Review

A review of biomass densification

systems to develop uniform
feedstock commodities
for bioenergy application†
Jaya Shankar Tumuluru, Christopher T. Wright, J. Richard Hess and Kevin L. Kenney, Idaho National
Laboratory, Idaho Falls, ID, USA

Received April 8, 2011; revised June 28, 2011; accepted July 1, 2011
View online October 6, 2011 at Wiley Online Library (wileyonlinelibrary.com); DOI: 10.1002/bbb.324;
Biofuels, Bioprod. Bioref. 5:683–707 (2011)

Abstract: Developing uniformly formatted, densified feedstock from lignocellulosic biomass is of interest to
achieve consistent physical properties such as size and shape, bulk and unit density, and durability, which sig-
nificantly influence storage, transportation and handling characteristics, and, by extension, feedstock cost and
quality. A variety of densification systems are considered for producing a uniform format feedstock commodity
for bioenergy applications, including (i) pellet mill, (ii) cuber, (iii) screw extruder, (iv) briquette press, (v) roller press,
(vi) tablet press, and (vii) agglomerator. Each of these systems has varying impacts on feedstock chemical and
physical properties, and energy consumption. This review discusses the suitability of these densification systems
for biomass feedstocks and the impact these systems have on specific energy consumption and end-product
quality. For example, a briquette press is more flexible in terms of feedstock variables where higher moisture
content and larger particles are acceptable for making good quality briquettes; or among different densification
systems, a screw press consumes the most energy because it not only compresses but also shears and mixes
the material. Pre-treatment options like pre-heating, grinding, steam explosion, torrefaction, and ammonia fiber
explosion (AFEX) can also help to reduce specific energy consumption during densification and improve binding
characteristics. Binding behavior can also be improved by adding natural binders, such as proteins, or commercial
binders, such as lignosulfonates. The quality of the densified biomass for both domestic and international markets
is evaluated using PFI (United States standard) or CEN (European standard). Published in 2011 by John Wiley &
Sons, Ltd

Keywords: densification systems; biomass density; densification energy; biomass pre-treatment; biomass quality;
solid fuel standards

Correspondence to: Jaya Shankar Tumuluru, Biofuels and Renewable Energy Technologies Department, Energy Systems & Technologies Division,

Idaho National Laboratory, 2525 North Fremont Ave., Idaho Falls, ID 83415-2025 USA. E-mail: [email protected]
†
This article is a US Government work and is in the public domain in the USA.

Published in 2011 by John Wiley & Sons, Ltd 683

JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

Introduction of wood residues to form an upgraded fuel. Mani et al. 10

researched the compaction characteristics of lignocellulosic
ehind only coal and oil, biomass stands as the third- biomass using an Instron Universal Testing Machine.

B largest energy resource in the world.1 One of the

major limitations of using biomass as a feedstock
for bioenergy products is its low bulk density (wet basis),
Most existing literature on densification focuses largely
on understanding densification mechanisms and quality
attributes. This review provides context for these considera-
which typically ranges from 80–100 kg/m3 for agricultural tions in development of advanced biomass feedstock supply
straws and grasses and 150–200 kg/m3 for woody resources systems that meet biorefinery needs at a commodity scale. It
like wood chips and sawdust.2,3 The low densities of biomass identifies advantages and limitations of using different den-
often make the material difficult to store, transport, and sification systems to create advanced feedstocks with defi ned
interface with biorefi nery infeed systems. For example, when size, shape, and bulk flowability properties for bioenergy
low-density biomass is co-fired with coal, the difference in applications.
density causes difficulties in feeding the fuel into the boiler A variety of approaches is discussed for understanding the
and reduces burning efficiencies.4 One way to overcome role of densification in development of advanced uniform
this limitation is to increase biomass density, which has the feedstocks for bioenergy applications, including (i) mecha-
added benefit of increasing the material’s unit density as nisms of particle bonding during densification, (ii) different
much as ten-fold.5 densification systems such as pellet mill, briquette press,
The densification process is critical for producing a feed- cuber, tablet press, roller press, screw extruder and agglom-
stock material suitable as a commodity product. Densification erator, (iii) specific energy consumption of different densifi-
enables several advantages, including (i) improved handling cation systems, (iv) effects of densification process variables
and conveyance efficiencies throughout the supply system and on quality of the densified products and (v) effects of pre-
biorefinery infeed, (ii) controlled particle size distribution for treatments, such as grinding, pre-heating, steam explosion,
improved feedstock uniformity and density, (iii) fractionated torrefaction, and ammonia fiber explosion (AFEX process)
structural components for improved compositional quality, on densification process. Finally, advantages of particular
and (iv) conformance to pre-determined conversion technol- systems are discussed in relationship to bioenergy applica-
ogy and supply system specifications. tions and recommendations are made for future studies.
Common biomass densification systems have been adapted
from other highly efficient processing industries like feed, Mechanisms of bonding of particles during
food, and pharmacy, and include (i) pellet mill, (ii) cuber, densification
(iii) briquette press, (iv) screw extruder, (v) tabletizer, and The quality of densified biomass depends on strength and
(vi) agglomerator. Among these, the pellet mill, briquette durability of the particle bonds, which are influenced by
press, and screw extruder are the most common ones used a number of process variables, like die diameter, die tem-
for bioenergy production. The quality of densified biomass perature, pressure, binders, and pre-heating of the biomass
produced using these systems is evaluated with the existing mix. Tabil11 and Tabil and Sokhansanj12,13 suggested that the
international standards developed for pellet mill and bri- compaction of biomass during pelletization can be attrib-
quette press systems; there are no system-specific standards uted to elastic and plastic deformation of the particles at
developed for the others. higher pressures. According to their studies, the two impor-
A number of studies have been performed on densification of tant aspects of pelletization are (i) the ability of the particles
herbaceous and woody biomass using pellet mills and screw/ to form pellets with considerable mechanical strength, and
piston presses. Ndiema et al.6 examined the influence of die (ii) the ability of the process to increase density.
pressure on relaxation characteristics of briquetted biomass. The first is a fundamental behavior that details which
Adapa et al.7,8 studied pelletization of fractionated agricultural type of bonding or interlocking mechanism results in bet-
straws. Li and Liu9 investigated high-pressure densification ter densified biomass. Rumpf14 and Sastry and Fuerstenau15

684 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

Figure 1. Deformation mechanisms of powder particles under compression.

suggested that the mechanism of binding during agglomera- with a 70% inter-particle conformity. It is also important to
tion can be from the formation of solid bridges. These solid understand that the yield point of the material governs the
bridges are developed by chemical reactions and sintering rate of approach to the true density of the product. Because
solidification, hardening of the binder, hardening of the the loading is hydrostatic in character, the application of
melted substances, or crystallization of the dissolved mate- pressure will fracture the brittle particles. These processes
rials. The pressure applied during densification reduces may also result in mechanical interlocking. Figure 1 shows
the melting point of the particles and causes them to move the deformation mechanism of the powder particles under
toward one another, thus increasing the contact area and compression.20,21
changing the melting point to a new equilibrium level.16,17 Biological material behavior is more complex during load-
Presence of liquids, like water, during densification results in ing and may have different deformation characteristics com-
interfacial forces and capillary pressures that increase par- pared to powders compaction. Research on understanding
ticle bonding. Schineberger18 mentioned that the attraction the compaction behavior of biomass using the rheological
between the particles is due to the van der Waals electro- models that take into account the viscoelastic nature of the
static or magnetic forces, and is inversely proportional to the material is still in initial stages.22
distance between the particles, where larger distances have The chemical composition of biomass, which includes cel-
less attraction. lulose, hemicelluloses, protein, starch, lignin, crude fiber,
Mani et al.19 postulated three stages during densification fat, and ash, also affects the densification process. During
of biomass. First, particles are rearranged to form a closely compression at high temperatures, the protein and starch
packed mass where most of the particles retain their proper- plasticizes and acts as a binder, which assists in increasing
ties and the energy is dissipated due to inter-particle and the strength of the pellet.23–26 Lignin in the biomass at tem-
particle-to-wall friction. Second, the particles are forced peratures above about 140°C softens and improves the bind-
against each other and undergo plastic and elastic deforma- ing of the particles.7,12,13,27 Scanning electron microscopes
tion, which significantly increases the inter-particle contact; (SEMs) have been used to understand the solid-type bridges
particles become bonded through the van der Waals and formed during briquetting and pelleting of corn stover and
electrostatic forces. Third, a significant reduction in volume switchgrass.28 More studies at a micro level using techniques
results in the density of the material reaching the true den- like SEM and transmission electron microscope (TEM) will
sity of the component ingredients. By the end of the third be useful in understanding intra-particle cavities, material
stage, the deformed and broken particles can no longer properties, and process variable interactions on the quality
change their position due to a decreased number of cavities attributes of densified biomass.

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 685
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

Densification systems evenly to each of the two rolls. The feed distributor flights
spread the material across the face of the die. Then friction-
Pellet mill driven rolls force the feed through holes in the die as the die
Pelletization is a popular processing technique in feed and revolves. Cut-off knives mounted on the swing cover cut the
fuel manufacturing. In simple terms, pelleting converts finely pellets as they are extruded from the die, and finally the pel-
ground ingredients into dense, free-flowing, durable pel- lets fall through the discharge opening in the swing door.31
lets.29,30 A pellet has uniform product characteristics in terms Typical commercial units have two rollers to meet the
of size (length and diameter: 13–19 mm and 6.3–6.4 mm), high production rates in the range of 2.5–5 t/hr.29 Power
shape (cylindrical), and unit densities (1125–1190 kg/m3).5 consumption of the pellet mills falls within the range of
A pelletizer consists of a perforated hard steel die with one 15–40 kWh/ton.33
or two rollers (Fig. 2).29 By rotating the die and rollers, the
feedstock is forced through the perforations to form densified Briquette press
pellets.31 Briquetting is usually performed using hydraulic, mechani-
Pellet presses consist of two types: ring die and flat die. In cal, or roller presses. Unlike pellet mills, briquetting
both machines, the die remains stationary and the rollers machines can handle larger-sized particles and wider mois-
rotate. Some rotating die pellet mills are available in which the ture contents without the addition of binders. Grover and
rollers remain stationary during the production process.32 Mishra33 found that agricultural material briquettes can be
The pellet mill operation starts with incoming biomass formed at 22% moisture content using briquette machines.
flowing into the conditioner for the controlled addition of They also suggest that briquettes offer advantages, such as
steam. The steam softens the feed and partially gelatinizes (i) better feed handling characteristics, (ii) higher calorific
the starch to create more durable pellets. Most mills have value, (iii) improved combustion characteristics, (iv) reduced
one or more conditioning units mounted above the main particulate emissions, and (v) more uniform size and shape.
unit. From the conditioner, the feed is discharged over a per- In addition, briquettes can be used in furnaces where other
manent magnet and into a feed spout leading to the pellet- solid fuels like wood pellets are used.33 A typical schematic
ing die. Inter-elevator fl ights in the die cover feed the mash of a mechanical or hydraulic press is shown in Fig. 3.34
During briquetting, the moisture in the material forms
steam under high pressure, hydrolyzing the hemicellulose
and lignin into lower molecular carbohydrates, lignin prod-
ucts, sugar polymers, and other derivatives.33 These prod-
ucts, when subjected to heat and pressure in the die, act as
adhesives and bind the particles together.33 Further addition
of heat helps in relaxing the biomass fibers and softens the
structure.33 Briquettes produced using a hydraulic press have
uniform shape and size, typically 40 × 40-mm cylinders, and
unit densities in the range of 800–1000 kg/m3.35

Hydraulic piston press

Hydraulic piston presses are commonly used as
briquetting machines. The output is lower compared to
mechanical presses because the movement of the cylinder
is slower. The required pressure in the hydraulic press
is produced by a specially designed hydraulic cylinder
that releases the compressed briquette once the required
Figure 2. Working process of a pellet mill die. pressure is reached. The pressure is adjusted using a

