Fuel 224 (2018) 111–120

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Valorization of citrus wastes by fast pyrolysis in a conical spouted bed T

reactor
Jon Alvareza, Bahar Hooshdaranb, Maria Cortazara, Maider Amutioa, Gartzen Lopeza,
⁎
Fabio B. Freirec, Masoud Haghshenasfardb, Seyyed Hossein Hosseinid, Martin Olazara,
a
University of the Basque Country UPV/EHU, Department of Chemical Engineering, P.O. Box 644, E48080 Bilbao, Spain
b
Department of Chemical Engineering, Isfahan University of Technology, Isfahan, Iran
c
Department of Chemical Engineering, Federal University of São Carlos, C.P. 676, 13565-905 São Carlos, SP, Brazil
d
Department of Chemical Engineering, Faculty of Engineering, University of Ilam, Ilam, Iran

A R T I C LE I N FO A B S T R A C T

Keywords: The fast pyrolysis of the juice squeezing derived orange waste has been carried out in a continuous pyrolysis
Fruit wastes bench-scale plant consisting of a conical spouted bed reactor (CSBR). A prior study performed in thermobalance
Orange waste and a kinetic model consisting of a multi-component mechanism allowed determining the contents of pectin
Pyrolysis (35 wt%), hemicellulose (16.6 wt%) and cellulose (17.1 wt%), but that of lignin could not be satisfactorily de-
Spouted bed reactor
termined as its degradation curve overlapped with other compounds such as sugars, proteins and fats. In the
Bio-oil
bench scale experiments, the bio-oil yields were very high in the 425–500 °C range (close to 55 wt%) due to the
Char
suitable features of the CSBR (high heat and mass transfer rates and short residence time of the volatiles), but
they are lower for higher temperatures due to the promotion of secondary cracking reactions. Compared to
lignocellulosic biomasses, the orange waste produced a bio-oil with more methanol and furfural and less phe-
nolic species, which is an encouraging fact for its stability and valorization by catalytic cracking or steam re-
forming. The high concentration of CO2 in the gas is a drawback for use for energy production. The char yield
(33–27 wt%) was high in the whole range of temperatures studied and its high carbon content (71–73 wt%) and
HHV (≈27 MJ kg−1) are suitable for use as fuel.

1. Introduction
Strategies for the production of bio-fuel and bio-based chemicals are
essential in order to gradually replace fossil fuel derived counterparts
and so reduce the environmental impacts associated with the emission
of greenhouse gases [1]. Amongst the different agro-forestry wastes
generated during harvesting or industrial processing [2,3], citrus fruit
wastes are a source of abundant, inexpensive and readily available re-
newable energy [4]. These wastes are composed of seeds, peels and
pulp and are the main by-product of juice industries, accounting for
about 50 wt% of the raw fruit processed, with their moisture content
being approximately 82 wt% [5]. Brazil is the world’s largest producer
of orange juice, accounting for approximately 53% of the orange juice
produced worldwide and 80% of the international trade in this product
[5]. Accordingly, a large amount of waste results from the juice pro-
cessing industry, which in 2011 reached about 9.3 million tons [6]. In
Spain, approximately 500,000 t of citrus wastes are produced per year,
with over 60% being derived from orange juice squeezing [7].
These wastes have no commercial applications, and they are usually
spread on soil areas adjacent to the orange juice industries, used as
compost, for animal feed or they are just burned in open fields [5,8].
Therefore, they are an unavoidable waste from the juice industry, in-
volving serious environmental concerns [4]. Currently, some bio-
chemical processes (anaerobic digestion, ethanol fermentation, re-
covery of flavonoids and chemicals) and thermochemical ones
(pyrolysis, gasification and/or combustion) have been proposed as en-
vironmentally friendly routes for the valorization of these wastes
[9,10]. Although each technique has its advantages and drawbacks,
pyrolysis, and specifically fast pyrolysis, has been widely used to con-
vert biomass and organic residues into diverse products (bio-oil, gas
and char) [11,12]. This technology is versatile, simple and has a low
investment cost, which allows decoupling its deployment on a moderate
scale in the regions where the raw material is available from the further
upgrading of the bio-oil in large-scale bio-refineries [3].
Most of biomass pyrolysis studies have been focused on lig-
nocellulosic wastes, which are made up of hemicellulose, cellulose and

⁎
Corresponding author.
E-mail address: martin.olazar@ehu.es (M. Olazar).

https://doi.org/10.1016/j.fuel.2018.03.028
Received 26 July 2017; Received in revised form 2 March 2018; Accepted 5 March 2018
Available online 17 March 2018
0016-2361/ © 2018 Elsevier Ltd. All rights reserved.
J. Alvarez et al. Fuel 224 (2018) 111–120

