Applied Energy 157 (2015) 508–516

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Effects of binders on the properties of bio-char pellets

Qiang Hu, Jingai Shao ⇑, Haiping Yang, Dingding Yao, Xianhua Wang, Hanping Chen
State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan 430074, China

h i g h l i g h t s

 Bio-char was densified with four binders of varying content (5–20%).

 The rheological characteristics of pellets were analyzed.
 Starch pellets showed good hydrophobicity, but low volume density and poor mechanical strength.
 NaOH pellets showed highest compressive strength and hydroscopicity among all pellets.
 Lignin and Ca(OH)2 pellets exhibited more desirable characteristics for use as biofuels.

a r t i c l e i n f o a b s t r a c t

Article history: Bio-char, produced from biomass pyrolysis, can be pelletized to improve its undesirable characteristics,
Received 3 March 2015 such as low bulk and energy densities, poor transportation and storage properties, and troublesome to
Received in revised form 22 April 2015 co-fire with coal. In this study, rice husk char was compressed into pellets with four kinds of binders (lig-
Accepted 8 May 2015
nin, starch, calcium hydroxide and sodium hydroxide). The compressive process, mechanical strength,
Available online 1 June 2015
basic fuel properties, and combustion characteristics were investigated to elucidate the effect of binders
on the properties of bio-char pellets. Results showed that starch pellets had good hydrophobicity, but low
Keywords:
volume density and poor mechanical strength. The softening and morphological transition of lignin
Bio-char pellet
Binders
during compression may account for the high elastic modulus and good bonding of lignin pellets. The
Mechanical properties low saturated moisture content of Ca(OH)2 pellets is mainly ascribed to its hydration. NaOH pellets
Combustion characteristics showed the highest compressive strength among all pellets, and also exhibited the highest moisture
uptake that may worsen the handling and storage treatment of bio-char. Compared with raw bio-char,
bio-char pellets had a lower ignition temperature, wider temperature interval, and higher oxidation
activity according to the thermogravimetric analysis. The lignin and Ca(OH)2 pellets showed lower com-
pression energy consumption and moisture uptake, enhanced mechanical strength and promoted com-
bustion performance, which demonstrated more desirable properties for utilization as biofuels.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction carbon content, excellent grindability, porosity, and species homo-

geneity [3,7], showed the potential to be used in firing plants, bar-
The development of biomass has been an important issue for becues, steelmaking, and soil amendment to replace conventional
the past several decades and would remain to be attractive in the coal partially and mitigate the environment burden. Therefore, it
future due to its clean, renewable, and carbon–neutral properties is very attractive to explore efficient ways to upgrade bio-char into
[1]. Among the plentiful conversion and utilization technologies high value-added bioenergy.
applied to biomass, pyrolysis is one of the most promising ways Whereas, the application of bio-char suffers from its low bulk
to convert biomass into bio-char, pyrolysis gas, and liquid oil density, low energy density, and high transportation and storage
[2,3]. Many upgrading methods, such as catalytic reforming, filtra- cost [8]. It has been reported that densification can improve feed-
tion, and purification [4,5], have been investigated extensively to stock uniformity and enhance the handling and conveyance effi-
improve the effective utilization of pyrolysis gas and bio-oil ciencies [9]. In addition, with pretreatment technologies such as
despite the ineluctable shortcomings faced in application [6]. torrefaction [10,11], steam treatment [12], and SO2-catalyzed
Bio-char, with the properties of high calorific value, concentrated steam pretreatment [13], the resulted densified biomass has con-
siderable density, desirable hardness, and good hydrophobicity.
⇑ Corresponding author. Tel.: +86 27 87542417; fax: +86 27 87545526. Low rank coal or biomass can be compressed into pellets (bri-
E-mail address: jashao@hust.edu.cn (J. Shao). quettes) with enhanced density and considerable mechanical

http://dx.doi.org/10.1016/j.apenergy.2015.05.019
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
 Q. Hu et al. / Applied Energy 157 (2015) 508–516 509