686 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

Figure 3. Mechanical or hydraulic piston press.

regulator to maintain consistency. Briquettes have a unit

density lower than 1000 kg/m³ because of limited pressure.
The typical production capacities of these machines are
in the range of 50–400 kg/hr. However, these machines
can tolerate higher moisture contents than the usually
accepted 15% for mechanical piston presses. 36

Mechanical piston press

The mechanical briquetting press develops a compression
force of approximately 2000 kg/cm2 to obtain high qual-
ity briquettes with high unit densities (>1000 kg/m3) and
without the addition of binders. Mechanical piston presses
are typically used for large-scale production, ranging from
200–2500 kg/hr. Energy loss in the machine is limited, and
the output in relation to power consumption is optimal. The
operating life of a mechanical press is considerably longer
than hydraulic presses. Generally, a mechanical press gives a
better return on investment than a hydraulic press.36

Tabletizer
A tabletizer tightly presses biomass with a hydraulic motor
and ram in a 4 to 6-in. diameter cylindrical mold, reduc-
ing the material from about 10 to 2-in. (smaller than most
biomass briquettes) (Fig. 4). 37 The application of about
20 000 psi in the mold is sufficient to force the material to
Figure 4. Energy tablet-making machine for biomass.
adhere together without adding binders. Long, coarse-cut
feedstocks are favorable in the process, as they stick together
more easily. Tablet densities average 55 lb/ft 3 compared to and for energy density. Suitability of the tableting process
bale at 10 lb/ft 3 and pellets at 45 lb/ft 3. However, the tableting has not been evaluated for power plant or gasification proc-
process uses more energy than pelletization. The tablets have ess feedstocks. Research is ongoing to determine the energy
not been tested extensively for various biomass resources requirements for making tablets and the scale-up process to

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 687
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

be followed in the case of large-scale production and product side. The distance between the two rollers, referred to as the
application in the areas of co-firing and gasification.38 gap, depends on many factors such as the type of biomass,
the particle size, the moisture content, and the addition of
Cuber binders.41 Design parameters that play a major role on the
The cuber die ring and press roller (wheel) are similar to the quality of the densified product are diameter of the rollers,
die ring of a pelleter ( Fig. 5).39 An auger moves the chopped gap width, roller force, and shape of the die. Typical bulk
biomass uniformly toward the openings in the die ring. As densities range from 450 to 550kg/m3.42
the material leaves the auger fl ight, the heavy press wheel
forces the feed through the die openings in the ring. The Screw extruder
pressures in a cuber range from 24 to 34 MPa. The natural Extrusion brings small particles <4 mm close together so
binders in chopped biomass, the high pressure of the press that the forces acting between them become stronger, pro-
wheel, and heat generated by forcing biomass through dies viding more strength to the densified bulk material. During
help bond the cubes. An adjustable deflector around the extrusion, biomass moves from the feed port, with a rotat-
outside of the die ring breaks the cubes in lengths of 50 to ing screw, through the barrel and against a die, resulting in
75 mm.39,40 Cubing operators often find it necessary to add a a significant pressure gradient and friction due to shearing
binder to increase cube durability. Typical binders used are of the biomass.33 The combined effects of wall friction at the
bentonite, hydrated lime, starch, lingo-sulfonates (by-prod- barrel, internal friction in the material, and high rotational
uct from pulp and paper operation), agro colloids, and other speed (~600 rpm) of the screw, increase the temperature of
commercial binders.39,40 the biomass. The heated biomass is further forced through
the extrusion die to form briquettes or pellets. External heat
Roller press using band or tape heaters is provided if the heat generated
Roller presses consist of two rollers of the same diameter, within the system is not sufficient to reach a pseudoplastic
rotating horizontally in opposite directions on parallel axes state for smooth extrusion.33 Figure 7 shows the typical
(Fig. 6).41 Ground biomass, when forced through the gap extruder, with different zones for processing of biomass.43
between the two rollers, is pressed into small pockets, form-
ing the densified product. Because the rotation of the rollers
is in opposite directions, the biomass is drawn in one side
and the densified product is discharged out the opposite

Figure 5. Press wheel and die arrangement in a cuber mill. Figure 6. Roller press mill.

688 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

Figure 7. Extruder for biomass or polymer processing.

Processing of biomass using screw extruder occurs in Table 1. General specification of extrudate
four stages: (i) solids conveyance; (ii) initial compression; produced by the Shimada SPMM 850 extrusion
(iii) final compression; and (iv) discharge.33 During solids press.145
conveyance, ground biomass is partially compressed and Raw material prior to extrusion (hard or soft wood)
packed, and maximum energy is required to overcome par- Moisture content (%) 8
ticle friction. During initial compression, biomass particles Average particle size (mm) 2–6

become relatively soft and lose their elastic nature due to Unit density (kg/m3) 200

high temperature (200–250°C), resulting in the formation Extruded logs

of local bridges and interlocking particles. The biomass Moisture content (%) 4
3
also absorbs energy from friction so that it may be heated Unit density (kg/m ) 1400

and mixed uniformly through its mass. Smaller particle Calorific value (kcal) 4870
Ash content (%) 0.35–0.5
sizes (2–4 mm depending on die diameter) are normally
preferred during extrusion as they lead to better binding of
the materials.33,44 During final compression, biomass enters
the tapered die, where moisture is further evaporated due consists of a rotating chamber fi lled with balls of varying
to temperatures on the order of 280°C, helping to increase sizes and fed with powder and often a binder. The rotation
the compression of the material. Finally, during discharge, of the agglomerator results in centrifugal, gravitational,
the pressure throughout the material normalizes, resulting inertial, and frictional forces. These forces press the
in a uniform extruded log. The high temperatures used dur- smooth rolling balls against the powder, helping them to
ing extrusion result in charring of the material and make stick together and the particle sizes to grow.47 Different
it more suitable for burning or co-firing applications. The types of agglomerators are drum, pan, conical, and plate
physical properties of the cylinder-shaped extruded product shaped.
are given in Table 1. A granulation agglomerator involves the following steps:
(i) fine raw material is continually added to the pan and wet-
Agglomerator ted by a liquid binder; (ii) the disc rotates causing the wetted
Agglomeration is a method of increasing particle size by fines to form small, seed-type particles (nucleation); and (iii)
gluing powder particles together. Th is system is used with the seed particles ‘snowball’ by coalescence into larger parti-
a variety of powders such as hydrated lime, pulverized cles until they discharge from the pan.48
coal, iron ores, fly ash, cement, and others. The application For any particular agglomerator, the main process param-
of agglomeration for biomass is limited.45,46 The most eters are the ball residence time (depending on powder feed
commonly used method is tumbling agglomeration, which rate, acting volume, and pan-tilt angle) and proper rolling

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 689
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

action (depending on scraper position, binder premixing, work performed during densification can be shown for both
and pan-tilt angle). Mort49 suggested that agglomeration is processes using Eqn (1):50
also a function of feedstock variables, such as particle size, x
w = A ∫0 P dx (1)
distribution and shape, porosity, and surface chemistry, as
well as process variables such as fluidization, residence time, where P is the applied pressure, x is the sample thickness,
temperature, and application energy (Fig. 8).49 He also con- and A is the cross-sectional area of the die. In the compres-
cluded that the addition of binders plays a significant role in sion apparatus, the density (D) at each point is calculated
the quality of the agglomerated powders. Typically, agglom- from as follows:50
erated materials are spherical with diameters ranging from
4–6 mm, depending on the residence time of the material in D = m/xA (2)
the agglomerator.
where m is the sample mass.
Winter51 postulates a power law or exponential relation to
Specific energy requirements for describe the specific energy (En) with respect to throughput:
densification
En = aM0b (3)
The specific energy requirements for biomass densification
depend on the system used, process variables like where M 0 is the mass flow rate, and a and b are con-
temperature and pressure, feedstock variables like moisture stants that vary with density and depend on die and feed
content, particle size and distribution, and biochemical characteristics. It was observed that the specific energy
composition like starch, protein, fat and other lignocellulosic requirements decreased from 180–8 kWh/metric ton
components.50 Most densification processes involve both over a throughput range of 0.05–1.3 metric tons per hour
compression and extrusion work. Extrusion requires more (MTPH).
energy than compression because the material has to Winter51 also calculated the specific energy required
overcome the friction during compression and pushing. The for pelletization. The material that passes through the

Figure 8. Agglomeration as a function of material properties and process parameters.

690 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

pellet die has to accomplish the following processes: pre- and bearing losses associated with commercial equipment
compress the loose feed, deform the feed as it enters the (Table 3). Given this fact, laboratory results likely represent
die, and balance the die frictional force as the pellet passes lower specific-energy requirements for densification. The
the die. The frictional force (F) is related to die length and specific-energy consumption for both compression and
diameter by: extrusion can be reduced by a factor of about two by pre-
heating the biomass to 200–225°C before densification.50
F = F0 exp 4μL
____ (4)
D This extra heating prior to densification may require about
where F0 is the initial frictional force, L is the die length; D is 1.8 J/g-C. However, electrical power and equipment costs
the die diameter, and μ is a constant. may be reduced due to lower pressure requirements and
The initial static friction that the pellet must overcome is reduced die wear from improved lubricity. Furthermore, the
greater than that which must be overcome when the pellet fuel value or energy content may increase due to complete
begins to flow. Stopping and starting the flows significantly water removal and pre-pyrolysis, such that pellets made at
increases the specific energy requirements. According to 225°C have an energy content of 20.2 J.50 Table 3 shows the
Reed et al., 50 there are three types of pressure applications in effect of temperature and extrusion rate on refuse-derived
commercial densification processes: (i) compression in a die; fuel (RDF).
(ii) extrusion through a constriction; and (iii) shear of pre- Mewes52 concluded that only 37–40% of the input energy
compacted material to produce heat and flow under pres- is required to compress the material; the remaining energy
sure. Table 2 shows the comparison of energy requirements is required to overcome friction during compression.
for commercial densification systems.50 Mohsenin and Zaske53 observed that increasing moisture
The required compression-specific energy is lower by content reduced the energy required to reach a specific
a factor of 2 to 10 compared to commercial compression density. O’Dogherty and Wheeler, 54 with barley straw in
machines because the measurement does not include motor a circular die, noted an energy requirement of 5–25 MJ/t,
depending on wafer density. Aqa and Bhattacharya55
observed that when densifying preheated (115°C) sawdust,
Table 2. Comparison of reported energy energy inputs to a briquetting machine motor, die heaters,
requirements for commercial densification
and overall system were reduced by 54, 30.6, and 40.2%,
apparatus with laboratory results.50
respectively. The specific energy consumption for different
Material Unit kWh/ kWh/
Density tonne ton
(g/cc)
Table 3. Specific energy consumption during
Compression
extrusion of RDF.50
In laboratorya Sawdust 1.0 4.0 3.6
Energy
Sawdust 1.2 6.6 6.0
Commercialb Sawdust ~1.2 37.4 34.0
kWh/tonne kWh/ton
a
Temperature (°C)
Extrusion
25 7.76 7.06
In laboratoryc Municipal 1.0 7.76 7.06
Solid 93 6.09 5.54
Waste
149 6.23 5.67
Commerciald Municipal 1.0 16.4 14.9
204 4.45 4.05
Solid
b
Waste Extrusion rate (cm/min)
Sawdust 1.0 36.8 33.5 5 7.76 7.06
Note: 10 10.93 9.95
a
2.5-cm pellet
b 20 10.90 9.92
From specifications of 150-hp Hausmann briquettor no. FH
2/90/200 for 8-cm diameter log. Note:
c a
1.2-cm pellet made at 25°C. Extrusion rate 5 cm (2 in.)/min
d b
Data supplied by California Pellet Mill Corp. At 225°C.

