Biomass and Bioenergy 117 (2018) 115–123

Contents lists available at ScienceDirect

Biomass and Bioenergy

journal homepage: www.elsevier.com/locate/biombioe

Research paper

Biochar production through slow pyrolysis of different biomass materials: T

Seeking the best operating conditions
Joan J. Manyàa,b,∗, Manuel Azuarab,c, José A. Mansob
a
Aragón Institute of Engineering Research (I3A), University of Zaragoza, crta. Cuarte s/n, Huesca, E-22071, Spain
b
Technological College of Huesca, University of Zaragoza, crta. Cuarte s/n, Huesca, E-22071, Spain
c
Institute of Nanoscience of Aragón (INA), University of Zaragoza, crta. Cuarte s/n, Huesca, E-22071, Spain

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

Keywords: The effect of three operating parameters (peak temperature, pressure, and pyrolysis atmosphere) during the slow
Biochar pyrolysis of three biomass sources (corn stover, vine shoots, and two-phase olive mill waste) was extensively
Residual biomass analyzed. A 2-level full factorial design of experiments was adopted to assess the effect of the above-mentioned
Pressurized slow pyrolysis factors on the potential stability of biochar as well as the yields of the main pyrolysis products. To evaluate the
Carbon sequestration potential
effect of the biomass feedstock, the design was divided into three blocks (one per biomass feedstock). Results
Produced gas
from the statistical analyses indicated that the properties of biochar related to its potential stability were mainly
affected by the peak temperature and, to a lesser extent, the biomass feedstock. A significant increase in the yield
of produced gas was observed when pressure was raised. This increase in the total gas yield was in part due to a
higher release of CO, CH4, and H2. Using a pyrolysis atmosphere of CO2 (instead of N2) did not result in any
remarkable change in neither the distribution of the pyrolysis products nor the potential stability of biochar.
However, when CO2 was used as carrier gas, a significant increase in the yield of CO, at the expense of produced
CO2, was observed. The findings reported herein suggest that processing biomass through pressurized slow
pyrolysis under CO2 atmosphere is interesting to simultaneously obtain two valuable products: a biochar with an
appropriate carbon sequestration potential, and a produced gas with an appropriate composition for energy
recovery purposes.

1. Introduction and transportation requirements. There is increasing uncertainty about

the potential of such systems for commercial exploitation. Thus, de-
Concerns about climate change and improvement of soil quality in a veloping more efficient production processes and novel alternative or
sustainable manner have generated much interest in biochar, a form of added-value uses for biochar can lead to a more competitive tech-
charred organic matter derived from biomass, which when applied to nology.
soil in a deliberate manner has the potential to improve carbon se- Pyrolysis is a thermochemical decomposition process during which
questration as well as soil productivity [1]. Biochar could enhance the biomass is heated at moderate temperature (typically 350–650 °C) in
capacity of soils for holding water and nutrients and, thus, reducing the the absence of oxygen. The process results in the formation of three
use of chemical fertilizers. Nevertheless, the effects of biochar and the main products: charcoal (i.e., biochar), a volatile matter which can
requirements on its properties for soil amendment and carbon seques- further be condensed to liquid phase (bio-oil), and the remaining so-
tration purposes have not fully been established [2]. The influence of called “non-condensable” gases, like CO, CO2, CH4, and H2. Depending
numerous parameters (such as biochar production process and oper- upon the residence time, peak temperature, and heating rate the pyr-
ating conditions, soil properties, and biomass feedstock) on the carbon olysis process is sub-divided into three categories: slow, intermediate
sequestration potential as well as soil-biochar interactions is still not and fast pyrolysis. Slow pyrolysis is usually preferred to produce bio-
well understood. char, with gas as co-product. It is a relatively simple and robust process
The economic feasibility of biochar systems for agronomic benefits which can be applicable to small-scale and farm-based production of
depends on numerous factors including the biomass feedstock selection biochar [3,4].

∗
Corresponding author. Aragón Institute of Engineering Research (I3A), Technological College of Huesca, University of Zaragoza, crta. Cuarte s/n, Huesca, E-
22071, Spain.
E-mail address: [email protected] (J.J. Manyà).

https://doi.org/10.1016/j.biombioe.2018.07.019
Received 29 January 2018; Received in revised form 18 May 2018; Accepted 25 July 2018
Available online 01 August 2018
0961-9534/ © 2018 Elsevier Ltd. All rights reserved.
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

Fig. 1. Schematic of the pressurized pyrolysis device: (1) fixed-bed pyrolysis reactor, (2) pyrolysis liquid condensation system, (3) volumetric gas meter, and (4)
micro-GC.