strength, thus improve the logistics of pellet fuel and benefit the without water and binder, in terms of energy saving and quality
following utilization processes. Therefore, it is potentially possible upgrading in industry. However, to the best of authors’ knowledge,
to treat bio-char using densification technology in order to over- there are very limited studies focused on the effects of binders on
come the inherent disadvantages mentioned above and enable a the properties of bio-char after densification. This raise two ques-
wider range of application. tions when densification of bio-char for the promising quality bri-
Densification, a well proved technology [14–16], shows the quette bioenergy. One is why binders are needed to form bio-char
splendid practical prospects in converting coal, biomass, and tor- into bio-char pellets in order to enhance the quality. The other is
refied biomass into pellets (briquettes). It was found that the mate- how binders affect the pelletization process of bio-char and the
rials properties (e.g. particle size, shape, moisture content, and properties of the derived pellets. The solution of these questions
composition) and processing conditions (e.g. compressing speed, will then guide for binders selection and application when com-
temperature, pressure, and die size and shape) were the main fac- pacting bio-char in the industrial production process.
tors affecting the densification results [9,11,17]. Poddar et al. [18] Concerns of the description above, the representative organic
found that the density of lignocellulosic biomass pellet increased binders, lignin and starch, as well as the inorganic binders,
radically at first and then gradually with the increase of the applied Ca(OH)2 and NaOH, were selected for use in this study. The densi-
pressure. Jiang et al. [19] recommended the optimum moisture fication process was described and the properties of the derived
content was 10–15% for co-pelletization of sewage sludge and bio- bio-char pellets were characterized. The effects of binders on den-
mass. Peng et al. [10] pointed out that the quality of torrefied pel- sification of bio-char were investigated and the mechanical proper-
lets was improved at the die temperature of 230 °C and moisture ties, basic fuel properties, and combustion characteristics of the
content of about 10%. Fine torrefied sawdust particles (less than pellets derived from different binders were explored. Especial
0.25 mm) was demonstrated to produce pellets with prior water attention was paid to explore the optimum densification condition
hydrophobicity and hardness compared with coarse ones [17]. with the binders, which may be helpful for future binder selection
Furthermore, the densification, hardening, and activation methods and application in the bio-char densification industry.
were developed to produce high-value added coal/biomass bri-
quettes [11,20,21]. Han et al. [14] and Massaro et al. [22] found
that high energy density and low ash content briquettes could be 2. Materials and methods
obtained with the densification of low rank pulverized coal. Diez
et al. [23] established a method for compressing 2.1. Materials
carbon-containing wastes into briquettes for metallurgical coke
production. Rubio et al. [21] pressed char and pitch at 125 MPa The raw material of rice husk bio-char was collected from the
to form briquettes, followed by curing, carbonization, and activa- biomass pyrolysis poly-generation demonstration plant in E’zhou,
tion to obtain briquettes with enhanced surface areas and pore Hubei province, China. The results of the proximate and ultimate
volumes. analysis of the bio-char are shown in Table 1. The bio-char sample
Compared with raw biomass, the derived pyrolytic bio-char has has more fixed carbon and ash content, lower O content and higher
different features, such as decreased moisture and volatile matter calorific value compared with the original rice husk sample.
content, increased friability, cracked fiber component, and Bio-char was ground in a blade mill (XY-1000A, Songqing
enhanced surface area and pore structure [3,7], which inevitable Hardware Company, China) for 3 min and then the particle size
changed the forming characteristics during densification process distribution was analyzed using a laser particle-size analyzer
and even might impede the deformation of particles and reduce (MS2000, Malvern Instruments Ltd., U.K.). The results are shown
their adherence among each other, thus making it difficult to form in Fig. 1. It was found that the bio-char particles were uniform in
pellets. For this very reason, a few studies have been conducted on size with an average particle size of about 0.1 mm, and about
the densification behavior of pyrolytic bio-char. Bazargan et al. [24] 98% of the ground bio-char was less than 0.3 mm in size, indicating
pointed out that compacting palm kernel shell bio-char without the good grindability of the bio-char as previously reported [7,26].
addition of water resulted in weak bonding products. And Bio-char was dried in an oven at 105 °C overnight and kept in a
bio-char pellets formed with water and soluble starch showed pro- desiccator before further analysis. Two organic binders (alkaline
moted tensile crushing strength, impact resistance, and water lignin and starch) and two inorganic binders (Ca(OH)2 and NaOH)
resistance compared with no-binder pellets. Kong et al. [8] investi- were selected. Alkaline lignin, a dark brown powder, was pur-
gated the conversion of sawdust into bio-char pellets through chased from Sigma–Aldrich Co., Ltd (USA). The starch, Ca(OH)2,
pyrolysis and densification. Results showed that the mixture of and NaOH were purchased from the Tianjin Chemical Reagent
Ca(OH)2 and lignin added in bio-char had water resistant effect Factory (China). All the chemicals used in this study were analyti-
on reducing moisture adsorption percentage during storage of cal grade and were not further purified.
bio-char pellets. Haykiri-Acma et al. [25] found that sulfide liquor,
molasses, and Linobind could be used as binding agents to improve 3. Experimental method
the durability of the briquettes when pelletizing carbonized brown
seaweed. Accordingly, binder addition assisted bio-char particles The densification of bio-char was carried out in a universal
bonded together and improved the densification characters of material testing machine (CMT5205, MTS, China) with binder
bio-char, especially for which was hard to form strong pellets and water added at room temperature. The pellet device of the

Table 1
Proximate and ultimate analysis of the rice husk and bio-char (dry material).

Sample Proximate analysis (d, wt.%) Ultimate analysis (d, wt.%) HHV (MJ/kg)
V A FC C H N S Oa
Rice husk 67.70 16.22 16.08 41.78 5.35 0.30 0.08 36.31 15.44
Bio-char 7.29 33.19 59.52 56.46 2.14 1.14 0.63 6.44 19.69
a
Calculated by difference.
510 Q. Hu et al. / Applied Energy 157 (2015) 508–516

Mass fraction in corresponding range (%)

8
(a) (b)

Cumulative mass fraction (%)

100
6

4
50

0
0

0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000

Particle size (µm) Particle size (µm)

Fig. 1. Particle size distribution of bio-char after 3 min of grinding: (a) mass fraction in the stated size range and (b) cumulative mass fraction smaller than the stated size.

Fig. 2. Schematic representation of the pelletization system: (a) experimental device system and (b) structure of the mold.

testing machine (Fig. 2a) can be operated with the real-time com- Dr DF=S H0
E¼ ¼ ¼ k ð1Þ
pressive load and displacement recorded by sensors. The machine De DH=H0 S
contains a removable girder in the middle and a fixed pedestal at
the bottom. The girder is controlled by the machine accurately so where E (MPa) is the elastic modulus, and Dr (MPa), De, DF (N), DH
that it can move down at a constant speed. A mold with mixed (mm) are the gradient of stress, strain, force and height of pellets,
bio-char was placed inside between the girder and the pedestal respectively. H0 (mm) is the initial height when the bio-char was
for compression into a single pellet. The mold is composed of a originally in the mold, S (mm2) is the cross-sectional area of pellets,
compressive bar (20 mm diameter), a die sleeve (20.20 mm inside and k is the slope of the relevant curve from the
diameter), two spacers (20 mm diameter and 10 mm in length), displacement-compress load graph. E, an index that exhibits the
and a pedestal as shown in Fig. 2b. complexity of the compressed bio-char, reflects the incremental
Initially, the binder and water were added into the bio-char pro- force needed to produce the corresponding strain.
portionally and the mixtures were thoroughly stirred. The dosage Referred to previous studies [27–30], the relaxation data during
of the binders varied from 5% to 20% (w/w). The amount of water the relaxation time of 3 min were normalized and linearized and
added was 20% (w/w). The water reintroduced into the bio-char represented as a straight line:
before compression enhanced the salutary briquette strength
through capillary binding effects [22]. Then, the mixtures were r0  t
¼k þk t ð2Þ
filled into the mold cylinder and placed horizontally on the fixed r0  rðtÞ 1 2
pedestal of the densification machine. The sample was compressed
by moving the girder down at a speed of 5 mm/min until the com- where r0 (MPa) is the initial stress, rðtÞ (MPa) is the stress after
pressive load reached 30 kN. The die was then held for 3 min at a time t (s) at relaxation, while k1 and k2 are constants. Then the
constant displacement for relaxation. Finally, the bio-char pellet obtained k2, the visco-elastic slope, was used to calculate asymp-
of 20 mm diameter was ejected from the mold and stored at totic modulus (EA, MPa) based on the following equation:
55 °C. At least five replicates were made at each condition.
 
