Journal of Cleaner Production 215 (2019) 1095e1102

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: [Link]/locate/jclepro

Energy use of Jatropha oil extraction wastes: Pellets from biochar and
Jatropha shell blends
Valeria Ramírez a, Jaime Martí-Herrero b, c, *, Michelle Romero a, Daniel Rivadeneira a
a
National Institute of Energy Efficiency and Renewable Energy (INER), Calle In~ aquito N35-37 St., & Juan Pablo Sanz St., Quito, EC170102, Ecuador
b
Centre Internacional de M etodes Numerics en Enginyeria (CIMNE), Building Energy and Environment Group, Edifici GAIA (TR14), C/Rambla Sant Nebridi
22, 08222, Terrassa, Barcelona, Spain
c
Universidad Regional Amazo nica Ikiam, Vía Tena-Muyuna, Km.7, Tena, Napo, Ecuador

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

Article history: The aim of this paper is to evaluate the use of biochar (SCB) produced through the pyrolysis of Jatropha
Received 21 March 2018 seed cake in combination with Jatropha shell waste (JSh) to form pellet combustion fuel. Mixtures using
Received in revised form different proportions of SCB and JSh (0, 25, 50, 75 and 100% each), were prepared, with each proportion
10 January 2019
tested using two particle sizes and two different quantities of water added. Combustion characteristics
Accepted 12 January 2019
Available online 16 January 2019
and mechanical durability were analyzed to identify the best blend, for which a thermogravimetric
analysis was made. Pellets composed of 50% JSh and 50% SCB, with 25% of additional water and 4 mm
particle size, presented the greatest mechanical durability (96.83%), and higher heating value
Keywords:
Biomass pellet
(22.14 MJ kg
1). Combustion of pellets made with these characteristics in an industrial burner showed
Jatropha shell average temperatures ranging from 300 to 350  C, with a maximum temperature of 460  C. The pro-
Jatropha seed cake duction cost of the selected pellet is estimated at 0.11 USD kg
1, making the energy cost of the pellet
Biochar [0.005 USD MJ
1] approximately 28% of the energy cost of LPG, demonstrating the commercial viability
Co-combustion of the selected pellet as feasible.
Waste to energy © 2019 Elsevier Ltd. All rights reserved.

1. Introduction oil in a mixture with diesel (Forson et al., 2004). Jatropha curcas is a
shrub native to the American tropics and is cultivated in tropical
The search for new sources of energy is growing day by day, and subtropical regions. The plant is most commonly found in
with biomass representing an attractive alternative, especially in Mexico and Central America, but is also widely distributed in wild
countries with limited economic resources, as properly-managed or semi-cultivated forests throughout Latin America, Africa, India
biomass can provide a sustainable low-cost fuel supply (Cuiping and South-East Asia (Pandey et al., 2012).
et al., 2004). In Ecuador Jatropha curcas is not used as an energy crop, but is
Among biomass resources, several types of agro-industrial rather employed as live fencing in rural areas, mainly in the coastal
waste can be used for the benefit of industry and the environ- province of Manabí. However, as part of the Ecuadorian govern-
ment (Hoogwijk et al., 2003). The use of biomass for energy pur- ment’s "Renewable Energy for the Galapagos Islands" program, the
poses generally involves pretreatment systems that improve the project "Production of pinion oil for a pilot project of electric power
energy characteristics of the feedstocks. These procedures can be generation in Galapagos" focuses on replacing the diesel fuel
physical or chemical depending on the specific need or interest in commonly used for electric generation with pure Jatropha curcas
maximizing the energy content. oil. For this project, Jatropha curcas fruit is harvested in continental
Several studies have been carried out on the use of Jatropha Ecuador and processed to extract the seeds’ oil, which then is sent
curcas as an energy resource, either by refining the plant’s seed oil to the Galapagos Islands as a clean electric generation fuel. Ac-
into a biodiesel fuel (Koh and Mohd. Ghazi, 2011), or by burning the cording to data from the Manabí Provincial Council, in 2008 the
province had 7000 km of Jatropha curcas live fences, which can
produce between one and ten tons of seeds per hectare, depending
 nica Ikiam, Vía Tena-
on the age of the plants, soil conditions, and the use of irrigation.
* Corresponding author. Universidad Regional Amazo
Muyuna, Km.7, Tena, Napo, Ecuador. From the extraction of Jatropha curcas oil in Manabí alone,
E-mail address: jaimemarti@[Link] (J. Martí-Herrero). approximately 200,000 kg of waste were produced in 2018. This

[Link]
0959-6526/© 2019 Elsevier Ltd. All rights reserved.
1096 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102

