were dried to very low moisture contents of <2% (w.b.) in a laboratory oven at 90 °C, for 8 h. Micrographs were taken of raw biomass and pelletised samples to understand the modification of lignin in raw and pelleted biomass.

of >90 °C, and lower feedstock moisture content of 30-33% (w.b.) maximised the unit, bulk, and tapped densities to about 1100-1150, 600-620, and 660-680 kg m~3, respectively (see Figs. 5-7). Increasing the moisture content to 38% (w.b.) and

preheating to 30 °C minimised the unit, bulk, and tapped densities to about 850, 420, and 480 kg m~3, respectively. Surface plots (not shown) were drawn to understand the effect of die speed and feedstock moisture content on unit, bulk, and

Fig. 3 - Ground and pelletised corn stover. Note: Expt No.1: FMC: 28% (w.b.); DS: 40 Hz; PHT: 30 °C; Expt No.2: FMC: 38% (w.b.); DS: 40 Hz; PHT: 30 °C; Expt No.3: FMC: 28% (w.b.); DS: 60 Hz; PHT: 30 °C; Expt No.3: FMC: 38%; (w.b.) DS: 60 Hz; PHT: 30 °C. FMC: Feedstock moisture content; DS: Die speed; PHT: Preheating temperature.

Fig. 4 - Pellet moisture content at 40 Hz die speed. Note: FMC: feedstock moisture content (%, w.b.); PHT: preheating temperature (° C).

tapped densities and indicated that higher die speeds of about 60 Hz and a lower feedstock moisture content of about <33% (w.b.) maximised the densities to 1200, 620, and 680 kg m~3, respectively.

increasing the preheating temperatures to >70 °C, and lowering the die speed to about 40 Hz, minimised the durability values to about 83%. Table 6 indicates the trends of the process variables for maximisation of density and durability of the pellets produced.

Scanning electron microscopy

The SEM photographs of raw and pelleted corn stover were used to understand the morphological changes in lignin during pelleting (see Figs. 9 and 10). According to Angles, Ferrando, Farriol, and Salvado (2001), lignin in its natural form appears as droplets or agglomerated small spheres on the surface (see Fig. 9). Figure 10 indicates that lignin observed in the pelletised biomass was basically in two states: a) cross-linked and b) agglomerated in the pelleted biomass. Cross-linked lignin indicated that lignin reached its glasstransition temperature, was joined to multiple surfaces during pelletisation, and became a bonding mechanism when the pellet cooled. Agglomerated lignin indicates that the glasstransition temperature was reached by the lignin, but the lignin was balled up instead of developing cross-links with the other chemical components in the biomass.

3.6. Durability (%)

Figure 8 shows the interaction effect of preheating and feedstock moisture content on durability. Increasing preheating and reducing feedstock moisture content reduced the durability, whereas feedstock moisture content of 33-34% (w.b.) and higher preheating temperatures of >70 °C maximised the durability values. A surface plot (not shown) drawn between die speed and feedstock moisture content indicated that maximum durability values of about 96% are achievable at feedstock moisture contents of 28-34% (w.b.) and die speeds in the range of 56-60 Hz at preheating temperature of 70 °C. Lowering the feedstock moisture content to about 28% (w.b.),

4. Discussion

4.1. Pellet moisture content

According to Koullas and Koukios (1987), water present in wheat straw helps in binding to produce more durable pellets. In this study, the pelleting of corn stover in the moisture range of 28-38% (w.b.) showed some promising results in terms of the pellet physical properties. Higher feedstock moisture content resulted in higher moisture in the pellets produced. Mani et al. (2003 and 2006) indicated that moisture in the biomass improves densification characteristics by increasing bonding via van der Waal's forces. The same authors

Table 4 - Experimental runs and measured pellet physical properties.

Expt. No. Process variables Pellet properties

FMC (% (w.b.)) DS (Hz) PHT (°C) UD (kg m~3) SD BD (kg m~3) SD TD (kg m~3) SD D (%) SD