In spite of the relatively simplicity of the slow pyrolysis process, the stability of biochar. Among them, two relatively easy techniques have
yields and properties of products (including those of biochar) can be recently gained attention: (1) the recalcitrance index (R50), which was
largely affected by the operating conditions (e.g., peak temperature, proposed by Harvey et al. [15] and is based on the relative thermal
residence time of the vapor phase, pressure, and particle size). In ad- stability of a given biochar to that of graphite; and (2) a method de-
dition, the nature of the biomass source also plays a key role in de- veloped at the UK Biochar Research Center, which is based on the
termining the yield and properties of produced biochar. oxidation of biochars using H2O2 to accelerate the “aging” and, hence,
An interesting option to improve the carbonization efficiency (i.e., the oxidative loss of carbon [16,17].
fixed-carbon yield [5]) is to increase pressure. In this sense, most of the The specific aim of this study is to determine the effect of certain
previous studies on pressurized pyrolysis reported on an increase in operating conditions (peak temperature, pressure at constant gas re-
both the biochar and gas yields, at the expense of the organic con- sidence time, and pyrolysis environment) on several properties of bio-
densable fraction [6–8]. In these studies, however, the effect of a char produced through slow pyrolysis of three different sources: corn
moderate pressure (0.5–5.0 MPa) was measured without keeping con- stover, vine shoots, and two-phase olive mill waste. Special attention is
stant the residence time of the inert gas within the reactor. In order to focused on how the operating conditions can affect the properties of
investigate the true effect of pressure, some researchers conducted biochar related to its potential stability (i.e., carbon sequestration po-
pressurized pyrolysis experiments at constant gas residence time tential). In addition, the yield and composition of the pyrolysis gas was
[4,9–12]. Results from these alternative studies indicated that the effect also determined as a function of the tested operating conditions.
of pressure on the biochar yield is still unclear, since negligible [11,12],
positive [9] and event negative [4,10] correlations were found.
2. Experimental section
Another interesting option to improve the efficiency of the slow
pyrolysis process, especially in terms of economics saving, is replacing
2.1. Materials
the use of a relatively expensive inert gas (e.g., N2) with CO2 coming
from residual flue gases [13]. In other words, the flue gas generated
Three waste biomass materials, all of them generated in the pro-
after combustion of pyrolysis gas can be used as pyrolysis gas en-
vince of Huesca (Spain), were used in the present study: (i) corn stover
vironment. Nevertheless, research is required to understand the effects
(CS) containing corncob (15.5 wt%), leaf (4.3 wt%) and stalk (80.2 wt
of modifying the pyrolysis environment (i.e., from pure N2 to a flue gas
%); (ii) two-phase olive mill waste (TPOMW), which was sun-dried in
containing CO2) on the distribution of the pyrolysis products as well as
the field for several months at the premises of an extra-virgin olive oil
on the properties of produced biochar (e.g., the properties related to its
factory; and (iii) vine shoots (VS) supplied by a local winery.
carbon sequestration potential). In this sense, results from a preliminary
Proximate analyses were performed in quadruplicate according to
study [14] indicated that the use of CO2 instead of N2 did not sig-
ASTM standards (D3173 for moisture, D3174 for ash, and D3175 or
nificantly affect the properties of biochar related to its carbon seques-
volatile matter), whereas elemental analyses were carried out using a
tration potential (e.g., fixed-carbon content and molar H:C and O:C
CHNS analyzer from Leco Corporation (USA). In addition, the ash
ratios).
composition expressed as weight percentages of the equivalent oxides
In addition to the fixed carbon content and molar H:C and O:C ra-
was measured using an ADVANT’XP + XRF spectrometer from Thermo
tios, several approaches have been proposed to estimate the long-term
ARL (Switzerland).

116
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

2.2. Experimental devices and procedures

The fixed-bed pyrolysis reactor consisted of a cylindrical and ver-

tical reactor (140 mm inside diameter and 465 mm long) made of
Sandvik 253 MA stainless steel (EN 1.4835; X9CrNiSiNCe21-11-2). The
reactor was heated by two electric resistances of 2.1 kW. A basket of 4 L,
made of Monel 400 alloy (Ni65Cu33Fe2) wire mesh, was used to put
the biomass into the reactor. The temperature inside the bed was
measured using four thermocouples placed in a thermowell at different
heights. More details regarding the configuration of the reactor are
available elsewhere [4,12,14]. The pressure of the system was con-
trolled through a back-pressure regulator. A schematic diagram of the
experimental setup is given in Fig. 1.
The mass flow rate at STP conditions of the carrier gas (N2 or CO2)
was adjusted as a function of the absolute pressure (0.1–1.0 MPa) to
keep the real flow rate of the carrier gas within the reactor (at the
highest temperature) at a constant value of 1.85 L min−1. After each
experiment, the condensed fraction was recovered directly from the
condensers without using any solvent as wash liquid. The glass traps
were weighted before and after each experimental run to determine the
total mass of liquid. The content of water in the condensable fraction
was measured using a Karl-Fischer volumetric titrator from Metrohm
(Switzerland). The composition of the gas fraction (N2, CO2, CO, CH4,
C2Hx and H2) was determined using an Agilent micro-GC equipped with
two analytical columns: a PoraPlot Q (using He as carrier gas) and a
Molsieve 5 A (using Ar as carrier gas). The mass of produced gas was
calculated from the N2 mass balances.
The initial sample mass of biomass, which depends on the bulk
density of the feedstock, was approximately of 250, 400, and 750 g for
CS, VS, and TPOMW; respectively. Particle size distributions were de-
termined by sieving according to ISO 3310-1 standard. Fig. 2 displays
the differential and cumulative distributions for each biomass feed-
stock. The median (d50) and surface-volume mean (calculated as the
harmonic mean of the cumulative mass fraction [18]) diameters for
each biomass feedstock are also given in Fig. 2. In the case of CS and VS,
we used large particles with the aim of avoiding the high energy-con-
suming milling pre-treatments. Furthermore, using large biomass par-
ticles can lead to an improvement in carbonization efficiency (i.e.,
biochars with higher fixed-carbon contents) as a consequence of an
enhancement of the secondary charring reactions at intra-particle level
[19].
The void-volume fractions in the pyrolysis reactor were estimated to
be in the range of 0.85–0.90. These estimates were calculated using the
experimentally measured bulk densities (268, 111, and 70 kg m−3 for
TPOMW, VS, and CS; respectively) and the particle densities available
in the literature (1394, 1250, and 1170 kg m−3 for TPOMW [20], VS
[21], and CS [22]; respectively). Given that the void-volume fraction
was almost constant regardless of the biomass feedstock, an almost
constant gas-hourly space velocity (GHSV) of 17.2 h−1 can be assumed
for all the experiments.
Pyrolysis experiments were conducted according to a 2-level full Fig. 2. Particle size distributions of the biomass sources used in the present
factorial design with three factors: peak temperature (400–600 °C), study: (a) CS, (b) VS, and (c) TPOMW. The d50 (median) and surface-volume
mean (harmonic mean [18]) values are shown in the corresponding plot.
absolute pressure (0.1–1.0 MPa), and carrier gas (measured as the mole
fraction of CO2 in a mixture N2/CO2, and ranging from 0 to 0.95). Three
replicates at the center point (500 °C, 0.55 MPa, and a CO2 mole frac- heated at a linear heating rate of 5 °C min−1 (highest
tion of 0.48) were conducted for each biomass feedstock to simulta- temperature = 600 °C) under an atmosphere of N2 at a flow rate of
neously estimate the experimental error and the overall curvature effect 200 mL (STP) min−1.
[23]. Three blocks (one per each feedstock) were added to the design in
order to analyze the effect of the biomass source on the average values. 2.3. Characterization of the pyrolysis products
All the statistical calculations were conducted using R software (version
3.3.3). The matrix of the design of experiments is shown in Table S1 The mass yields of biochar, water, condensable organics and pro-
(Supplementary Data). duced gas (ychar, ywater, yorg and ygas, respectively) were calculated in a
In order to get an estimate of the concentration of the main biomass dry-ash-free (daf) basis. Biochar samples were characterized by prox-
constituents, additional pyrolysis tests were conducted in an atmo- imate and elemental analyses according to the same procedures de-
spheric thermobalance (a MK2 microbalance with a readability of scribed in Section 2.1. The mass fraction of fixed carbon in the biochar
0.1 μg from CI Precision, UK). Initial sample masses of 25 mg were (xFC) was calculated from the proximate analysis data, whereas the