residual biomass consists of the Jatropha fruit’s shell, which is goes through a pressing procedure to extract the oil. This process
removed in filter presses before oil extraction, and the seed cake, produces two energy-rich waste products, seed cake and shell
which is the solid fraction of the seed remaining following the oil waste, of which only seed cake is currently used. The seed cake
extraction process. Neither the shell nor seed cake have a current undergoes a pyrolysis process for the production of synthesis gas
disposal management mechanism, leading to residual waste and bio oil. The processing capacity of the pyrolyzing furnace is
problems. 30 kg h
1, converting 70% of the seed cake to biochar and 30% to
Jatropha seed cake does not require densification (pelletizing) synthesis gas and bio-oil.
for the generation of heat (Kratzeisen and Müller, 2013), and can be The pyrolysis reactor uses a set of electrical resistance heaters
pyrolyzed in order to produce bio-diesel. This pyrolysis produces with a total power of 900 W, which are responsible for supplying
biochar that is a "stable, homogeneous, clean, and high-caloric fuel the heat required for the pyrolysis process. The reactor takes
(25e30 MJ kg
1)" (Yi et al., 2013). On the other hand, the Jatropha approximately 1 h to reach 450  C without biomass load (7  C
shell (fruit peel) represents a high-volume biomass that requires min
1). When the reactor reaches this temperature, the heaters are
densification to reduce transportation costs and facilitate subse- automatically turned off and a conveyor screw is activated, feeding
quent combustion, as recommended by (Mobini et al., 2014) for the the biomass through the reactor. The reactor has an automatic
case of wood pellet production. control system, which is responsible for turning on and off the
In Ecuador, management of the waste (seed cake) from the resistance heaters to maintain the temperature within ± 10  C of
extraction process of Jatropha curcas oil is being piloted using py- the set temperature. Once the reactor is loaded with biomass, the
rolysis to generate bio-diesel and biochar. The pyrolysis reactor resistors can remain off for an approximate time of 2 h. The gases
consists of a conveyor screw covered by a jacket of electrical re- produced in the process of thermal decomposition displace the air
sistors which raise the internal temperature of the biomass (seed inside the reactor, creating an oxygen-free atmosphere and giving
cake) circulating through the screw. The temperature reached in- rise to pyrolysis. As the volume of the gases increases, they are
side the reactor is 450  C. At this temperature, gases are generated directed to separators and condensers, which recover the con-
that displace the air inside the reactor, creating an anoxic atmo- densable bio-oil in liquid form and transfer non condensable gases
sphere. The gases generated are transferred by pipes to separators to a torch to be burned(See. Fig. 1).
and condensers in which the condensable gases (bio-oil) are
retained and the non condensable gases are combusted with a 2.2. Materials
torch. The remaining solids from the process (biochar) are cooled
and stored. The biomass used in this research was the pyrolyzed residue of
This research seeks to explore the opportunity for creating the waste resulting from the extraction of Jatropha curcas oil, also
pellets from Jatropha shell waste together with the biochar pro- called seed cake biochar (SCB), and the peel of the fruit of the same
duced in the pyrolysis of seed cake, taking advantage of the lignin plant, called Jatropha shell (JSh).
content of the shell which acts as a natural binder for the pelletizing The seedcake was pyrolyzed (at 450  C for 25 min) in order to
process, and the higher heating value of the biochar. Pelletizing remove part of the volatile content and increase its energy poten-
increases the energy density of the residual biomass through the tial. The JSh was not processed, with the intention of maintaining
compaction of the residues, and can lower transportation and the lignin content present to preserve its natural binding function
disposal costs if the pellets are used as a fuel source in domestic and for the pelletizing process. The raw materials were crushed and
industrial boilers (Mobini et al., 2014; Stelte et al., 2011). sieved to create two particle sizes, 4 mm and 2 mm, in order to
Recognizing the benefits available from both materials, research evaluate the influence of particle size on the pellet compaction
on the effect of mixing biomass and coal on the mechanical dura- (Franke and Rey, 2006).
bility and combustion characteristics of pellets has been previously
conducted (Gil et al., 2010b; Guo and Zhong, 2018). Published re- 2.3. Blend setup process
sults on the combustion behavior of pellets from the blending of
biochar and coal (Li et al., 2018) also exist, as do results on the The mixing process was carried out in order to evaluate the
performance of biochar pellets obtained from the pyrolysis of corn behavior of the pellets using varying SCB-JSh mixtures (100, 75, 50,
straw pellets (Xing et al., 2018), but the use of biochar blended with
biomass in pellets has not yet been studied.
In recognition of need to improve waste management in Jatro-
pha oil production and the gap in existing publications, the current
research demonstrates the feasibility of energy use for two waste
products. Various blends of the biochar and Jatropha shell were
evaluated to obtain the best-performing and most economically-
attractive pellet.

2. Methodology

2.1. Oil extraction plant and pilot pyrolysis plant

The sole Jatropha oil extraction plant in Ecuador is located at the

INIAP-Portoviejo Experimental Station (Manabí, Ecuador). A pilot
plant, built next to the oil extraction facility, uses the Jatropha seed
cake (waste product from the oil extraction process) to produce
bio-oil through pyrolysis. The processing capacity of the oil
extraction plant is 200 kg of Jatropha seed per hour, obtaining an Fig. 1. Schematic of the Jatropha oil extraction plant process, and the linked pilot plant
approximate yield of 30% oil and 70% seed cake. waste-to-energy processes. Novel processes under investigation appear in the
The process starts with dehulling to obtain the seed, which then emboldened box at center.
 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102 1097

