Energy for Sustainable Development 43 (2018) 1–14

Contents lists available at ScienceDirect

Energy for Sustainable Development

Pilot scale evaluation of fuel pellets production from garden

waste biomass
Priyabrata Pradhan a, Amit Arora a, Sanjay M. Mahajani b,⁎
a
Bio-Processing Laboratory, Centre for Technology Alternatives for Rural Areas, Indian Institute of Technology Bombay, Mumbai 400076, India
b
Chemical Engineering Department, Indian Institute of Technology Bombay, Mumbai 400076, India

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

Article history: Pelletization of garden waste, without additional binder, was investigated to produce high quality fuel pellets for
Received 21 June 2017 energy utilization. The influence of pelletization parameters viz. feedstock moisture content (5, 15, 25, and 35%),
Revised 29 November 2017 milling size (25.4 and 6.25 mm) and die size (12 and 15 mm) on pellet quality and pelletization process was stud-
Accepted 29 November 2017
ied. The results showed that the studied parameters had significant effect on pellet quality. A reduction in average
Available online xxxx
durability value (95.0% to 92.5%) was observed when moisture content of garden waste increased from 5% to 15%.
Keywords:
Appropriate regression models were also developed for each quality attribute by using multiple linear regres-
Biomass feedstock sions. Eventually, a feedstock moisture content of 5 ± 1%, milling size of 6.35 mm and die size of 15 mm were
Garden waste found to offer standard quality pellets with optimum throughput capacity. Scanning electron microscopy
Pelletization image analysis showed a closer agglomeration of biomass particles when feed materials were pelletized at low
Pellet quality moisture content. The equilibrium moisture content due to adsorption for garden waste pellet was found to be
Sustainable energy 14.6% which was quite low as compared to other feedstocks. Furthermore, we deduced from the combustion
test that garden waste pellets may be conveniently used in a residential cookstove. In a nutshell, pelletization
of garden waste biomass has been demonstrated at pilot scale in this study.
© 2017 International Energy Initiative. Published by Elsevier Inc. All rights reserved.

Introduction
Densification of biomass through pelletization allows one to use
carbon neutral fuel with higher energy density thereby reducing the
transportation, handling and storage costs (Tumuluru, Wright, Hess, &
Kenney, 2011; Miao, Grift, & Ting, 2014). Fuel pellets have various
applications, ranging from residential stoves to full-scale power plants
(Holm, Henriksen, Hustad, & Sørensen, 2006; Moya, Rodríguez-Zúñiga,
Tenorio, Valdez, & Valaert, 2015; Chen et al., 2016). The commonly
used biomass feedstock is wood and the market of wood biomass pellets
has shown an exponential rise of 0.5 million tonnes (8.5 Peta Joule) to
6.6 million tonnes (120 Peta Joule) between 2000 and 2010 (Lamers,
Junginger, Hamelinck, & Faaij, 2012); further the production is projected
to increase to 45.2 million tonnes in 2020 (Guo, Song, & Buhain, 2015).
The growing demand of fuel pellets has stimulated search for alternate
lignocellulosic biomass. After wood wastes, agricultural residues are
being considered as potential feedstock available in substantial quanti-
ties. Several researchers have investigated pelletization of agricultural
residues using binders such as starch, wood powder, lignosulphate etc.
to produce highly durable pellets (Serrano, Monedero, Lapuerta, &
Portero, 2011; Mediavilla, Esteban, & Fernández, 2012; Said, Abdel
⁎ Corresponding author.
E-mail addresses: ppradhan@iitb.ac.in (P. Pradhan), aarora@iitb.ac.in (A. Arora),
sanjaym@che.iitb.ac.in (S.M. Mahajani).
Daiem, Garcia-Maraver, & Zamorano, 2015; Stasiak et al., 2017). Howev-
er, external binder may add to the production cost of the pellets (Jiang
et al., 2016). In contrast, garden waste i.e. leaf litters from various tree
species, has natural binder such as lignin (Kaliyan & Morey, 2010)
which favours solid bridge formation during the pelletization process
and hence, the additional cost of binder can be avoided.
In recent times, garden waste is disposed off in open dumps and
landfills, which raises socio-environment problems. According to an es-
timate, the cities in China namely, Beijing and Shanghai, generate about
2.5 million tons of garden waste per year (Shi, Ge, Chang, Shao, & Tang,
2013) and in New Delhi, India, the garden waste shares up to 23% of
total landfill waste i.e. 1150 tonnes per day (Chakraborty, Sharma,
Pandey, & Gupta, 2013). Thus, garden waste, a potentially huge under-
utilised biomass can be considered as an alternate guaranteed biomass
feedstock to pellet manufacturers relying on wood and agricultural
residues (Lamers et al., 2012). Therefore, there is a need to study the
pelletizer performance and quality of pellets produced using garden
waste which in itself is a unique raw material when its chemical compo-
sition is compared with standard pellets made from wood sawdust or
homogeneous agricultural residues such as wheat straw, and corn
stover (Theerarattananoon et al., 2011).
Various pellet quality standards have been prescribed to maintain ho-
mogeneity of pellets both at national and international markets
(Table A.1). These pellet properties depend on the type of feedstock,
type and quantity of additives, machine specific parameters and expertise