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 691
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

Table 4. Specific energy consumption data for and briquetting system variables, including both process
different biomass materials.10 and feedstock variables, plays an important role in achieving
Materials Densification Specific Source the desired density, durability, and quality.63 For example,
unit type energy Shaw64 identified that process variables (die temperature,
consumption pressure, and geometry), feedstock variables (moisture
(kWh/t)
content and particle size/shape), and feedstock composition
Sawdust Pellet mill 36.8 Reed &
Bryant56 (protein, fat, cellulose, hemicelluloses, and lignin) play an
Municipal Pellet mill 16.4 Reed & important role in the quality of the densified biomass.
Solid Waste Bryant56
Bark + Pellet mill 30–45 Miles & Process variables
wood Miles57
Straws + Pellet mill 37–64 Miles & Temperature
binders Miles57 Quality attributes like durability and bulk density of densified
Straws Pellet mill 22–55 Neale58 biomass are significantly influenced by temperature. Hall and
Grass Pellet mill 33–61 Shepperson Hall65 found that for a given moisture content, the pressure
& Marchant59
Switchgrass Pellet mill 74.5 Jannasch
required to obtain a certain wafer density of Bermuda grass
et al.60 and alfalfa was reduced by the addition of heat in the die. In
Alfalfa Pellet mill 30 Tabil and addition, the upper limits of moisture content at which a cer-
Sokhansanj12
tain pressure was able to produce a specific wafer density was
Straws + Cubing 75 Miles and
binders machine Miles57 increased by the addition of heat. Smith et al.,66 in their study
Grass Cubing 28–36 Balk61 of briquetting wheatstraw, found that the degree of compac-
machine tion and dimensional stability went up as the temperature was
Cotton trash Cubing 60 Miles and
increased from 60 to 140°C. They also found that briquette
machine Miles57
Hay Cubing 37 Miles and
expansion decreased when the die temperature was between
machine Miles57 90 and 140°C. They further observed that briquettes were
Sawdust Piston press 37.4 Reed et al.50 surface-charred and slightly discolored at temperatures above
Straws Screw press 150–220 Carre et al.62 110°C due to chemical degradation. Tabil11 found that pel-
Grass Piston press 77 Shepperson leting temperatures >90°C significantly improved durability
& Marchant59
values of alfalfa pellets. They concluded that it is necessary
Straws + Ram extruder 60–95 Miles &
binders Miles57 to precondition the grinds to above 90°C to promote better
bonding of particles and to produce good durable pellets.
biomass material densified using different densification Kaliyan and Morey67 used the glass transition temperature
systems is shown in Table 4. Among these different systems, of the lignin to understand the densification behavior. Their
the screw press consumes the most energy because it studies included three different temperatures: two within the
involves not only compression but other forces like shearing glass transition temperature (75 and 100°C) and one outside
and mixing. The pellet mill consumes the least energy. (150°C). The durability values of the densified biomass outside
The chemical composition of biomass and methods of pre- the glass transition temperature were lower compared to ones
treatment before densification also significantly influence within the range.
the specific energy consumption.
Pressure
Pressure plays an important role on the quality of pellets
Densification system variables
made from agricultural biomass. Yaman et al.68 in their study
Pellet and briquette presses are commonly used systems to of fuel briquettes from olive refuse and paper mill waste sug-
create a uniform feedstock commodity with specific charac- gested that there is an optimum briquetting pressure above
teristics for bioenergy applications. Controlling the pelleting which may result in fractures due to dilation. High pressures

692 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

and temperatures during densification may develop solid cal analysis of rice husks to establish a multiple correlation
bridges by a diff usion of molecules from one particle to equation in the form of:
another at the points of contact, which increases density. Li
Y = a0 + a1P + a2T (5)
and Liu9 observed that compression of oak sawdust at pres-
sure application rates of 0.24 to 5.0 MPa/s had a significant where Y is the percent volume expansion; T (°C) and P (kg/m2)
effect on the dry density of the product. For compaction of are the die temperature and pressure, respectively; and a0, a1
biomass waste materials like waste paper, Demirbas et al.69 and a2 are constants.
observed that increasing the pressure from 300 to 800 MPa
Die geometry and speed
on biomass with ~7% moisture (w.b.) initially increased the
Die geometry refers to the size and shape of the die. These
density sharply, from 0.182 to 0.325 g/ml, and then further
dimensions affect the amount of material that can be pel-
increased it slightly to 0.405 g/ml. Butler and McColly70
leted and the energy required for compression. Die geometry
observed that the density of pellets is directly proportional to
also influences product properties like moisture content,
the natural logarithm of the applied pressure and increasing
bulk density, and durability. The L/D (length to diameter)
the pressure increased the unit density significantly.
ratio of the pellet die can be a good metric for the degree of
Retention or hold time and relaxation time compression during pelletization. An increase in the length
The quality of briquettes is influenced by the reten- of the pellet die increases the pelleting pressure, whereas an
tion or hold times of the materials in the die.13 However, increase in the diameter of the pellet die decreases the pellet-
Al-Widyan et al.71 found that the retention times between ing pressure. Hence, the dimensions of the die and the press
5 and 20 s did not have a significant effect on olive cake channels in the matrix have a strong influence on determin-
briquette durability and stability. Li and Liu9 found that the ing the pressure needed to press pellets through the matrix.74
hold time for oak sawdust had more effect at lower pressures Butler and McColly70 found that for a constant mass of mate-
than at higher pressures. At the highest pressure (138 MPa), rial, pellet density and length were greater for smaller diameter
the effect of holding time was negligible. They also observed chambers at a given pressure. Tabil and Sokhansanj13 studied
that the holding time had little effect on the expansion rate. the effect of process parameters like steam conditioning, die
A 10-s holding time could result in a 5% increase in log geometry, L/D ratio, die speed, and particle sizes of the biomass
density, whereas at holding times longer than 20 s, the effect and found that at higher conditioning temperatures (>95°C)
diminished significantly. the durability of the pellets increased. They also concluded that
In general, relaxation time impacts the density of materi- the durability of the pellets improved when a smaller die with
als. Final relaxed density of briquetted fuel and the relaxa- higher L/D ratios was used. Hill and Pulkinen75 reported that
tion behavior following removal from the die depend on the durability of alfalfa pellets increased by about 30–35% at an
many factors related to die geometry, the magnitude and L/D ratio between 8 and 10. Heffiner and Pfost76 evaluated the
mode of compression, the type and properties of the bio- effect of three die sizes (4.8 × 44.5, 6.4 × 57.2, and 9.5 × 76.2
mass material, and storage conditions.64 Many studies on mm) on durability, finding that pellets produced on the small-
high-pressure compaction of biomass materials indicate that est die had the best durability values. In their study of distillers’
upon removal of the material from the die, the density of dried grains with solubles (DDGS), Tumuluru et al.77 found
the product decreases with time to a final relaxed density.64 that a larger die diameters of 7.2 mm produced less durable
For most feed materials, the rate of expansion is highest DDGS pellets compared to a smaller die diameter of 6.4 mm,
just after the removal of pressure and decreases with time both with and without the addition of steam.
until the particle attains constant volume.57,62 The relaxation
characteristics, which are mainly measured by the percent- Feedstock variables
age of elongation and increase in voidance, depend on many Moisture content
factors related to the feed material and storage conditions, Moisture content plays an important role on pellet forma-
such as relative humidity.72 Shrivastava et al.73 used statisti- tion.78 Moisture in the biomass facilitates starch gelatinization,

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 693
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

protein unfolding, and fiber solubilization processes during Table 5. Optimum particle size distribution
densification. Steam-treated biomass is superior to raw bio- for producing quality pellets from agricultural
mass because the additional heat modifies the physiochemi- biomass.83
cal properties to the extent that binding between particles Sieve size (mm) Material retained on sieve
is enhanced, resulting in improved densification quality.25 3.0 ≤1%

Mani et al. 10,79

observed that moisture in the biomass during 2.0 ≤5%

densification increases the bonding via van der Waal’s forces, 1.0 ≈20%

thereby increasing the contact area of the particle. They also 0.5 ≈30%

found that low moisture biomass (5–10%) resulted in denser, 0.25 ≈24%
<0.25 ≥20%
more stable, and more durable briquettes compared to higher
moisture biomass (15%). Li and Liu9 recommended an opti-
mum moisture content of ~8% to produce high-density bri- In the case of briquette presses, bigger particles sizes (>6
quettes. They also recommended a moisture content of 5–12% mm) are desirable, leading to better interlocking of the par-
to produce good quality logs in terms of density and long- ticles and increasing the durability. Using wheat, oats, barley,
term storage properties from hardwood, softwood, and bark. and canola, Song et al.35 indicated that particle sizes between
Densification at optimum moisture content coupled with 19.05–31.15 mm resulted in good quality briquettes using a
temperature may result in increased lignin melting and hydraulic piston press. They also concluded that larger particle
improve the binding characteristics. Kaliyan and Morey78 sizes during briquetting help in interlocking particles and pro-
suggest that moisture in biomass affects the glass transition duce a more durable briquette.
temperatures during densification. They have found that at
optimum moistures of 10–15% in corn stover, the glass tran- Biomass composition
sition temperature decreased and resulted in better binding Feedstock composition contributes significantly to the
at lower temperatures of 70–90°C. Chirife and Del Pilar80 quality of densified materials. Raw biomass has both low
observed that increase in moisture content significantly molecular weight and macromolecular compositions. Low
decrease the glass transition temperatures of lignin, starch, molecular weight substances include organic and inorganic
and gluten. The effect of biomass moisture content on densifi- matter, while macromolecular substances include cellulose,
cation can be three-fold: (i) lowers the glass transition temper- hemi-cellulose, and lignin.84 Understanding the major com-
ature; (ii) promotes solid bridge formation; and (iii) increases positional changes that take place during biomass processing
the contact area of particles by van der Waal’s forces. can be useful in understanding their compaction behavior.
Thomas et al.85 identified some of the important ingredients
Particle size, shape, and distribution that influence pellet quality, including starch, protein, non-
In general, the density and durability of pellets is inversely starch polysaccharides (NSP), sugar, fat, fiber, inorganic
proportional to the particle size because smaller particles matter, and water. Tables 6 and 7 show the composition of
have greater surface area during densification. MacBain81 some agricultural and woody biomass. Wood is shown to
and Payne82 concluded that medium or fine-ground mate- have higher lignin content than other biomass materials,
rials are desirable in pelleting because they have greater and straws are shown to have a certain percentage of protein
surface area for moisture addition during steam condition- content, both of which can promote binding.
ing, which increases starch gelatinization and promotes
binding. They also reported that a certain percentage of fines Starch
to medium particle sizes improves pelleting efficiency and Starch is a D-glucose polymer with branched (amylopectin)
reduces pelleting cost. However, very small particles can lead or un-branched (amylose) chains.86 Its behavior is
to jamming of pellet mills and affect production capacity. mainly controlled by the gelatinization it undergoes at
Table 5 indicates the particle size distribution for producing high processing temperatures. Starch granules at high
good quality pellets.83 temperatures and moistures influence the binding properties