25 and 0% each), particle size of the solid components (2 and 4 mm) analytical balance (Ohaus model PA214) with an accuracy of
and amount of added water prior to the pelleting process (15% and 0.0001 g (initial mass). The sample was then introduced into a
25%). The mixtures were carried out in two stages. In stage 1, a dry durabilimeter (Bioenergy model Tumbler 1000), which uses a
mixture of 30 kg of SCB-JSh was prepared according to the per- rotating chamber at 50 ± 2 rpm, and underwent a single 500-
centages indicated in Table 1. In stage 2, 10 kg of each resulting SCB- rotation cycle. The sample was then sieved (3.15 mm) and the
JSh mixture was taken and the indicated quantity of water was mass of the sample retained by the sieve (final mass) was recorded.
added in order to evaluate the effects from the combination of The mechanical durability value corresponds to the percentage of
water, lignin (from the Jatropha shell) and heat (produced by the the initial mass remaining in the sample after the sieving of the fine
friction of the rollers of the pelletizer) on the mechanical durability particles generated by the collision and friction of the pellets
of the pellet. occurring inside the rotating chamber.
To homogenize the mixture in stage 1, a semi-industrial mixer The thermal characterization and the higher heating value for
with a capacity of 120 L was used, while in stage 2 a metal container both the raw materials and the mixtures were performed by ther-
and a 30 cm spatula were used for manual mixing. In both stages, mogravimetric (TG), derivative thermogravimetric (DTG) and
the mixing time was 1 min. calorimetric analysis. The thermogravimetric analysis was carried
The details of the 20 different mixtures obtained are presented out using a thermogravimetric analyzer (Shimadzu model TGA-50),
in Table 1. with a flow rate of 50 cm3 min
1 at a heating rate of 15  C min
1,
ranging from room temperature to 900  C. Approximately 5 mg of
2.4. Pelletizing process the sample were employed for each measurement. A calorimeter
pump (IKA model C2000 basic), was used to measure the higher
The pelletizer used for this process was a semi industrial heating value, and the procedure described in BS-EN-14918 (British
pelletizer (Chengda Machinery model KL 150B), with a capacity of Standard) was followed.
120 kg h
1 and a fixed mold, producing pellets 9 mm in diameter
and 2 cm in length. The pelletizer has movable rollers, which exert 2.6. Combustion in a continuous feed burner
the necessary pressure for the compaction and formation of the
pellet. The pellet that showed the best characteristics (durability and
Indirect preheating of the pelletizer was necessary prior to higher heating value) was tested in a semi-industrial continuous
beginning pellet production. A mixture of 10% food-grade oil by feed biomass burner to evaluate the maximum temperature
mass and 90% of the mixture to be pelletized was introduced into reached during the combustion and validate the results obtained in
the pelletizer and recirculated during 15 min, heated by the friction the DTG.
of the rollers on the matrix, in accordance with the recommenda- The burner used has a horizontal combustion chamber; pellets
tion of the user manual of the pelletizer. Once the preheating are fed into a hopper at the top of the device, and the air is supplied
procedure was completed, each blend was pelletized. by a hole in the side. Three type-k thermocouples were placed in
the burner to monitor temperatures, as shown in Fig. 2.
2.5. Characterization of pellets The burner was preheated to maintain stable operation of its
mechanical components during pellet testing. An air flow of
To evaluate the durability or mechanical strength of the pellets, 208 m3 h
1, and pellet feed rate of 2.6 kg h
1, were set to evaluate
the procedure established in “BS EN 15210-1 Determination of the temperature reached through the combustion of the pellet.
mechanical durability of pellets” (British Standard) was applied. For After combustion started, temperatures were recorded for an
this analysis, a pellet sample of 500 ± 10 g, previously sieved with a approximate period of 13 min, with a measurement frequency of 1
3.15 mm mesh, was taken. The sieved sample was weighed on an s
1.

2.7. Economic analysis

Table 1
Characteristics of the blends employed in the preparation of pellets.
The data available from the pilot plant were used to calculate the
Blends of JSh and SCB þ water Mesh Size Proportion Additional water cost of pellet production, which include capital investment, oper-
%SCB %JSh %W ation and maintenance services, and expenses for pellet produc-
tion. The pilot plant is located adjacent to the Jatropha oil extraction
(100SCB) 25 W 4 mm 100 0 25
(25JShþ75SCB) 25 W 75 25 25 plant, so transportation, storage and other raw material costs were
(50JShþ50SCB) 25 W 50 50 25
(75JShþ25SCB) 25 W 25 75 25
(100JSh) 25 W 0 100 25

(100SCB) 25 W 2 mm 100 0 25
(25JShþ75SCB) 25 W 75 25 25
(50JShþ50SCB) 25 W 50 50 25
(75JShþ25SCB) 25 W 25 75 25
(100JSh) 25 W 0 100 25

(100SCB) 15 W 4 mm 100 0 15
(25JShþ75SCB) 15 W 75 25 15
(50JShþ50SCB) 15 W 50 50 15
(75JShþ25SCB) 15 W 25 75 15
(100JSh) 15 W 0 100 15

(100SCB) 15 W 2 mm 100 0 15
(25JShþ75SCB) 15 W 75 25 15
(50JShþ50SCB) 15 W 50 50 15
Fig. 2. Schematic of the semi industrial burner. a) Power supply/pellet hopper; b) Air
(75JShþ25SCB) 15 W 25 75 15
intake and c) Combustion chamber. T1, T2 and T3 indicate the position of the type-k
(100JSh) 15 W 0 100 15
thermocouples. The burner has a diameter of 165 mm and a length of 300 mm.
1098 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102

Table 2
Proximate and ultimate analysis, and higher heating values of the raw materials.