lignin. The chemical composition of citrus juice wastes differ from Table 1
lignocellulosic ones, as citrus peels have also an significant content of Properties of the orange wastes used.
pectin in addition to cellulose, hemicellulose and lignin [13]. Further-
Ultimate analysisa (wt%)
more, they also contain other constituents, such as sugars, fats and Carbon 42.7
proteins, in a significant amount [14]. Pectin is a primary constituent in Hydrogen 6.4
the cell wall structure of terrestrial plants, and basically consists of 4- Nitrogen 1.0
linked D-galacturonic acid units with several degrees of methylation of Oxygenb 47.6b
Proximate analysisa (wt%)
the carboxyl group [15]. According to the relevant literature, pectin Volatile matter 74.1
decomposes at lower temperatures (between 150 and 250 °C) than Fixed carbon 23.6
hemicellulose or cellulose, which decompose between 220 and 315 °C Ash 2.3
and between 315 and 400 °C, respectively [16–18]. Therefore, the Moisture (wt%) 1.5
HHV (MJ kg−1) 19.4
thermal degradation behavior and the products derived from the fast
pyrolysis of pectin-containing wastes, especially those in the bio-oil, Chemical analysis of ashes (wt%)
SiO2 0.29
will be different from those commonly obtained in the pyrolysis of
Al2O3 0.33
lignocellulosic materials [19]. In fact, Kim et al. [13,16] predicted that Fe2O3 0.09
the higher content of pectin (16–43 wt%) and lower of lignin (0.8–9 wt MgO 4.78
%) in citrus wastes than in woody biomasses leads to larger amounts of CaO 29.47
methanol and lower of aromatics in the bio-oil. Na2O 1.98
K2 O 30.9
Although some papers deal with the pyrolysis of citrus peel wastes,
TiO2 0.02
most of these studies have been performed in thermogravimetric P2O5 8.34
equipment [4,13,16], semibatch/fixed bed reactors [2,10,19,20] or SO3 3.46
rotary kilns [21]. However, none of them have been performed in
a
fluidized bed or conical spouted bed reactors (CSBR), which can op- b
On a dry basis.
Calculated by subtraction.
erate in continuous mode, under isothermal regime, with very high heat
and mass transfer rates between phases, without segregation and low
residence time of the volatiles (conditions required for maximizing bio- 2. Materials and methods
oil production and minimizing gas formation [22,23]). The CSBR
technology is an alternative to conventional fluidized beds, and per- 2.1. Orange wastes characterization
forms well in the fast pyrolysis of different types of lignocellulosic
wastes [3,24–26], sewage sludge [23], as well as other waste materials, Orange wastes were collected from an orange juice dispenser. The
such as tyres [27] or plastics [28]. The main advantage of the CSBR is samples were dried in an oven for 24 h at 60 °C in order to reduce the
the fact that the vigorous cyclic movement of the solid allows handling moisture content and at the same time avoid the release of some ex-
particles that are bigger than those allowed in fluidized beds, of irre- tractives. After the drying process, the water content in the orange peel
gular texture, fine materials and sticky solids, with no agglomeration was reduced to 1.5 wt%. Next, these wastes were ground in a mill
and segregation problems [29]. Accordingly, the costs derived from (Retsch ZM 100) and sieved in order to remove particles finer than
orange waste grinding and milling are considerably reduced. Further- 1 mm, which may easily elutriate from the reactor without being
more, this reactor has a simple design (no distributor plate) and re- completely reacted. The main properties of the orange wastes are
quires lower volumes than fluidized beds for the same throughput [30]. summarized in Table 1. The ultimate and proximate analyses have been
It should be remarked that few papers deal with a qualitative carried out in a LECO CHNS-932 elemental analyzer and in a TGA
composition of the bio-oil derived from orange/tangerine peel, but Q500IR thermogravimetric analyzer, respectively. Table 1 also includes
none of them have determined quantitatively the concentration of each the ash chemical composition determined by X-ray fluorescence (model
compound in the liquid. For example, Lopez-Velazquez et al. [4] de- AXIOS, PANalytical).
tected by simultaneous TGA-FTIR analysis the presence of carboxylic Moreover, the analysis of the data obtained by TGA enabled the
acids, aldehydes/ketones, alcohols and phenolic compounds in the determination of the main chemical constituents of this waste (pectin,
volatiles evolved from orange waste pyrolysis. Kim et al. [9] observed hemicellulose and cellulose and lignin + proteins + fats). The kinetic
large amounts of acids, ketones and aldehydes, a low amount of phenols model implemented is based on a multi-component mechanism, which
and the presence of some nitrogen species in the bio-oil from the non- describes the volatile formation by means of four concurrent (in-
catalytic pyrolysis of tangerine residue. Likewise, they also identified dependent and parallel) reactions corresponding to the decomposition
certain tangerine peel pyrolytic products by Py-GC/MS and the major of the four main biomass pseudo-components. According to Boluda-
areas observed in the chromatograms corresponded to acetic acid and Aguilar et al. [7], pectin decomposes between 180 and 300 °C, hemi-
methanol [19]. Besides, the same authors [13] concluded in a later cellulose between 200 and 300 °C and cellulose between 315 and
study that Citrus Unshui peel fast pyrolysis produced larger amounts of 430 °C, whereas the remaining components (lignin, proteins and fats)
furans, ketones and methanol than wood powders [13]. decompose in a wider temperature range, 150–700 °C. Similar decom-
This study focuses on the valorization of orange waste by con- position temperatures for each pseudo-component were reported by
tinuous flash pyrolysis in a CSBR with the objective of maximizing the Kim et al. [16,19] (210, 270 and 368 °C for pectin, hemicellulose and
bio-oil yield at temperatures between 425 and 600 °C. A quantitative cellulose, respectively). Given that lignin decomposes in a wide range of
analysis of the bio-oil composition has been carried out in order to temperatures, and fats, sugars and proteins overlap in the DTG curve,
assess different bio-oil upgrading routes. In addition, the characteristics one of the pseudo-components is assigned to the sum of lignin, sugar
of the gas and char (and their possible applications) are also evaluated and protein content in the orange waste. Details about the approx-
in order to improve the overall economy of the process for a successful imation considered and on the multi-component mechanism have been
industrial implementation. Given that the composition of orange wastes reported in a previous paper [18], in which the authors successfully
differs from that of typical biomasses, this study also provides a determined the concentration of each pseudo-component (hemi-
methodology for the quantification by thermogravimetry of the main cellulose, cellulose and lignin) in forestry residues from the Portuguese
chemical components of these types of wastes (pectin, hemicellulose, Central Inland Region following the same procedure.
cellulose and lignin), which is based on a previous one developed by Hence, the overall kinetic model studied considers four parallel
Amutio et al. [18,31] for other biomass materials. independent reactions of first order as follows:

112
J. Alvarez et al.
4
⎛− 1 ⎞ ⎛ dw ⎞ = wi−w∞,i ⎤
⎜ ⎟ ∑ ci k i ⎡
⎝ w0−w∞ ⎠ ⎝ dt ⎠ ⎢ w0,i−w∞,i ⎥
i=1 ⎣ ⎦
with conversion being defined as follows:
w −w
X= 0
w0−w∞
Thus, the kinetic Eq. (1) can be expressed as;
dX
= c1 k1 (X∞,1−X1) + c2 k2 (X∞,2 −X2 ) + c3 k3 (X∞,3−X3) + c4 k 4 (X∞,4 −X 4 )
dt
(3)
Where Xi and X∞,i are the conversion at t time and the final con-
version of each pseudo-component (defined according to Eq. (2)), and ci
is the mass concentration of pectin, hemicellulose, cellulose and lignin/
proteins/fats in the orange waste. Besides, as the kinetic constants
follow the Arrhenius empirical equation, the frequency factor (k0i) and
activation energy (Ei) for each pseudo-component have been de-
termined. Accordingly, this methodology allows calculating ci, k0i and
Ei for each pseudo-component by fitting the experimental results of
DTG curve to the kinetic model proposed in Eq. (3). The theoretical
DTG curve is obtained by minimizing the following error objective
function
L
∑ [(DTG )calculated−(DTG )experimental ]2
j=1
EOF =
L (4)
where L is the number of experiments points used in the fitting.
The comparison between the experimental results and those calcu-
lated with the kinetic model for the four pseudo-components is dis-
played in Fig. 1. The temperatures of 205 °C, 273 °C, 350 °C and 405 °C
are assigned to the maximum decomposition rates of pectin, hemi-
cellulose, cellulose and lignin/proteins/sugars, respectively As ob-
served, the kinetic model proposed suitably fits the experimental results
and although there are certain deviations at temperatures higher than
600 °C, the simplicity of the methodology makes it especially inter-
esting for determining the approximate composition of this waste.
The kinetic parameters, as well as the mass concentration of each
pseudo-component considered in the model, are depicted in Table 2.
The values of ci have been provided on a dry basis. Thus, the orange
waste used in this study is approximately made up of 35.3% pectin,
16.6% hemicellulose, 17.1% cellulose, with the remaining 28.7% being
lignin, proteins and soluble sugars, as well as some fats.
These values and the decomposition peaks are consistent with those
reported by Boluda-Aguilar et al. [7] and Kim et al. [16], who
pectine
hemicellulose
cellulose
lignin/sugars/proteins
calculated
experimental
Fig. 1. Comparison of experimental results (points) and those estimated using the kinetic
model (thick line) for the evolution of DTG with temperature.
Fuel 224 (2018) 111–120
Table 2
Values of frequency factor, k0, activation energy, E, and concentrations of the four
(1)
pseudo-components in the pyrolysis of the orange waste.
Log K0 (min−1) E (kJ mol−1) Composition (wt%)
Pectin 6.95 56.1 35.3
(2) Hemicellulose 10.49 84.3 16.6
Cellulose 19.71 157.7 17.1
Lignin/sugars/proteins 2.08 36.3 28.7
determined the chemical composition of tangerine peel and Citrus Un-
shui peel respectively, by thermogravimetry. Furthermore, the kinetic
parameters obtained are within the range of those obtained by other
authors with other types of biomass materials [18,32]. Moreover, ac-
cording to the literature, [7,14] the lignin content in orange peels is
usually between 1 and 10 wt%, whereas sugars and proteins are in the
8–16 wt% and 6–9 wt% range, respectively. Thus, the sum of lignin,
sugars and proteins is in the same range as that reported in this work,
which accounted for 31 wt%.
In addition, Fig. 1 also reveals that most of the volatile matter has
been released at temperatures ranging from 400 °C to 550 °C, thus
suggesting that the minimum temperature for pyrolysis should range
between 400 and 425 °C for converting most of the fruit waste into
volatiles for subsequent recovery of potentially useful products (bio-oil
and gases) [8].
2.2. Experimental equipment and procedure
The continuous bench-scale plant used for the fast pyrolysis of or-
ange waste is shown in Fig. 2. The plant was designed and fine-tuned
for the pyrolysis of different solid wastes [23,25,27].
The plant consists of the following elements: (1) solid and (2) gas
feeding system, (3) pyrolysis reactor, (4) a device for retaining the fine
particles elutriated from the reactor, (5) a liquid collection device and
(6) a system for gaseous product analysis. The feeding system consists
of a vessel equipped with a vertical shaft connected to a piston placed
below the orange waste, allowing its continuous feeding (up to
200 g h−1). Besides, as the piston rises, an electric engine is activated to
vibrate the whole system and facilitate the feeding of the waste material
into the reactor. Moreover, the pipe that connects the feeding system
with the reactor is cooled by tap water to avoid the degradation of the
feed prior to entering the reactor, and therefore the blockage of the
entire system.
Nitrogen has been used as fluidizing agent and its flow rate is
controlled by a mass flow meter that allows feeding up to 30 L/min.
Prior to entering the reactor, the gas is heated to the reaction tem-
perature by means of a preheater.
The plant’s main component is the CSBR which has a lower conical
section joined to an upper cylindrical one. The detailed design and
Fig. 2. Schematic diagram of the pyrolysis bench scale plant.
113
J. Alvarez et al. Fuel 224 (2018) 111–120