https://doi.org/10.1016/j.esd.2017.11.005
0973-0826/© 2017 International Energy Initiative. Published by Elsevier Inc. All rights reserved.
2 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14
of operator (for manually operated machines). Among the feedstock pa-
rameters, moisture content is an important parameter as it regulates the
friction between the compression channel and the feedstock (Stelte et al.,
2011a). There is substantial data available on the desired moisture con-
tent range for woody biomass, e.g. 11.2% for sugar maple particles
(Nguyen, Cloutier, Achim, & Stevanovic, 2015) and 10% for olive tree
pruning residues (Carone, Pantaleo, & Pellerano, 2011). High moisture
pelletization with various types of biomass feedstocks were also evaluat-
ed by researchers; for example 40% moisture content for compost (Zafari
& Kianmehr, 2013), 11 to 41% for wood sawdust (Poddar et al., 2014), 38%
for corn stover (Tumuluru, 2015), and 60–80% for crop species like
S. bicolor, P. purpureum and A. donax (Aragón-Garita, Moya, Bond,
Valaert, & Filho, 2016). Further, for a city where the rainy season prevails
almost 4 months a year, coupled with high relative humidity region (for
example, Mumbai in India), it is difficult to get dry garden waste through-
out the year. Therefore, in the present investigation on garden waste, a
wide range of the moisture content (5–35%) has been considered.
Milling size of feedstock is also important as larger size results in more
voids in the press channel, and reduces bonds interlocking significantly
(Kirsten, Lenz, Schröder, & Repke, 2016). Mani, Tabil, and Sokhansanj
(2006) reported the ideal milling size of 3.2 mm at 12% moisture content
to produce corn stover pellets. Next, the die geometry, i.e. die hole diam-
eter and channel length, also affects the throughput capacity of pellet
mill, pellet quality and other process requirements (Theerarattananoon
et al., 2011). Holm et al. (2006) developed a model to explain the influ-
ence of channel length and die size on pelletization pressure. Tumuluru
et al. (2011) also reviewed the influence of die geometry on the quality
of the densified material. They reported that the durability of pellets im-
proved with smaller die of higher L/D ratios. Theerarattananoon et al.
(2011) reported that use of a thicker die (44.5 mm instead of 31.8 mm
die thickness) resulted in a significant increase in pellet durability for
wheat straw pellets, corn stover pellets, and sorghum stalk pellets. Fur-
ther, a single pellet press cannot simulate all aspects such as die choking,
fines generation, and throughput capacity of a large scale pelletization
plant. Thus, a pilot scale study is needed to examine the effects of
above process parameters before scale up scenarios can be evaluated.
Only few systematic studies in pelletization of biomass are reported in lit-
erature (Mediavilla et al., 2012; Monedero, Portero, & Lapuerta, 2015;
Jackson, Turner, Mark, & Montross, 2016; Kirsten et al., 2016) and to
the best of our knowledge, pelletization of garden waste as a feedstock
has still not attracted attention of researchers in this area.
The overall objective of this study was to evaluate the fuel pellets
production from garden waste at pilot scale. Specific objectives were
(1) to investigate the influence of pelletization parameters on pellets
quality and pelletization process; (2) to examine the combustion
behaviour of produced fuel pellets in a residential cookstove.
The article is organized as follows: First we describe the experimen-
tal protocol used to assess the feedstock characterization, pellet quality
and pelletizer performance, including micro-structural analysis, mois-
ture adsorption test and combustion behaviour. This is followed by
the development of a correlation between the pelletization parameters
and pellet quality using statistical approach. This section also includes
results obtained from scanning electron microscopy analysis, moisture
adsorption test and combustion experiment.
Materials and methods
Feedstock
Fresh Garden waste was collected from the Indian Institute of Tech-
nology, Bombay (IIT B) campus, Mumbai, India. The biomass was
comprised of leaf litters of Mango (Mangifera indica), Raintree (Siamese
cassia), Jackfruit (Artocarpus heterophyllus), and other sticks and grasses.
Around 3 tons of the mixed waste was prepared and homogeneity of
the feedstock was maintained throughout the experiment. The typical
composition of mixed garden waste was: Mango leaves (15.5%), Raintree
leaves (34.7%), Jackfruit leaves (29.3%), grasses (2.2%) and miscellaneous
matter containing sticks and barks (18.3%).
Characterization
Proximate analysis was performed using a thermogravimetric
analyser (TGA) (Model: Perkin Elmer-SII). The elemental composition
(C, H, N, S, and O) of biomass was determined by a CHNS (O) analyser
(Make: Thermo finnigan, Italy; Model: FLASH EA 1112 series). The lig-
nocellulosic weight percentages were estimated from neutral detergent
fiber (NDF), acid detergent fiber (ADF) and acid insoluble lignin (AIL)
content of biomass. The NDF and ADF were estimated using Van-soest
detergent fiber analytical techniques (Van Soest, Robertson, & Lewis,
1991) and AIL was determined by following NREL protocol for structural
carbohydrates and lignin (Sluiter et al., 2008). The higher heating value
(HHV) of biomass was obtained using an oxygen bomb calorimeter
(Model: IKA C 200).
Process description
A process flow diagram of the pelletization process from garden
waste is described in Fig. 1. Soils and stones present in garden waste
were separated using a soil separator. Two stationary hammer mill
type shredders i.e. a coarse shredder (Make: CRPL, mill screen size:
25.4 mm, power requirement: 15 hp, capacity: 160–180 kg h−1) and a
fine shredder (Make: CRPL, mill screen size: 6.35 mm, power require-
ment: 7.5 hp., capacity: 80–100 kg h−1) were used to reduce particle
size. During the experiment, moisture content in the garden waste
was raised by sprinkling water over the biomass and it was lowered
by using a rotary drier as per requirement. The conditioned garden
waste was then kept overnight to reduce the non-uniformity of
moisture. Garden waste (bulk density of 80–120 kg m−3) was fed to
the pelletizer using a screw conveyor to ensure uniform rate of feeding
during the run. Before each test, 5 kg of oily biomass was used to preheat
the die. Oily biomass was prepared by mixing 500 ml of furnace oil with
5 kg of garden waste. The die was preheated to 80–90 °C due to friction-
al heat by flushing the oily material in through the channel and subse-
quently the biomass was fed to pelletizer for smooth operation. Just
before the completion of each run, the oily biomass was again fed to
the pelletizer to prevent choking of press channel in the next run.
Each run was performed to densify 50 kg of garden waste which lasted
for about an hour and pellet samples were collected at an interval of
15 min. Detailed specifications of the pelletizer are given in Table 1.
Measurements of pellet quality
The pellets produced were allowed to cool at ambient temperature
before measuring quality parameters for a representative sample as
per the ASABE Standards S269.4 (ASABE, 2007).
Moisture content (%)
The moisture content was measured using a digital moisture meter
(Make: Amprobe; Model: MT10; Measurement range: 8–60%; Resolu-
tion: 0.1%; Accuracy: ± 2%) after 2 h once the pellets are made. The
moisture contents of samples were also measured intermittently using
oven drying method (105 °C for 24 h). Measurements were repeated
twice to report the average moisture content.
Pellet diameter and length (mm)
The pellet diameter and length were measured using a vernier
calliper (Make: Aerospace, Range: 0–150 mm, Accuracy: 0.02 mm)
and average value was calculated from ten random samples of each
run. For this, the pellets were sampled in every 15 min during a run.
 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14 3

Fig. 1. Process flow diagram of a garden waste pelletization process.

Particle density (kg m−3) was tumbled in a dust tight tumbler rotated at 50 rpm for 10 min. The
Particle density was calculated as the ratio of the pellet mass to pellet experiment was repeated twice to report the average value of durability.
volume. The mass of ten randomly selected individual pellets was
measured using a digital weighing balance. The pellet volume was cal- Hardness (kgf)
culated from the measured dimensions. The average particle density Hardness is defined as the maximum crushing resistance of a pellet.
was reported from ten measurements for each run. Hardness was determined using a Texture analyser (Make: Stable Micro
Systems; Model: TA.XT Plus; Force capacity: 100 kgf). A pellet was kept
on a horizontal platform of a Texture analyser and a stainless steel probe
Durability (%)
of 5 mm diameter was used to pierce the pellet at a specific speed.
Durability is expressed as the percentage ratio of the mass of pellets
The minimum force required to crush a pellet was determined and the
retained on the sieve after tumbling to the mass of pellets before
average hardness was reported based on the two tests for a given run.
tumbling. A tumbler was made as per the specification given in ASABE
Standards (ASABE, 2007). A representative sample of 500 g for each run
Measurements of performance parameters
Table 1
Detailed specifications of a flat die pelletizer. Throughput capacity (kg h−1)
The throughput capacity of pellet mill may be defined as the
Particulars Specification
maximum amount of pellets produced per hour without any choking
Make Gangotree
in the machine. This parameter was measured by weighing the pellets
Rated power (hp) 10
Output capacity (kg h−1) 60–90
produced in every 15 min during a continuous run. The average value
was determined by taking two measurements in each run.
Rollers
Width (mm) 35
Diameter (mm) 120 Fines at mill gate (%)
In a pellet mill, it is common to realize some fines along with the
Die specifications (die 1)
Channel length (mm) 40 pellets. The fines percentage is defined as the ratio of weight of fines
Die hole diameter (mm) 12 generated to the weight of biomass fed in a trial of 15 min during a
continuous run. The average value was determined by taking two
Die specifications (die 2)
Channel length (mm) 40 measurements in each run.
Die hole diameter (mm) 15

Overall dimensions Statistical analysis

Length (mm) 980
Breadth (mm) 630 Total 16 runs were performed using full factorial design of experi-
Height (mm) 1325 ment to investigate the influence of moisture content of feed (MCF),
Net machine weight (kg) 200
milling size (S), and Die size (D) on the pellet quality. The MCF values
4 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14