Sample Moisture (%) Proximate Analysis (% weight, db) Higher heating value (MJ kg
1) Ultimate Analysis (% weight, db)

Volatile matter Fixed Carbona Ash C H N S Oa

Jatropha seed cake (SC) 4.08 73.7 19.06 7.24 17.04 52.12 6.91 5.01 0.7 28.02
Seed cake biochar (SCB) 9.28 40.02 46.62 13.36 28.3 61.31 3.55 3.77 0.38 17.63
Jatropha Shell (JSh) 10.57 71.52 17.64 10.84 14.37 40.8 5.9 1.53 0.43 40.5
a
Calculated by difference; db: dry basis.

not considered. This economic analysis was used to determine the of a solid biofuel, the SCB will contribute its higher heating value,
energy cost (USD MJ
1) of the pellets, and these data were while the JSh will contribute a high oxygen content, facilitating
compared to the corresponding energy cost for Liquefied Petroleum both the ignition and subsequent combustion of the fuel.
Gas (LPG).

3. Results and discussion

3.1. Characteristics of raw materials 3.2. Mechanical durability

Proximate and elemental analysis along with the higher heating Pellets made with the 4 mm biomass presented the greatest
values of seed cake (SC), SCB and JSh samples are presented in mechanical durability, except for the mixture (50JSh þ 50SCB)
Table 2. Non-pyrolyzed Jatropha residues (SC and JSh) retain a low 15 W, as can be seen in Table 3.
ash content (less than 11%), and a fixed carbon level of less than Pellet blends that contained an amount greater than or equal to
20%, values that are in the ranges (ash: 1.5e19% and fixed carbon: 75% SCB in the mixture, regardless of particle size and water
11.3e22%), previously reported for these wastes (Kongkasawan quantity, did not form pellets due to compaction problems. This
et al., 2016; Singh et al., 2008). For SCB the ash content is 13.36%, implies an upper limit for the amount of SCB in the mixture for
and fixed carbon content is 47%, higher values than those of the pelleting to be viable.
original substrate. This is due to the elimination of 45.45% of the The mechanical durability of the pellets range from 89 to 97%,
volatile matter through pyrolysis of the Jatropha seed cake, similar to pellets made from better quality biomass, such as wood
increasing the final concentration of ash and fixed carbon. (Verma et al., 2012). Other previous research (Gil et al., 2010b)
The concentration of volatile solids in SC (73.7%) is higher than evaluated the addition of coal to other biomass for pelletizing, and
in SCB (40.02%). Volatile matter favors combustion by facilitating also found that an appropriate proportion of anthracite coal (be-
ignition, but an excess of volatile solids also generates large tween 5 and 15%) would maintain high durability values, approx-
amounts of gases during the combustion process at relatively low imately 93% in the best-case scenario. The present research shows
temperatures (Jenkins et al., 1998). Excessive gas production as a similar trends; the higher the biochar proportion, the lower the
consequence of combustion can cause negative environmental durability of the pellets. Compared to prior coal-biomass mixtures,
impacts (Ezzati and Kammen, 2001). the present research shows that more biochar (if equated with coal)
The elemental analysis of SC produced values of C, H, N, O and S, can be incorporated while maintaining superior mechanical dura-
similar to those reported in prior investigations of Jatropha residues bility, with values up to 50% of SCB in the blend. BS EN 15210-1
(Sharma et al., 2016). In the case of JSh, the H and C content was (British Standard) indicates 96% as an acceptable durability value
similar to that observed in other investigations (Pambudi et al., for pellets. This value represents the lower limit when choosing
2017). This is not the case however with the values of N and S, pellets based on quality. Fig. 3 and Table 3 demonstrate that pellet
with concentrations up to 50% higher than those presented in other mixtures with 50% SCB present durability values higher than the
investigations. These concentrations may be explained by varia- limit established by the standard, with particle size of 4 mm and an
tions in the characteristics of the soil where Jatropha was culti- additional water proportion of 25%.
vated, and by the application of fertilizer to neighboring crops. Following selection based on mechanical durability, the next
Compared to SCB, the JSh has a higher proportion of oxygen and criteria for which pellet selection was based on was higher heating
hydrogen (Table 2). These characteristics reduce the higher heating value.
value of JSh, since the energy contained in the carbon-hydrogen-
oxygen and carbon bond is inferior to that of carbon-carbon bond Table 3
(Gil et al., 2010b). On the other hand, the higher oxygen content of Mechanical durability of pellets.
the JSh indicates a higher thermal reactivity than the SCB (Yi et al.,
Sample Mechanical Durability, %
2013). In addition, the sulfur content of the SCB is 0.38%, while the
Mesh Size: 4 mm Mesh Size: 2 mm
SC contains 0.7%, revealing a reduction in sulfur concentration
during the pyrolysis process. Other compounds are formed in the (100SCB) 25 Wa e e
gaseous and liquid fractions during the pyrolysis process, favoring (25JShþ75SCB) 25 Wa e e
(50JShþ50SCB) 25 W 96.83 89.91
the clean combustion behavior of the selected material pellets.
(75JShþ25SCB) 25 W 97.11 96.88
The higher heating values for SC and JSh were similar, at (100JSh) 25 W 96.91 96.31
17.04 MJ kg
1 and 14.37 MJ kg
1 (dry basis) respectively. These
(100SCB) 15 Wa e e
values are similar to those reported in other studies (Kongkasawan (25JShþ75SCB) 15 Wa e e
et al., 2016). The SCB presented a higher heating value of (50JShþ50SCB) 15 W 92.80 95.36
(28.3 MJ kg
1), demonstrating it to be an appropriate component (75JShþ25SCB) 15 W 96.99 96.42
for improving the combustion characteristics of the pellets. (100JSh) 15 W 97.08 95.20
a
It is expected that when mixing JSh and SCB for the production No values recorded because the blend did not form pellets.
 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102 1099

Fig. 3. Mechanical durability of pellets made with Seed Cake Biochar (SCB) and Fig. 4. Higher heating value and mechanical durability of pellets using mixtures
Jatropha Shell (JSh) mixtures with 25% and 15% water addition; and 2 mm or 4 mm described in Table 1.
particle size.