1 28 40 30 1180 60 569 5 627 6 95 0.44

2 38 40 30 1065 80 518 28 556 23 94 0.23

3 28 60 30 1125 43 626 7 678 5 95 0.13

4 38 60 30 998 47 483 21 530 11 93 0.6

5 28 40 110 1194 56 572 15 619 14 76 0.85

6 38 40 110 813 75 497 6 553 7 94 0.38

7 28 60 110 1158 74 618 12 681 6 83 0.15

8 38 60 100 1021 19 524 10 576 7 96 0.04

9 28 50 70 1115 65 604 13 667 9 87 0.12

10 38 50 70 892 15 399 4 445 15 91 0.71

11 33 40 70 1116 12 582 13 650 15 96 0.49

12 33 60 70 1156 78 582 12 636 3 94 0.13

13 33 50 30 1082 52 604 8 654 21 94 0.32

14 33 50 110 1136 40 613 10 675 6 95 0.12

15 33 50 70 1122 53 566 10 618 4 95 0.42

Note: FMC: feedstock moisture content (% (w.b.)); DS: die speed (Hz); PHT: preheating temperature (°C); UD: unit density (kg m 3); BD: bulk density (kg m~3); TD: tapped density (kg m~3); D: durability (%); SD: standard deviation.

Table 5 - Response surface models for the pellet physical properties.

Pellet properties R2

Unit density (kg m~3) Bulk density (kg m~3) Tapped density (kg m~3) Durability (%) -459.12 + 258.95x1 - 96.22x2 + 1.343x3 - 4.478x12 + 0.706x22 - 0.004x32 + 0.58x1x2 - 0.172x1x3 + 0.092x2x3 -2043.51 + 182.41x1 - 2.95x2 - 4.62x3 - 2.74x12 + 0.12x22 + 0.02x32 - 0.28x1x2 + 0.02x1x3 + 0.02x2x3 -2100.21 + 193.17x1 - 4.67x2 - 4.88x3 - 2.92xa2 + 0.14x22 + 0.02x32 - 0.29xax2 + 0.03x1x3 + 0.02x2x3 -81.18 + 13.10x1 - 0.66x2 - 0.97x3 - 0.20x12 + 0.01x22 + 0.0003x32 - 0.015x1x2 + 0.021x1x3 + 0.0031x2x3 0.92 0.89 0.88 0.91

Note: x1 feedstock moisture content (% (w.b.)); x2: die speed (Hz); x3: preheating temperature (°C).

concluded that the chemical composition of the biomass also plays an important role on densification behaviour. Moisture and temperature can result in the chemical modification of protein, starch, and lignin in the biomass, thereby facilitating

better binding. Studies conducted by Kaliyan and Morey (2009b) suggest that the glass-transition temperature of lignin is lowered by increasing the moisture content and better binding is facilitated at lower temperatures (70-90 °C).

Fig. 5 - Effect of preheating temperature and feedstock moisture content on the unit density.

Fig. 7 - Effect of preheating temperature and feedstock moisture content on the tapped density.

Fig. 6 - Effect of preheating temperature and feedstock moisture content on bulk density.

Fig. 8 - Effect of preheating temperature and feedstock moisture content on durability.

Table 6 - Process variables trends for maximisation of pellet physical properties based on surface plots.

Pellet physical property Predicted value Objective

Process variables

Feedstock moisture content (% (w.b.))

Die rotational speed (Hz)

Preheating temperature (°C)

Unit density (kg m~3) Bulk density (kg m~3) Tapped density (kg m~3) Durability (%)

1000-1200 550-600 600-660 92-96

Maximise Maximise Maximise Maximise

30-33 30-33 30-33 33-34

>50 >50 >50 >50

90-110 90-110 90-110 >70

4.2. Density

In general, feedstock moisture content of the biomass affects properties like densities, angle of friction, specific-heat capacity, force deformation characteristics, and thermal conductivity. According to Mani et al. (2003, 2006) unit and bulk densities of pelletised biomass depend upon material moisture, particle size, process temperature, and pressure. In addition, higher feedstock moisture content along with a larger particle size minimises unit and bulk density. This is corroborated by the present results, where increasing the feedstock moisture content lowered the unit, bulk, and tapped densities. The decrease in density with increase in feedstock moisture content can be due to moisture flash-off due to heat generated in the die. According to Mohsenin and Zaske (1976) and Faborode (1989) briquettes made from high feedstock moisture expand thereby resulting in higher residual stresses in the briquettes after ejection from the die causing elastic spring back or expansion. Studies conducted by Mani, Tabil, and Sokhansanj (2004) on the compaction of four biomass

grinds of wheat and barley straws, corn stover, and switchgrass at different feedstock moisture contents of 5, 10, and 15% (w.b.), respectively, indicated that a lower feedstock moisture content produced briquettes with higher densities as compared to those with a higher feedstock moisture content. Rehkugar and Buchele (1969) reported that the density of forage pellets decreases with increasing feedstock moisture content in the range of 6-25% (w.b.). The trends observed in the present study corroborate with observations from Mani et al. (2004) and Rehkugar and Buchele (1969) where lower feedstock moisture content resulted in higher unit, bulk, and tapped densities. In this study, the other process variable that influenced unit, bulk, and tapped densities is preheating temperature. Increasing the preheating temperature to 110 °C with a lower moisture content of 28% (w.b.) and a die speed to 60 Hz maximised unit, bulk, and tapped densities. These conditions might have allowed the material to exit the die quickly, with less flash-off of moisture, thereby resulting in maximum unit, bulk, and tapped densities by minimising the diametrical expansion of the pellet during extrusion. In their