The compressive force–displacement-time data for the com- r0 1
EA ¼  1 ð3Þ
paction and relaxation of the bio-char in the universal tester were e k2
recorded for rheological analysis. The mechanical energy consump-
tion associated with compacting was calculated by integrating the where e is strain of the pellets during the relaxation process
force–displacement curve. The elastic modulus of the samples dur- (dimensionless or m m1). EA is an empirical index of solidity that
ing the compression period was determined from the slope of each is the ability of the compressed powder to sustain unrelaxed stres-
curve of the stress–strain graph [27] and expressed as: ses [27].
 Q. Hu et al. / Applied Energy 157 (2015) 508–516 511

3.1. Characterization of bio-char pellets binder content differed for the various types of binders. The com-
pressive strength of the lignin and Ca(OH)2 pellets increased with
Proximate analysis of the pellets was carried out according to increasing binder content, when the quantity of binder was less
standard of the Test Methods for Densified Biofuels (NY/T than 10% (w/w). This is in contrast to cases where the binder
1881-2010). Ash content was obtained as the percentage of residue amounts exceeded 10% (w/w), indicating that stronger pellets
remaining after the oxidation of the pellets at 550–600 °C for 4 h. formed after the appropriate quantity of binder was introduced.
Volatile content was defined as the mass loss of the dried sample Compared with binder amounts, binder type is more sensitive to
after cracking at 900 °C for 7 min in an anoxic atmosphere. The compressive strength, as seen in Fig. 3. The compressive strength
calorimetric values were measured by oxygen bomb calorimeter of no-binder pellet was 0.65 MPa. Except for starch pellets, whose
(Parr 6300, USA). compressive strength was not that different from no-binder pellet,
The moisture absorption behavior of the pellets was explored in the other pellets showed better mechanical hardness. For example,
a constant temperature and humidity incubator (HWS-150, for pellets densified with lignin and Ca(OH)2 (10% (w/w)), the
Sumsung Laboratory Instrument Company Ltd., China). At the mechanical strength increased to 3.82 and 5.55 MPa, respectively.
beginning of the testing process, the pellets were dried in the oven NaOH pellets had the highest compressive strength exceeding
at 105 °C for 24 h, and then put into the incubator at a temperature 10 MPa.
of 30 °C and a constant relative humidity of 70%. The mass of the The promoted mechanical strength of the lignin pellets (varying
pellets was measured every 30 min for the first 4 h, followed by from 2.77 to 3.82 MPa) was mainly related to the bonding and
every 120 min for the next 6 h. Finally, the mass of the pellets transformation process within the bio-char mixture. The bonding
was measured every 10 h until the mass reached constant. forces involved in the bio-char pellets densified with lignin mainly
The mechanical strength of the bio-char pellets densified with related to the attraction and cohesion forces that include hydrogen
different binders was measured in the universal material testing bonds, Van Der Waals forces, and mechanical interlock [10,14]. The
machine with a cylinder drill (diameter of 5 mm). Bio-char pellets softening and morphological transition of the lignin during the
were placed horizontally on the fixed pressure plate and a com- compression binds the bio-char particles like a bridge, as reported
pressive force was applied to the side of the cylindrical pellets. by Kaliyan and Morey and Reza et al. [32,33]. Ca(OH)2 pellets
The moving drill was run at a compression rate of 2 mm/min and showed good mechanical strength, due to the ionic bond produced
the load required to break or fracture was recorded in order to cal- from the dissolution and reaction of Ca(OH)2 and the hardening
culate the ultimate compressive strength of the pellet [15,31]. and strengthening effects produced by the hydration of Ca(OH)2
Combustion characteristics of the pellets were investigated [8]. The strong ionic force of NaOH captures the moisture and
using the thermogravimetric analyzer (TGA) (Pyris1 TGA, retains it within the pellets, which may be responsible for the
PerkinElmer Instruments, USA). The pellets were broken into small hardness of the NaOH pellets. Since rice husk ash contains 98% sil-
pieces and a small unitary mono-block (approximately 30 mg) was ica, sodium silicate, which acts as a cementing agent, probably
used for each pellet as the experimental sample. The combustion formed during the mixing and compressing process when NaOH
experiments were carried out at atmospheric pressure with an was added into the bio-char [34,35].
air flow rate of 100 mL/min and the temperature was regulated Fig. 4 illustrates the pelletization energy consumption for den-
as follows: (a) heating from room temperature to 105 °C at a rate sification with different binders and their quantity variation. The
of 10 °C/min and holding for 10 min to remove water, and (b) con- highest energy saving was observed for NaOH pellets when 1 g of
tinuous heating up to 800 °C at 10 °C/min and holding for 10 min bio-char mixture was compressed at room temperature and
to make sure the combustion was complete. 30 kN maximum compression load. Binder content had a more
noticeable effect than binder type on the compression energy.
The values of energy consumption for all pellets were reduced by
4. Results and discussion 10.66% (from 10.79 to 9.64 J/g) when the binder quantity increased
from 5% to 20% (w/w). Moreover, a good linear relationship is found
4.1. Mechanical properties and energy consumption between energy consumption and binder content with the correla-
tion coefficients above 0.90, as shown in Fig. 4. Large amounts of
Fig. 3 showed the compressive strength of the pellets as a func-
tion of the binder content. It was observed that the effects of the
11.0
Lignin Lignin
Compressive energy consumption (J/g)