(5 ± 1, 15 ± 1, 25 ± 2, and 35 ± 2%) are coded as 1, 2, 3, 4 respectively
and similarly, S values (6.35 and 25.4 mm) as 1, 2 and D values (12 and
15 mm) as 1, 2 (Table 2).
The influence of the parameters, MCF, S and D was analysed using
the statistical software Minitab version 17 by analysis of variance
(ANOVA). The relationship between predictor variables (MCF, S, and
D) and the responses (quality parameters) were evaluated by multiple
linear regressions. The regression model in Eq. (1) describes the regres-
sion coefficients (β0–β10), input parameters (MCF, S, D), their quadratic
terms (MCF2, S2, D2) and interaction terms (MCF × S, MCF × D, S × D,
MCF × S × D) with response variable Z.
Z ¼ β0 þ β1  MCF þ β2  S þ β3  D þ β4  MCF 2 þ β5  S2 þ β6
 D2 þ β7  MCF  S þ β8  MCF  D þ β9  S  D þ β10
 MCF  S  D ð1Þ
The models for all the responses were formulated separately and
linear, quadratic and interaction terms, which were not statistically
significant (p N 0.05) were excluded from the models. These models
were further used to study the main and interaction effects to under-
stand the influence of the process variables on the pellet quality.
Scanning electron microscopy (SEM)
Pellets obtained in four different runs (Run 2, 8, 12, 16) were selected
for their analysis through a SEM. The pellets produced from garden waste
were transverse fractured manually, and SEM images were obtained to vi-
sualize the morphological changes of biomass particles after pelletization.
All samples were wrapped from backside using an aluminium foil to en-
able mounting and then attached to metal stubs. Images were obtained
with a microscope (Quanta 200, FEI) operated at 15.0 kV of acceleration
voltage and 106 μA of beam current. Back scattered electron detector
was used for image analysis.
Moisture adsorption test
Moisture adsorption test is a measure of effect of air relative humid-
ity on the pellet produced. Pellets were dried in an oven at 105 °C for
24 h and then placed in a climatic chamber (Pooja Lab Equipments,
India; Model: PLE 106 ICH) at 30 °C and 90% of relative humidity
according to the procedure described in literature (Li et al., 2012;
Puig-Arnavat, Shang, Sarossy, Ahrenfeldt, & Henriksen, 2016). The
moisture content due to adsorption was measured as the weight differ-
ence between the initial dry weight and equilibrated weight after mois-
ture adsorption in the climatic chamber. The weight of pellets was
measured every 20 min for the first 2 h followed by every 30 min for
the next 4 h. After 6 h, the weight was measured by a few hours until
there was no significant change in the weight.
Combustion experiment
Combustion experiment was carried out on fuel pellets produced
along with other biomasses (wood chips and fine shredded garden
waste) using a residential biomass cookstove (Make: Gangotree Eco
Tech Pvt. Ltd.; Model: Samrat). The forced draft cookstove shown in
Fig. 2 had a combustion space of 230 mm diameter and 270 mm
depth with a grate at the bottom. The construction and operational
details of a similar cookstove can be found elsewhere (Varunkumar,
Ranjan, & Mukunda, 2012). The cookstove was filled with test fuels
and kept on a digital weighing balance to observe the change in weight
loss with respect to time while the temperature was monitored using
a K type thermocouple. The experiments were conducted in two
replicates (90 min each) for each biomass and consistency in cookstove
operation was maintained throughout according to the procedure
described by Larsson, Lockneus, Xiong, and Samuelsson (2015). The
combustion residual of different test fuels was analysed by visual
inspection and weighing.
Results and discussion
Characterization of garden waste
The data obtained from the characterization of garden waste is pre-
sented in Table 3. The volatiles percentage in garden waste was 67.1 ±
1.2% which strongly influences the thermal decomposition behaviour of
the produced pellet (Molina-Moreno, Leyva-Díaz, & Sánchez-Molina,
2016). The carbon content and hence the calorific value (16.5 ±
0.1 MJ kg−1) of garden waste was comparable with any other biomass
fuels such as agricultural and forestry crops (Tenorio, Moya, Filho, &
Valaert, 2015a). Among chemical components, the presence of lignin
(23.3 ± 0.6%) in garden waste might enhance the pellet quality and
pelletization process because lignin acts a natural binder.

Table 2
Pelletization parameters and level codes for the experimental design.

Run Pelletization parameters Level codes

MCFa Sb Dc MCF S D

(%) inch (mm) (mm)

1 5±1 ¼ (6.35) 12 1 1 1
2 5±1 ¼ (6.35) 15 1 1 2
3 5±1 1 (25.4) 12 1 2 1
4 5±1 1 (25.4) 15 1 2 2
5 15 ± 1 ¼ (6.35) 12 2 1 1
6 15 ± 1 ¼ (6.35) 15 2 1 2
7 15 ± 1 1 (25.4) 12 2 2 1
8 15 ± 1 1 (25.4) 15 2 2 2
9 25 ± 2 ¼ (6.35) 12 3 1 1
10 25 ± 2 ¼ (6.35) 15 3 1 2
11 25 ± 2 1 (25.4) 12 3 2 1
12 25 ± 2 1 (25.4) 15 3 2 2
13 35 ± 2 ¼ (6.35) 12 4 1 1
14 35 ± 2 ¼ (6.35) 15 4 1 2
15 35 ± 2 1 (25.4) 12 4 2 1
16 35 ± 2 1 (25.4) 15 4 2 2
a
MCF: Moisture content of feed.
b
S: Milling size.
c
D: Die size. Fig. 2. Schematic of a residential cookstove with fuel pellets inside.
 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14 5

Table 3 Table 5
Characterization of garden waste and fuel pellet. Regression models for pellet quality parameters.

Property Analytical method (1) Garden waste (2) Fuel pellet Serial Models R2

Moisture contenta ASTM E790 10.0 ± 0.1 10.1 ± 1.7 Model 1 Moisture content = −4.1 + 8.4 MCF 0.95
Model 2 Diameter = 8.5–0. 4 MCF + 2.6 D + 0.5 MCF × D 0.98
Proximate analysis (% of dry fuel) Model 3 Length = 51.8–3.6 MCF − 3.1 D 0.79
Volatile matter ASTM E897 67.1 ± 1.2 66.7 ± 0.7 Model 4 Particle density = 1628–11.1 MCF + 22 D−142 MCF × D 0.95
Fixed carbon by difference 18.4 ± 2.2 18.6 ± 0.8 Model 5 Durability = 97.2 + 8.7 MCF−3.7 S−0.1 D−3.1 MCF × 0.99
Ash content ASTM E830 14.5 ± 3.2 14.5 ± 1.5 MCF−1.4 MCF × D
Ultimate analysis (% of dry fuel) Model 6 Hardness = 46.1–9.7 MCF + 15.1 S−12.1 D 0.82
Carbon ASTM E777 38.0 ± 1.8 40.0 ± 0.3 Model 7 Throughput capacity = −5.6 + 8.9 MCF + 22.2 D 0.61
Hydrogen ASTM E777 4.4 ± 0.1 5.5 ± 0.6 Model 8 Fines generated = −38.6 + 37.7 S + 20.7 D−17.9 S × D 0.61
Nitrogen ASTM E778 1.1 ± 0.0 0.7 ± 0.2
Sulphur ASTM E775 nil nil
Oxygen (diff.) by difference 42.0 ± 0.5 39.3 ± 0.6
pellet moisture was noticeably influenced by feed moisture according to
Ash content ASTM E830 14.5 ± 3.2 14.5 ± 1.5
Table 4; because when the feed moisture is high, it cools down the die
Chemical analysis (wt% on dry basis) due to evaporation, thereby inhibiting further evaporation at reduced
Hemicellulose Van Soest method 7.9 ± 1.3 –
temperature which results in high moisture pellets (Monedero et al.,
Cellulose Van Soest method 12.0 ± 0.4 –
Ligninb NREL/TP-510-42,618 23.3 ± 0.6 – 2015). The results indicated that after pelletization the pellet moisture
HHVc (MJ kg−1) ASTM E711 16.5 ± 0.1 16.8 ± 0.1 was reduced by about 1 to 6% of the feed moisture present. Serrano
a
As received basis (%).
et al. (2011) also found similar trend while densifying barley straw in
b
Acid insoluble lignin. a semi-industrial pellet mill and noticed that high feed moisture with
c
Higher heating value. the cell structure makes particles harder to compress. The pellets pre-
pared using garden waste with low level of MCF (5 and 15%) were char-
acterized by a shiny surface while those prepared at high level of MCF
Pellet quality and pelletizer performance (25 and 35%) had cracked surfaces (Fig. 4). These cracks can be
interpreted as lack of inter particle bond among the biomass particles.
The pellet quality for different runs (Table 4) was assessed statisti- This might be due to the reduction of natural binder's influence as
cally using the model. The regression models for different quality well as a possible presence of solid-liquid biphasic mixture wherein,
parameters and their R2 values are presented in Table 5. The models there is a complete absence of intermolecular forces (Zafari &
showed that feed moisture content and die size were the major influen- Kianmehr, 2013). To meet the ISO and European standards, the pellet
tial pelletization parameters. Results obtained in the present work were moisture content should be below 10%. All the norms set for moisture
compared with the standard values established for non-woody pellet content of pellet are fulfilled by Run 1, Run 2, Run 3 and Run 4 wherein,
norm to select ideal operational conditions to produce high quality the moisture content of feed was below 10%.
pellets (Table A.1).
Pellet diameter
Small diameter pellets offer uniform combustion rate as compared
Moisture content to the large diameter pellets due to increase in exposed surface area
The moisture content of pellet influences the calorific value, (Said et al., 2015). The average diameter of pellet produced was 11.5
combustion behaviour, combustion efficiency (Obernberger & Thek, and 15.5 mm for die sizes of 12 and 15 mm, respectively. From Fig. 3B
2004) and also creates conducive environment for mold growth it can be noticed that the pellet diameter value has rapidly increased
(Lehtikangas, 2001). The pellet moisture content was significantly (p b after level 2 (i.e. when MCF = 15%) and this could be attributed to swell-
0.05) affected by feed moisture content (Model 1, Table 5). The decrease ing of biomass during high moisture pelletization. The swelling or
in pellet moisture at a particular feed moisture may be attributed to va- puffing of pellets may be due to release of water vapour from the
porization of water, which depends on the residence time of biomass in- warm pellets. Said et al. (2015) also reported slight increase in diameter
side the die during the pelletization process as shown in the Fig. 3A. The of rice straw pellets after their release from the die. The diameter of

Table 4
Pellets quality values (response variables) obtained from different runs.