3.3.2. Individual combustion performance of the raw materials

3.3. Energy and thermal characteristics
(SCB, JSh) and the best-quality pellet [(50JSh þ 50SCB) 25 W]
To evaluate the mixture of SCB and JSh in the formation of the
3.3.1. Pellet energy content
pellet during the combustion process, both the selected mixture
In general, pellets with a greater proportion of SCB in the
and its individual components were subjected to a thermogravi-
mixture register higher heating values, as can be observed in
metric analysis using an oxidizing atmosphere.
Table 4.
The derivative thermogravimetry (DTG) curves for the samples
The energy content of pellet mixtures with equal proportions of
of JSh, SCB and the mixture (50JSh þ 50SCB) 25 W, using an at-
JSh and SCB but with different particle sizes does not vary signifi-
mosphere of air, are presented in Fig. 5.
cantly (under 2.2%) when the amount of added water is 25%, and
The thermogravimetric analysis realized using DTG profiles
varies up to 5.9% when water content is 15%, as can be seen in
extends over a temperature range of 25  Ce700  C. It can be
Table 4. This stability is expected, because the energy content de-
observed that the JSh undergoes three stages of mass loss, while
pends on the composition of the combustion matter and not on the
SCB experiences only two.
particle diameter. As expected, the greater the SCB content, the
In the case of the JSh, an initial mass loss (stage A) occurs be-
greater the higher heating value of the pellet, up to the 50% SCB
tween temperatures of 25  C and 105  C. This loss corresponds to
composition which represents the upper limit for this material in
the evaporation of the water contained in the biomass and depends
the mixture in order to meet the established mechanical durability
on the moisture content of the sample. Therefore, this stage is
parameters. This mixture, 50% SCB 50% JSh, registers values in the
prominent in JSh due to its higher moisture content, while SCB
range of 19.96e22.14 MJ kg
1, which is less than the higher heating
shows almost no mass loss over this temperature range.
value of pure SCB found by the current research (28.3 MJ kg
1), or
Two additional stages of mass loss, one occurring with tem-
that reported (25e30 MJ kg
1) in previous studies (Yi et al., 2013).
peratures ranging from 180  C to 380  C (stage B), and the other
However, this mixture permits material densification through the
with temperatures from 380  C to 480  C (stage C), are observed in
pelletization process.
the DTG curve of JSh. The loss of mass in stage B is due to the
The mechanical durability results (Table 3) and the higher
oxidative degradation of the sample, as light organic volatile solids
heating values (Table 4) for each pellet mixture are shown simul-
are released and then burned. The loss of mass in stage C is due to
taneously in Fig. 4. Most of the pellets register a durability between
combustion of the remaining coal (Gil et al., 2010a; Guo and Zhong,
96 and 98%, which are acceptable values according to the norm BS
2018). In the case of SCB, stage B and C take place in the temper-
EN 15210-1 (British Standard). Fig. 4 shows that the pellets elabo-
ature ranges of 220  Ce420  C and 420  Ce540  C, respectively.
rated with (50JSh þ 50SCB) 25 W and sieved with a mesh size of
The temperature ranges for stages B and C in the case of the JSh
4 mm, present the greatest mechanical durability and higher
are similar to those observed for other types of biomass, whereas in
heating value.
the case of SCB, a similarity can be observed in the temperature
range of stage C to that of anthracite coal (Gil et al., 2010a). The DTG

Table 4
Higher heating values of JSh and SCB pellets.

Sample Higher heating value (MJ kg
1)

Mesh Size: 4 mm Mesh Size: 2 mm

(100SCB) 25 Wa e e
(25JShþ75SCB) 25 Wa e e
(50JShþ50SCB) 25 W 22.14 21.91
(75JShþ25SCB) 25 W 15.94 16.24
(100JSh) 25 W 13.93 14.24

(100SCB) 15 Wa e e
(25JShþ75SCB) 15 Wa e e
(50JShþ50SCB) 15 W 21.12 19.96
(75JShþ25SCB) 15 W 16.74 16.80
(100JSh) 15 W 14.97 14.35
Fig. 5. DTG curves for pellets elaborated with the mixture (50JSh þ 50SCB) 25 W, and
a
No values recorded because the mixture did not form pellets. their raw materials with an air flow of 50 cm3 min
1 and a heating rate of 15  C min
1..
1100 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102

Table 5
Temperature interval, mass loss, residue, peak temperature and maximum rate of mass loss (DTGmax), for the combustion of individual materials (JSh, SCB) and the blend.