694 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

Table 6. Chemical composition of selected physical quality of the pellets compared to denatured proteins.
agricultural straws.27 Tabil11 reported an improvement in the binding properties of
Composition Barley Canola Oat Wheat the material if sufficient natural proteins are present during
(% DMa) straw straw straw straw pelletization. Sokhansanj et al.90 identified that feed material,
Protein 3.62 6.53 5.34 2.33 which contain higher proportions of starch and protein, will
Fat 1.91 0.69 1.65 1.59 produce more durable and higher quality pellets than biomass
Starch 0.11 0.34 0.12 2.58 containing only cellulosic material. They also concluded that
Lignin 17.13 14.15 12.85 13.88 the optimum moisture content for pelleting cellulosic materials
Celluloseb 33.25 42.39 37.60 34.20 is 8–12%, whereas for starch and protein materials (mostly ani-
c
Hemicellulose 20.36 16.41 23.34 23.68
mal feeds), the optimum moisture can range up to 20%.
Ash content 2.18 2.10 2.19 2.39
a
DM – dry matter. Lipid/Fat
b
Cellulose percentage is calculated indirectly from percentage ADF
Fat content in biomass acts as a lubricant during pelletiza-
and lignin (%ADF-%lignin).10
c
Hemicellulose percentage is calculated indirectly from percentage tion, increasing throughput, and reducing pelleting pres-
Neutral detergent fiber (NDF) and acid detergent fiber (ADF): (%NDF
-%ADF).10
sure.11 However, higher fat content can hinder binding.
Briggs et al.23 found that increased oil content produced
lower quality pellets since fat is hydrophobic and tends to
Table 7. Biochemical composition of herbaceous
and woody biomass.84,147 interfere with particle binding during pelletization.
Plant material Lignocellulosic content (%) Cellulose
Hemicelluloses Cellulose Lignin Cellulose is an organic, polysaccharide compound (C6H10O5)
Orchard grass 40.0 32.0 4.7 consisting of a linear chain of several hundred to over ten
(medium
maturity) thousand β(1
4) linked D-glucose units.91,92 Cellulose forms
Rice straw 27.2 34.0 14.2 crystalline microfibrils that are surrounded by amorphous cel-
Birch wood 25.7 40.0 15.7 lulose inside plant cells.93 The structural integrity of cellulose
Scots pine 28.5 40 27.7 is produced by hydrogen bonding that occurs between the
Spruce 30.6 39.5 27.5 glucose monomers.94 According to Nelson and Cox95 cellulose
Eucalyptus 19.2 45 31.3 is considered to be an abundant source of carbon in biomass.
Silver birch 32.4 41 22 Semi-crystalline structure and highly hydrogen bonded cellu-
lose itself is not a suitable adhesive, but this limitation can be
of many foods and feeds. Gelatinization of starch is an overcome by heat treatment in the drying range, making the
irreversible process and influenced by densification process cellulose molecule more flexible.96
variables like heat, water, shear, and residence time.87
Hemicellulose
During pelletization, starch not only acts as a binder but also
Hemicellulose is any of several heteropolymers (matrix
as a lubricating agent, helping to ease the flow of materials
polysaccharides), such as arabinoxylans, present along with
through the die. In the pharmaceutical industry, starch is
cellulose in almost all plant cell walls. While cellulose is
widely used as a binder or fi ller in tablet formulations.88
crystalline, strong, and resistant to hydrolysis, hemicellulose
Protein has a random, amorphous structure with little strength. It is
Protein that is heated during the densification process under- easily hydrolyzed using a dilute acid or base as well as many
goes denaturization, leading to the formation of new bonds and hemicellulase enzymes. The amorphous structure of hemi-
structures with other available proteins, lipids, and starches, celluloses – which is easily hydrolyzed or dissolved in alkali
helping to improve the binding capacity.85,89 According to solution – results from branching. Some researchers believe
Briggs et al.23 and Wood,24 increasing the protein content that natural bonding may occur due to the adhesive prod-
increases the pellet durability. Raw protein improves the ucts produced by degradation of hemicellulose.33

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 695
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

Lignin
Lignin is a complex chemical compound most commonly
derived from wood and an integral part of the secondary
cell walls of plants and some algae.97,98 Lignin is a random
network polymer with a variety of linkages based on phenyl
propane units.99 The lignin molecule provides many struc-
tural purposes, such as, acting like glue, to the cellulose fib-
ers. Lignin plays a crucial part in conducting water in plant
stems. The polysaccharide components of plant cell walls
are highly hydrophilic and thus permeable to water, whereas
lignin is more hydrophobic, which helps improve storage
behavior.
Lignin helps in building solid bridges at elevated tempera- Figure 9. Pellet durability versus lignin in pure species.
tures and plays a significant role in biomass densification.
Lignin is the component that permits adhesion in the wood
structure and acts as a rigidifying and bulking agent. It is, in
general, believed that highly lignified wood is more durable
and therefore a good raw material for many applications. It is
also an excellent fuel, because lignin yields more energy when
burned than cellulose. The presence of lignin in plant mate-
rials allows pelletization without adding binders (Table 7).
Van Dam et al.26 reported that lignin exhibits thermosetting
properties at working temperatures of >140°C and acts as an
intrinsic resin, producing more durable pellets.
It is believed that higher lignin levels lead to more durable
pellets because lignin acts as the ‘glue’ that binds particles
together. However, Lehtikangas100,101 reported a loose corre- Figure 10. Pellet durability versus lignin in mixed species.

lation between lignin content and pellet durability. Similarly,

need to be specified as part of the final product. The most
Wilson102 concluded that there is no consistent relationship
commonly used binders in pellet making are lignosulpho-
between lignin content and pellet durability for hard and
nates (Wafolin), or sulfonate salts made from the lignin
soft woods, but that a mixture of woody biomass with higher
in pulp mill liquors.104,105 Lignosulfonates, considered the
lignin content gave less durable pellets compared to pure
most effective binders, are used in animal feeds.63,106 The
samples. Figures 9 and 10 show the effect of different lignin
general quantity to include for effective binding ranges
and moisture contents on pellet durability in pure and mixed
from 1–3%. Bentonite, or colloidal clay, is commonly used
species.102 Bradfield and Levi103 reported that when lignin
as a binder in feed pelleting and is made up of aluminum
plus extractives content increased above a threshold level of
silicate composed of montmorillonite. As mentioned pre-
34% in wood samples, the pellet durability decreased.
viously, proteins are natural binders that are activated
Binders used in biomass densification through interactions with other biomass compositions,
Binders improve the cohesive characteristic of biomass by such as lipids and starches, and the heat produced in the
forming a gel with water, helping produce a more durable dies. Some agricultural biomass, like alfalfa, has a high
product. Binders also help reduce the wear on production protein content and can be used as a binder to improve the
equipment and increase the abrasion resistance of the durability of pellets made from lower lignin content bio-
fuel. In general, binders are allowed in a fuel feedstock but mass materials.

696 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

Pre-treatment of biomass then released, causing the material to expand rapidly.113 This
process produces significant physical, chemical, and struc-
Pre-treatment plays an important role in densification
tural changes in the biomass and makes lignin more avail-
because it prepares lignocellulosic biomass for different den-
able for binding during pelletization.112 In addition, steam
sification systems. Pre-treatment helps reduce the specific
explosion breaks the lignin down into low-molecular weight
energy consumption and produce different high-quality den-
products that retain their basic structure and are moderately
sified products for different end-use applications. In general
reactive. Mosier et al.114 postulated that the compression
pre-treatment improves the quality attributes (higher dura-
and compaction characteristics of biomass can be improved
bility and bulk and energy densities), storage and handling
through steam explosion pre-treatment. According to
characteristics, and transportation logistics. Some promising
Zandersons et al.,99 the activation of lignin and changes in
pre-treatment methods for bioenergy applications include
cellulose structure during steam explosion help form new
(i) grinding, (ii) pre-heating/steam conditioning, (iii) steam
bonds, which in turn create more durable pellets.
explosion, (iv) torrefaction, and (v) AFEX. Integrating pre-
Steam explosion also has benefits in terms of enzymatic
treatment with a densification process can help address many
hydrolysis. Lignin is extensively depolymerized by cleav-
storage, handling, and transportation logistics challenges.107
age of the β-aryl-ether bonds, making it soluble in alkaline
Grinding solutions or certain organic solvents. In addition, hemicel-
Prior to densification, biomass is ground to a certain par- lulose is partially broken down, making it soluble in water
ticle size. This grinding partially breaks down the lignin, and allowing it to condense with lignin, thereby increasing
increases the specific area of the materials, and contributes the lignin content. The major effect of steam explosion is the
to better binding. Peleg,108 Peleg and Mannheim,109 and large increase in the accessibility of cellulose to enzymatic
Mani et al.10 concluded that particle size has a significant hydrolysis.115–118 Kaar et al.119 noted that steam explosion
effect on the binding characteristics and the mechanical requires little or no chemical input, making it more envi-
properties of pellets. Fine powders have advantages because ronmentally friendly than chemical treatment methods.
they have a higher number of contact points, more exposed Thus, steam explosion is a beneficial pre-treatment option
surface area, and greater surface energy per unit of weight because it causes hemicelluloses to become more water solu-
regardless of their physical and chemical characteristics. ble and makes cellulose and lignin more accessible through
depolymerization.120
Pre-heating/steam conditioning
Pre-heating biomass before densification is widely used as it Torrefaction
results in a higher quality product. Most commercial pellet Torrefaction is the slow heating of biomass in an inert
or briquette producers use pre-heating to form more stable environment to a maximum temperature of 300°C.121–123
and dense product.110,111 Aqa and Bhattacharya55 indicate Torrefaction removes most of the smoke-producing com-
that pre-heating could increase the throughput of densifica- pounds and other volatiles, resulting in a final product that
tion and reduce the energy required per kilogram of prod- has approximately 70% of the initial weight and 80–90% of
uct formed. Steam conditioning is a process where steam the original energy content.124,125 The major decomposition
is added to the biomass to make the natural binder, lignin, reactions affect the hemicelluloses, and, to a lesser degree,
more available during densification.112 It is postulated that the lignin and cellulose.126,127
by disrupting lignocellulosic biomass materials via steam Torrefaction helps to develop a uniform feedstock and
conditioning will improve the compression characteristics of improves binding during pelletization by increasing the
the biomass. number of available lignin sites, breaking down the hemi-
cellulose matrix, and forming fatty unsaturated structures,
Steam explosion resulting in bulk densities of 750–850 kg/m3 and energy
Steam explosion is a technique where high pressure steam densities exceeding 20GJ/m3.128,129 Bergman130 indicated that
is introduced into a reactor for a short period of time and torrefaction results in weakened biomass polymers (i.e., less

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 697
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