Run Moisture content Diameter Length Particle density Durability Hardness Throughput Fines
−3 −1
(%) (mm) (mm) (kg m ) (%) (kgf) (kg h ) (%)

1 3.4 11.3 (0.1)a 43.8 (5.7) 1482 (58) 96.2 34.0 36.4 3.1
2 4.5 14.7 (0.2) 39.2 (5.0) 1397 (56) 97.7 24.5 58.8 5.9
3 4.3 11.4 (0.2) 46.2 (8.9) 1473 (39) 94.8 52.1 15.0 11.8
4 3.8 14.7 (0.3) 41.7 (5.4) 1302 (60) 91.4 54.5 45.3 1.1
5 13.6 11.4 (0.2) 44.1 (8.2) 1326 (18) 98.3 27.0 58.2 0.9
6 13.0 15.2 (0.1) 40.7 (8.2) 1002 (47) 92.3 18.8 45.0 4.7
7 14.0 11.3 (0.1) 45.1 (7.6) 1467 (26) 91.0 55.1 18.7 40.8
8 11.5 14.6 (0.2) 40.1 (6.8) 1262 (56) 89.4 32.2 39.0 10.3
9 18.0 11.7 (0.3) 36.8 (3.0) 1132 (86) 87.3 15.5 39.9 1.4
10 21.4 16.1 (0.7) 34.6 (3.8) 755 (98) 82.7 10.1 82.5 4.2
11 21.5 11.5 (0.2) 35.2 (4.6) 1151 (53) 83.2 24.9 44.5 12.2
12 24.0 15.5 (0.2) 32.1 (5.5) 778 (69) 80.0 13.1 63.6 6.3
13 26.0 11.9 (0.3) 36.4 (5.3) 1062 (35) 72.6 16.3 41.9 2.0
14 34.0 16.7 (0.3) 33.6 (5.0) 409 (65) 67.0 3.9 85.8 3.3
15 31.0 11.5 (0.6) 31.1 (6.3) 1041 (85) 70.1 34.2 55.5 21.9
16 26.0 16.6 (0.5) 32.4 (7.2) 566 (36) 64.6 4.7 67.3 8.4
a
Number enclosed in parenthesis are standard deviation for n = 10.
6 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14
 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14
Fig. 4. Garden waste pellets at different feed moisture content (MCF); viz. (A) 5 ± 1% (Run 2), (B) 15 ± 1% (Run 8), (C) 25 ± 2% (Run 12), and (D) 35 ± 2% (Run 16).
pellet produced was close to the die size (Table 4); as expected, the re-
gression model showed the highest coefficient for D (Model 2, Table 5).
Furthermore, the contour plots for selected quality parameters are
shown in Fig. 5 wherever, an interaction effect exists. The contour plot
(Fig. 5A) reveals the significance of interaction parameter MCF × D
wherein the slope of the contour lines increased with increase in die
size. That means the influence of feed moisture content on pellet diam-
eter is more at larger die size. As per the European and ISO standards,
the pellet diameter is limited to ‘die diameter ± 1.00 mm’. As expected,
the pellet diameter values are not satisfactory as per the ISO in case of
Run 10, Run 14 and Run 16 because these runs were performed beyond
25% MCF.
Length
Uniform length pellets help in easier flow of material inside the
reactor during combustion or gasification processes, and longer pellets
can be more easily broken (Said et al., 2015). The length to diameter
ratio of pellet is a very significant factor as far as pneumatic feeding
(Obernberger & Thek, 2004) and choking of the pellet mill are con-
cerned (Holm et al., 2006). Visual observation of the pellets illustrated
the difference in length (Fig. 4). Only MCF and D had significant effect
(p b 0.05) on pellet length which is also reflected in regression analysis
(Model 3, Table 5). The maximum and minimum lengths of pellet
produced were 46.2 mm (Run 3) and 31.1 mm (Run 15), respectively
(Table 4), where runs were conducted at moisture contents of 5% and
35%, respectively. This indicates that higher moisture could be negatively
affecting compaction which inhibits the particles agglomeration and re-
duces the pellet length (Fig. 3C). Zamorano, Popov, Rodriguez, and
Garcia-Maraver (2011) found similar trends for agricultural and forest
Fig. 3. Influence of moisture content of feed (MCF), milling size (S), and die size (D) on various pellet quality parameters such as (A) Moisture content, (B) Diameter, (C) Length, (D) Particle
density, (E) Durability, (F) Hardness, (G) Throughput capacity, (H) Fines generated.
7
lopping residues studied over the moisture content of 4.7 to 6.5%. Howev-
er, Serrano et al. (2011) reported opposite results while experimenting on
barley straw with MCF of 6 to 8%. The average pellet length reduced from
39.8 to 36.8 mm with an increase in die size (D) from 12 to 15 mm
(Fig. 3C), which is corroborated with the negative coefficient of D in
Model 3 (Table 5). As per the European standards, the pellet length should
be less than four times of the ‘die diameter’ (Table A.1). The European
norm set for the pellet length is fulfilled by all the runs in the present
work.
Particle density
Extremely high particle density influences the combustion and
torrefaction behaviours as the lack of voids restrict the oxygen move-
ment inside the pellet (Cao et al., 2015) and also shows a longer burnout
time (Obernberger & Thek, 2004). The average values of particle density
were 1414, 1265, 955, and 770 kg m−3 at corresponding feed moistures
of 5, 15, 25, and 35% (Fig. 3D). Similar trend was observed by Fasina
(2008) in the case of peanut hull pellets at the moisture content in the
range 4.2–21.2%. Serrano et al. (2011) reported reduction in particle
density of barley straw pellets from 1.4 ± 0.1 to 1.3 ± 0.1 g cm−3 at
straw moisture contents ranging over 12–20.2%. This could be due to
the volumetric expansion of the pellet at higher feed moisture content.
In the present investigation, the observed maximum and minimum
values of particle density were 1482 kg m−3 and 409 kg m−3 in Run 1
and Run 14, respectively (Table 4). Here, the Run 1 was conducted at
5% feed moisture content using 12 mm die while the Run 14 was per-
formed at 35% moisture with 15 mm die size. The regression model
showed that milling size (S) did not have significant effect on particle
density (Model 4, Table 5). Mani et al. (2006) also reported that the
8 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14

Fig. 5. Contour plots for various pellet quality parameters such as (A) Influence of MCF × D on diameter, (B) Influence of MCF × D on particle density, (C) Influence of MCF × D on durability,
(D) Influence of S × D on fines generated. [MCF: Moisture content of feed (%); S: Milling size (mm); D: Die size (mm)].