25 Starch
Starch
Ca(OH) 2 Ca(OH) 2
NaOH
Compressive strength (MPa)

20 NaOH
10.5

15

10.0
10
2
Lignin: y=10.837-0.041x; R =0.9687
2
5 Starch: y=10.832-0.024x; R =0.9614
9.5 Ca(OH)2: y=10.880-0.033x; R2 =0.9014
2
0 NaOH: y=10.881-0.059x; R =0.9902

0 5 10 15 20 0 5 10 15 20
Binder content (%) Binder content (%)

Fig. 3. Mechanical strength of the bio-char pellets as a function of the binder Fig. 4. Energy consumption of the bio-char pellet as a function of binder type and
content. content.
512 Q. Hu et al. / Applied Energy 157 (2015) 508–516

energy were consumed to rearrange particles, break larger ones, relaxation of the pellets during 3 min are shown in Table 2. The
and reduce the interspace when impacting the mixture before elas- k2 of the binder pellets was less than that of the no-binder pellet.
tic compression [36]. Binders contributed to the cementation of As the binder content increased from 5% to 20% w/w, the
bio-char particles, which led to the reduction in energy costs for visco-elastic slope of k2 decreased. This indicated that the addition
the deformation and plasticization period [27]. of binders could decrease the degree of plastic flow after compres-
Weak mechanical hardness of pellets will lead to high breakage sion. The more binders used, the higher the degree to which plastic
levels during handling, transportation, and storage. It would also flow decreased [27]. In this study, the asymptotic modulus is an
reduce the efficiency of a pellet stove or burner, to some extent indicator of the degree of stress relaxation of the bio-char pellets
[10]. The improved mechanical strength and reduced compression under constant displacement [30]. A higher asymptotic modulus
energy use of the bio-char pellets is attractive for end-users. of the pellets means higher stress residue remains in the pellets,
Considering the compressive strength and energy costs for the four and the more rigid of the pellets can be inferred. Therefore, a high
types of pellets in this study, starch pellets had the worse strength asymptotic modulus tends to disperse through the plastic flow of
properties and did not exhibit obvious energy saving advantages. the solid matrix and realign the inter-particle bridges during the
Consequently, starch can be excluded as a potential binder from storage and handling process [29]. Good correlation of the relax-
the point of view of energy saving and hardness strengthening. ation curves was found using Eq. (2), with a correlation coefficient
The rheological properties of the bio-char pellets not only higher than 0.99 (Table 2), which showed credible viscoelastic rela-
reflect the relationship between stress and strain that guides the tionships in the relaxation process. Pellets densified with binders
compression treatment procedure, but also shows the solidity of showed lower EA than no-binder pellet, also it decreased constantly
the pellets when sustaining unrelaxed stresses. The maximum as binders content increased, indicating that binders would con-
elastic modulus (Emax) of the bio-char pellets densified with vari- tribute to the fastening of the bio-char particles. And with more
ous binders was calculated using Eq. (1), as shown in Table 2. It binders added, the degree of stress relaxation increased and the
clearly showed that among all the pellets, lignin pellets had the rigid of the pellets reduced.
highest elastic modulus (from 1207 to 1379 MPa), which was also
higher than the no-binder pellet (1165 MPa). No significant differ- 4.2. Basic fuel properties of bio-char pellets
ences were found among the elastic modulus values for the pellets
with the other three binders (starch, Ca(OH)2 and NaOH). It is pos- The proximate analyses of the bio-char pellets are presented in
tulated that amorphous lignin, with a low morphological transition Table 3. With organic binders (lignin and starch), bio-char pellets
temperature, had a high modulus of elasticity at this compressive had higher volatile matter, but lower ash content and fixed carbon;
condition [27,33]. The elastic modulus reflects the change of com- while pellets with inorganic binders (Ca(OH)2 and NaOH) had
pressive stress corresponding to the increase in strain. The higher enhanced ash content but less fixed carbon compared with that
the elastic modulus of the pellet, the higher the force needed to of organic binders. Apparently, organic binders contributed to the
compress the bio-char particles and the harder it would be to com- increase of volatile content. The residues increased after the oxida-
press the bio-char into pellets. Moreover, the elastic modulus has tion reaction at 550–600 °C with the addition of inorganic binders.
been reported to be dependent on the chemical composition of As the quantity of added binder increased, the ash content also
the material [27]. The great disparity in the elastic modulus values increased. The internal moisture was defined by measuring weight
has been ascribed to the large variation in material. The large loss when pellets were heated from 55 to 105 °C. The internal
increase in the elastic modulus with binder quantity increases moisture content of no-binder pellet was 1.24%, and there was a
from 15% to 20% (w/w) would have caused a relatively large varia- notable increase in pellets densified with binders (except for
tion in the pellets properties. Therefore, binder quantities of less Ca(OH)2) (Table 3). For example, the internal moisture content
than 15% would be recommended. increased by 33.06%, 14.52%, and 252.42% for lignin (1.65 wt.%),
For the relaxation period, the linearized relaxation data were starch (1.42 wt.%), and NaOH pellets (4.37 wt.%) when binders
used to fit a straight line based on Eq. (2). The estimated k2 was were added at 10%, respectively. The higher internal moisture con-
used in Eq. (3) to calculate EA. The results obtained from the tent of NaOH pellets implied that good cementation occurred and
more hydrogen bonds formed, which are speculated to have higher
mechanical strength, when binders were introduced into the
Table 2 bio-char. This is in agreement with the compressive strength
Mechanical parameters of elastic modulus and asymptotic modulus from compres- results in Fig. 3.
sion and relaxation.
The volume density of the pellets is related to their transporta-
Binders Added content Emax (MPa) k2 Correlation EA (MPa) tion costs and energy density. Compared with the raw bio-char
(w/w) coefficients (523 kg/m3), no-binder pellet had much higher volume density
No 0 1164.93 11.62 0.9976 765034.34 and it was further increased when lignin, Ca(OH)2, and NaOH were
Lignin 5% 1206.61 9.63 0.9973 478230.90 added as binders. NaOH pellets exhibited the highest densities
10% 1244.31 8.09 0.9969 376321.42 ranging from 1163 to 1298 kg/m3 when compared to all other pel-
15% 1284.19 7.18 0.9975 283063.53 lets with binder quantities raised from 5% to 20% (w/w). The high
20% 1379.45 6.17 0.9980 292384.89
volume density of NaOH pellets was linked to the reaction of
Starch 5% 1164.86 9.23 0.9978 384467.16 NaOH with the inorganic constituents in bio-char and the subse-
10% 1126.08 8.38 0.9983 306018.48
15% 1125.68 7.57 0.9984 294483.94
quent strong binding with bio-char particles; this may have also
20% 1332.07 6.92 0.9983 243578.79 contributed to the high mechanical strength. Starch pellets showed
Ca(OH)2 5% 1125.34 10.15 0.9967 559718.22
the lowest volume density ranging from 1100 to 1080 kg/m3 with
10% 1225.59 9.01 0.9960 383780.56 the varied starch content, indicating the less adaptation possibility
15% 1174.26 7.86 0.9951 320525.98 for use as a binder in this study. In addition, the energy density of
20% 1273.57 6.95 0.9943 290087.12 raw bio-char material was 10.30 GJ/m3. Through pelletization, the
NaOH 5% 1131.31 8.96 0.9966 237828.98 energy density of the pellets increased significantly for both
10% 1099.59 8.64 0.9963 358052.07 no-binder and binder pellets (in the range from 18.99 to
15% 1105.66 8.41 0.9951 466169.44
23.09 GJ/m3). The enhanced volume and energy densities of
20% 1255.26 7.50 0.9958 389555.07
bio-char pellets implies that pelletization is an effective method
 Q. Hu et al. / Applied Energy 157 (2015) 508–516 513