fibrous and more plastic) and catalyzes chemical modifica- Unit and bulk density (kg/m3)
tions that lead to more fatty structures, which act as binding Unit density and bulk density are important parameters for
agents during densification. In addition, the lignin content storage and transportation. Several researchers have found
increases typically 10–15% as the devolatilization process dur- that these parameters are greatly influenced by the material’s
ing torrefaction leads to degradation of hemicellulose. Studies moisture content and particle size, and the process pressure
of densification of torrefied biomass at 250°C indicated that and temperature.10,134 Generally they found that materials
the pressure and energy required for densification can be with higher moisture and larger particle sizes reduce the unit
reduced by a factor of two and the throughput increases by and bulk density of the product, while higher process temper-
two times compared to raw biomass densification using a pel- atures and pressures increase the unit and bulk density. Rhen
let mill.56,128,130 These researchers also indicated that heating et al.134 also found that high dry unit density corresponds to
the material to temperatures >250°C during densification is high compression strength. Tumuluru et al.135 in their article
not recommended as it leads to heavy devolatilization. on pelleting DDGS, supported the conclusions that both unit
and bulk density is dependent on feed moisture and die tem-
Ammonia fiber explosion (AFEX) perature, where a maximum unit density of 1200 kg/m3 and
AFEX pre-treatment of the biomass (ammonia fiber explo- bulk density of 700 kg/m3 is achievable at temperatures of
sion) uses aqueous ammonia at elevated temperatures and about 100°C and feed moistures of about 5–7%.
pressures131 to produce higher hydrolysis yields for many
herbaceous feedstocks. Th is process reduces lignin and Durability index (%)
removes some hemicellulose while decrystallizing cellulose The durability index is a quality parameter defi ned as the
in the biomass. The major advantage of this process is little ability of densified materials to remain intact when handled
biomass degradation.132 The process offers other advantages during storage and transportation. Thus, pellet durability
like elimination of a separate liquid phase and the possibility is its physical strength and resistance to being broken
of very high solids loading. The resulting dark black product up. Durability or abrasive-resistance measurements help
may offer improved densification characteristics because simulate either mechanical or pneumatic handling forces
it opens the cellulosic structure and makes more lignin to help or control feed quality. Different types of equipment
sites available for binding. Eranki et al.107 in their study on (Holmen tester, tumbling can, Ligno tester and Dural tester)
advanced biomass processing depots evaluated densifying are used to test durability.78
AFEX products to solve storage and transportation logistics. Moisture increase durability when water soluble com-
pounds, such as sugar, starch, soda ash, sodium phosphate,
potassium salt, and calcium chloride, are present in the
Physical attributes of densified biomass feed.78 High starch content acts as a binder and increases
Moisture content (%) durability. However, native starch has less binding capacity
The optimum final moisture content of densified biomass is than gelatinized starch, where moisture and heat accelerate
very important and greatly depend on process conditions like the process.26,85,136,137 Protein will plasticize with heat and
initial moisture content, temperature, and pressure. Higher moisture and act as a binder, increasing the durability of the
moisture in the final product results when the initial moisture products.23,138 Furthermore, high fat content will result in
is greater than 15%. Mani et al.10 observe that initial moisture low durability, as fat acts as a lubricant between the feed par-
>15% and pressure >15 MPa has a negative effect on the final ticles and die wall.23,139–142
briquette quality where cracks occur. Lower moisture in the Lignin, at elevated temperatures (140°C), acts as a binder
pellets (<5%) can result in revenue loss as pellets tend to break and increases durability. However, Bradfield and Levi103
up, creating more fines during storage and transportation. observe that when the lignin content and other extractives
Pellets with high moisture content can be subject to spoilage increase to more than 35%, the durability values decrease.
due to microbial decomposition, resulting in significant dry They postulate that the auto-adhesive nature of lignin and
matter loss during storage and transportation.133 other extractives decreases at higher concentrations due to

698 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

their excessive mastic nature. Pre-heating or steam condi- Table 8. Standards recommended for measuring
tioning increases the activity of inherent binders like lignin densified biomass quality.
and starch, thus producing more durable pellets. Pre-heating S. No Pellet Common Pellet Fuel Institute
temperatures are usually restricted to 300°C to limit the Quality European (PFI)b
decomposition of the biomass.78 Steam conditioning also Standard
(CEN)a
helps to release and activate natural binders and lubricants
1 Moisture CEN/TS ASTM E 871 Standard
in the feed, thus increasing starch gelatinization, protein content 15414-1:2010 Test Method for Moisture
denaturation and pellet durability.78 Finally, particle size and Analysis of Particulate
Wood Fuels
process variables, such as die dimensions, L/D ratios, and
2 Bulk CEN/TS ASTM E 873 Standard
rotational speeds, also influence durability values.19,78,79 density 15401:2010 Test Method for Bulk
Density of Densified
Particulate Biomass
Percent fines (%)
Fuels.
The presence of fines in the densified product is not desirable, 3 Durability CEN/TS Kansas State University
especially when co-firing with other fossils fuels. Fines are 15639:2010 - Mechanical Durability
of Feed Pellets, Call
generated during transportation and storage by the break- Number: LD2668.T4 1962
down of the densified product. Pellets processed under subop- Y68) for assessing the
durability of residential/
timal conditions, such as lower moisture lower temperatures, commercial densified
and with less desirable chemical compositions or with insuffi- fuel products, with the
exception that the screen
cient die size and roller speeds, are less durable and can result size used in determining
durability has been modi-
in more fines in the final product. Once the percent fines cross
fied to be a 1/8-inch (3.17
the storage threshold value in silos, spontaneous combustion mm) wire screen sieve.
and dust explosion problems can occur. Tumuluru et al.77 in 4 Percent 3.15 mm 1/8-inch (3.17 mm) wire
fines screen screen sieve
their studies on the effect of storage temperature on the qual-
5 Calorific EN ASTM E 711 Standard
ity of wood pellets found that higher storage temperatures value 15400:2011 Test Method for Gross
(30–50°C) increase the percent fines by more than 1%. Calorific Value of Refuse-
Derived Fuel by the Bomb
Calorimeter
Calorific value (MJ/kg)
Note: Unit density is not a standard followed by PFI and CEN,
In general, the caloric value of pellets and briquettes but the American Society of Agricultural and Biological Engineers
(ASABE) has a standard procedure (ASAE S269.4) for measuring
depends on process conditions like temperature, particle
unit density of pellets and briquettes.
size, and feed pre-treatment. Generally, pellets with higher Source:
a
density have higher caloric value. The typical calorific val- CEN/TC 343 - Published standards;
b
PFI Standard Specification for Residential/Commercial Densified
ues of wood- and straw-based pellets range from 17–18 MJ/ Fuel, October 25, 2010.
kg.77,143 Many researchers have observed that pre-treatment
processes like steam explosion or torrefaction prior to densi- standards discuss specifications for densified fuel for residen-
fication increases the calorific value to 20–22 MJ/kg.113,144,145 tial and commercial applications, but do not specify whether
Table 8 indicates the methods recommended by European it is in a pellet, briquettes, or densified log form. On the other
Common Standard for Solid Fuel (CEN) and Pellet Fuel hand, the European Committee for Standardization146 pre-
Institute (PFI in North America) for measurement of the pared testing methods and technical specifications for solid
physical properties discussed. biofuels specifically for pellets and briquettes. Tumuluru et
al.133 have reviewed the existing PFI and CEN standards for
International standards for densified solid pellets and briquettes in their article on biomass densification
fuels technologies for bioenergy applications. Further standards
The standards for densified biomass application as a solid need to be developed for densified biomass produced using
fuel in USA are given by the PFI and in Europe by CEN. PFI other densification systems, which can include (i) a cuber,

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 699
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

(ii) an agglomerator, (iii) a tablet press, and (iv) a screw press, The specific energy consumption of different densification
which can help develop a consistent uniform feedstock com- systems varies depending on the different unit operations
modity product for energy applications. involved, like compression, pushing, shearing, and mixing.
The systems that involve more compression and pushing con-
Discussion sume more energy because they depend on the dimensions
of the pressing channel. About 40% of the energy is required
A comparison of various densification systems in terms of
for compressing the material and the remaining energy is
feedstock properties, specific energy consumption, process-
required for overcoming friction during compression. Among
ing additives, and suitability of the densified material for
the different densification systems, a screw press consumes the
different end-use applications is shown in Table 9. All densi- most energy because it not only compresses but also shears
fication systems reviewed in this study help in developing an and mixes the material, whereas a pellet mill or cuber con-
advanced uniform feedstock with bulk flow characteristics sume the least, depending on the material processed.
for bioenergy applications. Even though the unit densi- Densification process variables like temperature, residence
ties of pellets, briquettes, and cubes are similar (1.0–1.2 kg/ time and application pressure play a vital role on the binding
m3), pellets have higher bulk densities and offer advantages behavior. Higher temperatures of >200°C during densifica-
in terms of storage, handling, and transportation logistics. tion can lead to charring of the densified biomass, rendering
On the other hand, briquette presses can handle bigger it unsuitable for some conversion processes. Knowing the
particle size and higher moisture contents, giving them an end use of the material will help determine the appropriate
advantage over other densification systems like pellet mills, temperature the biomass is exposed to. Another important
screw presses, and agglomerators. Table 9 shows that screw- variable that influences the quality of the densified biomass is
pressed material is more suitable for co-fi ring and combus- retention time, where higher holding times of 5–20 s improve
tion because the biomass is carbonized during densification, the density of the pellet or briquettes. Of course, a balance
whereas the pellet, roller, and piston-pressed materials are between holding times and machine capacity will have to be
more suitable for biochemical and thermochemical conver- determined. Finally, higher pressures lead to higher product
sion processes. densities and are proportional to the natural logarithm of the
The use of roller presses, tabletizers, and agglomerators applied pressure. However, higher pressures often require
for energy applications is still in the early stages of research, higher operating costs and higher machine wear requiring a
and more detailed studies in terms of process and feedstock trade-off in cost and product density.
variables are needed to understand the suitability of these Among the feedstock variables, moisture content and
systems. Data on energy consumption is also not readily particle size have the greatest influence on the densifica-
available. Literature from other industries provides some tion process. Moisture can lower the glass transition tem-
indication of the promise of these systems. For example, perature, promote solid bridge formation, and increase the
though there is no specific information on the properties contact area of particles. Lowering the glass transition tem-
of agglomerates made from biomass, their suitability men- perature of the biomass by managing the moisture content
tioned in Table 9 takes into account physical properties like is a good way to densify biomass with less recalcitrance for
size, shape, and bulk density determined for the pharmaceu- the conversion process. In the case of particle size, differ-
tical industry. Finally, binders will play an important role ent systems support different particle sizes. A pellet mill
in some, if not all, of the densification systems, particularly requires smaller particles because binding depends on the
agglomerators and roller presses, which some studies indi- contact area between the particles, and briquette presses
cate will require the addition of a binder to improve the require larger particle sizes because the material bonds by
durability and bulk density. The extent of the role of binders, interlocking. Thus, managing the material properties to suit
however, needs further evaluation. Thus, thorough research the densification equipment will be crucial for getting the
on the densified material properties and the effect of process right quality of feedstock product and managing the cost of
variables on these densification systems is needed. the system.

700 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

Table 9. Comparison of different densification equipments.2,5,19,33,38,40,42,44,78,148

Pellet mill Piston Roller press Cuber Tabletizer Screw Agglomorator
press press
Optimum 10–15 10–15 10–15 15–25 10–15 4–8 No information
moisture con-
tent of the raw
material (%)
Particle size <3 6–12 <4 12–16 <20 2–6 0.05–0.25
requirements
(mm)
Addition of Not required Not required Required Required Not required Not required Required
binder
Shape Cylindrical Cylindrical Generally Elliptical Cylindrical Cylindrical Cylindrical Spherical
(depends upon the
shape of the die)
Dimensions 4.8–19.1 (dia); 32 (dia) x 25 Almond shaped 33 x 33 cross 100–150 (dia) Length: 1940 2–6 (dia)
(mm) 11–19 (length) (thick) briquettes dimen- section and 50 (length) Width: 750
12.7 to 25.4 sions: 31.75 25.4 to 101 Height: 1310
(length) (length) × 20.32 (length) (Smaller dies
(width) × 11.16 produces
(depth) (depends smaller
upon the shape of extruded
the die) logs)
Wear of con- High Low High Low Low High Low
tact parts
Output from Continuous In strokes Continuous Continuous In strokes Continuous Continuous
machine
Specific 16.4–74.5 37.4–77 29.91–83.1 28–75 High energy 36.8–150 No information
energy con- requirements
sumption (Still under
(kWh/ton) research)
Through puts 5 2.5 5–10 5 0.5–1 0.5–1 No information
(ton/hr)
Unit density 1.1–1.2 <0.1 No information 0.8 1.2 1–1.4 No information
(g/cm3)
Bulk density 0.65–0.75 0.4–0.5 0.48–0.53 0.45–0.55 0.6–0.7 0.5–0.6 0.4–0.5
(g/cm3)
Maintenance Low High Low Low Low Low Low
Combustion Very good Moderate Moderate No No Very good No information
performance information information
of briquettes
Carbonization Not possible Not possible Not possible Not possible Not possible Makes good Not possible
of charcoal charcoal
Suitability in Suitable Suitable Suitable Suitable Suitable Suitable Suitable
gasifiers
Suitability for Suitable Suitable Suitable Suitable Suitable Suitable Suitable
cofiring
Suitability for Suitable Suitable Suitable Suitable Suitable Not suitable No information
biochemical
conversion
Homogeneity Homogeneous Not Not homogeneous Not Not Homogenous Homogenous
of densified homogenous homogeneous homogeneous
biomass