influence of milling size on particle density produced from wheat straw
was not significant. The negative coefficient of MCF × D indicates lack of
bonding and inter-particle interactions with simultaneous increase in
MCF and D. Thus the higher particle density can be obtained at lower
MCFs and small Ds (Fig. 5B). This might be due to the decrease in pellet-
ization pressure with increase in die hole diameter and feed moisture
content. The results of earlier studies indicate that the ideal value for a
high quality pellet is 1200 kg m−3 (Said et al., 2015), although the ISO
standard does not include any specific value for this property. Run 1,
Run 3 and Run 7 did not fulfil the upper limit of particle density
(1400 kg m− 3) as per the French recommendation and hence were
not considered as the best operating conditions.
Durability
Durability represents the resistance of pellets towards abrasion.
High pellet durability reduces problem in fuel feeding systems and
also reduces dust emissions during handling. The durability was
most markedly influenced by feed moisture content as can be seen in
Fig. 3E. The values of durability measured were 95.0, 92.8, 83.3 and
68.6% for the feed moisture contents of 5, 15, 25, and 35%, respectively.
These results support the findings by Monedero et al. (2015) for poplar
and pine sawdust, where a reduction in durability (93.9% to 92.5%) was
observed when higher moisture content (from 12% to 18%) feedstock
was tested. Furthermore, Larsson, Thyrel, Geladi, and Lestander (2008)
supported similar trend for reed canary grass beyond 15% moisture
content. As mentioned before, the excessive feed moisture reduces the
die temperature which inhibits the formation of agglomerate. The pres-
ence of free moisture causes cohesive forces between biomass particles
and high durability can be obtained at sufficiently high feed moisture
(Theerarattananoon et al., 2011). However, further increase in moisture
results in disintegration of pellets due to reduction in cohesive forces.
The average durability of pellets slightly reduced from 86.8 to 83.1%
with increase in milling size (S) from 6.35 to 25.4 mm. This might be
due to decrease in inter-particle distance which results in reduction of
 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14
short-range forces, which contribute to bonding without solid bridge
formation (Kaliyan & Morey, 2010). In contrast, Serrano et al. (2011)
did not notice any significant effect of milling size on durability of barley
straw pellets. The durability of pellets influenced by all pelletization
parameters is reflected in Model 5 (Table 5). The contour plot indicates
that lower die size and lower feedstock moisture content resulted in
higher durability (Fig. 5C). As per the European and the ISO standards
the pellet durability should be ≥97.5%. In this experiment, only Run 2
(durability: 97.7%) and Run 5 (durability: 98.3%) fulfilled the above
standards (Table 4). Here, Run 2 was performed at low feed moisture
(5% MCF) and 15 mm die size whereas, for Run 5 the conditions were
15% MCF and 12 mm die. It can be inferred from the results that highly
durable pellets can be obtained at slightly increased moisture by using a
smaller die. Here, the durability of garden waste pellet produced may be
improved further using additional binder.
Hardness
Hardness measures the crushing strength of a pellet. The hardness
values of pellets were found to be 41.3, 33.3, 15.9, and 14.8 kgf for the
feed moisture contents of 5, 15, 25, and 35%, respectively (Fig. 3F). The
decrease in pellet hardness is due to poor binding ability of particles at
high feed moistures. High moistures may lead to incompressibility of
biomass wherein, the voids are mostly filled with water. The regression
model shows that the linear terms (MCF and D) have inverse relation-
ship and mill size (S) has a direct relationship with hardness (Model
6, Table 5). An increase in surface temperature of pelletizer was noticed
when runs were performed at higher milling size (S). At elevated
temperature and pressure, glass transition happens where amorphous
materials (lignin and hemicellulose) change their state from glassy to
soft rubbery and form solid bridge after cooling (Kaliyan & Morey,
2010). That could be the reason why the average hardness increased
Fig. 6. Scanning electron microscope images of the transversal fracture (i.e. cross-section) of garden waste pellets at different feed moisture content (MCF); viz. (A) 5 ± 1% (Run 2), (B) 15 ± 1%
(Run 8), (C) 25 ± 2% (Run 12), and (D) 35 ± 2% (Run 16).
9
from 18.8 to 33.8 kgf with an increase in milling size (S) from 6.35 to
25.4 mm whereas, average hardness decreased from 32.4 to 20.2 kgf
with an increase in die size (D) from 12 to 15 mm (Fig. 3F). Said et al.
(2015) reported the ideal hardness value for high quality pellets to be
22 kgf.
Throughput capacity
The measured values of throughput capacity were 38.9, 40.2, 57.6,
and 62.6 kg h−1 for the feed moisture contents of 5, 15, 25, and 35%,
respectively (Fig. 3G). Ciolkosz et al. (2015) also obtained similar
trend during pelletization of switchgrass at moisture ranged from 5 to
30%. The possible reason for this trend is that moisture might be acting
as a lubricant, perhaps reducing the friction between the biomass parti-
cles or between biomass and die wall. In addition, the increase in feed
moisture resulted in a decrease in glass transition temperature and
thus increased the flowabilty of hemicelluloses and other extractives,
which resulted in reduction of pelletization pressure (Stelte et al.,
2011a). Therefore, it helped the garden waste to pass through the
press channel easily thereby giving high capacity. A recent paper by
Jackson et al. (2016) showed that miscanthus, switchgrass and wheat
straw resulted in high production capacity at 25% moisture content
whereas, the pellet production rate of corn stover decreased with in-
creased moisture. Run 14 (high MCF, low S, and high D) resulted highest
throughput capacity (85.8 kg h−1) whereas, in Run 3 (low MCF, high S,
and low D) pellet production capacity reduced to 15 kg h−1 (Table 4).
Furthermore, at increased throughput, problems like failure to form
pellets or poor quality pellets were encountered. This finding shows a
scope for optimisation for pellet quality with respect to throughput
capacity. The average throughput capacity increased from 38.8 to
60.9 kg h−1 with change in die size from 12 to 15 mm (Fig. 3G), and
the reason could be the high flow rate of garden waste through larger
10 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14
die size (that means higher cross sectional area). The regression model
showed that die size (D) has the highest coefficient value and positively
correlated (Model 7, Table 5). The milling size (S) and other interaction
parameters in the regression model were found to be insignificant (p N
0.05) as expected. Nevertheless, the low R2 value (= 0.61) indicates
throughput capacity cannot be explained solely on the basis of varia-
tions in the pelletization parameters.
Fines generated at mill gate
The fines formation in a pelletizer during the operation is a practical
problem, which was also observed in the present case. It may be noted
that fine formation (where feedstock simply passes through the die
without getting pelletized) results in instability in machine and fluctua-
tions in power demand. The fines formation increased from 3.2–14.1%
with increase in milling size (Fig. 3H), which is reflected in the highest
positive coefficient value for milling size (S), however, fines formation
showed no statistical correlation with MCF (Model 8, Table 5). The
fines generated decreased from 11.8 to 5.5% with change in die
size from 12 to 15 mm however, due to presence of interaction term
(S × D) the influence of die size (D) alone could not be explained. The
contour plot shows that the use of large die size (high D) and fine
biomass feed (small S) may reduce the fine generation (Fig. 5D). Fur-
thermore, a considerable influence of die size was observed at higher
milling size (large S). This might be due to better packing of feed in a
comparatively larger die size. A further shredding of feed could avoid
fines production in small die size. Further investigation is required to
prove this point. In the present investigation, the fine generation of
above 10% was observed in Runs 3, 7, 11, 15 (Table 4). These runs
were performed mostly with coarse biomass feed (S = 25.4 mm)
and small die size (D = 12 mm). It can be hypothesized that the dy-
namic plug is not able to form during those runs that resist the biomass
Fig. 7. Scanning electron microscopy (magnification at 600×) images of the transversal fracture (i.e. cross-section) of garden waste pellets at different feed moisture content (MCF); viz.
(A) 5 ± 1% (Run 2), (B) 15 ± 1% (Run 8), (C) 25 ± 2% (Run 12), and (D) 35 ± 2% (Run 16).
flow to develop sufficient back pressure. We observed that, during
those runs, rollers were often acting as a grinder while creating a dusty
and noisy environment.
In brief, pellets produced in Run 2 fulfils the ISO standard and other
norms, and at the same time, gave better throughput with less fines
generation during the trial. Therefore, the pellets produced in Run 2
(possibly the best) were characterized and the results are presented in
Table 3 (in the second column). The scanning electron microscopy anal-
ysis and moisture adsorption test were performed on pellets produced
in Run 2, 8, 12, 16 (all runs were taken using 15 mm die) to study the
microscopic characteristics.
Scanning electron microscopy (SEM) analysis
SEM images showed a significant influence of moisture content on
the degree of agglomeration in the pellets (Fig. 6). The pellet produced
at low moisture content showed the presence of smooth edge and tight-
ly packed surface (Fig. 6A and B) while the existence of voids and cracks
observed at increased moisture content indicating poor adhesion
among biomass particles (Fig. 6C and D). This might be due to the
presence of excess water that prevents bonding, reduce pelletization
temperature and release of water after pelletization resulting in cracks.
Stelte et al. (2011b) also observed significant gaps and voids among ad-
jacent particles in a straw pellet when compressed at 20 °C. Castellano
Gomez, Fernandez, Esteban, and Carrasco (2015) also observed that
high moisture content results in lower temperature and pressure during
pelletization. This in combination with the reduced frictional force
resulted in voids and big gap among the biomass particles. The rough
edges of pellet at higher moisture content might be due to spilling out
of water towards edges at high pressure during pelletization. Thus, the
increase in voids and cracks resulted in reduction of pellet durability
 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14 11