Table 3
Proximate analysis, and volume and energy densities of pellets densified with various binders.

Binder* Added content (w/w) Volatile matter (%) Ash (%) Fixed carbon (%) Internal moisture (%) HHV Volume density Energy density
(MJ/kg) (kg/m3) (GJ/m3)
No 0 6.87 33.65 59.48 1.24 19.60 1123 22.01
Lignin 5% 9.54 33.26 57.20 1.54 19.31 1145 22.11
10% 11.01 32.37 56.62 1.65 19.36 1141 22.09
15% 10.48 31.69 57.83 1.89 19.49 1131 22.04
20% 13.43 31.13 55.44 2.01 19.34 1134 21.93
Starch 5% 9.96 32.82 57.22 1.31 19.35 1100 21.29
10% 12.93 31.26 55.81 1.42 19.40 1091 21.17
15% 15.62 29.78 54.60 1.38 19.45 1087 21.14
20% 17.83 28.73 53.44 1.62 19.38 1080 20.93
Ca(OH)2 5% 10.86 35.72 53.42 1.28 18.43 1123 20.70
10% 10.93 36.65 52.42 1.22 18.08 1136 20.54
15% 13.48 38.57 47.95 1.16 17.15 1153 19.77
20% 15.74 40.21 44.05 1.13 16.20 1172 18.99
NaOH 5% 8.96 34.81 56.23 2.60 18.99 1163 22.09
10% 10.37 34.95 54.68 4.37 18.65 1210 22.57
15% 12.45 36.30 51.25 4.74 17.89 1255 22.45
20% 14.83 35.41 49.76 4.83 17.79 1298 23.09
*
No binder refers to bio-char densified with only 20% water, without a binder.

8
No binder (a) No binder
5% Lignin 6 (b)
5% Starch
10% Lignin 10% Starch
6 15% Lignin 15% Starch
Moisture uptake (%)
Moisture uptake (%)

20% Lignin 20% Starch

4

2
2

0 0

0 500 1000 1500 2000 0 500 1000 1500 2000

Exposure time (min) Exposure time (min)
6
No binder (c) No binder (d)
5% Ca(OH) 2 15 5% NaOH
10% Ca(OH) 2 10% NaOH
Moisture uptake (%)

15% Ca(OH) 2 15% NaOH

Moisture uptake (%)

4 20% Ca(OH) 2 12
20% NaOH

2 6

0
0
0 500 1000 1500 2000 0 500 1000 1500 2000
Exposure time (min) Exposure time (min)

Fig. 5. Moisture uptake rate of the bio-char pellets compressed with different binders at a relative humidity of 70% and a temperature of 30 °C: (a) lignin pellets, (b) starch
pellets, (c) Ca(OH)2 pellets and (d) NaOH pellets.

