Predicting the total product distribution from biomass gasification: An approach to

thermochemical biorefineries' design and optimization

Abstract

Introduction
In the context of circular economy and carbon neutrality, biorefineries present an attractive,
and sustainable way to produce various renewable products from biomass. However, while
significant progress has been made, implementation of cost-effectively large-scale
biorefineries is still a challenge due the technological immaturity. Process systems
engineering could play a significant role in the synthesis, optimization and up-scaling
technologies in biorefineries; however, more research is still needed to improve the
mechanism that occur on reactor scale (Bong, Fan et al. 2022).
Biomass can be converted to energy via multiple conversion processes classified as
biochemical and thermochemical processes. Unlike biochemical methods that are use slow
acting microbes, the thermochemical conversion technologies such as such as pyrolysis and
gasification have the ability thermally breakdown all types of dry biomass into valuable
products and within a short time. Among the thermochemical methods, pyrolysis stands out
as most important and intermediate step of the other two thermochemical processes i.e.,
gasification and combustion processes. Pyrolysis has been used as a preliminary step to other
thermochemical processes such as combustion, gasification, and reforming. Therefore, it is of
primary importance to know the composition of the different products formed during this pre-
step since it may have a strong influence on the overall process composition (Abou Rjeily,
Cazier et al. 2023).
The bio-oil is complex viscous liquid mixture and can be composed of more than 300 organic
compounds, and its composition is highly influenced by the nature of biomass feedstock and
operating conditions of pyrolysis(Martínez-Lera and Pallarés Ranz 2016, Abou Rjeily,
Cazier et al. 2023). Nevertheless, the pyrolysis process is not restricted to the main
decomposition of the biomass since secondary reactions such as tar cracking or bio-oil
pyrolysis can take place between the volatile species at high temperatures (Tao, Dong et al.
2015, Abou Rjeily, Cazier et al. 2023). The organic liquid includes large, thermally unstable
molecules need higher temperatures and more steam to effectively gasify the carbonaceous
deposits formed by thermal decomposition (Morf, Hasler et al. 2002). Besides, biomass
composition and temperature, tar cracking to a large extent also depends on the vapor
residence time in the reactor (Martínez-Lera and Pallarés Ranz 2016).
In biomass gasification, prediction of tar in syngas remains a challenge in process design.
Moreover, tar is a very complex material, and its composition depends on many parameters
such as temperature, type of gasifying agent and reactor designs (Zhang and Pang 2017).
According to the review by Mutlu and Zeng (Mutlu and Zeng 2020), kinetic modeling of tar
formation stands as one of the challenges and there is still no consensus in the literature.
While some studies have attempted to model based on assumptions and tar representations,
comparison cannot be possible due to the various exclusions (Ahmed, Salmiaton et al.
2015).
In most studies that involve modeling of gasification processes, it either assumed that in the
first step the atomic components (C, H, O, N, S) transform in to volatile components ( H 2,
CO,CO2, N2 and Carbon) in an RYield type model reactor followed by Gibbs minimization or
a reactor with the kinetics of steam reforming (Visconti, Miccio et al. 2015).
Sierra et al. (Sierra Jimenez, Ceballos Marín et al. 2021) implemented kinetic and Gibbs
equilibrium models from literature to simulate fast pyrolysis and gasification in a fixed-bed
and fluidized bed reactors in Aspen Plus software. While, a detailed decomposition kinetics
was used to describe the fast pyrolysis model, the Gibbs equilibrium models were used to
describe the gasification process. Kaushal and Tyagi (Kaushal and Tyagi 2017) proposed a
sub model for generation and cracking of tar during biomass gasification in bubbling
fluidized gasifier . Even though the model was capable of predicting the yield of tar for a
wide variety of biomasses and fluidizing agents, the composition of tar components could not
be determined. Moreover, the gas phase reactions were defined by Gibbs equilibrium.
While other researchers have attempted to describe biomass devolatilization and evolution of
tar using empirical models, followed by kinetic models for the gasification of the pyrolytic
products, their application is only limited to the specific reactor and reaction conditions
(Abdelouahed, Authier et al. 2012).
Kinetic models provide more realistic insight into the various mechanisms of biomass
devolatilization and gasification, that reveals the relative influence of gasifier design and
operating parameters that important for the optimization of the process (Buragohain,
Chakma et al. 2013).
Even process modeling has been frequently applied to study the gasification process,
consecutive kinetics of biomass devolatilization and gasification reactions to provide detailed
product composition is rarely studied if any. Moreover, the use of lumped kinetic data widely
used in the literature can affect the reliability of the gasification models including
devolatilization and char reactions (Janajreh, Adeyemi et al. 2021).
To predict the detailed product distribution of biomass pyrolysis and gasification, i.e., gas,
liquid, and solid products. This work offers a new wood gasification model for fixed bed
reactors, characterized by a detailed definition of homogeneous reactions, considering both
generation and evolution of created tar through thermal cracking and steam reforming. Scot
pine wood is selected as the biomass sample. The effect of temperature will be studied by at
reactor temperatures of 750, 850 and 930 oC.
The model was based on the extension of the multi-step kinetic mechanism of biomass
decomposition earlier proposed by Faravelli and co-authors to include decomposition of
extractives, secondary and steam gasification reactions of primary pyrolysis products. The
kinetics has been implemented in AspenPlus ® to predict the complete product yields and
composition of slow pyrolysis and steam gasification of wood biomass. The model was
validated using experimental data obtained from the slow pyrolysis and steam gasification
(Steam/Biomass ~0.21) of pine wood (< 1mm) in a Fixed-Bed Batch reactor at temperatures
of 750, 850 and 930 oC.
Material and methods
Elemental and proximate analysis
Pine wood was characterized in terms of proximate, and elemental analysis. Details on the
equipment used for elemental and proximate analysis are found in ref. (Konist, Valtsev et al.
2015, Jarvik, Sulg et al. 2020).
Experimental setup
The slow pyrolysis and steam gasification experiments were performed in a laboratory-scale
fixed bed reactor, described in more detail in (Konist, Valtsev et al. 2015, Jarvik, Sulg et al.
2020). It is equipped with control (temperature control by means of thermocouples and gas
flow by means of mass flow controllers).
The sweep gas flow (Argon) was maintained at 100 mL/min throughout the experiment, and
steam to biomass ratio of 0.21 used to study steam gasification at all temperatures (750,850
and 930 oC). For each temperature, the experiments were conducted at least three times to
ensure reproducibility of the results. The average and standard deviation values of the
experimental results (gas composition, tar, and char) were calculated.
The solid residence time was determined using an FTIR, and 8 minutes was considered
sufficient to enable complete devolatilization of the biomass. The vapor residence time is
directly related to the operating conditions such as the carrier gas flow rate and the reactor
volume.