Sample Temperature interval ( C) Mass loss (%) Residue (%) Peak temperature ( C) DTGmax (% s
1)

Stage B Stage C Stage B Stage C Stage B Stage C Stage B Stage C

JSh 180e380 380e480 46.2 20.6 16.5 270 415 0.16 0.44
SCB 220e420 420e540 23.1 61.8 13.7 330 460 0.06 0.28
(50JShþ50SCB)25 W 200e340 340e520 24.3 47.5 14.2 280 413 0.08 0.37

curve of the mix (50JSh þ 50SCB) 25 W shares the characteristics of properly feed pellets from the vertical hopper. This experience
its raw materials. Stage B and C take place in the temperature demonstrates that optimizing the design of the burner in necessary
ranges of 200  Ce340  C and 340  C to 520  C, respectively. to fully exploit the potential of the selected pellet under semi in-
Table 5 shows the combustion temperature range, mass loss, dustrial conditions.
final post-combustion residue, maximum temperature and The average and maximum temperatures reached in the burner
maximum mass loss rate (DTGmax), corresponding to the two main are in the respective ranges of the peaks of stage B for JSh (270  C)
stages of mass loss (stages B and C). The initial temperature in stage and C for SCB (460  C). The maximum temperatures reached in the
B and the final temperature in stage C were considered as tem- burner was 460  C, while the maximum temperature in the DTG for
perature values at which the rate of mass loss was 0.005% s
1 the mixture (50JSh þ 50SCB) 25 W was 500  C (Fig. 5). This implies
(Rubiera et al., 1999). a difference of 8% comparing the value obtained in the laboratory
Table 5 shows that JSh experiences a loss of volatile solids (stage and the value tested in the semi industrial burner, which can be
B) at a lower temperature (180  C) than the SCB (220  C), and reg- explained by burner heat losses and the lack of optimization of the
isters a higher percentage of mass loss during this stage (46.2%). burner for the pellet type.
Thus, combustion of JSh is dominated by stage B processes, occur-
ring between 180  C and 380  C. In the case of the SCB, stage C 3.5. Economic analysis results
dominated the combustion with a mass loss of 61.8% taking place in
the range of 420e540  C. In order to estimate the energy cost of the selected pellet, the
The blend (50JSh þ 50SCB) 25 W offers advantages from both cost per kg of pellet was first estimated using the annual cost of
materials’ combustion characteristics, since the loss of mass in pilot plant operation (considering credit on investment, labor,
stage B (24.3%) is lower than that in stage C (47.5%), allowing the electricity and water supply costs) and the plant’s annual pellet
greater mass to be burned at higher temperatures (340e520  C). production. Then, the energy cost of the selected pellet was
This benefit also occurs with SCB, however, as this substrate goes calculated using the higher heating value and compared to the
through a previous pyrolysis process, the easily combusted volatile respective value for LPG.
solids are lost, which may impede ignition in the combustion Number of pellets per year. To determine the amount of raw
process. material available, an average value was calculated based on the
The JSh generated a larger amount of unburned residual mate- residual biomass production of the Jatropha oil extraction plant
rial (16.5%) compared to SCB and the blend (50JSh þ 50SCB) 25 W. If over the last five years. The result obtained was 63,924 kg per year
the combustion residues of JSh and SCB from Table 5 are compared of seed cake biochar (SCB), and 55,910 kg per year of Jatropha Shell
with the ash content detailed in Table 2, the more complete com- (JSh). Also, a 15% mass loss was observed during the pelletization
bustion of SCB is evident. The percentage of SCB combustion waste process.
(13.7%) is much closer to the percentage of SCB ash (13.36%, Table 2) To manufacture pellets with the chosen formulation
when compared to JSh. Therefore, SCB will help to achieve a better (50JSh þ 50SCB) 25 W, the use of the entire JSh (55,910 kg) and an
combustion of the blend. equivalent amount of SCB (55,910 kg) would be necessary. Thus,
As the temperature at the point of maximum rate of mass loss is taking into account losses during pelletization (15% by mass), the
observed to be inversely proportional to the reactivity (Gil et al., final production of pellets was estimated at 95,047 kg per year.
2010a), SCB proves to be the least reactive material. However, Annual operational costs. The main equipment at the production
when SCB is combusted in combination with JSh, the mixture’s plant includes: crusher, screw conveyor, mixer and pelletizer, cor-
reactivity is not very different from the reactivity of pure JSh, the responding to an investment of 15,000 USD. Estimating a 10% in-
most reactive material. This behavior is observed for both stages B terest loan paid over 5 years, this translates to an annual cost of
and C. In general, it can be observed that the reactivity of a mixture
decreases with the reduction of JSh levels. Subsequently, the main
mass loss stage occurs during a higher range of temperatures,
indicating greater thermal efficiency (Liu and Han, 2015).

3.4. Combustion in a semi-industrial burner

The combustion of the selected pellet (50JSh þ 50SCB) 25 W in a

horizontal flow burner reached temperatures above 350  C during
the first 7 min of combustion, and subsequently stabilized at
around 300  C, as shown in Fig. 6. This temperature decrease is due
to the design of the burner, which is not optimized for this type of
pellet. The accumulation of ash hampers continuous pellet feeding
into the combustion chamber. This is because the ash evacuation
grid’s mesh is too small for this type of pellet ash. If air flow is Fig. 6. Temperature evolution in the semi-industrial burner when fed with the
increased to displace the accumulated ash, it becomes difficult to selected pellet (50JSh þ 50SCB) 25 W
 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102 1101