at high moisture content while the loose packing at the edges resulted
in greater throughput capacity.
A bigger magnification (600×) revealed that the biomass particles at
high moisture content preserved their shape and structure after pellet-
ization, while biomass materials were hugely deformed at low moisture
content (Fig. 7). This might be due to higher inter-diffusion of lignin
among biomass particles and therefore higher durability. Although,
the SEM technique was used in the present study, Tenorio, Moya,
Filho, and Valaert (2015b) used X-ray densitometry technique to visual-
ize the surface cracks and pellet quality instantly.

Moisture adsorption test

Moisture adsorption is an important parameter while considering pel-

let storage. The pellets obtained in Run 2 showed the lowest moisture
content due to adsorption (14.6%) and took longer time to reach that
point, among all the pellets tested after 90 h in the chamber (Fig. 8).
The shiny surface of pellets produced during Run 2, possibly due to
higher content of waxes, extractives and lignin on the pellets surface,
resulted in more hydrophobic characteristics. However, the pellets
produced at higher feed moisture (Run 8, 12, 16) showed early and
higher moisture content percentage due to higher water adsorption.
The reason could be due to the presence of cracks and voids in the pellet.
Note that, in above Runs, pellet were produced at different feed mois-
ture contents using a particular die (15 mm). The moisture content
due to adsorption of pellets produced from various feedstock has been
reported in other literatures; for example 19.72% and 17.98% for Tritica-
le and Sorghum respectively (Puig-Arnavat et al., 2016) and 20.73% and
13.6% for sawdust and torrefied sawdust (Li et al., 2012). So, the value
for moisture content due to adsorption of the garden waste pellets
favours good pellet storage potential as compared to those obtained
from other feedstocks.

Combustion results

The combustion behaviour of biomass fuels in a residential cookstove

can be seen in Fig. 9. From the ‘weight loss rate vs time’ curves for all feed-
stocks, it is evident that the rate of burning of wood chips was significant-
ly higher (7.2 kg min−1) and early when shredded biomass showed a
maximum of 1.7 kg min−1 and 2.9 kg min−1 for fuel pellets. This is pos-
sibly due to the difference in volatile content (66.7% for pellets vs 75.9%
for wood) and particle density (1397 ± 56 kg m−3 for pellets vs 580 ±
116 kg m−3 for wood) of both the biomasses. Comparing the ‘weight
loss vs time’ curve for all tested biomass fuels, the amount of residual

Fig. 9. Combustion behaviour of biomasses; (A) Wood chips, (B) Shredded biomass, and
(C) Fuel pellets.

obtained after 90 min of operation was found to be much higher

Fig. 8. Moisture content of produced pellets at different Runs (30 °C and 90% relative
humidity).
(49 wt%) in case of shredded biomass and the same was significantly re-
duced for fuel pellets (18.2 wt%). The reason could be due to the higher
combustion temperature obtained in case of fuel pellets compared to
that of shredded biomass. But still the amount of residual was higher
than those obtained from wood chips (8 wt%) and that was due to less
ash content of wood. Agglomeration of ash in case of fuel pellets
12 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14

Table 6
Combustion experiment results.

Assortment Unit Wood chips Shredded biomass Fuel pellets

Fuel characteristics
Moisture content % 11.0 ± 0.1 10.0 ± 0.1 10.1 ± 1.7
Size (length × diameter) mm × mm 100 × 40 6.35 (milling) 40 × 15
Bulk density kg m−3 255.0 ± 7.1 177.5 ± 3.5 617.0 ± 4.2
True density kg m−3 580 ± 116 – 1397 ± 56
Higher heating value MJ kg−1 15.5 ± 0.3 16.5 ± 0.1 16.8 ± 0.1
Ash content in fuel % 4.3 ± 0.3 14.5 ± 3.2 14.5 ± 1.5

Combustion characteristics
Combustion residual % of fuel 8.5 48.7 18.2
Ash content in residual % of residual 83.5 ± 1.2 68.0 ± 2.4 91.0 ± 1.6
Clinker material % of ash 0 0 20.8
Maximum combustion temperature °C 870 546 1150
Maximum rate of combustion wt% min−1 7.2 1.7 2.9
Combustion environment Clean Smoky Clean

combustion was also observed during the test. Certainly, high ash content
of fuel pellets may create hindrance in industrial application. Thus,
blending of garden waste with low ash and high energy content biomass
can be considered to produce fuel pellets of acceptable quality in terms of
ash content and heating value. The details of all biomass fuels tested
and the residual obtained after combustion experiment are presented in
Table 6. Low amount of unburned material indicates effective combus-
tion. Smoke, dust and fly ash emission along with intermittent
flame were observed during the combustion of shredded biomass.
Combustion of fuel pellets and wood appeared to be clean. Limousy,
Jeguirim, Labbe, Balay, and Fossard (2015) also did a performance and
emissions study using spent coffee ground and wood chip logs in a
residential stove.
Conclusions
The pellets that meet quality standards were successfully produced
from garden waste which has been demonstrated at pilot scale for the
first time. The pellets produced in Run 2 fulfilled the ISO standard and
other norms, and at the same time, gave better throughput with less
fines generation during the trial. Hence, the pelletization conditions
for Run 2 may be considered to be the best i.e., feed moisture content:
5 ± 1%, milling size: 6.35 mm, and die size: 15 mm. The pellet quality
under such conditions were: pellet moisture content: 4.5%, diameter:
14.7 mm, length: 39.2 mm, particle density: 1397 kg m−3, durability:
97.7%, hardness: 24.5 kgf. The following conclusions can be highlighted
from the results:
• Pellet moisture reduced by 1 to 6% compared to that of feed moisture
content (5–35%) after pelletization. Pellets produced at 5 and 15% feed
moisture content showed shiny surface, while those produced at 25
and 35% showed cracked surface.
• Pellet diameter increased by 4.6% and length reduced by 18.4% as the
average feed moisture content increased from 15 to 25%. The average
pellet length also reduced by 7.5% with an increase in die size from 12
to 15 mm.
• Feed moisture content and die size had a noticeable effect on particle
density and durability, with high quality pellet production occurring
at average feed moisture content of 5%.
• It was observed that throughput capacity increased by 43.2% with an
increase in average feed moisture from 15 to 25% whereas, the fines
generation increased by 3.2 to 14.1% with an increase in milling size
from 6.35 to 25.4 mm.
• Furthermore, SEM images confirmed inter-diffusion of lignin inside
the biomass particle for the pellets produced at low moisture content,
thereby imparting a higher durability. The moisture adsorption tests
revealed the hygroscopicity of fuel pellets.
• Combustion of fuel pellets appeared to be clean and efficient, and
comparable to that of wood chips.
Thus, garden waste can be considered as an alternate feedstock for
fuel pellets production and simultaneously, it can be an option to
address the waste management problem as well. However, more tests
on large scale energy utilization, and techno-economic analysis for this
approach should be performed to ascertain its feasibility at an industrial
scale.
Acknowledgements
This research has been funded by the Tata Centre for Technology and
Design, Indian Institute of Technology Bombay, India, through project
DGDON 422. Priyabrata Pradhan (Roll no: 144350002) is also thankful
to the Ministry of Human Resource Development, Government of India
for providing Ph.D. fellowship.