117
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

molar H:C and O:C ratios were calculated from the elemental analysis Table 2
data. Summary statistics for the regression models based on the data given in Table
Temperature-programmed oxidation (TPO) of the biochars was S2 (values in brackets correspond to the p-values resulting from t-tests; sig-
conducted using the same thermobalance described above. nificant terms are marked in bold).
Approximately 10 mg of sample was heated in N2-diluted air (100 mL Coefficient Response variable
STP min−1) from room temperature to 950 °C at a linear heating rate of
10 °C min−1. The R50 index was then calculated from the TPO data ychara xFCb H:C ratio O:C ratio R50 index Stable C

using the following equation: β0 0.350 0.813 0.437 0.104 0.542 79.6
Block 1 (CS) −0.015 0.026 −0.039 0.009 0.007 6.09
R50, x = T50, x / T50, g (1) (0.000) (0.001) (0.000) (0.176) (0.165) (0.000)
Block 2 (VS) −0.007 −0.024 0.030 0.001 −0.015 −6.93
where T50,x and T50,g are the temperatures corresponding to 50% of (0.034) (0.002) (0.000) (0.935) (0.004) (0.000)
mass loss of biochar and graphite, respectively. Block 3 0.022 −0.002 0.009 −0.010 0.008 0.842
Direct oxidation of biochars was performed according to the pro- (TPOMW) (0.000) (0.739) (0.172) (0.153) (0.093) (0.373)
cedure described by Cross and Sohi (Edinburgh stability tool) [17]. β1 (T) −0.030 0.055 −0.141 −0.043 0.027 6.52
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Briefly: biochar samples containing 0.1 g of C were milled to a fine
β2 (P) −0.009 0.022 −0.014 −0.010 0.006 −0.375
powder in a ball mill and then treated in a test tube with 7 mL of an (0.001) (0.001) (0.016) (0.092) (0.168) (0.630)
aqueous solution of 5% H2O2, initially at room temperature and then at β3 (CO2) −0.001 0.004 −0.038 −0.005 0.003 0.325
80 °C for 48 h. The samples were then dried in an oven at 105 °C (0.580) (0.487) (0.002) (0.409) (0.493) (0.676)
overnight. The “stable C” was expressed as the percentage of the initial β12 (T·P) −0.002 0.011 −0.004 0.000 −0.001 −0.292
(0.514) (0.069) (0.472) (0.940) (0.813) (0.708)
0.1 g of C that remains after oxidation. β13 (T·CO2) −0.002 −0.002 0.001 0.001 0.002 0.025
(0.535) (0.719) (0.806) (0.916) (0.630) (0.974)
3. Results and discussion β23 (P·CO2) 0.001 0.000 0.002 0.002 0.001 0.083
(0.801) (0.971) (0.650) (0.717) (0.781) (0.915)
β123 (T·P·CO2) 0.002 −0.001 0.001 −0.001 0.000 0.067
Results from proximate, elemental and ash analyses are summarized (0.472) (0.854) (0.852) (0.856) (0.943) (0.932)
in Table 1. The experimental results obtained from the factorial design Overall −0.017 0.025 −0.041 −0.042 −0.003 3.23
of experiments are given in Supplementary Data (Tables S2 and S3). Curvature (0.001) (0.027) (0.001) (0.001) (0.695) (0.039)
Adj. R2 0.874 0.805 0.957 0.703 0.604 0.808
The mass-balance closures for all the pyrolysis tests (above 95%) are
given in Table S2. The mass yields of the different pyrolysis products a
Mass fraction in daf basis.
(ychar, ywater, yorg and ygas) were then calculated attributing the error in b
Mass fraction.