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 701
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

Pre-treatment of biomass before densification can play an Even though densification of biomass has been in prac-
important role in densification as well and can improve the tice for a long time, there are still research gaps that need
binding characteristics of biomass that is low in lignin con- to be addressed in order to understand the interaction
tent. Some of the commonly used pre-treatment processes are of feedstock, process variables, and pre-treatment meth-
grinding, pre-heating, steam explosion, steam conditioning, ods on the quality of the densified biomass. The follow-
torrefaction and AFEX. Pre-treating biomass prior to densi- ing research areas can help to advance the science of
fication improves properties like durability, bulk and energy densification:
density, and calorific value and reduces the specific energy
1. Explore the mechanisms of densification using SEM and
consumption. Other promising methods of improving the
TEM techniques.
binding characteristics include addition of natural or synthetic
2. Study the agglomeration technique by modifying mate-
binders. Lignocellulosic biomass, which does not bind easily,
rial properties.
can be improved by adding either natural or commercial bind-
3. Evaluate the effect of torrefaction, steam explo-
ers like protein or lignosulfonates. Note that pre-treatment
sion, and AFEX pre-treatment methods on material
methods are generally required to produce a quality product.
properties, densification behavior, and specific energy
International organizations like CEN and PFI have estab-
consumption.
lished standards to evaluate the quality of densified products
4. Study glass transition temperatures of both raw and pre-
like pellet and briquettes for solid fuel applications, but
treated biomass in relation to densification processes.
standards still need to be developed for the other systems,
5. Examine process variable effects on quality attributes
such as the cuber, screw press, table press, and roller press in
and specific energy consumption from tablet press, roller
relation to energy applications.
press, and agglomerator systems.
6. Develop new standards for densified biomass produced
Conclusion using cuber, tabletizer, roller press, and agglomerator
Among the technologies discussed in this review, the pellet systems.
mill and briquette press are most commonly used for biomass
densification and make the product suitable for biochemi- Acknowledgements
cal, thermochemical, and co-firing applications. In terms of
The authors acknowledge Leslie Park Ovard for her valu-
energy consumption, a screw press consumes the most and
able contribution to the manuscript and Lisa Plaster, Huan-
pellet mill the least. Densification system variables, which
chen Tang, Gordon Holt, and Allen Haroldsen from INL’s
include process variables (temperature, residence time and
R&D Publication Support Service for editorial and graphics
application pressure), feedstock variables (moisture content
support.
and particle size), and biomass composition (protein and
This work is supported by the US Department of Energy,
lignin), have the greatest influence on binding characteristics.
under DoE Idaho Operations Office Contract DE-AC07-
The densification behavior of low-lignin-content biomass
05ID14517. Accordingly, the US government retains and the
material can be improved by pre-treating using steam condi-
publisher, by accepting the article for publication, acknowl-
tioning, steam explosion, torrrefaction, or AFEX. Addition of
edges that the US Government retains a nonexclusive, paid-
either natural or commercial binders is also a good alternative
up, irrevocable, world-wide license to publish or reproduce
to improve the binding characteristics of low-lignin-content
the published form of this manuscript, or allow others to do
biomass while reducing the specific energy during densifica-
so, for US government purposes.
tion. The important quality attributes of the densified biomass
includes durability index, bulk density, moisture content,
US Department of Energy Disclaimer
percent fine, and calorific value, which are evaluated using the
existing international standards developed by PFI in the USA This information was prepared as an account of work
and CEN in Europe. sponsored by an agency of the US government. Neither the

702 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

US government nor any agency thereof, nor any of their 12. Tabil LG and Sokhansanj S, Compression and compaction behavior of
alfalfa grinds, Part 1: Compression behavior. Powder Handling Process
employees, makes any warranty, express or implied, or
8(1):17–23 (1996).
assumes any legal liability or responsibility for the accuracy, 13. Tabil LG and Sokhansanj S, Compression and compaction behavior of
completeness, or usefulness of any information, appara- alfalfa grinds, Part 2: Compaction behavior. Powder Handling Process
tus, product, or process disclosed, or represents that is use 8(2):117–122 (1996).
14. Rumpf H, The strength of granules and agglomerates, in Agglomeration,
would not infringe privately owned rights. References herein
ed by Knepper WA. Interscience Publishers (1962), New York, USA, pp.
to any specific commercial product, process, or service by 379–418.
trade name, trademark, manufacturer, or otherwise, does 15. Sastry KVS and Fuerstenau DW, Mechanisms of agglomerate growth in
not necessarily constitute or imply its endorsement, recom- green pelletization. Powder Technol 7:97–105 (1973).
16. York P and Pilpel N, The tensile strength and compression behavior
mendation, or favoring by the US government or any agency
of lactose: Four fatty acids and their mixture in relation to tableting. J
thereof. The views and opinions of authors expressed herein
Pharm Pharmacol 25:1–11 (1973).
do not necessarily state or reflect those of the US govern- 17. Pietsch WB, Size enlargement methods and equipments, part 2:
ment or any agency thereof. Agglomerate bonding and strength, in Handbook of Powder Science
and Technology, ed by Fayed ME and Otten L. Van Nostrand Reinhold
References Co., New York, USA, pp. 231–252 (1984).
1. Bapat DW, Kulkarni SV and Bhandarkar VP, Design and operating expe- 18. Schineberger GL, Understanding Adhesives. Hitchcock Publishing Co,
rience on fluidized bed boiler burning biomass fuels with high alkali ash, Wheaton, IL, USA, pp. 3.6–3.7 (1971).
in Proceedings of the 14th International Conference on Fluidized Bed 19. Mani S, Tabil LG and Sokhansanj S, Compaction Characteristics of
Combustion, ed by Preto FDS. ASME Publishers, New York, USA, pp. Some Biomass Grinds. AIC 2002 Meeting, CSAE/SCGR Program,
165–174 (1997). Saskatoon, Saskatchewan, Canada, July 14–17 (2002).
2. Sokhansanj S and Fenton J, Cost benefit of biomass supply and 20. Comoglu T, An overview of compaction equations. J Fac Pharm Ankara
preprocessing: BIOCAP (Canada) research integration program syn- 36(2):123–133 (2007).
thesis paper. [Online]. Available at: http://www.biocap.ca/rif/reprt/ 21. Denny PJ, Compaction equations: A comparison of the Heckel and
Sokhansanj_S.pdf [June 22, 2011]. Kawakita equations. Powder Technol 127:162–172 (2002).
3. Mitchell P, Kiel J, Livingston B and Dupont-Roc G, Torrefied biomass: 22. Muñoz G and Herrera P, Multidimensional Modeling of Agricultural
A foresighting study into the business case for pellets from torrefied Fibrous Materials in Densification: Compression Stage. ASAE Paper
biomass as a new solid fuel, in All Energy 2007. University of Aberdeen, No. 023151. American Society of Agricultural and Biological Engineers
ECN, Doosan Babcock, and ITI Energy. (ASABE), St Joseph, MI, USA (2002).
4. Van Loo S and Koppejan J, Handbook of Biomass Combustion and 23. Briggs JL, Maier DE, Watkins BA and Behnke KC, Effects of ingredients and
Co-Firing. Earthscan Publication Ltd, London, UK (2008). processing parameters on pellet quality. Poult Sci 78:1464–1471 (1999).
5. Tumuluru JS, Wright CT, Kenney KL and Hess JR, A review on biomass 24. Wood JF, The functional properties of feed raw materials and the effect
densification technologies for energy applications. [Online]. Tech. Report on the production and quality of feed pellets. Anim Feed Sci Technol
INL/EXT-10-18420, Idaho National Laboratory, Idaho Falls, Idaho, USA 18:1–17 (1987).
(2010). Available at: http://www.inl.gov/bioenergy [June 22, 2011]. 25. Thomas M, van Zuilichem DJ and van der Poel AFB, Quality of pelleted
6. Ndiema CKW, Manga PN and Ruttoh CR, Influence of die pressure animal feed, part 2: Contribution of processes and its conditions. Anim
on relaxation characteristics of briquetted biomass. Energy Convers Feed Sci Technol 64:173–192 (1997).
Manage 43:2157–2161 (2002). 26. van Dam JEG, van den Oever MJA, Teunissen W, Keijsers ERP and
7. Adapa PK, Bucko J, Tabil LG, Schoenau G and Sokhansanj S, Pelleting Peralta AG, Process for production of high density/high performance
Characteristics of Fractionated Suncure and Dehydrated Alfalfa binderless boards from whole coconut husk, part 1: Lignin as intrinsic
Grinds. ASAE/CSAE North-Central Intersectional Meeting, Saskatoon, thermosetting binder resin. Ind Crops Prod 19(3):207–216 (2004).
Saskatchewan, Canada, September 27–28 (2002). 27. Adapa PK, Tabil LG and Schoenau G, Compression characteristics

8. Adapa PK, Schoenau GJ, Tabil LG, Sokhansanj S and Crerar B, Pelleting of selected ground agricultural biomass. Ag Eng Int: CIGR Ejournal,

of Fractionated Alfalfa Products. ASAE Annual International Meeting, Manuscript1347, 11 (2009).

Las Vegas, Nevada, USA, July 27–30 (2003). 28. Kaliyan N and Morey RV, Natural binders and solid bridge type bind-

9. Li Y and Liu H, High pressure densification of wood residues to form an ing mechanisms in briquettes and pellets made from corn stover and

upgraded fuel. Biomass Bioenerg 19:177–186 (2000). switchgrass. Bioresource Technol 101:1082–1090 (2010).

10. Mani S, Tabil LG and Sokhansanj S, Specific energy requirement for 29. Feed Machinery, Pellet Mill description. [Online]. Available at: http://

compacting corn stover. Bioresour Technol 97:1420–1426 (2006). www.feedmachinery.com/glossary/pellet_mill.php [June 22, 2011].

11. Tabil LG, Binding and pelleting characteristics of alfalfa. PhD disserta- 30. Finney KN, Sharifi VN and Swithenbank J, Fuel pelletization with a

tion, Department of Agricultural and Bioresource Engineering, University binder, part 1: Identification of a suitable binder for spent mushroom

of Saskatchewan, Canada (1996). compost-coal tailing pellets. Energ Fuel 23:3195–3202 (2009).