Fig. 9 - SEM image of ground corn stover.

Fig. 10 - SEM image of pelletised corn stover. Note: Feedstock moisture content (33% (w.b.)), die speed: 60 Hz, and preheating temperature: 30 °C.

studies on the briquetting of corn stover and switchgrass, Kaliyan and Morey (2009b) indicated that an increase in feedstock moisture content subsequently increased the expansion of the briquettes and reduced the density.

4.3. Durability

Pellet durability indicates the ability of the pellets to withstand physical disintegration. It can also be defined as the ability of the pellets to withstand compression and resist impacts during storage and transportation. Kaliyan and Morey (2009a) pointed out that feed particle sizes, biomass constituents, feed moisture content, feed conditioning temperature/ preheating of feed, added binders and densification equipment variables all affect densified product durability. In general, biomass feedstock moisture content helps to develop intermolecular and interfacial forces, which bind the particles by pressure and temperature applications (Tumuluru et al., 2011). In the case of biomass, which is a biopolymer, in addition to the forces mentioned above, there are other reactive mechanisms that make binding plausible. Kaliyan and Morey (2009b) in their studies on the pelleting of corn stover and switchgrass reported that increasing the moisture from 10 to 20% (w.b.) reduced the glass-transition temperature of lignin to about 70 °C. Our studies indicate that corn stover has the ability to pelletise at high moisture content (>28% (w.b.)), which can be due to lowering of glass-transition temperature of lignin and other chemical components, such as starch due

thermal and mechanical degradation of corn stover in the pellet mill in presence of diluent like water. SEM image (see Fig. 10) of pelleted corn stover has corroborated the above fact. Also protein and starch in the corn stover might have got gelatinised and denatured due to preheating temperatures and frictional heat in the die and might have aided in binding by formation of protein and starch complexes. In their studies on pelleting of barley straws, Serrano et al. (2011) indicated that the highest durability values of 95.5% were achieved at moisture content values of 19-23% (w.b.). According to the same authors, reducing moisture content to lower than 19% did not result in pellet formation. These results corroborate the present results, where strong positive correlation was found with respect to feedstock moisture content and preheating temperatures.

4.4. Advantage of high moisture pelleting

The results indicate that it is possible to make corn stover pellets at high feedstock moisture content in the range of 28-38% (w.b.) by varying other process variables (e.g., preheating and die rotational speed). Yancey, Tumuluru, and Wright (2013) analysed the cost of various unit operations in the pelletisation of lodgepole pine using a process demonstration unit and indicated that a rotary dryer consumes about 70% of the energy to dry loose biomass from initial moisture content of 30% (w.b.) to final moisture content of 15% (w.b.), whereas pellet mill consumes about 7%. The major

disadvantages of using rotary dryer are: (a) it requires high quality heat, (b) it has greater particulate emissions, (c) it has greater volatile organic compounds (VOC) emissions, (d) has a greater fire hazard, (e) a large footprint, and (f) it is difficult to control the material moisture content (Worley, 2011). In the high moisture feedstock pelleting process, the biomass is preheated to approximately 110 ° C for short durations (typically 5 min) and is further pelleted. The key advantages of this approach are utilisation of short preheating temperatures, the process heat developed in the die, and further cooling, to produce partially dry pellets. These partially dried pellets can be dried to safe moisture content of <9% (w.b.) using cost and energy efficient dryers (such as grain or belt dryers). Low temperature belt or grain driers are highly efficient drying systems which have been used in agricultural and forest industries for products such as wood chips, bark, sawdust, as well as all other pourable goods. Working with low drying temperatures (60-100 °C), a belt dryer dries the product gently and homogeneously. Other decisive advantages of low temperature drying are low VOC emissions and a reduced risk of fire and explosion. Further advantage of drying pellets is reduced tendency for material to become entrained in the air flow due to higher densities. Our future studies will be aimed at (a) understanding the glass-transition temperature of lignin at different feedstock moisture contents; (b) measuring the net specific energy consumption for pelleting and, (c) completing energy balance study of high moisture pelleting process (HMPP).