Table 1 the mass-balance closure to minor inaccuracies in determining the mass

Proximate, elemental and XRF analyses of the biomass sources (CS, TPOMW, of produced gas.
and VS).
Proximate CS TPOMW VS
3.1. Effects on the yield of biochar and its potential stability
Moisture (wt. %) 7.27 ± 0.31 12.8 ± 0.12 7.97 ± 0.68
Ash (wt. % in dry basis) 2.70 ± 0.20 2.27 ± 0.11 1.08 ± 0.05 For the response variables which are related to the biochar yield as
Volatile matter (wt. % in dry 86.6 ± 0.11 87.5 ± 1.95 74.0 ± 1.19 well as carbon sequestration potential (i.e., ychar, xFC, molar H:C and
basis)
O:C ratios, R50 index, and “stable C”), Table S2 shows the data obtained
Fixed carbon (wt. % in dry basis) 10.7 ± 0.49 10.2 ± 1.13 24.9 ± 1.91
for each experimental run (a total of 11 per block).
a
Elemental (wt. % in daf basis) b
CS TPOMW VS The structure of the regression model, which was used as an ap-
proximation for statistical testing, was the following:
C 44.4 ± 0.31 50.0 ± 0.48 47.1 ± 0.14
H 5.60 ± 0.04 6.21 ± 0.17 5.29 ± 0.09 yˆ = β0 + β1 T + β2 P + β3 CO2 + β12 T ·P + β13 T ·CO2
N 0.43 ± 0.01 2.29 ± 0.09 0.66 ± 0.05
S 0.45 ± 0.05 0.40 ± 0.07 0.56 ± 0.13 + β23 P·CO2 + β123 P·T ·CO2 (2)

Inorganic matter (wt.% of ash) CS TPOMW VS where β0, βi, βij, and βijk are the intercept, linear, 2-way interaction, and
3-way interaction coefficients; respectively.
CaO 30.7 ± 0.23 43.3 ± 0.25 58.3 ± 0.25
K2O 9.85 ± 0.15 33.7 ± 0.24 18.4 ± 0.12
Table 2 lists, for each response variable, the estimated regression
MgO 3.45 ± 0.17 1.65 ± 0.14 6.66 ± 0.14 coefficients for coded factors (−1, +1) and associated p-values (from t-
SiO2 31.4 ± 0.23 4.34 ± 0.10 5.73 ± 0.08 tests), as well as the adjusted R2 values. In the event that the regression
Fe2O3 6.49 ± 0.12 2.91 ± 0.08 3.51 ± 0.11 coefficient for the overall curvature term becomes significant, a second-
Al2O3 4.85 ± 0.12 1.60 ± 0.07 2.57 ± 0.07
order regression model (including pure quadratic terms for each factor)
P2O5 4.13 ± 0.10 4.33 ± 0.10 1.24 ± 0.04
PbO 2.50 ± 0.08 0.55 ± 0.06 0.26 ± 0.02 is needed in order to improve the accuracy of model predictions. In
S (inorganic) 1.94 ± 0.07 4.26 ± 0.01 0.60 ± 0.03 addition, the regression coefficients estimated for the blocks are also
Cl (inorganic) 0.59 ± 0.03 0.37 ± 0.02 0.48 ± 0.02 presented in Table 2. The regression coefficient for a given block re-
MnO 0.53 ± 0.03 0.31 ± 0.03 0.53 ± 0.03 presents the difference between the mean of the response for this block
ZnO 0.24 ± 0.02 0.14 ± 0.02 0.32 ± 0.02
SnO2 0.45 ± 0.03 0.33 ± 0.03 0.26 ± 0.02
and the overall mean of the response.
TiO2 0.59 ± 0.03 0.20 ± 0.02 0.34 ± 0.02 To get a graphical interpretation of the results, the normal plots of
CuO 0.08 ± 0.02 0.26 ± 0.03 0.09 ± 0.01 standardized effects (for a significance level of 0.05) for the above-
MoO3 0.28 ± 0.02 0.25 ± 0.02 0.02 ± 0.01 mentioned response variables are shown in Fig. 3. From these plots, and
La2O3 0.23 ± 0.04 0.31 ± 0.04 NDc
also from the statistical data listed in Table 2, it can be concluded that
Y2O3 0.26 ± 0.02 ND ND
the peak temperature was the most influential factor on the potential
a
Dry-ash-free. stability of biochar, since its effect was statistically significant for all the
b
Oxygen is calculated by difference. response variables analyzed in this section. As expected, an increase in
c
Not detected. the peak temperature led to an increase in xFC, R50 index and “stable C”

118
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

Fig. 3. Normal plots of standardized effects (α = 0.05) for (a) ychar, (b) xFC, (c) molar H:C ratio, (d) molar O:C ratio, (e) R50 index, and (f) “stable C”.

as well as a decrease in both the H:C and O:C molar ratios. The direction deduced that an increase in the absolute pressure led to a significant
of these significant effects is consistent with the idea that any of these increase in xFC and a significant decrease in the molar H:C ratio.
variables related to the potential stability of biochar can be taken as a However, the improvement of these indicators of the potential stability
rough indicator of the carbon sequestration potential. of biochar was quantitatively much lower than that observed for an
It should be emphasized that the significant effects of the peak increase in the peak temperature. In regard to the effect of the pyrolysis
temperature were found from the experimental data obtained using atmosphere, a slightly significant decrease in the molar H:C ratio was
three different biomass sources (CS, VS, and TPOMW). In other words, observed when the carrier gas was mainly composed of CO2.
increasing the peak temperature from 400 to 600 °C led to biochars with It is also interesting to highlight that the curvature term was sig-
higher carbon sequestration potential, regardless of the biomass pro- nificant for all response variables with the exception of the R50 index.
cessed in the present study. This finding is not contradictory to the fact This finding means that an augmentation to a central composite design
that the regression coefficients for the block terms were significant in should be considered in further studies in order to define more accu-
several cases (see Table 2). This means that, as expected, the potential rately the location of the optimal conditions (via Response Surface
stability of biochar is dependent on the type of biomass feedstock. As an Methodology) [23]. In regard with this, it can be possible that the op-
example, the percentages of “stable C” and fixed-carbon contents (xFC) timal value for a given factor (e.g., the peak temperature that minimizes
obtained for corn stover-derived biochars were, on average, sig- the molar H:C ratio) could be lower than the highest level (e.g., below
nificantly higher than those obtained for the other two biomass sources. 600 °C for the peak temperature factor and/or below 1.0 MPa for the
From the statistics shown in Table 2 and Fig. 3, it can also be absolute pressure factor).