Appendix A. Appendix

Table A.1
Various standards and guidelines for high quality pellets.

Austrian Swedish German Italian French European International

standards standards standards standards recommendations standards standards

Quality parameter ÖNORM M 7135a,b SS 187120a,b DIN EN 15270b CTI-R04/05 A.2b ITEBE (Stove)b CEN/TS 14588b ISO 17225-6c,d

Diameter, D (mm) 4–10 b25 –e 6 ± 0.5–8 ± 0.5 6±1 D12: D ≤ 12 ± 1.0 D±1
Length, L (mm) b5D b4D b5D b5D 10–30 b4D 3.15 ≤ L ≤ 40.00
Particle density (kg dm−3) b1.12 – b1.2 – 1.2–1.4 – –
Bulk density (kg m−3) – N600 – 600–720 N650 – ≥600
Durability (%) N97.7 N99.2 N97.7 N99 – ≥97.5 ≥97.50
Moisture content (%) b10 b10 b10 b10 b10 ≤10 ≤10.00
Ash content (%) b0.5 b0.7 b0.5 b1.5 b10 ≤0.7 –
Heating value (kcal kg−1) N4302 N4039 N4302 N3870 N4052 ≥4042 –
 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14 13

Table A.1 (continued)

Austrian Swedish German Italian French European International

standards standards standards standards recommendations standards standards

Quality parameter ÖNORM M 7135a,b SS 187120a,b DIN EN 15270b CTI-R04/05 A.2b ITEBE (Stove)b CEN/TS 14588b ISO 17225-6c,d

N (%) b0.3 – b0.3 ≤0.3 b0.3 – –

S (%) b0.04 b0.08 b0.04 ≤0.5 b0.08 ≤0.05 –
Cl (%) b0.02 b0.03 b0.02 – b0.3 – –
Additives (%) b2 – b2 – – b2 Informative
As (mg kg−1) – – b0.8 – – – –
Cd (mg kg−1) – – b0.5 – – – –
Cr (mg kg−1) – – b8 – – – –
Cu (mg kg−1) – – b5 – – – –
Hg (mg kg−1) – – b0.05 – – – –
Pb (mg kg−1) – – b10 – – – –
Zn (mg kg−1) – – b100 – – – –
a
Obernberger & Thek, 2004.
b
Garcia-Maraver, Popov, & Zamorano, 2011.
c
Said et al., 2015.
d
Kirsten et al., 2016.
e
−: No norms available.