for promoting the properties of bio-char, in spite of the small storage and long distance transportation as reported before [37].
increase in ash content and the decrease in calorific value. However, with binder addition, the water adsorption process was
Fig. 5 shows the measured moisture uptake rate of the bio-char extended, and it took about 20 h for most tested pellets to reach
pellets made with four kinds of binders in various quantities rang- saturation. Furthermore, the saturated moisture content of all pel-
ing from 0% to 20% (w/w) at 30 °C and 70% relative humidity. It can lets densified with binders was higher than 3 wt.%; for NaOH, the
be seen that no-binder pellet reached saturation in 10 h. The value exceeded 10%. This indicated that the binders introducing
hydrophobicity of no-binder pellet is relatively strong given the may lead to the improvements in hygroscopicity due to the
saturated moisture content of around 2 wt.%. This is adequate for addition of hydrophilic elements. The saturated moisture content
514 Q. Hu et al. / Applied Energy 157 (2015) 508–516

of the bio-char pellets also consistently increases with increasing thermogravimetric (TG) analysis under an oxidizing atmosphere.
binder content, as seen in Fig. 5a (lignin pellets) and Fig. 5b (starch For the pellets, the thermal analysis was carried out only on those
pellets), because of the presence of oxygen-containing functional that presented good compressive strength (10% w/w of binder
groups from the binders [14]. Additionally, it was found that starch used). The raw bio-char and no-binder pellet were introduced for
pellets had less saturated moisture content (varying from 3.69 to comparison. TG and derivative TG (DTG) curves depicting the com-
4.42 wt.%) compared with lignin pellets (varying from 4.69 to bustion process of raw bio-char and bio-char pellets prepared from
6.48 wt.%). As for the pellets densified with inorganic binders, the different additives are presented in Fig. 6. The characteristic com-
saturated moisture content of Ca(OH)2 pellets was between 3.37 bustion parameters of each curve are summarized in Table 4. The
and 3.62 wt.% when the content of Ca(OH)2 increased from 5% to ignition temperature was obtained using the method reported by
20% (w/w). The hydration of Ca(OH)2 may contribute to the low Zhao et al. [38]. The temperature interval of the combustion pro-
moisture absorption of Ca(OH)2 pellets [8]. The corresponding sat- cess was defined as the temperature range of the combustion pro-
urated moisture content of NaOH pellets varied from 8.22 to cess of bio-char/bio-char pellets. The initial and final temperatures
16.35 wt.%. The high hygroscopicity of the NaOH pellets is mainly of combustion were considered as the temperature values at which
due to the high water absorbing properties of the mineral alkali. 1% of dry material burned and 99% of combustion material com-
The saturated moisture content of the pellets increased simultane- pletely burned out, respectively.
ously with the increase in NaOH content. For the TG curves, the first weight loss near 100 °C occurred due
to the evaporation of moisture. The higher and longer peak for the
4.3. Analysis of combustion characteristics NaOH pellets in the dewatering process related to a large release of
water, as seen in the DTG curves, representing a higher amount of
In order to evaluate the effects of the type of binder on the internal moisture in NaOH pellets. This is consistent with the
combustion characteristics, the pellets were subjected to results in Table 3. The combustion of bio-char and bio-char pellets

100 (a) Raw bio-char 100 (b) No-binder pellet

Mass loss rate (wt.%/ °C)

0.0 0.0

Mass loss rate (wt.%/°C)

Weight (wt.%)

Weight (wt.%)

80 80

60 -0.5 60 -0.5

40 40

-1.0 -1.0
0 200 400 600 800 0 200 400 600 800
Temperature (°C) Temperature (°C)

100 (c) Lignin-pellet 100 (d) Starch-pellet

0.0 0.0
Mass loss rate (wt.%/ °C)

Mass loss rate (wt.%/ °C)

Weight (wt.%)

80 80
Weight (wt.%)

60 -0.5 60 -0.5

40 40

-1.0 -1.0
0 200 400 600 800 0 200 400 600 800
Temperature (°C) Temperature (°C)

100 (e) Ca(OH) 2 -pellet 100 (f) NaOH-pellet

Mass loss rate (wt.%/°C)

0.0
Mass loss rate (wt.%/°C)

0.0
Weight (wt.%)

80 80
Weight (wt.%)

60 -0.5 -0.5
60

40
40
-1.0 -1.0
0 200 400 600 800 0 200 400 600 800
Temperature (°C) Temperature (°C)

Fig. 6. TG and DTG curves of (a) raw bio-char, (b) no-binder pellet, (c) pellet with 10% lignin, (d) pellet with 10% starch, (e) pellet with 10% Ca(OH)2 and (f) pellet with 10%
NaOH in an air flow rate of 100 mL/min at a heating rate of 10 °C/min.
 Q. Hu et al. / Applied Energy 157 (2015) 508–516 515

Table 4
Characteristic combustion parameters of the pellets derived from raw bio-char and the corresponding bio-char pellets.

Fuel Ignition temperature (°C) Temperature interval (°C) Temperature at maximum Maximum weight loss Residue (%)
weight loss rate (°C) rate (%/°C)
Raw bio-char 419 374–504 452 0.90 40.20
No-binder pellet 363 337–506 388 0.60 36.79
Lignin pellet 334 241–502 360 0.79 35.07
Starch pellet 355 234–512 387 0.68 33.98
Ca(OH)2 pellet 366 242–667 388 0.76 39.56
NaOH pellet 332 278–754 358 0.53 39.39