dimensions of the CSBR have been described elsewhere [25]. It is no- Table 3
teworthy that this technology allows using draft tubes of different de- FID response factors for the more representative compounds in the bio-oil.
signs and dimensions in order to optimize the spouting regime char-
Compound Molar mass (g Formula Response factor
acteristics, improve its stability and, at the same time, reduce the mol−1)
fluidizing agent requirements [33]. The design and main dimensions of
that used in the present study are shown elsewhere [27]. Methanol 32 CH4O 3.5
Acetaldehyde 45 C2H5O 2.2
The product stream exiting the reactor is a combination of vapors,
Acetone 58 C3H6O 2.2
aerosols and polar molecules together with the N2 used as fluidizing Acetic acid 60 C2H4O2 4.6
agent. In order to retain the fine solids elutriated from the reactor, this Propanoic acid 74 C2H6O2 4.0
stream passes through a high-efficiency cyclone and a 25 µm sintered Pyridine 79 C5H5N 1.2
steel filter. Both are located in a hot box heated to 290 °C in order to Phenol 94 C6H6O 1.2
Furfural 96 C5H4O2 2.4
avoid the condensation of heavy compounds. Subsequently, the gases
3-Methyl-2-cyclopentenone 96 C6H8O 1.5
circulate through a volatile condensation system consisting of a double 2-Furanmethanol 98 C5H6O2 2.4
shell tube condenser and a 60 µm stainless steel filter, both cooled by 2-Pentanone 98 C5H10O 1.4
tap water. It is noteworthy that high nitrogen flow rates, as well as the 2-Methyl-phenol 108 C7H8O 1.1
1,2-Benzenediol 110 C6H6O2 1.8
formation of non-condensable gases, pose significant heat and mass
Hydroquinone 110 C6H6O2 1.6
transfer resistance and reduce partial pressure, which means a reduc- 3-Methyl-2,5-Furandione 112 C5H4O3 2.0
tion in the collection efficiency of the liquid product in tube-shell 5-(Hydroxymethyl)-2-furfural 126 C6H6O3 2.6
condensers. In order to maximize the heat transfer rate and improve Levoglucosan 162 C6H10O5 5.0
bio-oil collection efficiency, the cooling area was dramatically in-
creased by inserting a 60-µm stainless steel filter, which plays the role
of an impingement device. Thus, the gas collides with the cool surface compounds. Table 3 shows the FID response factors experimentally
of the filter, which promotes the retention of fine droplets and aerosols determined for the most representative compounds identified in the
in the gaseous stream. Moreover, the steel filter is located inside a 1 L bio-oil. As observed, the response factors depend on the nature of the
vessel where the bio-oil condensed is collected. The liquid falls then by compound (type of bond between carbon and oxygen), as well as on
gravity to a glass vessel throughout the run length and is finally their carbon/hydrogen/oxygen ratio. For example, acetone has a re-
weighed and stored for its characterization. In order to ensure retention sponse factor of 2, but a ketone, such as 2-pentanone, has a lower one
of all the bio-oil, a coalescence filter has been placed subsequent to the (1.4) due to its lower content of oxygen than acetone. A fact to note is
bio-oil collection vessel. the high response factor for levoglucosan due to its high content of
In order to study the effect of temperature on product yields and oxygen. This means that all saccharides with similar carbon/hydrogen/
composition, runs have been carried out at 425, 500 and 600 °C. Each oxygen ratio have a response factor close to 5. In the case of phenol, the
experiment has been carried out in continuous mode by feeding response factor (1.2) is rather low and slightly lower for 2-methyl-
1 g min−1 of orange waste. The bed was made up of 100 g of sand phenol (1.1) because it content of oxygen is lower. In view of these
(0.4–0.6 mm), which guaranteed a fast heating of the feedstock and facts, response factors have been assigned to the compounds in the bio-
isothermicity of the bed. The fluidizing gas (N2) flow rate was con- oil depending on their nature (chemical family) and their carbon/hy-
siderably reduced compared to that used in previous works [25,26] by drogen/oxygen ratio, based on the values of response factors obtained
using a draft tube, i.e., only 7 L min−1 were required for ensuring stable experimentally for the major compounds detected in the bio-oil. These
spouting, which corresponds to 1.2 times that for minimum spouting. values are set out in Table S1.
In each run, 1 g min−1 of orange waste has been fed continuously The water content in the bio-oil has been measured by Karl-Fisher
for 50 min to obtain reproducible results in the analysis of the gaseous titration (ASTM D1744). The surface area and pore volume of the chars
products. Furthermore, each run has been repeated three times in order have been determined from nitrogen adsorption-desorption isotherms
to obtain significant results with high confidence interval. The mass obtained in a Micromeritics ASAP 2010. Char has also been char-
balance has been closed based on the microGC analysis and on the in- acterized by means of elemental and proximate analyses in a LECO
formation obtained by weighing the bio-oil collected in the vessel and CHNS-932 elemental analyzer and in a TGA Q500IR thermobalance,
the char collected through the lateral outlet, as well as the char col- respectively. The higher heating value of the char was determined using
lected in the cyclone and reactor at the end of the continuous run, an isoperibolic bomb calorimeter (Parr 1356).
which allowed attaining a balance closure of around 93%.
3. Results and discussion
2.3. Product analysis
3.1. Product yields
The analysis of the gaseous stream has been conducted in-line by
means of a micro-GC (Varian 4900), which allows the identification and Orange waste pyrolysis products have been grouped into three dif-
quantification of previously calibrated gaseous products, by means of ferent fractions: gas, liquid (bio-oil) and solid (char). Fig. 3 shows the
three independent modules (Molecular sieve, Porapak and Plot influence of temperature on product distribution. As observed, bio-oil is
Alumina). In order to determine the gas yield and validate the mass the major product in the whole range of temperatures studied with the
balance, an internal standard (cyclohexane, not formed in the process) maximum yield at 500 °C (54.9 wt%). However, the yield at 425 °C is
has been introduced downstream the condensation system with a con- almost the same (54.6 wt%), which is a great advantage from an eco-
stant flow of 0.005 ml min−1. nomic point of view, given that a similar amount of bio-oil can be
The identification of the liquid fraction collected has been carried produced with lower energy requirements. Gas yield increases with
out in a gas chromatograph/mass spectrometer (GC/MS, Shimadazu temperature from 12 wt% at 425 °C to 24 wt% at 600 °C, whereas that
UP-2010S), while the quantification of the identified compounds has of char decreases from 33 wt% to 27%, respectively. At higher tem-
been performed in a chromatograph (Agilent 6890) equipped with a peratures, excess heat to that required for maximum bio-oil yield is
flame ionization detector (FID). Given that the bio-oil is made up of a supplied, and therefore orange waste devolatilization and secondary
great number of oxygenated compounds and, unlike hydrocarbons, reactions like cracking are enhanced, thus favoring gas formation at the
their FID response is not proportional to mass, a calibration has been expense of bio-oil and char.
performed to calculate the FID device’s response factors to oxygenated Although huge amounts of orange wastes are discarded all over the

114
J. Alvarez et al. Fuel 224 (2018) 111–120

Fig. 4. Influence of temperature on gas product composition.