119
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

Regarding the biochar yield (ychar), the peak temperature was also
the most influential factor. As expected and in agreement with a
number of previous studies [6,7,24–28], a decrease in the production of
biochar was associated to an increase in the peak temperature. The
biochar yield was also significantly affected by an increase in the ab-
solute pressure, resulting in a slightly decrease in ychar. This finding,
which was already observed in previous studies [4,10], can be ex-
plained by two main reasons. First, pressurized pyrolysis can cause an
enhancement of the steam gasification reaction kinetics leading to non-
negligible reaction rates, even at relatively low temperatures. Second,
an increase in the mass flow rate of carrier gas for pressurized pyrolysis
can lead to a reduction of the concentration of volatiles within the re-
actor (i.e., dilution effect), possibly resulting in a lower extent of the
secondary charring reactions. No significant effect of the composition of
the carrier gas on the yield of biochar was found.
As can be seen from the regression coefficients for the block terms
listed in Table 2, the biochar yield was also dependent on the type of
biomass feedstock. Higher mass yields of biochar than average were
obtained when TPOMW was pyrolyzed. This fact could be explained by
differences in the concentration of the biomass constituents (hemi-
celluloses, cellulose and lignin). Thermogravimetric measurements
under dynamic heating conditions were conducted for the three bio-
mass sources in order to estimate the contents of the biomass con-
stituents. The experimental differential thermogravimetric (DTG)
curves were deconvoluted into three peaks using the “Peak Analyzer”
tool implemented in OriginPro version 9.0 (OriginLab, USA). These
three peaks are associated to the devolatilization of hemicelluloses
(peak 1), cellulose (peak 2), and lignin (peak 3). Fig. 4 shows the de-
convoluted DTG curves obtained for the three biomass sources (CS, VS,
and TPOMW). From Fig. 4, it can be deduced that the apparent lignin
content (which is associated to peak 3) for TPOMW is considerably
higher than that determined for CS and VS. This is consistent with the
fact that lignin is the biomass constituent which leads to the highest
char yield [29]. In summary, the composition of the biomass source in
terms of its main lignocellulosic biomass constituents appears to be a
key parameter in explaining the differences observed in the biochar
yield.
Regarding the possible effects of alkali and alkaline earth metal
species AAEMs, the differences in the content of Ca and K between the
three biomass sources (deduced from Table 1) were probably too low to
observe any quantitative effect on the biochar yield. Nevertheless, as
evidenced in Fig. 4, the temperature corresponding to the highest de-
volatilization rate (for the thermogravimetric pyrolysis tests) was de-
pendent on the biomass feedstock (303 °C for TPOMW, 330 °C for CS,
and 353 °C for VS). This fact could be related to the role of the AAEMs,
which can promote the biomass decomposition at lower temperatures.
In this sense, it should be noted that the temperature for the highest
devolatilization rate decreased as the content of Ca plus K in the bio-
mass feedstock increased (1.17, 0.752, and 0.563 g for each 100 g of
raw biomass for TPOMW, CS, and VS; respectively).

Fig. 4. Deconvoluted DTG curves for the pyrolysis of (a) corn stover (CS), (b)
3.2. Effects on the yield of pyrolysis gas and its composition
vine shoots (VS), and (c) two-phase olive mill waste (TPOMW).

From analyzing the statistical results listed in Table 3 (regression

models) and shown in Fig. 5 (normal plots of standardized effects), it pyrolysis vapors within the reactor.
can be concluded that the absolute pressure plays a key role in de- Concerning the favorable effect of the absolute pressure on the
termining the yield of the produced pyrolysis gas as well as the yield of produced gas at the expense of the yield of water (ywater), an inter-
the individual gaseous species analyzed in the present study. pretation can be envisaged to explain this finding. The vapor pressure of
The yield of total gas (ygas) significantly increased when the peak volatiles increases with raising the absolute pressure, leading to an
temperature was increased from 400 to 600 °C (at the expense of the acceleration of the cross-linking reactions and the subsequent higher
yield of biochar) or the absolute pressure was raised from 0.1 to production of char and gas at relatively low process temperatures [30].
1.0 MPa (at the expense of the yield of produced water). The positive Nevertheless, above a certain temperature, the extent of the steam-char
effect of the peak temperature on the gas yield was certainly expected gasification reaction could be non-negligible at high pressure [31],
because of the higher extent of devolatilization processes. In addition, especially when the content of AAEMs in the biomass feedstock is non-
an increased temperature can result in a promotion of the secondary negligible. In other words, the formation of char could be enhanced
cracking reactions, given the relatively high residence time of the with the increasing pressure, mainly due to the restricted transport of