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 703
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

31. Leaver R. The Pelleting Process. Andritz Publications, USA (1970). 49. Mort PR, Scale-up and control of binder agglomeration process-flow
32. Samuel MA, The chemistry and technology of cereals as food and feed, and stress fields. Powder Technol 189:313–317 (2009).
2nd ed. Springer, New York, USA (1991). 50. Reed TB, Trezek G and Diaz L, Biomass densification energy require-
33. Grover PD and Mishra SK, Biomass briquetting: Technology and prac- ments, in Thermal Conversion of Solid Wastes and Biomass. American
tices, in Regional Wood Energy Development Program in Asia, Tech. Chemical Society, Washington DC, USA, pp. 169–177 (1980).
Report GCP/RAS/154/NET. Food and Agricultural Organization of the 51. Winter E, Fundamental considerations for preparing densified refuse-
United Nations, Bangkok, Thailand (1996). derived fuel. U.S. Environmental Protection Agency Tech. Report EPA-
34. Hassan A, Biomass Briquetting in Sudan: A feasibility study. [Online]. 600/S2-81-180, Mun Env Res Lab, Cincinnati OH (1981).
United States Agency for International Development (USAID) (2010). 52. Mewes E, Berechung der druckverteilung an stroh-und heupressen
Available at: http://www.fuelnetwork.org/index.php?option=com_ (Calculation of the pressure distribution in straw and hay balers).
docman&task=doc_download&gid=297 [June 22, 2011]. Landtechnische Forschung 9(6):160–170 (1959).
35. Song Y, Tumuluru JS, Iroba KL, Tabil LG, Xin M and Meda V, Material 53. Mohsenin N and Zaske J, Stress relaxation and energy requirements in
and operating variables affecting the physical quality of biomass bri- compaction of unconsolidated materials. J Agric Eng Res 11:193–205
quettes, in Proceedings of the XVIIthWorld Congress of the International (1975).
Commission of Agricultural and Biosystems Engineering (CIGR). 54. O’Dogherty MJ and Wheeler JA, Compression of straw to high densities
Canadian Society for Bioengineering (CSBE/SCGAB), Québec City, in closed cylindrical dies. J Agric Eng Res 29:61–72 (1984).
Canada, June 13–17 (2010). 55. Aqa S and Bhattacharya SC, Densification of preheated sawdust for
36. C.F. Nielsen A/S, Briquetting presses and total solutions. [Online]. energy conservation. Energy 17(6):575–578 (1992).
Available at: http://www.cfnielsen.com/ [June 22, 2011]. 56. Reed TB and Bryant B, Densified biomass: A new form of solid fuel.
37. Vanengelenhoven J, Energy tablet-making machine for biomass. Solar Energy Research Institute Report #SERI-35, Golden, CO (1978).
[Online]. US Department of Energy. Available at: http://www.columbi- 57. Miles TR and Miles Jr., TR, Densification systems for agricultural
amissourian.com/multimedia/graphic/2010/07/05/tabletizer-compacts- residues, in Thermal Conversion of Solid Wastes and Biomass,
biomass-fuel/ [June 22, 2011]. American Chemical Society, Washington, D. C., U. S. A., 179–191
38. Gibson L, Energy tablets. [Online]. Available at: http://www.biomass- (1980).
magazine.com/articles/5086/energy-tablets [June 22, 2011]. 58. Neale MA, Straw compaction research. Agricultural Engineer, 4(4):
39. Sokhansanj S and Turhollow AF, Biomass densification: Cubing opera- 126–130 (1986).
tions and costs for corn stover. App Eng Agr 20(4):495−499 (2004). 59. Shepperson G and Marchant WTB, Production of grass and alfalfa
40. Sokhansanj S and Turhollow AF, Biomass densification: Operations cubes using an experimental ring die press, in Proceedings of
and costs, in Proceedings of Bioenergy 2002 – Bioenergy for the the 2nd International Green Crop Drying Conference, Saskatoon,
Environment, Boise, ID, USA, Sept 22–26 (2002). Saskatchewan, Canada, 264–270 (1978).
41. Yehia KA, Estimation of roll press design parameters based on the assess- 60. Jannasch R, Quan Y, and Samson R, A process and energy analysis of
ment of a particular nip region. Powder Technol 177:148–153 (2007). pelletizing switchgrass: Final report [Online]. Available at: http://www.
42. Kaliyan N, Morey RV, White MD and Doering A, Roll press briquetting reap-canada.com/online_library/feedstock_biomass/11%20A%20
and pelleting of corn stover and switchgrass. Transactions of the ASABE Process.pdf2001 (2001) [August 15, 2011].
52(2):543–555 (2009). 61. Balk WA, Energy requirements for dehydrating and pelleting coastal
43. Brewer J, Scientific principles: Materials science and technology Bermuda grass, Trans of the ASAE, 7: 349–351, 355 (1964).
teacher’s workshop. [Online]. Available at: http://matse1.mse.uiuc.edu/ 62. Carre J, Hebert L, Lacrosse L and Schenkel Y, Briquetting agricultural
polymers/prin.html [June 22, 2011]. and wood residues: Experience gained with a heated die cylindri-
44. Olsen J, NAEMI Biomass & Business Training Workshop, Spokane, WA, cal screw press, in Handling and Processing of Biomass for Energy:
USA, May15–19 (2006). Report and Proceedings, ed by Keller P, CNRE Bulletin (FAO), no. 18,
45. Beaudequin F, Mignard B, Boillot M, Girard P and Floret J-F, (1987) Rome, Italy, pp. 45–52. European Cooperative Networks on Rural
Agglomeration of charcoal, in Bioenergy 84, ed by Egneus H. et al. Energy: First Workshop on Handling and Processing of Biomass for
Elsevier Science, Barking, UK (1985). Energy, Hamburg, Germany, Sept. 14–15, 1987.
46. Reynieix M, Plancon M and Mohamed FA, Pelletization of vegeta- 63. MacMahon MJ, Additives for physical quality of animal feed, in
ble charcoal smalls, in Biomass for Energy and Industry: 4th E.C. Manufacturing of Animal Feed, ed by Beaven DA. Turret-Wheatland Ltd,
Conference, ed by Grassi G. Elsevier Applied Science, New York, USA Herts, England, pp. 69–70 (1984).
(1987). 64. Shaw M, Feedstock and process variables influencing biomass den-
47. Siemons RV, Ahmed Dr. and Hood H, Cotton stalk charcoal agglomera- sification. PhD dissertation submitted to Department of Agricultural
tion in the Sudan, Boiling Point, 198900. [Online]. Available at: http:// and Bioresource Engineering, University of Saskatchewan, Saskatoon,
www.hedon.info/View+Article&itemId=10900 [June 22, 2011]. Saskatchewan, Canada (2008).
48. Snow RH, Allen T, Ennis BJ and Litster JD, Size reduction and size 65. Hall GE and Hall CW, Heated-die wafer formation of alfalfa and bermud-
enlargement, in Perry’s Chemical Engineers Handbook, 7th edition, ed agrass. Trans ASAE 11:578–581 (1968).
by Perry RH, Green DW and Maloney JO, McGraw-Hill (1997), New York, 66. Smith IE, Probert SD, Stokes RE and Hansford RJ, The briquetting of
U. S. A., pp. 20 -189. wheat straw. J Agric Eng Res 22:105–111 (1977).

704 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

67. Kaliyan N and Morey R, Densification characteristics of corn stover and 87. Thomas M, Huijnen PTHJ, van Vliet T, van Zuilichem DJ and van der
switchgrass, ASABE Paper No. 066174, ASABE Annual International Poel AFB, Effects of process conditions during expander processing
Meeting, Portland, OR. ASAE, St Joseph, MI, USA, July 9–12, (2006). and pelleting on starch modification and pellet quality of tapioca. J Sci
68. Yaman S‚ahan M, Haykiri-açma H, S‚es‚en K and Küçükbayrak S, Food Agric 79:1481–1494 (1999).
Production of fuel briquettes from olive refuse and paper mill waste. 88. Alebiowu G and Itiola OA, Compression characteristics of native and
Fuel Process Technol 68:23–31 (2000). pregelatinized forms of sorghum, plantain, and corn starches and the
69. Demirbas A, S‚ahin-Demirbas‚ A and Hilal-Demirbas‚ A, Briquetting prop- mechanical properties of their tablets. Drug Dev Ind Pharm 28(6):663–
erties of biomass waste materials. Energ Source 26:83–91 (2004). 672 (2002).
70. Butler JL and McColly HF, Factors affecting the pelleting of hay. Agric 89. Nyanzi FA and Maga JA, Effect of processing temperature on detergent-
Eng 40:442–446 (1959). solubilized protein in extrusion-cooked cornstarch/soy protein subunit
71. Al-Widyan MI, Al-Jalil HF, Abu-Zreig MM and Abu-Handeh NH, Physical blends. J Agric Food Chem 40:131–133 (1992).
durability and stability of olive cake briquettes. Can Biosyst Eng 44: 90. Sokhansanj S, Mani S, Bi X, Zaini P and Tabil LG, Binderless pelletization
341–345 (2002). of biomass, ASAE Paper No. 056061, ASAE Annual International Meeting,
72. Wamukonya L and Jenkins B, Durability and relaxation of sawdust and Tampa, FL, USA, July 17–20, 2005. ASAE, St Joseph, MI, USA (2005).
wheat-straw briquettes as possible fuels for Kenya. Biomass Bioenerg 91. Crawford RL, Lignin Biodegradation and Transformation. John Wiley
8(3):175–179 (1995). and Sons Inc., New York, USA (1981).
73. Shrivastava M, et al., Briquetting of rice husks under hot compression, 92. Updegraff DM, Semimicro determination of cellulose in biological mate-
in Proceedings of the International Agricultural Engineering Conference rials. Anal Biochem 32(3):420–424 (1969).
and Exhibition. Bangkok, Thailand, pp. 666–672 (1990). 93. Chen W, Lickfield GC and Yang CQ, Molecular modeling of cellulose in
74. Holm JK, Henriksen UB, Hustad JE and Sørensen LH, Toward an under- amorphous state, part I: Model building and plastic deformation study.
standing of controlling parameters in softwood and hardwood pellets Polymer 45:1063–1071 (2004).
production. Energy Fuels 20:2686–2694 (2006). 94. Goldstein IS, Composition of biomass, in Organic Chemicals from
75. Hill B and Pulkinen DA, A study of factors affecting pellet durability and pel- Biomass, ed by Goldstein IS. CRC Press In., Boca Raton, FL, USA, pp.
leting efficiency in the production of dehydrated alfalfa pellets: A special 9–18 (1981).
report, Saskatchewan Dehydrators Association, Tisdale, SK, Canada (1998). 95. Nelson DL and Cox MM, Lehninger Principles of Biochemistry. W.H.
76. Heffiner LE and Pfost HB, Gelatinization during pelleting. Feedstuffs, Freeman and Company, New York, USA (2005).
45(23):33 (1973). 96. Hon DNS, Cellulosic adhesives, in Adhesives from Renewable
77. Tumuluru JS, Sokhansanj S, Lim CJ, Bi X, Anthony L, Staffan M, Sowlati Resources, ed by Hemmingway RW and Conner AH. American
T and Ehsan O, Quality of wood pellets produced in British Columbia. Chemical Society, Washington DC, USA, pp. 289–304 (1989).
App Eng Agr 26(6):1013–1020 (2010b). 97. Lebo SE Jr, Gargulak JD and McNally TJ, Lignin, in Kirk-Othmer
78. Kaliyan N and Morey RV, Factors affecting strength and durability Encyclopedia of Chemical Technology. John Wiley & Sons, Inc., New
of densifi ed biomass products. Biomass Bioenerg 33(3):337–359 York, USA (2001).
(2009). 98. Martone PT, Estevez JM, Lu F, Ruel K, Denny MW, Somerville C and
79. Mani S, Tabil LG and Sokhansanj S, An overview of compaction of bio- Ralph J, Discovery of lignin in seaweed reveals convergent evolution of
mass grinds. Powder Handling Process 15(3):160–168 (2003). cell-wall architecture. Curr Biol 19(2):169–175 (2009).
80. Chirife J and Del Pilar M, Water activity, glass transition, and micro- 99. Zandersons J, Gravitis J, Zhurinsh A, Kokorevics A, Kallavus U and
bial stability in concentrated/semimoist food systems. J Food Sci Suzuki CK, Carbon materials obtained from self-binding sugar cane
59(5):921–927 (1994). bagasse and deciduous wood residues plastics. Biomass Bioenerg
81. MacBain R, Pelleting Animal Feed. American Feed Manufacturers 26:345–360 (2004).
Association, Arlington, VA, USA (1966). 100. Lehtikangas P, Quality properties of fuel pellets from forest biomass,
82. Payne JD, Improving quality of pellet feeds. Mill Feed Fert 161:34–41 Licentiate Thesis, Report nr 4. Department of Forest Management and
(1978). Products, Uppsala (1999).
83. Payne JD, Troubleshooting the pelleting process. [Online]. American 101. Lehtikangas P, Quality properties of pelletised sawdust: Logging resi-
Soybean Association Tech. Bulletin #MITA (P) NO. 044/11/96 (FT40- dues and bark. Biomass Bioenerg 20:351–360 (2000).
1997) (1997). Available at: www.asaimsea.com/download_doc. 102. Wilson TO, Factors affecting the wood pellet durability, MSc thesis.
php?file=FT40-97.pdf [June 22, 2011]. Department of Agricultural and Biological Engineering, Pennsylvania
84. Mohan D, Pittman CU and Steele PH, Pyrolysis of wood/biomass for State University (2010).
bio-oil: A critical review. Energy Fuel 20:848–889 (2006). 103. Bradfield J and Levi MP, Effect of species and wood to bark ratio on
85. Thomas M, van Vliet T and van der Poel AFB, Physical quality of pelleted pelleting of southern woods. For Prod J 34:61–63 (1984).
animal feed, part 3: Contribution of feedstuff components. Anim Feed 104. Tabil LG, Sokhansanj S and Tyler RT, Performance of different binders
Sci Technol 70:59–78 (1998). during alfalfa pelleting. Can Agric Eng 39(1):17–23 (1997).
86. Collado LS and Corke H, Starch properties and functionalities, in 105. Tabil LG and Sokhansanj S, Process conditions affecting the physical
Characterization of Cereals and Flours, ed by Kaletunç G and Breslauer quality of alfalfa pellets. Am Soc Agric Eng 12(3):345–350 (1996c).
KJ. Marcel Dekker, Inc., York, USA, pp. 473–506 (2003). 106. Anonymous, Special report, Binders. Milling 166(2):31–33 (1983).