Fig. 3. Effect of temperature on the distribution of gas, solid and liquid fractions in the
orange waste pyrolysis.
dehydrogenation and cracking reactions.
The LHV values for the gases obtained at different temperatures are
world, very few studies in the literature focus on the pyrolysis of this
shown in Fig. 4. The values have been determined following the
material. Overall, the bio-oil yield obtained in the CSBR is higher than
equation deduced by Lv et al. [40] for the gas obtained in the biomass
those reported for the pyrolysis of orange waste or similar citrus fruit
air/steam gasification. As observed, LHV values increase by increasing
wastes, which range from 35 to 53 wt% [2,4,13,20,34]. Most of these
temperature due to the increase in the concentration of CO, H2 and light
studies have been developed in fixed bed or semi-batch reactors.
hydrocarbons and the decrease in that of CO2. In view of the compo-
Therefore, the high yield of bio-oil and the low one of the gas fraction at
sition of this gaseous stream (high CO2 concentration), especially of
moderate temperatures evidences the efficiency of CSBR for the fast
that obtained at 425 and 500 °C, as well as its low LHV compared to
pyrolysis of orange waste due to both the high heat and mass transfer
other conventional gaseous fuels, the best option is to use this stream
rates promoting devolatilization reactions and the short residence times
for reducing fluidizing agent requirements, which usually involve
of the volatiles hindering secondary cracking reactions and so enhan-
considerable operational costs in fast pyrolysis process. The gas is sui-
cing gas formation.
table for burning to provide heat to the pyrolysis process, but only that
It is noteworthy that the char and gas yields obtained in this work
obtained by operating at 600 °C, whose LHV reached up to 8.5 MJ m−3.
are significantly higher (and consequently that of bio-oil lower) than
In a previous work in which biomass fast pyrolysis was performed in a
those obtained in previous studies in the pyrolysis of lignocellulosic
conical spouted bed at larger scale (25 kg/h pilot plant), part of the gas
biomasses, such as pinewood sawdust [24] and agro-forestry wastes
was recirculated and used as fluidizing gas for the reactor (which
[3,25], using the same technology. Given that the ash content and
minimized N2 requirements), whereas the excess of non-condensable
composition do not differ significantly in these biomasses, the differ-
gases were purged and burnt in a flare [30].
ences should be attributed to the composition of the orange waste,
Similar gas compositions were observed by other authors in the
mainly the presence of pectin. In fact, apart from the contribution of the
pyrolysis of orange/tangerine peel, with the major compound being in
lignocellulosic components to the formation of char (especially lignin
all cases CO2 at temperatures between 400 and 600 °C [20,34]. These
and hemicelluloses, as stated by Collard et al. [35]), several papers in
results also confirm that the presence of pectin promotes CO2 forma-
the literature report that pectin thermal decomposition under non-
tion, given that the fast pyrolysis (using the CSBR technology) of other
oxidative atmosphere leads to high char yields, around 20–25% at
lignocellulosic wastes, such as pinewood sawdust [24] or rice husk
temperatures higher than 600 °C [15,17,36–38], which explains the
[25], leads to a CO2 concentration of around 30–50% vol. at tem-
greater yield of solid residue. Moreover, pectin pyrolysis also leads to
peratures between 400 and 500 °C, which is almost half of that ob-
high yields of uncondensable gases, such as CO and CO2, which explains
served in this work. According to Ge et al. [15], pectin is basically made
the higher gas yields in this study [17,36,39]. As stated by Kim et al.
up of D-galacturonic acid units joined in chains by means of α-(1-4)
[16], the different composition of tangerine peel leads to a more spe-
glycosidic linkage. These uronic acids have carboxyl groups that are
cific pyrolysis kinetics, thereby leading to a different product distribu-
removed via decarboxylation reactions during fast pyrolysis, thereby
tion compared to other biomasses such as wood.
releasing higher amounts of CO2 than typical lignocellulosic biomasses,
in which hemicellulose is mainly responsible for CO2 formation
3.2. Gas composition [17,41].

Fig. 4 shows the gas composition and the lower heating value (LHV)
for each gas as a function of temperature. As observed, the gas stream
consists mainly of CO2 and CO, with the former being the major com-
pound in the whole range of temperatures studied (a concentration in
the 80–45% vol. range). CO concentration increases with temperature
due to the severity of devolatilization and secondary cracking reactions,
whereas that of CO2 decreases considerably because higher tempera-
tures enhance decarbonylation rather than decarboxylation [17,25]. In
addition, the gaseous fraction is also composed of small amounts of H2
and C1–C4 hydrocarbons, whose yield and concentration increase with
temperature (especially that of H2) due to the enhancement of
3.3. Bio-oil composition
Bio-oil is a dark brown liquid composed of a large number of oxy-
genated compounds. Therefore, in order to simplify the results, the
individual compounds have been grouped according to their functional
groups, Table 4. The liquid consists of alcohols, aldehydes, carboxylic
acids, furans and phenols, amongst others, which is a consequence of
the features of the components in the pectin, hemicellulose, cellulose
and lignin contained in the orange waste. The chemical composition of
the bio-oil has been determined based on the identification and quan-
tification of around 70 compounds, which account for 90 wt% of the

115
J. Alvarez et al. Fuel 224 (2018) 111–120

Table 4
Detailed liquid composition (wt%) at different temperatures.

Compounds (wt%) Formula 425 °C 500 °C 600 °C

Oxygenated compounds 43.21 47.34 42.95

Alcohols 5.27 4.74 5.18

Methanol CH4O 4.71 4.04 3.49

Cis-1,2-Cyclohexanediol C6H12O2 0.01 0.02 0.01
1,2-ethanediol, monoacetate C4H8O3 0.39 0.52 1.60
2-butanol, 3-methyl-, acetate C7H14O2 0.15 0.16 0.08

Carboxylic acids 6.62 6.06 5.35

Acetic acid C2H4O2 6.25 5.86 5.10
2-oxo-, methyl ester propanoic acid C4H6O3 0.21 0.05 0.15
Ethenyl ester propanoic acid C5H8O2 0.08 0.10 0.09
4-oxo-, methyl ester pentanoic acid C6H10O3 0.09 0.06 0.02

Ketones 9.53 13.98 13.45

Acetone C3H6O 3.59 6.61 7.53
3-Penten-2-one C5H8O 0.03 0.03 0.21
1-(Acetyloxy)-2-butanone C6H10O3 0.10 0.26 0.15
4-(Acetyloxy)-2-butanone C6H10O3 1.51 2.34 0.83
4-Hydroxy-4-methyl-2-pentanone C6H12O2 0.39 0.20 0.64
2-Butanone C4H8O 0.26 0.26 0.63
1-(Acetyloxy)-2-propanone C5H8O3 0.81 1.17 1.23
2,5-Hexanedione C6H10O2 0.01 0.02 0.01
3-Methyl, 3-penten-2-one C6H10O 0.01 0.01 0.01
3-Methyl-2-cyclopenten-1-one C6H8O 0.14 0.19 0.21
4-Methyl-6-hepten-3-one C8H14O 0.16 0.17 0.05
2-Hydroxy-3-methyl-2-cyclopenten-1-one C6H8O2 1.10 1.02 1.08
2,3-Dimethyl-2-cyclopenten-1-one C7H10O 0.29 0.43 0.16
2,4-Dimethyl-2-cyclopenten-1-one C7H10O 0.06 0.08 0.05
2,3-Dihydro-3,5-dihydroxy-5-methyl-4H-Pyran-4-one C6H8O4 0.15 0.17 0.05
2,3-Dihydro-3,5-dihydroxy-6-methyl-4H-Pyran-4-one C6H8O4 0.14 0.24 0.16
1-(1-cyclohexen-1-yl) ethanone C8H12O 0.12 0.16 0.13
2-Pentanone C5H10O 0.13 0.21 0.23
3-Ethyl-2-hydroxy-2-cyclopenten-1-one C7H10O2 0.24 0.23 0.02
3-Methyl-2,4-hexanedione C7H12O2 0.19 0.10 0.02
Maltol C6H6O3 0.10 0.09 0.07