5. Conclusions

The pelletisation data collected in the present study indicated that it is possible to densify high-moisture corn stover feedstock to produce a durable pellet by manipulation of process variables like preheating and die rotational speeds. Based on the present research regarding the understanding of the pel-letisation characteristics of corn stover at different feedstock moisture content values (28-38% (w.b.)), die rotational speeds (40-60 Hz), and preheating temperatures (30-110 °C), the following conclusions regarding their impact on pellet-quality attributes were drawn:

• Lower feedstock moisture content of 28% (w.b.) and higher preheating temperatures of 110 °C have reduced the final moisture content of the pellets to about 15% (w.b.). Preheating temperature, frictional heat developed in the die and further cooling has helped to produce partially dry pellets with moisture content of about 15% (w.b.).

• The other quality attributes of like unit, bulk and tapped density and durability of the pellets measured after drying the partially dry pellets to safe moisture content of <9% (w.b.) were in the range of 813-1180 kg m~3, 399-626 kg m~3 and 445-681 kg m~3 and 76-96%.

• ANOVA of the experimental data indicated that feedstock moisture and preheating influenced all the quality attributes.

• Reducing the feedstock moisture content to 30-33% (w.b.) and increasing the preheating temperatures to >90 °C increased the unit, bulk, and tapped density, whereas

feedstock moisture content of about 33-34% (w.b.) and preheating temperature of >70 °C produced more durable pellets.

• Response surface models developed for quality attributes with respect to pelleting process variables have adequately described the process with coefficient of determination values of >0.88.

• SEM studies indicate that binding is mainly due to cross-linking of lignin with the other chemical components in the biomass.

Author disclosure statement

No competing financial interests exist. This information was prepared as an account of work sponsored by an agency of the US government. Neither the US government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. References herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, do not necessarily constitute or imply its endorsement, recommendation, or favouring by the US government or any agency thereof. The views and opinions of the authors expressed herein do not necessarily state or reflect those of the US government or any agency thereof.

Acknowledgements

The author would like to acknowledge Craig C. Conner of Idaho National Laboratory and Richard McCulloch, Graduate Student, University of Utah for supporting the experimental work and SEM study. The authors would also like to acknowledge Leslie Park Ovard, Quinn Grover, Gordon Holt, David L. Combs, and Allen Haroldsen from Idaho National Laboratory's R&D Publications Support Team for their editorial and graphics assistance. This work was supported by the DOE, Office of Energy Efficiency and Renewable Energy under DOE Idaho Operations Office Contract DE-AC07-05ID14517. Accordingly, the U.S. government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains, a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes.

REFERENCES

Angles, M. N., Ferrando, F., Farriol, X., & Salvado, J. (2001). Suitability of steam exploded residual softwood for the production of binderless panels. Effect of the pre-treatment severity and lignin addition. Biomass and Bioenergy, 21(3), 211-224.

Aqa, S., & Bhattacharya, S. C. (1992). Densification of preheated sawdust for energy conservation. Energy, 17(6), 575-578.

ASABE Standards S269.4. (2007). Cubes, pellets, and crumbles -Definitions and methods for determining density, durability and moisture content. St. Joseph, Michigan, U.S.A.: ASABE.

Bas, D., & Boyaci, I. H. (2007). Modelling and optimization, part I: usability of response surface methodology. Journal of Food Engineering, 78, 836-845.

Bhattacharya, S. C. (1993). State-of-the-art of utilizing residues and other types of biomass as an energy source. RERIC International Energy Journal, 15(1), 1-21.

Bhattacharya, S. C., Sett, S., & Shrestha, R. M. (1989). State of the art for biomass densification. Energy Sources, 11, 161-182.

Brackley, A. M., & Parrent, D. J. (2011). Production of wood pellets from Alaska-grown white spruce and hemlock. Pacific Northwest Research Station, Portland, OR, U.S.A.: U.S. Department of Agriculture, Forest Service Department. General Technical Report PNW-GTR-845.

Braga, A. R. C., Gomes, P. A., & Kalil, S. J. (2012). Formulation of culture medium with agroindustrial waste for b-Galactosidase production from Kluyveromyces marxianus ATCC 16045. Food and Bioprocess Technology, 5(5), 1653-1663.