showed only one main DTG peak, which was due to the continuous Ca(OH)2 may have contributed to the low moisture absorption of
drop of mass in the TG curves in the bio-char combustion process. the Ca(OH)2 pellets. The highest moisture uptake of the NaOH pel-
From Fig. 6c and d, it is clear that a small peak occurred in the DTG lets could be possibly harmful to the handling and storage treat-
curves at about 240–310 °C, which corresponds to the releasing ment processes of bio-char.
and combustion of volatiles from the organic binders (lignin and Bio-char pellets had reduced ignition temperature, wider tem-
starch). And also from Fig. 6e and f, the volatilization and catalysis perature range, and higher oxidation activity compared with the
of the alkali metal may be responsible for the small peak in the raw bio-char. The releasing and combustion of volatiles from the
DTG curve at about 615–685 °C for Ca(OH)2 pellets and the contin- added organic binders led to a small peak before char combustion
uous drop of the TG curve for the NaOH pellets at temperatures on the TG curves of the organic pellets. The volatilization and catal-
above 600 °C, which also led to the high burnout temperatures of ysis of the alkali metal may have caused the high temperature
Ca(OH)2 and NaOH pellets as shown in Table 4. However, it was interval of the inorganic pellets.
confirmed that the alkali metal could cause serious fouling and With the added water content of 20% and binder content of 10%
slagging problems in the combustion boiler [39], NaOH should w/w, the lignin and Ca(OH)2 pellets showed lower compressive
not be recommended to make bio-char pellets for use as a fuel. energy consumption, acceptable moisture uptake, enhanced
Considering that the the maximum weight loss rate (%/°C) and mechanical strength, and promoted combustion performance,
the temperature at the maximum rate of mass loss are inversely which showed better potential for utilization as biofuels.
proportional to the reactivity of the fuel [26,40], the bio-char pel-
lets proved to be more reactive than raw bio-char particles as seen Acknowledgements
in Table 4. This indicates that pelletization promotes the oxidation
activity of the bio-char materials and makes them easier to burn as The authors wish to express sincere thanks for the financial
a fuel. And this was coincidence to the reduced ignition tempera- support from the National Basic Research Program of China
tures and the wider combustion temperature range of the pellets (2013CB228102), the National Natural Science Foundation of
compared with the raw bio-char. From Table 4, the residues of China (51476067), the Special Fund for Agro-scientific Research
organic pellets were lower than inorganic pellets and they all lower in the Public Interest (201303095), and the Independent
than the raw bio-char. All the residues are higher than the corre- Innovation Foundation of Huazhong University of Science and
sponding ash contents as described in Table 3. This is attributed Technology (2014TS118). The authors are also grateful for the
to the differences in the combustion system. The combustion in assistance on the experimental studies provided by the Analytical
the TG analyzer limits by contact area and oxygen concentration and Testing Center in Huazhong University of Science &
that requires a longer time to complete the burnout stage and Technology (http://atc.hust.edu.cn), Wuhan 430074, China.
reach the ash content of the materials [41].

References
5. Conclusions
[1] Carpenter D, Westover TL, Czernik S, Jablonski W. Biomass feedstocks for
renewable fuel production: a review of the impacts of feedstock and
In this study, rice husk bio-char was compressed into pellets pretreatment on the yield and product distribution of fast pyrolysis bio-oils
with four kinds of binders, namely lignin, starch, calcium hydrox- and vapors. Green Chem 2014;16:384–406.
ide, and sodium hydroxide, at room temperature and compressive [2] Huang YF, Chiueh PT, Kuan WH, Lo SL. Pyrolysis kinetics of biomass from
product information. Appl Energy 2013;110:1–8.
load of 30 kN in the universal material testing machine. The con- [3] Chen Y, Yang H, Wang X, Zhang S, Chen H. Biomass-based pyrolytic
clusions are summarized as follows: polygeneration system on cotton stalk pyrolysis: influence of temperature.
Bio-char pellets densified with binders showed lower compres- Bioresour Technol 2012;107:411–8.
[4] Yao D, Wu C, Yang H, Hu Q, Nahil MA, Chen H, et al. Hydrogen production from
sive energy consumption than no-binder pellet, due to the cemen- catalytic reforming of the aqueous fraction of pyrolysis bio-oil with modified
tation of the bio-char particles. The low mechanical strength and Ni–Al catalysts. Int J Hydrogen Energy 2014;39:14642–52.
volume density of the starch pellets indicated that starch was [5] Bridgwater AV. Review of fast pyrolysis of biomass and product upgrading.
Biomass Bioenergy 2012;38:68–94.
not suitable to be used as binders for the preparation of bio-char [6] Fatih Demirbas M. Biorefineries for biofuel upgrading: a critical review. Appl
pellets in this study. The softening and morphological transition Energy 2009;86(Supplement 1):S151–61.
of lignin during the compression may have resulted in the high [7] Abdullah H, Wu H. Biochar as a fuel: 1. Properties and grindability of biochars
produced from the pyrolysis of mallee wood under slow-heating conditions.
elastic modulus and good bonding of the lignin pellets. The asymp- Energy Fuels 2009;23:4174–81.
totic modulus was gradually reduced with the increase in the [8] Kong L, Tian S, Li Z, Luo R, Chen D, Tu Y, et al. Conversion of recycled sawdust
quantity of binder used from 5% to 20% (w/w), indicating that bin- into high HHV and low NOx emission bio-char pellets using lignin and calcium
hydroxide blended binders. Renewable Energy 2013;60:559–65.
ders can increase the degree of stress relaxation and reduce the
[9] Tumuluru JS, Wright CT, Hess JR, Kenney KL. A review of biomass densification
rigidity of the pellets. systems to develop uniform feedstock commodities for bioenergy application.
Pellets had increased volume density and enhanced energy den- Biofuels, Bioprod Biorefin 2011;5:683–707.
sity when compared with the raw bio-char particles. Bio-char pel- [10] Peng J, Bi H, Lim C, Sokhansanj S. Study on density, hardness, and moisture
uptake of torrefied wood pellets. Energy Fuels 2013;27:967–74.
lets densified with organic binders showed lower ash content and [11] Peng J, Bi H, Sokhansanj S, Lim J. Torrefaction and densification of different
more fixed carbon than the inorganic pellets. The hydration of species of softwood residues. Fuel 2013;111:411–21.
516 Q. Hu et al. / Applied Energy 157 (2015) 508–516