Furans 20.32 21.47 18.70

Furfural C5H4O2 7.83 12.66 11.85
5-Methylfurfural C6H6O2 1.62 1.62 0.97
5-Hydroxymethylfurfural C6H6O3 7.19 3.69 2.27
2-Furanmethanol C5H6O2 0.13 0.54 0.14
2(5H)-Furanone C4H4O2 0.76 1.03 1.27
3-Methyl-2,5-Furandione C5H4O3 0.04 0.02 0.01
3-Methyl-2(5H)-Furanone C5H6O2 0.44 0.18 0.05
5-Methyl-2(5H)-Furanone C5H6O2 0.06 0.07 0.01
2-Acetyl-5-methylfuran C7H8O2 0.04 0.06 0.06
Methyl 2-furoate C6H6O3 0.10 0.07 0.05
2,5-Dimethyl-4-hydroxy-3(5H)-Furanone C6H8O3 0.43 0.35 0.07
Butyrolactone C4H6O2 1.69 1.18 1.93

Phenols 1.65 1.71 1.10

Phenol C6H6O 0.23 0.17 0.09
4-Methyl-phenol C7H8O 0.45 0.60 0.45
2-Methyl-phenol C7H8O 0.05 0.10 0.08
3-Ethyl-phenol C8H10O 0.03 0.02 0.01
1,2-Benzenediol C6H6O2 0.36 0.34 0.13
Hydroquinone C6H6O2 0.38 0.34 0.21
2-Methyl-1,4-benzenediol C7H8O2 0.14 0.16 0.13

Saccharides 3.51 2.88 2.78

1,4:3,6-Dianhydro-alpha-d-glucopyranose C6H8O4 1.99 1.97 1.87
D-allose C6H12O6 0.25 0.26 0.28
Levoglucosan C6H10O5 1.09 0.44 0.57
Beta-d-glucopyranose C7H14O6 0.17 0.20 0.05