References
Aragón-Garita, S., Moya, R., Bond, B., Valaert, J., & Filho, M. T. (2016). Production and
quality analysis of pellets manufactured from five potential energy crops in the
northern region of Costa Rica. Biomass and Bioenergy, 87, 84–95. https://doi.org/10.
1016/j.biombioe.2016.02.006.
ASABE Standards, S269.4 (2007). Cubes, pellets, and crumbles – Definitions and methods for
determining density, durability, and moisture content. St. Joseph, MI: ASABE.
Cao, L., Yuan, X., Li, H., Li, C., Xiao, Z., Jiang, L., et al. (2015). Complementary effects
of torrefaction and co-pelletization: energy consumption and characteristics of
pellets. Bioresource Technology, 185, 254–262. https://doi.org/10.1016/j.biortech.
2015.02.045.
Carone, M. T., Pantaleo, A., & Pellerano, A. (2011). Influence of process parameters and
biomass characteristics on the durability of pellets from the pruning residues of
Olea europaea L. Biomass and Bioenergy, 35, 402–410. https://doi.org/10.1016/j.
biombioe.2010.08.052.
Castellano, J. M., Gomez, M., Fernandez, M., Esteban, L. S., & Carrasco, J. E. (2015). Study on
the effects of raw materials composition and pelletization conditions on the quality
and properties of pellets obtained from different woody and non woody biomasses.
Fuel, 139, 629–636. https://doi.org/10.1016/j.fuel.2014.09.033.
Chakraborty, M., Sharma, C., Pandey, J., & Gupta, P. K. (2013). Assessment of energy
generation potentials of MSW in Delhi under different technological options. Energy
Conversion and Management, 75, 249–255. https://doi.org/10.1016/j.enconman.
2013.06.027.
Chen, Y., Shen, G., Su, S., Du, W., Huangfu, Y., Liu, G., et al. (2016). Efficiencies and
pollutant emissions from forced-draft biomass-pellet semi-gasifier stoves:
Comparison of international and Chinese water boiling test protocols. Energy for
Sustainable Development, 32, 22–30. https://doi.org/10.1016/j.esd.2016.02.008.
Ciolkosz, D., Hilton, R., Swackhamer, C., Yi, H., Puri, V. M., Swomley, D., et al. (2015).
Farm-scale biomass pelletizer performance for switchgrass pellet production.
Applied Engineering in Agriculture, 31, 559–567. https://doi.org/10.13031/aea.31.
10803.
Fasina, O. O. (2008). Physical properties of peanut hull pellets. Bioresource Technology, 99,
1259–1266. https://doi.org/10.1016/j.biortech.2007.02.041.
Garcia-Maraver, A., Popov, V., & Zamorano, M. (2011). A review of European standards for
pellet quality. Renewable Energy, 36, 3537–3540. https://doi.org/10.1016/j.renene.2011.
05.013.
Guo, M., Song, W., & Buhain, J. (2015). Bioenergy and biofuels: History, status, and per-
spective. Renewable and Sustainable Energy Reviews, 42, 712–725. https://doi.org/10.
1016/j.rser.2014.10.013.
Holm, J. K., Henriksen, U. B., Hustad, J. E., & Sørensen, L. H. (2006). Toward an understanding
of controlling parameters in softwood and hardwood pellets production. Energy & Fuels,
20, 2686–2694. https://doi.org/10.1021/ef0503360.
Jackson, J., Turner, A., Mark, T., & Montross, M. (2016). Densification of biomass using a pilot
scale flat ring roller pellet mill. Fuel Processing Technology, 148, 43–49. https://doi.org/10.
1016/j.fuproc.2016.02.024.
Jiang, L., Yuan, X., Xiao, Z., Liang, J., Li, H., Cao, L., et al. (2016). A comparative study of bio-
mass pellet and biomass-sludge mixed pellet: Energy input and pellet properties.
Energy Conversion and Management, 126, 509–515. https://doi.org/10.1016/j.
enconman.2016.08.035.
Kaliyan, N., & Morey, R. V. (2010). Natural binders and solid bridge type binding mecha-
nisms in briquettes and pellets made from corn stover and switchgrass. Bioresource
Technology, 101, 1082–1090. https://doi.org/10.1016/j.biortech.2009.08.064.
Kirsten, C., Lenz, V., Schröder, H. W., & Repke, J. U. (2016). Hay pellets - The influence of
particle size reduction on their physical-mechanical quality and energy demand
during production. Fuel Processing Technology, 148, 163–174. https://doi.org/10.
1016/j.fuproc.2016.02.013.
Lamers, P., Junginger, M., Hamelinck, C., & Faaij, A. (2012). Developments in international
solid biofuel trade - An analysis of volumes, policies, and market factors. Renewable
and Sustainable Energy Reviews, 16, 3176–3199. https://doi.org/10.1016/j.rser.2012.
02.027.
Larsson, S., Lockneus, O., Xiong, S., & Samuelsson, R. (2015). Cassava stem powder as an
additive in biomass fuel pellet production. Energy & Fuels, 29(9), 5902–5908.
https://doi.org/10.1021/acs.energyfuels.5b01418.
Larsson, S. H., Thyrel, M., Geladi, P., & Lestander, T. A. (2008). High quality biofuel pellet
production from pre-compacted low density raw materials. Bioresource Technology,
99, 7176–7182. https://doi.org/10.1016/j.biortech.2007.12.065.
Lehtikangas, P. (2001). Quality properties of pelletised sawdust, logging residues and bark.
Biomass and Bioenergy, 20, 351–360. https://doi.org/10.1016/S0961-9534(00)00092-1.
Li, H., Liu, X., Legros, R., Bi, X. T., Lim, C. J., & Sokhansanj, S. (2012). Pelletization of torrefied
sawdust and properties of torrefied pellets. Applied Energy, 93, 680–685.
Limousy, L., Jeguirim, M., Labbe, S., Balay, F., & Fossard, E. (2015). Performance and emissions
characteristics of compressed spent coffee ground/wood chip logs in a residential stove.
Energy for Sustainable Development, 28, 52–59. https://doi.org/10.1016/j.esd.2015.07.002.
Mani, S., Tabil, L. G., & Sokhansanj, S. (2006). Effects of compressive force, particle size and
moisture content on mechanical properties of biomass pellets from grasses. Biomass
and Bioenergy, 30, 648–654. https://doi.org/10.1016/j.biombioe.2005.01.004.
Mediavilla, I., Esteban, L. S., & Fernández, M. J. (2012). Optimisation of pelletisation condi-
tions for poplar energy crop. Fuel Processing Technology, 104, 7–15. https://doi.org/10.
1016/j.fuproc.2012.05.031.
Miao, Z., Grift, T. E., & Ting, K. C. (2014). Size reduction and densification of lignocellulosic
biomass feedstock for biopower, bioproducts, and liquid biofuel production. In D. R.
Heldman, & C. I. Moraru (Eds.), Encyclopedia of agricultural, food, and biological
engineering (pp. 1545–1551). New York: Taylor and Francis. https://doi.org/10.
1081/E-EAFE2-120051298.
Molina-Moreno, V., Leyva-Díaz, J. C., & Sánchez-Molina, J. (2016). Pellet as a technological
nutrient within the circular economy model: Comparative analysis of combustion ef-
ficiency and CO and NOx emissions for pellets from olive and almond trees. Energies,
9(10), 777. https://doi.org/10.3390/en9100777.
Monedero, E., Portero, H., & Lapuerta, M. (2015). Pellet blends of poplar and pine saw-
dust: Effects of material composition, additive, moisture content and compression
die on pellet quality. Fuel Processing Technology, 132, 15–23. https://doi.org/10.
1016/j.fuproc.2014.12.013.
Moya, R., Rodríguez-Zúñiga, A., Tenorio, C., Valdez, J., & Valaert, J. (2015). Pellets evalua-
tion made from tropical-climate agricultural and forestry crops of Costa Rica with a
domestic stove. Waste and Biomass Valorization, 6, 1037–1046. https://doi.org/10.
1007/s12649-015-9420-1.
Nguyen, Q. N., Cloutier, A., Achim, A., & Stevanovic, T. (2015). Effect of process parameters
and raw material characteristics on physical and mechanical properties of wood
pellets made from sugar maple particles. Biomass and Bioenergy, 80, 338–349.
https://doi.org/10.1016/j.biombioe.2015.06.010.
Obernberger, I., & Thek, G. (2004). Physical characterisation and chemical composition of
densified biomass fuels with regard to their combustion behaviour. Biomass and
Bioenergy, 27, 653–669. https://doi.org/10.1016/j.biombioe.2003.07.006.
Poddar, S., Kamruzzaman, M., Sujan, S. M. A., Hossain, M., Jamal, M. S., Gafur, M. A., et al.
(2014). Effect of compression pressure on lignocellulosic biomass pellet to improve
fuel properties: Higher heating value. Fuel, 131, 43–48. https://doi.org/10.1016/j.
fuel.2014.04.061.
Puig-Arnavat, M., Shang, L., Sarossy, Z., Ahrenfeldt, J., & Henriksen, U. B. (2016). From a single
pellet press to a bench scale pellet mill - Pelletizing six different biomass feedstocks.
Fuel Processing Technology, 142, 27–33. https://doi.org/10.1016/j.fuproc.2015.09.022.
Said, N., Abdel Daiem, M. M., Garcia-Maraver, A., & Zamorano, M. (2015). Influence of
densification parameters on quality properties of rice straw pellets. Fuel Processing
Technology, 138, 56–64. https://doi.org/10.1016/j.fuproc.2015.05.011.
Serrano, C., Monedero, E., Lapuerta, M., & Portero, H. (2011). Effect of moisture
content, particle size and pine addition on quality parameters of barley straw
pellets. Fuel Processing Technology, 92, 699–706. https://doi.org/10.1016/j.
fuproc.2010.11.031.
Shi, Y., Ge, Y., Chang, J., Shao, H., & Tang, Y. (2013). Garden waste biomass for renewable
and sustainable energy production in China: Potential, challenges and development.
Renewable and Sustainable Energy Reviews, 22, 432–437. https://doi.org/10.1016/j.
rser.2013.02.003.
14 P. Pradhan et al. / Energy for Sustainable Development 43 (2018) 1–14
Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., et al. (2008). Determination
of structural carbohydrates and lignin in biomass: Laboratory analytical procedure (LAP).
(Revised July 2011).
Stasiak, M., Molenda, M., Banda, M., Wiacek, J., Parafiniuk, P., & Gondek, E. (2017). Me-
chanical and combustion properties of sawdust-straw pellets blended in different
proportions. Fuel Processing Technology, 156, 366–375. https://doi.org/10.1016/j.
fuproc.2016.09.021.
Stelte, W., Holm, J. K., Sanadi, A. R., Barsberg, S., Ahrenfeldt, J., & Henriksen, U. B. (2011a).
Fuel pellets from biomass: The importance of the pelletizing pressure and its depen-
dency on the processing conditions. Fuel, 90, 3285–3290. https://doi.org/10.1007/
s12155-011-9169-8.
Stelte, W., Holm, J. K., Sanadi, A. R., Barsberg, S., Ahrenfeldt, J., & Henriksen, U. B. (2011b).
A study of bonding and failure mechanisms in fuel pellets from different biomass re-
sources. Biomass and Bioenergy, 35, 910–918. https://doi.org/10.1016/j.biombioe.
2010.11.003.
Tenorio, C., Moya, R., Filho, M. T., & Valaert, J. (2015a). Quality of pellets made from agricul-
tural and forestry crops in Costa Rican tropical climates. Bioresources, 10(1), 482–498.
Tenorio, C., Moya, R., Filho, M. T., & Valaert, J. (2015b). Application of the X-ray densitom-
etry in the evaluation of the quality and mechanical properties of biomass pellets.
Fuel Processing Technology, 132, 62–73. https://doi.org/10.1016/j.fuproc.2014.12.040.
Theerarattananoon, K., Xu, F., Wilson, J., Ballard, R., Mckinney, L., Staggenborg, S., et al.
(2011). Physical properties of pellets made from sorghum stalk, corn stover, wheat
straw, and big bluestem. Industrial Crops and Products, 33, 325–332. https://doi.org/
10.1016/j.indcrop.2010.11.014.
Tumuluru, J. S. (2015). High moisture corn stover pelleting in a flat die pellet mill fitted
with a 6 mm die: physical properties and specific energy consumption. Energy
Science & Engineering, 3, 327–341. https://doi.org/10.1002/ese3.74.
Tumuluru, J. S., Wright, C. T., Hess, J. R., & Kenney, K. L. (2011). A review of biomass
densification systems to develop uniform feedstock commodities for bioenergy
application. Biofuels, Bioproducts and Biorefining, 5, 683–707. https://doi.org/10.
1002/bbb.324.
Van Soest, P. J., Robertson, J. B., & Lewis, B. A. (1991). Methods for dietary fiber, neutral
detergent fiber, and nonstarch polysaccharides in relation to animal nutrition.
Journal of Dairy Science, 74, 3583–3597. https://doi.org/10.3168/jds.S0022-
0302(91)78551-2.
Varunkumar, S., Ranjan, N. K. S., & Mukunda, H. S. (2012). Experimental and computation-
al studies on a gasifier based stove. Energy Conversion and Management, 53, 135–141.
https://doi.org/10.1016/j.enconman.2011.08.022.
Zafari, A., & Kianmehr, M. H. (2013). Factors affecting mechanical properties of biomass
pellet from compost. Environmental Technology, 35, 478–486. https://doi.org/10.
1080/09593330.2013.833639.
Zamorano, M., Popov, V., Rodriguez, M. L., & Garcia-Maraver, A. (2011). A comparative
study of quality properties of pelletized agricultural and forestry lopping residues.
Renewable Energy, 36, 3133–3140. https://doi.org/10.1016/j.renene.2011.03.020.