120
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

Table 3
Summary statistics for the regression models based on the data given in Table S3 (values in brackets correspond to the p-values resulting from t-tests; significant terms
are marked in bold).
Coefficient Response variable

ygasa ywatera yorga yCO2b yCOb yCH4b yH2b

β0 0.308 0.156 0.191 3.731 2.655 1.397 0.722

Block 1 (CS) 0.006 0.004 0.005 −0.035 0.315 0.070 0.094
(0.425) (0.566) (0.519) (0.580) (0.006) (0.148) (0.168)
Block 2 (VS) −0.041 0.042 0.001 −0.457 −0.405 −0.118 −0.108
(0.000) (0.000) (0.910) (0.000) (0.001) (0.020) (0.117)
Block 3 (TPOMW) 0.036 −0.045 −0.006 0.492 0.090 0.048 0.014
(0.000) (0.000) (0.450) (0.000) (0.397) (0.321) (0.839)
β1 (T) 0.035 0.002 −0.006 0.305 0.333 0.253 0.389
(0.000) (0.717) (0.386) (0.000) (0.001) (0.000) (0.000)
β2 (P) 0.036 −0.028 −0.002 0.372 0.420 0.261 0.244
(0.000) (0.000) (0.795) (0.000) (0.000) (0.000) (0.000)
β3 (CO2) 0.006 0.004 −0.001 −0.154 0.315 −0.040 −0.051
(0.326) (0.534) (0.852) (0.007) (0.001) (0.320) (0.366)
β12 (T·P) −0.005 0.000 0.000 −0.036 0.011 −0.036 0.026
(0.405) (0.994) (0.995) (0.492) (0.897) (0.361) (0.643)
β13 (T·CO2) 0.002 0.001 0.004 −0.080 0.044 0.026 −0.059
(0.724) (0.859) (0.592) (0.135) (0.617) (0.507) (0.294)
β23 (P·CO2) 0.000 0.000 −0.003 0.007 0.161 −0.072 0.014
(0.972) (0.920) (0.660) (0.893) (0.075) (0.077) (0.799)
β123 (T·P·CO2) 0.001 −0.004 −0.003 0.042 0.030 0.017 −0.006
(0.884) (0.413) (0.643) (0.426) (0.727) (0.665) (0.917)
Overall Curvature 0.009 −0.001 0.005 0.184 0.171 −0.037 0.247
(0.453) (0.939) (0.722) (0.077) (0.310) (0.623) (0.028)
Adj. R2 0.765 0.717 0.000 0.841 0.666 0.738 0.688

a
Mass fraction in daf basis.
b
In mmol g−1 daf feedstock.

volatiles. However, some carbon can simultaneously react with steam, the observed increases in the yields of the gaseous species can be at-
resulting in a decrease in the overall yields of biochar and water and a tributed to the higher decomposition of the biomass constituents,
parallel increase in the overall yield of pyrolysis gas. especially lignin, which decomposes over a very wide temperature
With regard to the effect of the biomass source on the yields of range (see Fig. 4). In addition, the enhancement of the thermal cracking
produced water and gas, Table 3 shows statistically significant regres- of the primary volatiles at high temperatures (leading to an additional
sion coefficients for both blocks 2 (associated to VS) and 3 (associated formation of permanent gaseous compounds) can also explain the ob-
to TPOMW). This fact could also be explained by differences in the served effect of the peak temperature.
content and activity of K plus Ca between the biomass sources. Given The increase in the yield of CO2 with the increased pressure could
that the AAEMs contained in the TPOMW feedstock seems to be highly be explained by a promotion of decarboxylation routes for both hemi-
active compared with the other biomass samples (see Fig. 4), a higher celluloses and cellulose, as suggested by Qian et al. [9] in an earlier
yield of gas (and a lower yield of water) than average was expected. The study on pressurized pyrolysis of rice husks. Moreover, additional rea-
same argumentation is valid to explain the lower-than-average gas sons for the effect of the absolute pressure on the yields of gaseous
yields (and higher-than-average yields of water) observed for the VS species can be gleaned from analyzing the data reported in Table 4. In
feedstock, which seems to contain the least active AAEMs. this sense, an increase in the absolute pressure from 0.1 to 1.0 MPa can
No statistically significant effects (neither for ygas nor for ywater) were favor the production of CO2 through reactions #2 (Boudouard) and #6
found for the pyrolysis atmosphere. This finding is apparently in dis- (reverse dry reforming). In addition, the observed increase in the yield
agreement with earlier observations from Pilon and Lavoie [13] (for of CH4 with the increased pressure can be explained by a thermo-
pyrolysis of switchgrass at 500 °C) and Azuara et al. [14] (for pyrolysis dynamic promotion of the reactions #3, #4 and #6. However, the yield
of vine shoots at 600 °C). In these two studies, a slightly significant of CO also increased with the absolute pressure, despite the fact that the
increase in the yield of gas under an atmosphere of CO2 was reported promotion of the reactions #2, #3, and #6 leads to a certain con-
for experiments conducted at atmospheric pressure. However, and as sumption of CO. This fact seems to confirm that the kinetics of the re-
revealed from the statistical data reported in Table 3, the effect of using action #5 (steam gasification) can definitely be enhanced by an in-
an atmosphere of CO2 on the yield of gas was not significant when the crease in the absolute pressure, leading to an additional formation of
experimental data obtained from the pyrolysis of three different bio- CO and H2 at the expense of a certain amount of char. The higher yield
mass sources were simultaneously analyzed. of H2 (when pressure was raised) can also be explained by a higher rate
Concerning the yield of condensable organic compounds (yorg), of the reaction #1 (water-gas-shift). In line whit this, Hla et al. [32]
there were no statistically significant effects of any factor (see Table 3). already reported an enhancement of the kinetics of the water-gas-shift
The lack of statistical influence of the peak temperature could be ex- reaction when pressure was increased.
plained by a higher extent of the thermal cracking reactions when this From the regression coefficients listed in Table 3, it can be seen that
factor was raised. This could balance the higher release of volatiles the pyrolysis atmosphere had a statistically significant effect on the
caused by a higher devolatilization of the biomass constituents at high yields of both CO2 and CO. When a pyrolysis atmosphere mainly
peak temperatures. composed of CO2 was used, a higher yield of CO at the expense of CO2
From the results shown in Table 3 and Fig. 5, it can also be con- was observed. This finding, which was already reported in previous
cluded that the cumulative yield of all the gaseous species analyzed was studies [13,14], could be explained by an enhancement of the reverse
significantly affected by the peak temperature and the absolute pres- Boudouard reaction under an atmosphere of carbon dioxide. Despite
sure. As has been mentioned above to explain the dependence of ygas, the fact that the rate of the reverse Boudouard reaction is very low at

121
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

Fig. 5. Normal plots of standardized effects (α = 0.05) for (a) ygas, (b) ywater, (c) yCO2, (d) yCO, (e), yCH4, and (f) yH2.