Nitrogenous compounds 0.24 0.51 0.54

Pyridine C5H5N 0.24 0.51 0.54

Hydrocarbons 0.02 0.02 0.01

2,3-Dimethyl-2,4-hexadiene C8H14 0.02 0.02 0.01

Unidentified 9.90 7.82 8.46

Water H2O 42.93 ± 2.21 40.81 ± 1.94 44.45 ± 2.84

Total 100.00 100.00 100.00

116
J. Alvarez et al.
total liquid.
Water is the major product in the bio-oil and its content does not
change significantly and remains in the 40–45 wt% range between 400
and 600 °C. The content of water is slightly higher than in the bio-oils
derived from pinewood sawdust [24] or rice husk [25] fast pyrolysis in
a CSBR, probably due to the presence of pectin. According to most of
the relevant literature, water comes from the initial moisture of the
material, and is formed by dehydration reactions of cellulose and
hemicelluloses, as well as secondary reactions of intermediate products
[35,42–44]. However, Aburto et al. [17] determined that water was
also one of the main compounds released during citrus pectin pyrolysis.
As observed in Table 4, furans are the organic functional group with
the highest concentration, with furfural being the most abundant
compound in the whole range of temperatures studied. Its yield peaks at
500 °C and then decreases notably at 600 °C, which is also observed in
the other compounds within aldehydes due to the secondary vapor-
phase reactions. Ge et al. [15] also observed a high formation of fur-
fural and 5-methylfurfural in the pyrolysis of pectin. According to these
authors, furfural is mainly derived from pectin decomposition, specifi-
cally from galacturonic acid subunits at the end chains of the pectin,
and they postulated a route based on the dehydration chemistry of
galacturonic acid. Furfural is also obtained (at lower extent) from the
levoglucosan formed by the pyrolysis of cellulose and hemicellulose as
well as from the cracking of 5-methylfurfural (the yield of both le-
voglucosan and 5-methylfurfural decreases at 500 °C whereas that of
furfural increases) [45]. Moreover, Kim et al. [16] determined that 5-
hydroxymethylfurfural is formed from the dehydration of sugars, such
as glucose and fructose, which are also contained in the orange peel.
Other furan derivatives are also formed from hemicellulose primary
decomposition at temperatures above 400 °C [46].
Ketones make up the second major organic group in the bio-oil and
the main compound is acetone with a maximum concentration of
7.53 wt% at 600 °C. It is well documented that ketones are formed by
condensation reactions involving the carbohydrate-derived fraction and
decomposition reactions involving miscellaneous oxygenates, such as
saccharides and furans, in the pyrolysis of lignocellulosic biomasses
[43,47]. Nevertheless, pectin also contributes to the formation of these
species. Thus, Zhou et al. [39] also detected the formation of acetone
and butanone during the pyrolysis of pectin. In addition, Kim et al. [16]
also reported the presence of 2,3-Dihydro-3,5-dihydroxy-5-methyl-4H-
Pyran-4-one (a bioactive component with antimicrobial, anti-in-
flammatory and anti-proliferative activities) in the pyrolysis of Citrus
unshiu (tangerine type) peel. Furthermore, Ge et al. [15] observed that
2-Hydroxy-3-methyl-2-cyclopenten-1-one is also formed in the thermal
decomposition of pectin.
Carboxilic acids are also a significant group and its content de-
creases by increasing temperature, from 6.62 to 5.35 wt%. Similarly to
standard bio-oils, acetic acid is the main compound in this group and it
is commonly formed by the fragmentation of acetyl substituents con-
tained in the hemicellulose [35,42,43] and, to a lesser extent, by the
secondary cracking of saccharides, such as levoglucosan. The presence
of these compounds confer upon the bio-oil an undesirable acidic
nature (pH between 2 and 4), which could lead to corrosion problems in
common structures during storage or upgrading processes [48].
Within the alcohols group, methanol is the compound with the
highest concentration. Methanol is formed from the cleavage of
methoxy groups in the uronic acid and methyl esters and ethers derived
from pectin [13]. Pectin is decomposed at lower temperatures than
other biomass constituents, and therefore the formation of this com-
pound is favored at 425 °C (4.71 wt%). At higher temperatures, the
generation of other compounds, as well as the severity of cracking re-
actions, reduces the concentration of methanol in the bio-oil [19]. Zhou
et al. [39] studied pectin decomposition by thermogravimetry and de-
termined that methanol formation peaked at 253 °C and then decreased
continuously with temperature. This result suggests that methanol
concentration would be higher working at temperatures lower than
117
Fuel 224 (2018) 111–120
425 °C. Nevertheless, as shown in Fig. 1, orange waste will not be fully
decomposed at these temperatures, thereby reducing notably the bio-oil
yield.
An interesting characteristic of this bio-oil is the low presence of
phenolic compounds, which are mainly linked to lignin fragmentation.
The concentration of these species in the whole range of temperatures
studied is lower than 2 wt%, which is much lower than in typical bio-
oils derived from lignocellulosic biomasses, whose content ranges be-
tween 5 and 20 wt%, depending on the feedstock [24,44]. This differ-
ence is attributed to the lower lignin content in the orange peel com-
pared to lignocellulosic wastes, thereby leading to a smaller phenolic
fraction. The same conclusion has been reported on the liquids derived
from citrus peel wastes [9,16].
The last functional group identified within the oxygenated com-
pounds corresponds to saccharides, whose content decreases with in-
creasing temperature due to the poor thermal stability of sugars. They
are normally formed at intermediate temperatures from cellulose de-
gradation [45], although hemicellulose also contributes to their pro-
duction [35]. According to Kim et al. [16], sugars like levoglucosan
seems to decompose by interactions with intermediates formed from
pectin and hemicellulose. It should be noted that this bio-oil also con-
tains pyridine, a nitrogenous compound derived from the proteins
present in the orange peel, whose amount is in the 0.25–0.50 wt%
range.
Table 5 shows the ultimate analysis of the bio-oils obtained at dif-
ferent temperatures. As observed, the content of C peaks at 500 °C and
then decreases due to the enhancement of secondary cracking reactions
promoting the formation of non-condensable gases. The high oxygen
content in all the bio-oils is mainly associated with the high water
content of the samples. It is noteworthy that the carbon content is
slightly lower than typical lignocellulosic bio-oils (usually in the
33–48 wt% range), whereas the content of oxygen is slightly higher
(usually between 40 and 55%) due to the higher content of water as it
was revealed in Table 4 [49]. The concentration of N is low in all bio-
oils and does not change significantly with temperature.
It should be noted that the liquid collected in the condensation
system makes up a stable single phase and does not undergo any phase
separation. Although the water content (around 40 wt%) is higher than
that determined by Oasmaa et al. [50] for phase separation (22–31 wt
%), the high content of methanol and low of hydrophobic compounds
(high molecular weight lignin derivatives) formed from the high con-
tent of pectin and low of lignin, respectively, improve the stability of
the bio-oil contained in only one phase. In order to attain a deeper
insight on the role of methanol and water in bio-oil stability and phase
separation, solvent fractionation procedure proposed by Oasmaa et al.
[51] should be performed. However, in view of the results obtained in
this study, it can be concluded that the quality of these bio-oils is much
higher than those with a high content of lignin derived aromatic
compounds [45]. Furthermore, the low content of phenols makes bio-
oil catalytic upgrading more appealing, as they are very stable com-
pounds (and therefore more refractory to cracking reactions), whereas
other family compounds, such as furans and ketones (the most abun-
dant organic groups), are more easily converted (oxygen can be re-
leased more easily by breaking the double bonds in which it is attached)
[47]. Furthermore, phenolic species are the main responsible for the
rapid deactivation by coke deposition of Ni based catalysts in the steam
Table 5
Ultimate analysis of the bio-oils obtained at different temperatures.
Ultimate analysis (wt%) 425 °C 500 °C 600 °C
C 31.71 34.90 33.76
H 8.82 8.52 9.12
O 59.42 56.49 55.04
N 0.04 0.09 0.08
J. Alvarez et al.
Table 6
Properties of the char samples obtained in the pyrolysis of orange waste at 425, 500 and
600 °C.
425 °C 500 °C
Ultimate analysis (wt%)
C 71.6 71.8
H 3.6 2.9
N 1.8 1.5
O 14.7 13.9
Proximate analysis (wt%)
Volatile matter 23.8 18.9
Fixed carbon 67.9 71.2
Ash 8.3 9.9
HHV (MJ/kg) 26.1 26.8
Surface characteristics
BET surface area (m2 g−1) 1.5 3.2
Pore volume (m3 g−1) 0.002 0.003