[12] Tooyserkani Z, Sokhansanj S, Bi X, Lim J, Lau A, Saddler J, et al. Steam treatment [28] Peleg M. Flowability of food powders and methods for its evaluation—a
of four softwood species and bark to produce torrefied wood. Appl Energy review. J Food Process Eng 1977;1:303–28.
2013;103:514–21. [29] Mani S, Tabil LG, Sokhansanj S. Effects of compressive force, particle size and
[13] Tooyserkani Z, Kumar L, Sokhansanj S, Saddler J, Bi XT, Lim CJ, et al. SO2- moisture content on mechanical properties of biomass pellets from grasses.
catalyzed steam pretreatment enhances the strength and stability of softwood Biomass Bioenergy 2006;30:648–54.
pellets. Bioresour Technol 2013;130:59–68. [30] Lam PY, Lam PS, Sokhansanj S, Bi XT, Lim CJ, Melin S. Effects of pelletization
[14] Han Y, Tahmasebi A, Yu J, Li X, Meesri C. An experimental study on binderless conditions on breaking strength and dimensional stability of Douglas fir pellet.
briquetting of low-rank coals. Chem Eng Technol 2013;36:749–56. Fuel 2014;117:1085–92.
[15] Kambo HS, Dutta A. Strength, storage, and combustion characteristics of [31] Liu Z, Quek A, Balasubramanian R. Preparation and characterization of fuel
densified lignocellulosic biomass produced via torrefaction and hydrothermal pellets from woody biomass, agro-residues and their corresponding
carbonization. Appl Energy 2014;135:182–91. hydrochars. Appl Energy 2014;113:1315–22.
[16] Duncan A, Pollard A, Fellouah H. Torrefied, spherical biomass pellets through [32] Kaliyan N, Morey RV. Natural binders and solid bridge type binding
the use of experimental design. Appl Energy 2012;101:237–43. mechanisms in briquettes and pellets made from corn stover and
[17] Peng J, Bi H, Sokhansanj S, Lim J. A study of particle size effect on biomass switchgrass. Bioresour Technol 2010;101:1082–90.
torrefaction and densification. Energy Fuels 2012;26:3826–39. [33] Reza MT, Lynam JG, Vasquez VR, Coronella CJ. Pelletization of biochar from
[18] Poddar S, Kamruzzaman M, Sujan SMA, Hossain M, Jamal MS, Gafur MA, et al. hydrothermally carbonized wood. Environ. Prog. Sustain. Energy 2012;31:
Effect of compression pressure on lignocellulosic biomass pellet to improve 225–34.
fuel properties: higher heating value. Fuel 2014;131:43–8. [34] Alvarez J, Lopez G, Amutio M, Bilbao J, Olazar M. Bio-oil production from rice
[19] Jiang L, Liang J, Yuan X, Li H, Li C, Xiao Z, et al. Co-pelletization of sewage husk fast pyrolysis in a conical spouted bed reactor. Fuel 2014;128:162–9.
sludge and biomass: the density and hardness of pellet. Bioresour Technol [35] Alvarez J, Lopez G, Amutio M, Bilbao J, Olazar M. Upgrading the rice husk char
2014;166:435–43. obtained by flash pyrolysis for the production of amorphous silica and high
[20] Nieto-Delgado C, Cannon FS, Paulsen PD, Furness JC, Voigt RC, Pagnotti JR. quality activated carbon. Bioresour Technol 2014;170:132–7.
Bindered anthracite briquettes as fuel alternative to metallurgical coke: full [36] Lam PS, Sokhansanj S, Bi X, Lim CJ, Melin S. Energy input and quality of pellets
scale performance in cupola furnaces. Fuel 2014;121:39–47. made from steam-exploded Douglas fir (Pseudotsuga menziesii). Energy Fuels
[21] Rubio B, Izquierdo M, Segura E. Effect of binder addition on the mechanical and 2011;25:1521–8.
physicochemical properties of low rank coal char briquettes. Carbon 1999;37: [37] Chen Y, Liu B, Yang H, Yang Q, Chen H. Evolution of functional groups and pore
1833–41. structure during cotton and corn stalks torrefaction and its correlation with
[22] Massaro M, Son S, Groven L. Mechanical, pyrolysis, and combustion hydrophobicity. Fuel 2014;137:41–9.
characterization of briquetted coal fines with municipal solid waste plastic [38] Zhao P, Ge S, Yoshikawa K. An orthogonal experimental study on solid fuel
(MSW) binders. Fuel 2013;115:62–9. production from sewage sludge by employing steam explosion. Appl Energy
[23] Diez MA, Alvarez R, Cimadevilla JLG. Briquetting of carbon-containing wastes 2013;112:1213–21.
from steelmaking for metallurgical coke production. Fuel 2013;114:216–23. [39] Bläsing M, Müller M. Investigation of the effect of alkali metal sorbents on the
[24] Bazargan A, Rough SL, McKay G. Compaction of palm kernel shell biochars for release and capture of trace elements during combustion of straw. Combust
application as solid fuel. Biomass Bioenergy 2014;70:489–97. Flame 2013;160:3015–20.
[25] Haykiri-Acma H, Yaman S, Kucukbayrak S. Production of biobriquettes from [40] Gil M, Oulego P, Casal M, Pevida C, Pis J, Rubiera F. Mechanical durability and
carbonized brown seaweed. Fuel Process Technol 2013;106:33–40. combustion characteristics of pellets from biomass blends. Bioresour Technol
[26] Liu Z, Balasubramanian R. A comparison of thermal behaviors of raw biomass, 2010;101:8859–67.
pyrolytic biochar and their blends with lignite. Bioresour Technol 2013;146: [41] Wang C, Wang F, Yang Q, Liang R. Thermogravimetric studies of the behavior
371–8. of wheat straw with added coal during combustion. Biomass Bioenergy
[27] Lam PS, Lam PY, Sokhansanj S, Bi XT, Lim CJ. Mechanical and compositional 2009;33:50–6.
characteristics of steam-treated Douglas fir (Pseudotsuga menziesii L.) during
pelletization. Biomass Bioenergy 2013;56:116–26.