3300 USD. The employment of one operator, paid Ecuador’s mini- project. Jaime Martí-Herrero would like to acknowledge the
mum wage (400 USD per month) (Ministerio de Trabajo ECUADOR, Prometeo Project of the Secretariat for Higher Education, Science,
2017), is estimated as sufficient to manage the volume of the Technology and Innovation of the Republic of Ecuador for funding
existing residual biomass and the processing capacity, leading to an part of his work, and the financial support of CIMNE via the CERCA
additional 4800 USD per year in expenses. Programme/Generalitat de Catalunya. The authors wish to
The electricity cost was estimated based on the total electricity acknowledge the collaboration of Ricardo Narvaez and Marco
consumption of the pilot plant (5616 kWh per month). The energy Heredia and the revision of the document by Samay Schütt, Samuel
consumption was 8.55 kWh h
1, considering only the pelletizing Schlesinger and Sophie Smith.
process, so operation 8 h d
1 results in a monthly consumption of
1368 kWh. The electricity cost for the industrial sector in Ecuador is
0.08 USD kWh
1 (Agencia de Regulacio  n y Control de Electricidad, References
2018), which results in 1313 USD per year of expenses from the n y Control de Electricidad, 2018. Tariff Schedule for Electricity
Agencia de Regulacio
pelletizing process. Distribution Companies. Public Electricity Service. Period: January - December
The consumption of water in the pelletizing process is low, but if 2018 (In Spanish) [WWW Document]. [Link]
cleaning activities are also considered, the consumption of water wp-content/uploads/downloads/2018/01/2018-01-11-Pliego-y-Cargos-
[Link].
per month at the plant is 67 m3. The water cost for the industrial Cuiping, L., Chuangzhi, W., Yanyongjie, Haitao, H., 2004. Chemical elemental char-
sector in the province of Manabí is 1.50 USD m
3 (Gobierno acteristics of biomass fuels in China. Biomass Bioenergy 27, 119e130. https://
Auto nomo Descentralizado Municipal del Canto  n Portoviejo, [Link]/10.1016/[Link].2004.01.002.
Ezzati, M., Kammen, D., 2001. Indoor air pollution from biomass combustion and
2016), which results in 1206 USD per year of expenses. acute respiratory infections in Kenya: an exposure-response study. Lancet
Cost of pellets: Considering all the expenses mentioned, the (London, England) 358, 619e624. [Link]
yearly cost of producing 95,047 kg of (50JSh þ 50SCB) 25 W pellets 05777-4.
Forson, F., Oduro, E., Hammond-Donkoh, E., 2004. Performance of jatropha oil
is estimated to be 10,619 USD, resulting in a production cost of 0.11 blends in a diesel engine. Renew. Energy 29, 1135e1145. [Link]
USD kg
1. [Link].2003.11.002.
Energy cost of pellets compared to LPG: The use of pellets in the Franke, M., Rey, A., 2006. Pelleting Quality. World Grain, pp. 78e79.
Gil, M.V., Casal, D., Pevida, C., Pis, J.J., Rubiera, F., 2010a. Thermal behaviour and
industrial sector would displace the consumption of LPG. LPG for kinetics of coal/biomass blends during co-combustion. Bioresour. Technol. 101,
the industrial sector in Ecuador costs 0.804950 USD kg
1 5601e5608. [Link]
(PetroEcuador, 2018) and presents a higher heating value estimated Gil, M.V., Oulego, P., Casal, M.D., Pevida, C., Pis, J.J., Rubiera, F., 2010b. Mechanical
at 44.3 MJ kg
1 (OLADE, 2017). Thus, the energy cost of LPG is 0.018 durability and combustion characteristics of pellets from biomass blends. Bio-
resour. Technol. 101, 8859e8867. [Link]
USD MJ
1. The (50JSh þ 50SCB) 25 W pellet has a cost of 0.11 USD 062.
kg
1, and a higher heating value of 22.5 MJ kg
1, so the energy cost Gobierno Auto  nomo Descentralizado Municipal del Canto n Portoviejo, 2016.
of this pellet is 0.005 USD MJ
1. These values demonstrate that the Ordinance that Establishes the Collection of the Fee for Providing the Public
Service of Drinking Water, Sewerage and Sanitation for the Canton Portoviejo
energy cost of (50JSh þ 50SCB) 25 W pellets represents only 28% of (In Spanish) [WWW Document]. URL. [Link]
the energy cost of the LPG heating. This estimation does not [Link]?id¼731.
consider the final energy delivered by the combustion of the LPG or Guo, F., Zhong, Z., 2018. Optimization of the co-combustion of coal and composite
biomass pellets. J. Clean. Prod. 185, 399e407. [Link]
the pellets, since this depends on the combustion efficiency of the 2018.03.064.
burners and thermal efficiency of other equipment, nor does it Hoogwijk, M., Faaij, A., van den Broek, R., Berndes, G., Gielen, D., Turkenburg, W.,
consider transportation costs. 2003. Exploration of the ranges of the global potential of biomass for energy.
Biomass Bioenergy 25, 119e133. [Link]
As the pellets’ energy cost is lower than that of LPG, the profit 5.
margin that can be obtained by substitution could be used for pellet Jenkins, B., Baxter, L., Miles, T., Miles, T., 1998. Combustion properties of biomass.
promotion, pellet burner facilities for users, or to provide price Fuel Process. Technol. 54, 17e46. [Link]
00059-3.
subsidies, in order to overcome adoption barriers related to users’
Koh, M.Y., Mohd Ghazi, T.I., 2011. A review of biodiesel production from Jatropha
unfamiliarity with pellets or other issues. curcas L. oil. Renew. Sustain. Energy Rev. 15, 2240e2251. [Link]
1016/[Link].2011.02.013.
Kongkasawan, J., Nam, H., Capareda, S.C., 2016. Jatropha waste meal as an alter-
4. Conclusions
native energy source via pressurized pyrolysis: a study on temperature effects.
Energy 113, 631e642. [Link]
Pellets made with a 50/50 mixture of Seed Cake Biochar and Kratzeisen, M., Müller, J., 2013. Suitability of Jatropha seed shells as fuel for small-
Jatropha Shell constitute a biofuel with good combustion charac- scale combustion units. Renew. Energy 51, 46e52. [Link]
RENENE.2012.08.037.
teristics. The Seed Cake Biochar contributes its high energy content, Li, Y.-H., Lin, H.-T., Xiao, K.-L., Lasek, J., 2018. Combustion behavior of coal pellets
which is reflected in a greater range of combustion temperatures blended with Miscanthus biochar. Energy 163, 180e190. [Link]
and a lower percentage of unburned material, in comparison with 1016/[Link].2018.08.117.
Liu, Z., Han, G., 2015. Production of solid fuel biochar from waste biomass by low
Jatropha shell by itself. The Jatropha shell provides reactivity for the temperature pyrolysis. Fuel 158, 159e165. [Link]
pellet blend with its high content of volatile matter, facilitating the 05.032.
ignition of the pellet. The studied mixture is recommended for Ministerio de Trabajo ECUADOR, 2017. Ministerial Agreement No. MDT-2017-0195
(In Spanish) [WWW Document]. [Link]
industrial use in large burners, as a low percentage of post- uploads/2018/01/[Link].
combustion residues are generated. The energy cost [USD MJ
1] Mobini, M., Meyer, J.-C., Trippe, F., Sowlati, T., Fro €hling, M., Schultmann, F., 2014.
of the selected pellet blend is 60% lower than that of LPG, making it Assessing the integration of torrefaction into wood pellet production. J. Clean.
Prod. 78, 216e225. [Link]
a feasible alternative. OLADE, 2017. Energy Statistics Manual (In Spanish).
Pambudi, N.A., Laukkanen, T., Syamsiro, M., Gandidi, I.M., 2017. Simulation of
Acknowledgements Jatropha curcas shell in gasifier for synthesis gas and hydrogen production.
J. Energy Inst. 90, 672e679. [Link]
Pandey, V.C., Singh, K., Singh, J.S., Kumar, A., Singh, B., Singh, R.P., 2012. Jatropha
Financial support to carry out this research was provided by the curcas: a potential biofuel plant for sustainable environmental development.
Instituto Nacional de Eficiencia Energe tica y Energías Renovables Renew. Sustain. Energy Rev. 16, 2870e2883. [Link]
and the Secretaría Nacional de Planificacio  n y Desarrollo (SEN- 2012.02.004.
PetroEcuador, 2018. Sales Prices at the Terminal Level for Qualified and Authorized
PLADES) through the "Estudio de alternativas de aprovechamiento Marketers at the National Level. Period of Validity: 01 to 03 May 2018. Exec-
energetico de biomasa residual del Proyecto Pin ~o
 n para Gala
pagos" utive Order No. 799 Y 352. Ecuador (In Spanish) [WWW Document]. https://
1102 V. Ramírez et al. / Journal of Cleaner Production 215 (2019) 1095e1102