Table 4 temperatures below 900 °C [33], the AAEMs present in the biomass
Main reactions involved during the release of the pyrolysis gas. feedstock as well as the nickel-rich alloy (Monel 400, from which the
No. Reaction Extent of reaction (kmol)a sample basket is made) can catalyze this reaction even at the tem-
peratures used in the present work.
600 °C and 0.1 MPa 600 °C and 1.0 MPa
4. Conclusions
1 H2O + CO ⇌ CO2 + H2 0.244 0.244
2 2CO ⇌ CO2 + C 0.715 0.908
3 3H2 + CO ⇌ CH4 + H2O 0.063 0.669 From the results presented and discussed above, the following
4 C + 2H2 ⇌ CH4 −0.212 0.532 conclusions can be drawn:
5 C + H2O ⇌ CO + H2 −0.136 −0.700
6 2H2 + 2CO ⇌ CO2 + CH4 0.148 0.705
(i) Pyrolysis peak temperature is confirmed as the most influential
a
Calculated using Aspen Plus V8.8; NRTL package and a Gibbs Reactor factor on the potential stability of biochar (i.e., carbon sequestra-
module. Stoichiometric coefficients were taken as initial molar flow rates (in tion potential), since its effect on all the variables related to the
kmol h−1) for all the species involved in the reaction. biochar stability was statistically significant. However, the curva-
ture term was significant for almost all these response variables,
suggesting that the optimal peak temperature (at which the carbon