Demirbas, A., Sahin-Demirbas, A., & Hilal-Demirbas, A. (2004). Briquetting properties of biomass waste materials. Energy Sources, 26, 83-91.

Faborode, M. O. (1989). Moisture effects in the compaction of fibrous agricultural residues. Biological Wastes, 28(1), 61-71.

Kaliyan, N., & Morey, R. V. (2009a). Factors affecting the strength and durability of densified products. Biomass and Bioenergy, 33(3), 337-359.

Kaliyan, N., & Morey, R. V. (2009b). Densification characteristics of corn stover and switchgrass. Transactions of ASABE, 52(3), 907-920.

Kaliyan, N., & Morey, R. V. (2010). Natural binders and solid bridge type binding mechanisms in briquettes and pellets made from corn stover and switchgrass. Bioresource Technology, 101, 1082-1090.

Khuri, A. Z., & Cornell, J. A. (1987). Response Surfaces, 149. New York, NY: Marcel Dekker.

Koullas, D. P., & Koukios, E. G. (1987). Briquetting of wheat straw. In FAO/CNRE workshop on handling and processing of biomass for energy, 14-15 September, Hamburg, Germany (pp. 53-60).

Larsson, S. H., Thyrel, M., Geladi, P., & Lestander, T. A. (2008). High quality biofuel pellet production from pre-compacted low density raw materials. Bioresource Technology, 99, 7176-7182.

Li, Y., & Liu, H. (2000). High-pressure densification of wood residues to form an upgraded fuel. Biomass and Bioenergy, 19, 177-186.

MacMahon, M. J. (1984). Additives for physical quality of animal feed. In D. A. Beaven (Ed.), Manufacturing of animal feed (pp. 69-70). Herts, England: Turret-Wheatland Limited.

Mani, S., Tabil, L. G., & Sokhansanj, S. (2003). An overview of compaction of biomass grinds. Powder Handling Processing, 15(3), 160-168.

Mani, S., Tabil, L. G., & Sokhansanj, S. (2004). Evaluation of compaction equations applied to four biomass species. Canadian Biosystems Engineering, 46, 3.55-3.61.

Mani, S., Tabil, L. G., & Sokhansanj, S. (2006). Specific energy requirement for compacting corn stover. Bioresource Technology, 97, 1420-1426.

Mohsenin, N., & Zaske, J. (1976). Stress relaxation and energy requirements in compaction of unconsolidated materials. Journal of Agricultural Engineering Research, 21(2), 193-205.

Ndiema, C. K. W., Manga, P. N., & Ruttoh, C. R. (2002). Influence of die pressure on relaxation characteristics of briquetted biomass. Energy Conversion and Management, 43, 2157-2161.

Nielsen, N. P. K., Gardner, D. J., Poulsen, T., & Felby, C. (2009). Importance of temperature, moisture content, and species for the conversion process of wood residues into pellets. Wood and Fiber Science, 41(4), 414-425.

Ovissipour, M., Kenari, A. A., Motamedzadegan, A., & Nazari, R. A.

(2012). Optimization of enzymatic hydrolysis of visceral waste proteins of yellowfin tuna (Thunnus albacares). Food and Bioprocess Technology, 5(2), 696-705.

Perlack, R. D., Wright, L. L., Turhollow, A. F., Graham, R. L., Stokes, B. J., & Erbach, D. C. (2005). Biomass as feedstock for a bioenergy and bioproducts industry: The technical feasibility of a billion-ton annual supply. Oak Ridge National Laboratory report, D0E/G0-102005-2135.

Pordesimo, L. O., Hames, B. R., Sokhansanj, S., & Edens, W. C. (2005). Variation in corn stover composition and energy content with cop maturity. Biomass and Bioenergy, 28(4), 366-374.

Ray, A. E., Hoover, A. N., Nagle, N., Chen, X., & Gresham, G. L.

(2013). Effect of pelleting on the recalcitrance and bioconversion of dilute-acid pretreated corn stover under low-and high-solids conditions. Biofuels, 4(3), 271-284.

Rehkugar, G. E., & Buchele, W. F. (1969). Bio-mechanics of forage watering. Transactions of the ASAE, 12, 1-8.

Rolfe, L. A., Huff, H. E., & Hsief, F. (2001). Effect of particle size and processing variables on the properties of an extruded catfish feed. Journal of Aquatic Food Product Technology, 10(3), 21-34.