[Link]/wp-content/plugins/download-monitor/download. dependency on the processing conditions. Fuel 90, 3285e3290. [Link]

php?id¼218. 10.1016/[Link].2011.05.011.
Rubiera, F., Arenillas, A., Fuente, E., Miles, N., Pis, J., 1999. Effect of the grinding Verma, V.K., Bram, S., Delattin, F., Laha, P., Vandendael, I., Hubin, A., De Ruyck, J.,
behaviour of coal blends on coal utilisation for combustion. Powder Technol. 2012. Agro-pellets for domestic heating boilers: standard laboratory and real
105, 351e356. [Link] life performance. Appl. Energy 90, 17e23. [Link]
Sharma, R., Sheth, P.N., Gujrathi, A.M., 2016. Kinetic modeling and simulation: py- 2010.12.079.
rolysis of Jatropha residue de-oiled cake. Renew. Energy 86, 554e562. https:// Xing, X., Fan, F., Jiang, W., 2018. Characteristics of biochar pellets from corn straw
[Link]/10.1016/[Link].2015.08.066. under different pyrolysis temperatures. R. Soc. Open Sci. 5, 172346. [Link]
Singh, R.N., Vyas, D.K., Srivastava, N.S.L., Narra, M., 2008. SPRERI experience on org/10.1098/rsos.172346.
holistic approach to utilize all parts of Jatropha curcas fruit for energy. Renew. Yi, Q., Qi, F., Cheng, G., Zhang, Y., Xiao, B., Hu, Z., Liu, S., Cai, H., Xu, S., 2013. Ther-
Energy 33, 1868e1873. [Link] mogravimetric analysis of co-combustion of biomass and biochar. J. Therm.
Stelte, W., Holm, J.K., Sanadi, A.R., Barsberg, S., Ahrenfeldt, J., Henriksen, U.B., 2011. Anal. Calorim. 112, 1475e1479. [Link]
Fuel pellets from biomass: the importance of the pelletizing pressure and its