reforming of raw bio-oil to produce hydrogen [52,53]. Besides, several
papers reported high conversions and H2 productions with low catalyst
deactivation when the feed is made up of model compounds, such as
methanol, furfural [54] and acetic acid [55], which are the major or-
ganic species in the orange peel derived bio-oil. Furthermore, the high
water content in this bio-oil has another advantage for steam reforming,
i.e., there is no need to separate the water from the bio-oil, as it hinders
catalyst deactivation by coke deposition and reduces the steam required
in the feed. It is clear that knowledge of the chemical composition of
pyrolysis liquids enables the choice of correct storage conditions and
appropriate future applications for the pyrolysis liquid.
3.4. Char characterization
The evaluation of char characteristics is of great relevance for de-
termining its possible applications and improving the economic aspects
of the fast pyrolysis process at large scale. Table 6 shows the ultimate
and proximate analyses, high heating value (HHV) and some surface
characteristics, such as BET surface area and pore volume. The ashes
are mainly retained in the carbonaceous matrix during pyrolysis and, as
observed in Table 6, their content increases with temperature due to the
promotion of devolatilization and cracking reactions leading to a higher
release of volatile matter (as observed in the proximate analysis).
Likewise, carbon content increases slightly from 71.6 wt% at 425 °C to
73.8 wt% at 600 °C, which positively affects the chars calorific values,
increasing from 26.1 to 27.5 MJ kg−1. Similar HHVs (in the
25–30 MJ kg−1 range) were obtained for the char from the pyrolysis of
orange peel [20] or citrus wastes [10] at temperatures between 300 and
600 °C. It is noticeable that the nitrogen content decreases slightly with
temperature and, interestingly, that of pyridine in the bio-oil (Table 4)
follows the opposite trend. Table 6 also shows the surface character-
istics of the chars, and almost of all of them have negligible BET surface
area, although it is slightly higher as the pyrolysis temperature in-
creased, which is due to pore formation by the release of volatiles.
Chars have been widely used as adsorbents (or precursors for acti-
vated carbon synthesis) [56,57], fertilizers (usually for soil amend-
ment) [58] or fuel for reducing the energy requirements in the pyrolysis
process [59]. In view of the results provided in Table 6, i.e., a high
carbon content, and therefore high HHV, with a low BET surface area
and low nitrogen content, this char is more suitable for use as fuel than
adsorbent or fertilizer. In fact, the values obtained for HHV are in the
same range as those from other solid fuels, such as lignite (28 MJ kg−1)
or coke (29 MJ kg−1).
However, in order to assess the char viability as fuel and char-
acterize its aromaticity, a conventional method in the literature lies in
the use of Van Krevelen diagram, which allows comparing the H/C and
O/C molar ratios of a given sample with those of the most common
Fuel 224 (2018) 111–120
600 °C
72.9
2.6
1.4
12.2
26.9
72.2
10.9
27.5
4.8
0.003
Fig. 5. Effect of temperature on the H/C and O/C molar ratios of the chars obtained in the
pyrolysis of orange waste (Van Krevelen diagram).
solid fuels [60,61]. As observed in Fig. 5, both H/C and O/C ratios
decrease by increasing temperature due to the more severe devolatili-
zation and depolymeration, with the consequence being carbon en-
richment in the chars [62], as well as the development of aromatic
structures [63]. Therefore, the increase in energy content by the pyr-
olysis process is explained by the increase in CeC bonds, which have
more energy than C-O and CeH ones. Moreover, a linear relationship
seems to be between these two ratios in the chars and those in the initial
orange peel, which is indicative that dehydration, decarboxylation and
decarbonylation reactions prevail during the pyrolysis process. This
observation is consistent with the gas yield shown in Fig. 3, where the
main products are CO2 and CO, and the results shown in Table 4, in
which water is the major product in the whole temperature range stu-
died [36]. It should be noted that the H/C and O/C molar ratios are
similar to those in certain coals, such as bituminous coal and anthracite,
and much lower than those in the initial biomass, which is an indication
of a lower formation of smokes and steam during the combustion, and
therefore a greater efficiency in the combustor [62]. Besides, according
to Laird et al. [59], NOx emissions from char combustion are compar-
able to those from the combustion of different coals, so there are
technologies available for the reduction and elimination of these pol-
luting gases. Hence, these chars have good perspectives for use as fuel
and, at the same time, reduce the energy requirements in the pyrolysis
process.
4. Conclusions
The conical spouted bed reactor (CSBR) has been proven to be a
suitable technology for the valorization of orange waste by continuous
fast pyrolysis, which is the main residue derived from juice manu-
facturing and whose disposal is becoming a serious concern in citrus
industries. The thermogravimetric analysis of this waste revealed the
presence of pectin (35.3 wt%), hemicellulose (16.6 wt%), cellulose
(17.1 wt%) and ashes (2.3 wt%), with the remaining content (28.7 wt
%) being associated with lignin, proteins, other sugars and fats.
The maximum yield of bio-oil (55 wt%, on a wet basis) is achieved
at 425–500 °C temperature range. The high bio-oil yield and low gas
yield obtained in this study compared to those in other ones dealing
with orange/tangerine peel pyrolysis are due to the characteristics of
the CSBR, i.e., high heat transfer rates and short residence times of the
volatiles. However, pectin containing wastes also lead to higher char
yields (between 27 and 32 wt%) than other woody biomasses (between
15 and 23 wt%) using the same technology.
The liquid obtained makes up a single phase and is composed of
water and oxygenated hydrocarbons. The organic fraction contains
118
J. Alvarez et al.
mainly furans, ketones, alcohols and carboxylic acids, with the major
compound in each group being furfural, acetone, methanol and acetic
acid, respectively. The low content of phenols is a highly relevant fact,
with is attributed to the low lignin content in the orange peel. These
features make highly attractive the catalytic upgrading of the bio-oil.
Moreover, given the HHV of the chars (around 27 MJ kg−1), a suitable
application lies in the supply of energy to the pyrolysis process, whereas
the gas obtained may be preferably used to reduce N2 requirements due
to its high CO2 content.
Acknowledgements
This work was carried out with financial support from the Ministry
of Economy and Competitiveness of the Spanish Government
(CTQ2016-75535-R (AEI/FEDER, UE) and CTQ2015-69436-R
(MINECO/FEDER, UE)), the Basque Government (IT748-13) and the
University of the Basque Country (UFI 11/39). Jon Alvarez also thanks
the University of the Basque Country UPV/EHU for his post-graduate
Grant (ESPDOC 2015). Maria Cortazar also thanks the Basque
Government for her training grant.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.fuel.2018.03.028.
References
[1] Anex RP, Aden A, Kazi FK, Fortman J, Swanson RM, Wright MM, et al. Techno-
economic comparison of biomass-to-transportation fuels via pyrolysis, gasification,
and biochemical pathways. Fuel 2010;89(Supplement 1):S29–35.
[2] Miranda R, Bustos-Martinez D, Blanco CS, Villarreal MHG, Cantú MER. Pyrolysis of
sweet orange (Citrus sinensis) dry peel. J Anal Appl Pyrolysis 2009;86:245–51.
[3] Amutio M, Lopez G, Alvarez J, Moreira R, Duarte G, Nunes J, et al. Flash pyrolysis of
forestry residues from the Portuguese Central Inland Region within the framework
of the BioREFINA-Ter project. Bioresour Technol 2013;129:512–8.
[4] Lopez-Velazquez MA, Santes V, Balmaseda J, Torres-Garcia E. Pyrolysis of orange
waste: a thermo-kinetic study. J Anal Appl Pyrolysis 2013;99:170–7.
[5] Rezzadori K, Benedetti S, Amante ER. Proposals for the residues recovery: orange
waste as raw material for new products. Food Bioprod Process 2012;90:606–14.
[6] Santos CM, Dweck J, Viotto RS, Rosa AH, de Morais LC. Application of orange peel
waste in the production of solid biofuels and biosorbents. Bioresour Technol
2015;196:469–79.
[7] Boluda-Aguilar M, García-Vidal L, González-Castañeda FdP, López-Gómez A.
Mandarin peel wastes pretreatment with steam explosion for bioethanol production.
Bioresour Technol 2010;101:3506–13.
[8] Lam SS, Liew RK, Lim XY, Ani FN, Jusoh A. Fruit waste as feedstock for recovery by
pyrolysis technique. Int Biodeterior Biodegrad 2016;113:325–33.
[9] Kim JW, Park SH, Jung J, Jeon J, Ko CH, Jeong K, et al. Catalytic pyrolysis of
mandarin residue from the mandarin juice processing industry. Bioresour Technol
2013;136:431–6.
[10] Volpe M, Panno D, Volpe R, Messineo A. Upgrade of citrus waste as a biofuel via
slow pyrolysis. J Anal Appl Pyrolysis 2015;115:66–76.