Santillan-Moreno, A., Martínez-Bustos, F., Castaño-Tostado, E., & Amaya-Llano, S. L. (2011). Physicochemical characterization of extruded blends of corn starch-whey protein concentrate-agave tequilana fiber. Food and Bioprocess Technology, 4(5), 797-808.

Serrano, C., Monedero, E., Laupuerta, M., & Portero, H. (2011). Effect of moisture content, particle size and pine addition on quality parameters of barley straw pellets. Fuel Processing Technology, 92, 699-706.

Shankar, T. J., & Bandyopadhyay, S. (2004). Optimization of

extrusion process variables using a genetic algorithm. Food and Bioprocess Technology, 82(2), 143-150.

Shankar, T. J., & Bandyopadhyay, S. (2005). Process variables during single screw extrusion of fish and rice flour blends. Journal of Food Processing and Preservation, 29(2), 151-164.

Shankar, T. J., Xingya, K., Sokhansanj, S., Lim, C. J., Bi, X., & Melin, S. (2007). Studies on off gassing during storage of wood pellets. ASABE Paper No. 071022. Minneapolis, Minnesota: ASABE. June 17-20.

Shankar, T. J., Xingya, K., Sokhansanj, S., Lim, C. J., Bi, X., & Melin, S. (2008). Effect of storage temperature on off-gassing and physical properties of wood pellets. ASABE Paper No.084248. Providence, Rhode Island: ASABE. June 29-July 2.

Shankar, T. J., Shahab, S., Bandyopadhyay, S., & Bawa, A. S. (2010). A case study on optimization of biomass flow during singlescrew extrusion cooking using genetic algorithm (GA) and response surface method (RSM). Food and Bioprocess Technology, 3(4), 498-510.

Sokhansanj, S., Mani, S., Bi, X. T., Zaini, P., & Tabil, L. (2005). Binderless pelletization of biomass. ASAE Paper No. 056061. St. Joseph, Mich: ASABE.

StatSoft, Inc.. (2010). STATISTICA (data analysis software system), version 9.1. www.statsoft.com.

Tabil, L. G., & Sokhansanj, S. (1996). Process conditions affecting the physical quality of alfalfa pellets. Transactions of the ASAE, 12(3), 345-350.

Tavakkoli, M., Hamidi-Esfahani, Z., & Azizi, M. H. (2012). Optimization of Corynebacterium glutamicum glutamic acid production by response surface methodology. Food and Bioprocess Technology, 5(1), 92-99.

Tumuluru, J. S., Sokhansanj, S., Bandyopadhyay, S., & Bawa, A. S. (2013). Changes in moisture, protein, and fat content of fish and rice flour coextrudates during single screw extrusion cooking. Food and Bioprocess Technology, 6(2), 403-415.

Tumuluru, J. S., Sokhansanj, S., Lim, C. J., Bi, X. T., Lau, A., Melin, S., et al. (2010). Quality of wood pellets produced in

British Columbia for export. Applied Engineering in Agriculture, 26(6), 1013-1020.

Tumuluru, J. S., Tabil, L. G., Opoku, A., Mosqueda, M. R., & Fadeyi, O. (2010). Effect of process variables on the quality characteristics of pelleted wheat distiller's dried grains with soluble. Biosystems Engineering, 105(4), 466-475.

Tumuluru, J. S., Wright, C. T., Hess, J. R., & Kenney, K. L. (2011). A review of biomass densification systems to develop uniform feedstock commodities for bioenergy applications. Biofuels Bioproducts and Biorefining, 5(6), 683-707.

Tumuluru, J. S., Wright, C. T., Kenney, K. L., & Hess, J. R. (2010). A review on biomass densification technologies for energy applications. Technical Report INL/EXT-10-18420, Idaho Falls, Idaho. Available at http://www.inl.gov/bioenergy Accessed 22.06.12.

Worley, M. (2011). Biomass drying technology update. BioPro Expo & Marketplace, Atlanta, GA, March 14-16, 2011 http://www. tappi.org/content/Events/11BIOPRO/19.2Worley.pdf. Page Accessed 17.10.13.

Yancey, N. A., Tumuluru, J. S., & Wright, C. T. (2013). Drying, grinding and pelletization studies on raw and formulated biomass feedstock's for bioenergy applications. Journal of Biobased Materials and Bioenergy, 7(5), 549-558.

Zafari, A., & Kianmehr, M. H. (2013). Factors affecting mechanical properties of biomass pellet from compost. Environmental Technology. http://dx.doi.org/10.1080/09593330.2013.833639 Accessed 17.10.13 (published online: 23rd September, 2013).