122
J.J. Manyà et al. Biomass and Bioenergy 117 (2018) 115–123

sequestration potential is maximized) could be lower than the eucalyptus wood charcoal for the steel industry: a statistical analysis approach, Fuel
highest value of 600 °C. Further studies are required to establish Process. Technol. 92 (2011) 1890–1897.
[9] Y. Qian, J. Zhang, J. Wang, Pressurized pyrolysis of rice husk in an inert gas
the most appropriate peak temperature as a function of the re- sweeping fixed-bed reactor with a focus on bio-oil deoxygenation, Bioresour.
sponse variable (e.g., “stable C”) for a given biomass feedstock. Technol. 174 (2014) 95–102.
(ii) On the basis of the statistical outcomes reported herein, the ne- [10] J.J. Manyà, S. Laguarta, M.A. Ortigosa, J.A. Manso, Biochar from slow pyrolysis of
two-phase olive mill waste: effect of pressure and peak temperature on its potential
gative effect of the absolute pressure on the yield of biochar is stability, Energy Fuels 28 (2014) 3271–3280.
confirmed. In other words, keeping constant the residence time of [11] J.J. Manyà, M.A. Ortigosa, S. Laguarta, J.A. Manso, Experimental study on the ef-
the gaseous phase within the pyrolysis reactor, an increase in fect of pyrolysis pressure, peak temperature, and particle size on the potential
stability of vine shoots-derived biochar, Fuel 133 (2014) 163–172.
pressure leads to a slightly decrease in the yield of biochar. [12] M. Azuara, B. Baguer, J.I. Villacampa, N. Hedin, J.J. Manyà, Influence of pressure
Nevertheless, the most significant effect of increasing the absolute and temperature on key physicochemical properties of corn stover-derived biochar,
pressure was the higher production of pyrolysis gas at the expense Fuel 186 (2016) 525–533.
[13] G. Pilon, J.M. Lavoie, Pyrolysis of switchgrass (Panicum virgatum L.) at low tem-
of the produced water.
peratures within N2 and CO2 environments: product yield study, ACS Sustain.
(iii) The use of a pyrolysis atmosphere manly composed of CO2 instead Chem. Eng. 1 (2013) 198–204.
of N2 did not result in any remarkable change in neither the dis- [14] M. Azuara, E. Sáiz, J.A. Manso, F.J. García-Ramos, J.J. Manyà, Study on the effects
tribution of the pyrolysis products nor the potential stability of the of using a carbon dioxide atmosphere on the properties of vine shoots-derived
biochar, J. Anal. Appl. Pyrolysis 124 (2017) 719–725.
produced biochar. However, working under an atmosphere of CO2 [15] O.R. Harvey, L.J. Kuo, A.R. Zimmerman, P. Louchouarn, J.E. Amonette,
resulted in a significant increase in the cumulative yield of CO at B.E. Herbert, An index-based approach to assessing recalcitrance and soil carbon
the expense of produced CO2, leading to an interesting improve- sequestration potential of engineered black carbons (biochars), Environ. Sci.
Technol. 46 (2012) 1415–1421.
ment in the composition of the produced gas. [16] K. Crombie, O. Masek, S.P. Sohi, P. Brownsort, A. Cross, The effect of pyrolysis
(iv) In summary, it seems reasonable to further explore the possibility conditions on biochar stability as determined by three methods, GCB Bioenergy 5
of processing biomass through pressurized slow pyrolysis under an (2013) 122–131.
[17] A. Cross, S.P. Sohi, A method for screening the relative long-term stability of bio-
atmosphere of CO2 (or the typical composition of an exhaust gas), char, GCB Bioenergy 5 (2013) 215–220.
since this process can lead to important benefits for scaling-up [18] M. Rhodes, Introduction to Particle Technology, second ed., John Wiley & Sons, Ltd,
purposes. In this context, further research is needed in determining Chichester, UK, 2008.
[19] L. Wang, O. Skreiberg, M. Gronli, G.P. Specht, M.J. Antal, Is elevated pressure re-
the optimal set of operating conditions for a given purpose (e.g., quired to achieve a high fixed-carbon yield of charcoal from Biomass? Part 2: the
carbon sequestration potential) and for a given biomass feedstock. importance of particle size, Energy Fuels 27 (2013) 2146–2156.
[20] G.B. García, M. Calero De Hoces, C. Martínez García, M.T. Cotes Palomino,
A.R. Gálvez, M.Á. Martín-Lara, Characterization and modeling of pyrolysis of the
Acknowledgements
two-phase olive mill solid waste, Fuel Process. Technol. 126 (2014) 104–111.
[21] I. Mediavilla, M.J. Fernández, L.S. Esteban, Optimization of pelletisation and
The authors wish to acknowl nish MINECO-DGI (Project ENE2013- combustion in a boiler of 17.5 kWth for vine shoots and industrial cork residue, Fuel
47880-C3-1-R). JJM also express his gratitude to the Aragon Process. Technol. 90 (2009) 621–628.
[22] S. Mani, L.G. Tabil, S. Sokhansanj, Grinding performance and physical properties of
Government (GPT group) and the European Social Fund for additional wheat and barley straws, corn stover and switchgrass, Biomass Bioenergy 27 (2004)
financial support. 339–352.
[23] D.C. Montgomery, Design and Analysis of Experiments, sixth ed., John Wiley &
Sons, Hoboken, NJ, 2005.
Appendix A. Supplementary data [24] C. Di Blasi, G. Signorelli, C. Di Russo, G. Rea, Product distribution from pyrolysis of
wood and agricultural residues, Ind. Eng. Chem. Res. 38 (1999) 2216–2224.
Supplementary data related to this article can be found at https:// [25] O. Mašek, P. Brownsort, A. Cross, S. Sohi, Influence of production conditions on the
yield and environmental stability of biochar, Fuel 103 (2013) 151–155.
doi.org/10.1016/j.biombioe.2018.07.019. [26] F. Ronsse, S. van Hecke, D. Dickinson, W. Prins, Production and characterization of
slow pyrolysis biochar: influence of feedstock type and pyrolysis conditions, GCB
References Bioenergy 5 (2013) 104–115.
[27] Y. Wang, Y. Hu, X. Zhao, S. Wang, G. Xing, Comparisons of biochar properties from
wood material and crop residues at different temperatures and residence times,
[1] F. Verheijen, S. Jeffery, A.C. Bastos, M. van der Velde, I. Diafas, Biochar Application Energy Fuels 27 (2013) 5890–5899.
to Soils – a Critical Scientific Review of Effects on Soil Properties, Processes and [28] L. Zhao, X. Cao, O. Mašek, A. Zimmerman, Heterogeneity of biochar properties as a
Functions, Office for the Official Publications of the European Communities, function of feedstock sources and production temperatures, J. Hazard Mater.
Luxembourg, 2010. 256–257 (2013) 1–9.
[2] J.J. Manyà, Pyrolysis for biochar purposes: a review to establish current knowledge [29] F.-X. Collard, J. Blin, A review on pyrolysis of biomass constituents: mechanisms
gaps and research needs, Environ. Sci. Technol. 46 (2012) 7939–7954. and composition of the products obtained from the conversion of cellulose, hemi-
[3] W. Song, M. Guo, Quality variations of poultry litter biochar generated at different celluloses and lignin, Renew. Sustain. Energy Rev. 38 (2014) 594–608.
pyrolysis temperatures, J. Anal. Appl. Pyrolysis 94 (2012) 138–145. [30] R. Ragucci, P. Giudicianni, A. Cavaliere, Cellulose slow pyrolysis products in a
[4] J.J. Manyà, D. Alvira, M. Azuara, D. Bernin, N. Hedin, Effects of pressure and the pressurized steam flow reactor, Fuel 107 (2013) 122–130.
addition of a rejected material from municipal waste composting on the pyrolysis of [31] K. Matsuoka, D. Kajiwara, K. Kuramoto, A. Sharma, Y. Suzuki, Factors affecting
two-phase olive mill waste, Energy Fuels 30 (2016) 8055–8064. steam gasification rate of low rank coal char in a pressurized fluidized bed, Fuel
[5] M.J. Antal, S.G. Allen, X. Dai, B. Shimizu, M.S. Tam, M. Gronli, Attainment of the Process. Technol. 90 (2009) 895–900.
theoretical yield of carbon from biomass, Ind. Eng. Chem. Res. 39 (2000) [32] S.S. Hla, G.J. Duffy, L.D. Morpeth, A. Cousins, D.G. Roberts, J.H. Edwards,
4024–4031. Investigation of the effect of total pressure on performance of the catalytic wa-
[6] E.S. Noumi, J. Blin, J. Valette, P. Rousset, Combined effect of pyrolysis pressure and ter–gas shift reaction using simulated coal-derived syngases, Catal. Commun. 11
temperature on the yield and CO2 gasification reactivity of Acacia wood in macro- (2009) 272–275.
TG, Energy Fuels 29 (2015) 7301–7308. [33] P. Lahijani, Z.A. Zainal, M. Mohammadi, A.R. Mohamed, Conversion of the
[7] J. Recari, C. Berrueco, S. Abelló, D. Montané, X. Farriol, Effect of temperature and greenhouse gas CO2 to the fuel gas CO via the Boudouard reaction: a review,
pressure on characteristics and reactivity of biomass-derived chars, Bioresour. Renew. Sustain. Energy Rev. 41 (2015) 615–632.
Technol. 170 (2014) 204–210.
[8] P. Rousset, C. Figueiredo, M. De Souza, W. Quirino, Pressure effect on the quality of

123