[11] Bridgwater AV. Review of fast pyrolysis of biomass and product upgrading. Biomass
Bioenergy 2012;38:68–94.
[12] Butler E, Devlin G, Meier D, McDonnell K. A review of recent laboratory research
and commercial developments in fast pyrolysis and upgrading. Renew Sustain
Energy Rev 2011;15:4171–86.
[13] Kim B, Kim Y, Jae J, Watanabe C, Kim S, Jung S, et al. Pyrolysis and catalytic
upgrading of Citrus unshiu peel. Bioresour Technol 2015;194:312–9.
[14] Rivas B, Torrado A, Torre P, Converti A, Domínguez JM. Submerged citric acid
fermentation on orange peel autohydrolysate. J Agric Food Chem 2008;56:2380–7.
[15] Ge S, Xu Y, Tian Z, et al. Pyrolysis study of pectin by tunable synchrotron vacuum
ultraviolet photoionization mass spectrometry. J Therm Anal Calor
2015;120:1399–405.
[16] Kim Y, Lee HW, Kim S, Watanabe C, Park Y. Non-isothermal pyrolysis of citrus
unshiu peel. Bioenergy Res 2015;8:431–9.
[17] Aburto J, Moran M, Galano A, Torres-García E. Non-isothermal pyrolysis of pectin:
a thermochemical and kinetic approach. J Anal Appl Pyrolysis 2015;112:94–104.
[18] Amutio M, Lopez G, Alvarez J, Moreira R, Duarte G, Nunes J, et al. Pyrolysis ki-
netics of forestry residues from the Portuguese Central Inland Region. Chem Eng
Res Design 2013;91:2682–90.
[19] Kim Y, Lee HW, Lee S, Kim S, Park SH, Jeon J, et al. Pyrolysis properties and
kinetics of mandarin peel. Korean J Chem Eng 2011;28:2012–6.
[20] Aguiar L, Márquez-Montesinos F, Gonzalo A, Sánchez JL, Arauzo J. Influence of
temperature and particle size on the fixed bed pyrolysis of orange peel residues. J
Anal Appl Pyrolysis 2008;83:124–30.
119
Fuel 224 (2018) 111–120
[21] Li AM, Li XD, Li SQ, Ren Y, Shang N, Chi Y, et al. Experimental studies on municipal
solid waste pyrolysis in a laboratory-scale rotary kiln. Energy 1999;24:209–18.
[22] Cui H, Grace JR. Spouting of biomass particles: a review. Bioresour Technol
2008;99:4008–20.
[23] Alvarez J, Amutio M, Lopez G, Barbarias I, Bilbao J, Olazar M. Sewage sludge va-
lorization by flash pyrolysis in a conical spouted bed reactor. Chem Eng J
2015;273:173–83.
[24] Amutio M, Lopez G, Artetxe M, Elordi G, Olazar M, Bilbao J. Influence of tem-
perature on biomass pyrolysis in a conical spouted bed reactor. Resour Conserv
Recycl 2012;59:23–31.
[25] Alvarez J, Lopez G, Amutio M, Bilbao J, Olazar M. Bio-oil production from rice husk
fast pyrolysis in a conical spouted bed reactor. Fuel 2014;128:162–9.
[26] Amutio M, Lopez G, Alvarez J, Olazar M, Bilbao J. Fast pyrolysis of eucalyptus
waste in a conical spouted bed reactor. Bioresour Technol 2015;194:225–32.
[27] Lopez G, Alvarez J, Amutio M, Mkhize NM, Danon B, van der Gryp P, et al. Waste
truck-tyre processing by flash pyrolysis in a conical spouted bed reactor. Energy
Convers Manage 2017;142:523–32.
[28] Artetxe M, Lopez G, Amutio M, Barbarias I, Arregi A, Aguado R, et al. Styrene
recovery from polystyrene by flash pyrolysis in a conical spouted bed reactor. Waste
Manage 2015;45:126–33.
[29] Aguado R, Prieto R, José MJS, Alvarez S, Olazar M, Bilbao J. Defluidization mod-
elling of pyrolysis of plastics in a conical spouted bed reactor. Chem Eng Process
2005;44:231–5.
[30] Fernandez-Akarregi AR, Makibar J, Lopez G, Amutio M, Olazar M. Design and
operation of a conical spouted bed reactor pilot plant (25 kg/h) for biomass fast
pyrolysis. Fuel Process Technol 2013;112:48–56.
[31] Amutio M, Lopez G, Aguado R, Artetxe M, Bilbao J, Olazar M. Kinetic study of
lignocellulosic biomass oxidative pyrolysis. Fuel 2012;95:305–11.
[32] Branca C, Albano A, Di Blasi C. Critical evaluation of global mechanisms of wood
devolatilization. Thermochim Acta 2005;429:133–41.
[33] Altzibar H, Lopez G, Bilbao J, Olazar M. Minimum spouting velocity of conical
spouted beds equipped with draft tubes of different configuration. Ind Eng Chem
Res 2013;52:2995–3006.
[34] Kim JW, Lee S, Kim S, Park SH, Jeon J, Park Y. The pyrolysis of waste mandarin
residue using thermogravimetric analysis and a batch reactor. Korean J Chem Eng
2011;28:1867–72.
[35] Collard F, Blin JA. Review on pyrolysis of biomass constituents: mechanisms and
composition of the products obtained from the conversion of cellulose, hemi-
celluloses and lignin. Renew Sustain Energy Rev 2014;38:594–608.
[36] Sharma RK, Wooten JB, Baliga VL, Hajaligol MR. Characterization of chars from
biomass-derived materials: pectin chars. Fuel 2001;80:1825–36.
[37] Chen B, Chen Z. Sorption of naphthalene and 1-naphthol by biochars of orange
peels with different pyrolytic temperatures. Chemosphere 2009;76:127–33.
[38] Einhorn-Stoll U, Kunzek H. The influence of the storage conditions heat and hu-
midity on conformation, state transitions and degradation behaviour of dried pec-
tins. Food Hydrocolloids 2009;23:856–66.
[39] Zhou S, Xu Y, Wang C, Tian Z. Pyrolysis behavior of pectin under the conditions that
simulate cigarette smoking. J Anal Appl Pyrolysis 2011;91:232–40.
[40] Lv PM, Xiong ZH, Chang J, Wu CZ, Chen Y, Zhu JX. An experimental study on
biomass air-steam gasification in a fluidized bed. Bioresour Technol
2004;95:95–101.
[41] Yang H, Yan R, Chen H, Lee DH, Zheng C. Characteristics of hemicellulose, cellulose
and lignin pyrolysis. Fuel 2007;86:1781–8.
[42] Demirbas A. Mechanisms of liquefaction and pyrolysis reactions of biomass. Energy
Convers Manage 2000;41:633–46.
[43] Mohan D, Pittman Jr. CU, Steele PH. Pyrolysis of wood/biomass for bio-oil: a cri-
tical review. Energy Fuels 2006;20:848–89.
[44] Oasmaa A, Fonts I, Pelaez-Samaniego MR, Garcia-Perez ME, Garcia-Perez M.
Pyrolysis oil multiphase behavior and phase stability: a review. Energy Fuels
2016;30:6179–200.
[45] Shen D, Jin W, Hu J, Xiao R, Luo K. An overview on fast pyrolysis of the main
constituents in lignocellulosic biomass to valued-added chemicals: structures,
pathways and interactions. Renew Sustain Energy Rev 2015;51:761–74.
[46] Branca C, Giudicianni P, Di Blasi C. GC/MS characterization of liquids generated
from low-temperature pyrolysis of wood. Ind Eng Chem Res 2003;42:3190–202.
[47] Jacobson K, Maheria KC, Kumar Dalai A. Bio-oil valorization: a review. Renew
Sustain Energy Rev 2013;23:91–106.
[48] Kan T, Strezov V, Evans TJ. Lignocellulosic biomass pyrolysis: a review of product
properties and effects of pyrolysis parameters. Renew Sustain Energy Rev
2016;57:1126–40.
[49] Chiaramonti D, Oasmaa A, Solantausta Y. Power generation using fast pyrolysis
liquids from biomass. Renew Sust Energy Rev 2007;11:1056–86.
[50] Oasmaa A, Czernik S. Fuel oil quality of biomass pyrolysis oils – state of the art for
the end users. Energy Fuels 1999;13:914–21.
[51] Oasmaa A, Kuoppala E, Solantausta Y. Fast pyrolysis of forestry residue. 2.
Physicochemical composition of product liquid. Energy Fuels 2003;17:433–43.
[52] Valle B, Aramburu B, Remiro A, Bilbao J, Gayubo AG. Effect of calcination/re-
duction conditions of Ni/La2O3-aAl2O3 catalyst on its activity and stability for
hydrogen production by steam reforming of raw bio-oil/ethanol. Appl Catal B
2014;147:402–10.
[53] Remiro A, Valle B, Aguayo AT, Bilbao J, Gayubo AG. Steam reforming of raw bio-oil
in a fluidized bed reactor with prior separation of pyrolytic lignin. Energy Fuels
2013;27:7549–59.
[54] Artetxe M, Alvarez J, Nahil MA, Olazar M, Williams PT. Steam reforming of dif-
ferent biomass tar model compounds over Ni/Al2O3 catalysts. Energy Convers
Manage 2017;136:119–26.
J. Alvarez et al.
[55] Bimbela F, Oliva M, Ruiz J, García L, Arauzo J. Hydrogen production by catalytic
steam reforming of acetic acid, a model compound of biomass pyrolysis liquids. J
Anal Appl Pyrolysis 2007;79:112–20.
[56] Bhatnagar A, Sillanpää M, Witek-Krowiak A. Agricultural waste peels as versatile
biomass for water purification – a review. Chem Eng J 2015;270:244–71.
[57] Alvarez J, Lopez G, Amutio M, Bilbao J, Olazar M. Physical activation of rice husk
pyrolysis char for the production of high surface area activated carbons. Ind Eng
Chem Res 2015;54:7241–50.
[58] Rebitanim NZ, Karim Ghani Wan Azlina Wan Ab, Rebitanim NA, Amran Mohd
Salleh M. Potential applications of wastes from energy generation particularly
biochar in Malaysia. Renew Sustain Energy Rev 2013;21:694–702.
[59] Laird DA, Brown RC, Amonette JE, Lehmann J. Review of the pyrolysis platform for
120
Fuel 224 (2018) 111–120
coproducing bio-oil and biochar. Biofuels Bioprod Biorefin 2009;3:547–62.
[60] Sharma RK, Wooten JB, Baliga VL, Lin X, Geoffrey Chan W, Hajaligol MR.
Characterization of chars from pyrolysis of lignin. Fuel 2004;83:1469–82.
[61] Azargohar R, Jacobson KL, Powell EE, Dalai AK. Evaluation of properties of fast
pyrolysis products obtained, from Canadian waste biomass. J Anal Appl Pyrolysis
2013;104:330–40.
[62] Abdullah H, Wu H. Biochar as a fuel: 1. Properties and grindability of biochars
produced from the pyrolysis of mallee wood under slow-heating conditions. Energy
Fuels 2009;23:4174–81.
[63] Azargohar R, Nanda S, Kozinski JA, Dalai AK, Sutarto R. Effects of temperature on
the physicochemical characteristics of fast pyrolysis bio-chars derived from
Canadian waste biomass. Fuel 2014;125:90–100.