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 705
JS Tumuluru et al. Review: Biomass densification technologies to develop uniform feedstock commodities

107. Eranki PL, Bals BD and Dale BE, Advanced regional biomass process- 126. Shafizedeh F, Pyrolytic reactions and products of biomass, in
ing depots: A key to the logistical challenges of the cellulosic biofuels Fundamentals of Biomass Thermochemical Conversion, ed by Overend
industry. BioFPR (under review) (2011). RP, Milne TA and Mudge LK. Elsevier, London, UK, pp. 183–217 (1985).
108. Peleg M, Flowability of food powders and methods for its evaluation—A 127. Williams PT and Besler S, The influence of temperature and heating rate
review. J Food Process Eng 1:303–328 (1977). on the slow pyrolysis of biomass. Renew Energ 7:233–250 (1996).
109. Peleg M and Mannheim CH, Effect of conditioners on the flow proper- 128. Lipinsky ES, Arcate JR and Reed TB, Enhanced wood fuels via torrefac-
ties of powdered sucrose. Powder Technol 7:45–50 (1973). tion, Fuel Chem Div Preprints 47(1):408–410 (2002).
110. Bhattacharya SC, Sett S and Shrestha RM, State of the art for biomass 129. Koukios EG, Progress in thermochemical, solid-state refining of biofu-
densification. Energ Source 11:161–182 (1989). els: From research to commericalisation, Adv Th Ch Biomass Conv, 2,
111. Bhattacharya SC, State-of-the-art of utilizing residues and other Bridgwater (1993).
types of biomass as an energy source. RERIC Int Energ J 15(1):1–21 130. Bergman PCA, Combined torrefaction and pelletisation: The TOP proc-
(1993). ess, ECN Report # ECN-C-05-073 (2005).
112. Liu C and Wyman CE, Partial flow of compressed-hot water through 131. Teymouri F, Laureano-Perez L, Alizadeh H and Dale BE, Optimization of
corn stover to enhance hemicellulose sugar recovery and enzymatic the ammonia fiber explosion (AFEX) treatment parameters for enzymatic
digestibility of cellulose. Bioresource Technol 96:1978–1985 (2005). hydrolysis of corn stover. Bioresour Technol 96(18):2014–2018 (2005).
113. Lam PK, Sokhansanj S, Lim CJ, Bi X and Melin S, Energy input and 132. Dale BE, Densifying and handling AFEX biomass: A cooperative
quality of pellets made from steam exploded Douglas Fir (pseudotsuga- research project. Northern Plains Biomass Economy. Fargo, ND, USA,
menziesii). J Energ Fuel (2011) in press. September 22 (2009).
114. Mosier N, Henderson R, Ho N, Sedlak M and Ladisch MR, Optimization 133. Tumuluru JS, Kuang X, Sokhansanj S, Lim CJ, Bi X and Staffan M,
of pH-controlled liquid hot water pretreatment of corn stover. Development of laboratory studies on the off-gassing of wood pellets.
Bioresource Technol 96:1986–1993 (2005). Can Bio Eng 52:8.1–8.9 (2010c).
115. Marchessault RH, Coulombe S, Hanai T and Morikawa H, Vinylic 134. Rhen C, Gref R, Sjöström M and Wästerlund I, Effects of raw material
moncmers from bioconversion of wood. 181st National Meeting of the moisture content: Densification pressure and temperature on some
American Chemical Society, Atlanta, GA, USA (1981). properties of Norway spruce pellets. Fuel Process Technol 87(1):11–16
116. Marchessault RH, Coulombe S, Morikawa H and Robert D, (2005).
Characterization of aspen exploded wood lignin. Can J Chem 135. Tumuluru JS, Tabil LG, Opoku A, Mosqueda MR and Fadeyi O, Effect of
1082(60):2372 (1981). process variables on the quality characteristics of pelleted wheat distill-
117. DeLong EA, Method of rendering lignin separable from cellulose and er’s dried grains with soluble (DDGS). Bio Eng 105(4):466–475 (2010d).
hemicellulose in lignocellulosic material and the product so produced. 136. Reece FN, Temperature, pressure, and time relationships in forming
Canada Patent 1096374 (1981). dense hay wafers, Trans of the ASAE 9:749–751 (1966).
118. Foody P, Optimization of steam explosion pretreatment, DOE Contract 137. Israelsen M, Busk J and Jensen J, Pelleting properties of dairy com-
AC02-79ET23050, Final Report. 1 B80 (1980). pounds with molasses, alkali-treated straw and other byproducts.
119. Kaar WE, Gutierrez CV and Kinoshita CM, Steam explosion of sugar- Feedstuffs 7:26–28 (1981).
cane bagasse as a pretreatment for conversion to ethanol. Biomass 138. Winowiski T, Wheat and pellet quality. Feed Management 39:58–64
Bioenerg 14(3):277–287 (1998). (1988).
120. Toussaint B, Excoffier G and Vignon MR, Effect of steam explosion 139. Stark CR, I. Pellet quality and its effect on swine performance; II.
treatment on the physico-chemical characteristics and enzymic Functional characteristics of ingredients in the formation of quality pellets.
hydrolysis of poplar cell wall components. Anim Feed Sci Technol PhD dissertation, Kansas State University, Manhattan, KS, USA (1994).
32:235–242 (1991). 140. Cavalcanti WB, The effect of ingredient composition on the physical
121. Felfli FF, Luengo CA and Beaton PA, Bench unit for biomass resi- quality of pelleted feeds: A mixture experimental approach. PhD disser-
dues torrefaction, in Biomass for Energy and Industry: Proceedings tation, Kansas State University, Manhattan, KS, USA (2004).
in International Conference, ed by Kopetz H. Wurzburg, Germany, pp 141. Richardson W and Day EJ, Effect of varying levels of added fat in broiler
1593–1595 (1998). diets on pellet quality. Feedstuffs 48:24 (1976).
122. Zanzi R, Ferro DT, Torres A, Soler PB and Bjornbom E, Biomass torre- 142. Angulo E, Brufau J and Esteve-Garcia E, Effect of a Sepiolite product on
faction, in The 6th Asia-Pacific International Symposium on Combustion pellet durability in pig diets differing particle size and in broiler starter
and Energy Utilization, Kuala Lumpur, May 20–22 (2002). and finisher diets. Anim Feed Sci Technol 63:25–34 (1996).
123. Bourgeois J and Doat J, Proceedings conference on bioenergy. = o rg G and Ay P, Increasing
143. Satyanarayana N, Tao Y, Glaser C, Hans-JC
Göteborg 3:153 (1985). the calorific value of rye straw pellets with biogenous and fossil fuel
124. Arcate JR, New process for torrefied wood manufacturing. Bioenergy additives. Energy Fuel 24:5228–5234 (2010).
Update 2 (2000) [Online]. Available at: https://www.bioenergyupdate. 144. Sadaka S and Negi S, Improvements of biomass physical and ther-
com/magazine/security/NL0400/bioenergy_update_april_2000.htm mochemical characteristics via torrefaction process. Environ Prog
[June 22, 2011]. Sustainable Energy 28(3):427–434 (2009).
125. Arcate JR, Global markets and technologies for torrefied wood in 2002. 145. Kleinschmidt C, Overview of international developments in torrefac-
Wood Energy 5:26–28 (2002). tion. Central European Biomass Conference 2011, Graz, Austria (2011).

706 Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb
Review: Biomass densification technologies to develop uniform feedstock commodities JS Tumuluru et al.

[Online]. Available at: http://www.bioenergytrade.org/downloads/grazk-

leinschmidtpaper2011.pdf [June 22, 2011]. J. Richard Hess
146. European Committee for Standardization/Technical Specification (CEN/ Richard Hess is currently Manager for the
TS), Solid biofuels: Fuel specification and Bradfield classes. Technical Biofuels and Renewable Energy Technolo-
specification CEN/TS 14961, Brussels, Belgium (2005). gies Department and Program Manager for
147. Sjostrom E, Wood chemistry, in Fundamentals and Applications, 2nd the Biomass Program at the Idaho National
Edition, Academic Press, San Diego, CA, USA, pp. 292. (1993). Laboratory (INL). His research interests are
148. Hancock BC, Colvin JT, Mullarney MP and Zinchuk AV, The relative centered on biomass feedstock production
densities of pharmaceutical powders, blends, dry granulations, and and supply systems for biorefining. He is also
immediate-release tablets. [Online]. Pharmaceutical Technology. (2003) the US National Task Leader for the International Energy Agency
Available at: http://www.pharmtech.com [June 22, 2011]. Task 40 Team, the National Bioenergy Center – Laboratory Repre-
149. Shimida Systems [Online]. Available at: http://www.shimada.co.uk/ sentative for the Idaho National Laboratory, and is a part of the US
index.php [July 20, 2011]. DoE Biomass Research and Development Initiative and Logistics
Team. He received his PhD in Agricultural Plant Sciences from
Utah State University.

Jaya Shankar Tumuluru

Dr Tumuluru is currently working as a Re-

search Scientist, Biofuels and Bioenergy
Department at the Idaho National Laboratory
(INL). His research is focused on biomass pre- Kevin L. Kenney
treatments and densification technologies. His
research interests include biomass processing Kevin L. Kenney is a Research Engineer for
and storage for bioenergy applications, extru- the Idaho National Laboratory (INL) Biomass
sion processing of food, feed and fuel pellets, experimental data Program. Kevin has been working in the
analysis using statistical methods, evolutionary and back propaga- INL Biomass Program for eight years, and is
tion algorithms. He received his PhD in Agricultural Engineering currently Team Lead of the Biomass Harvest,
from Indian Institute of Technology, Kharagpur, India in 2005. Collection and Storage Group; the Lead of
the Industrial Partnership Program for the INL
Biomass Program, and as is actively involved in program strategic
development. Kevin holds four patents relating to biomass har-
vesting equipment and technology.
Christopher T. Wright

Dr Wright is currently the Preprocessing and

Handling Group Lead at the Idaho National
Laboratory. His interests are applied fluid
dynamics, thermal hydraulics, and two-phase
flow. His research focuses on biomass size
reduction, thermal treatment, and advanced
densification techniques. His work also
includes biomass supply system techno-economic modeling and
systems analysis to assess the state of technology, identify barri-
ers, and guide R&D. He received his PhD in Mechanical Engineer-
ing from Utah State University in 2005.

Published in 2011 by John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 5:683–707 (2011); DOI: 10.1002/bbb 707