To investigate the tar cracking and gaseous secondary reactions, the vapor residence time was
estimated based on the sweep gas flow and volume of the tube reactor (Diameter 1 inch and 2
inch long). Since the reaction sample was positioned half of the reactor tube, a contribution
from only the first half of the tube was considered as the secondary reaction zone. Therefore,
at sweep gas glow rate of 100 mL/min, the vapor residence time was considered 40 seconds.
Gas composition
The volumetric concentration of combustible gases H 2, CO, and CH4 was measured using a
gas chromatography with thermoconductivity detector (GC-TCD) and a Gazohrom3101 gas
analyzer. The analyzer air (70 mL/min) as a carrier gas, and the vol.% was calculated using a
chromatogram function of Clarity software. The pyrolytic gas was collected in Tedlar bags as
it exited the batch reactor.
Mass balance

Initial amount of wood biomass ( mwood ). Considering the moisture content i.e., 8.5 wt.%, the
dry wood weight was determined ( mdry wood ). At the end of the experiment, the remain char is
weighed, ( mchar ) . At the beginning of the experiment, the gas outlet apparatus including the
outlet tubes, condenser gas tube, conical flask for tar collection were weighed. The apparatus
connections were tested for leaks before introducing biomass into the system. The weight
difference at the start and end of experiment provided the amount of tar or bio-oil trapped,
( mbio−oil ). The gases exiting the pyrolysis reactors were cooled by the condensate trap at -30 oC
temperature, and the condensed liquid is weighed. The weight of the gas was determined
from the global mass, mgas =mwood −mbio−oil−mchar [ Assuming the minimal losses]. The
experiments were repeated to enable reproducibility of the results.
Modeling in AspenPlus®
In this study, drying was modeled based on the kinetics of devolatilization of the water
(moisture) loosely bound in the biomass material. The multistep kinetic scheme proposed in
Faravelli et al. (Faravelli, Frassoldati et al. 2013) was modified to cater for extractives
(Debiagi, Pecchi et al. 2015), and secondary reactions (Ranzi, Debiagi et al. 2017a, Ranzi,
Debiagi et al. 2017b) (see; support information ). In the kinetic model , char was assumed to
only compose of carbon and ash. However, in this work, the composition of char after
reaction was assumed to consist of carbon, ash, and the unreacted solid components from
biomass i.e., Hemicellulose, cellulose, lignin, and extractives. While the batch type reactor in
AspenPlus® has been recommended for slow pyrolysis process (Peters 2015), the RCSTR
has been used in this study due to constant temperature conditions during the experiments.
The Peng-Robinson Equation of state with Boston -Mathias (PR-BM) modification has been
chosen to calculate the thermodynamic properties of the of the conventional components in
the model. The thermophysical properties of components including; 3-hydroxypropanal,
para-coumaryl alcohol, sinapyl aldehyde, xylosan , Lignin (Lig, Lig-C , Lig-CC, Lig -O ,
Lig-OH) , extractives (Tannin and triglyceride) , cellulose and activated cellulose , and
hemicellulose components were estimated using an approach proposed by Gorensek et al.
(Gorensek, Shukre et al. 2019)

Results and discussion

Biomass characterization
The elemental, proximate and biochemical composition of pine wood has been shown in
Table 1. The biochemical composition of scot pine was obtained from literature (Grønli
1996). The composition of lignin components ( Lig-C, Lig-H and Lig-O) , Tannin (TANN),
and tryglyceride (TGL) ) were estimated from lignin and extractives weigh fraction using the
approach proposed by Faravelli et al. (Faravelli, Frassoldati et al. 2010), and Debiagi et al.
(Debiagi, Pecchi et al. 2015). Direct estimation of the complete biochemical composition
from elemental analysis via Debiagi’s method could not be considered due to the over
prediction of cellulose component.
Table 1: Ultimate (dry basis) and proximate analysis (wet basis) of pine woods.

Moisture 8.5
Proximate analysis (wt.%) Volatile matter 85.2
Fixed carbon 14.5
Ash 0.3
C 50.1
Ultimate analysis(wt.%) H 6.6
(daf) N 0.19
S n.d.
O 43.1
Cellulose 41.0
Biochemical analysisa (wt.%) Hemicellulose 25.7
Lignin 28.6
Extractives 4.8
a
source: (Grønli 1996)
Table 2: Distribution of biochemical components (wt.%)

Biochemical composition (wt.%)

Cellulose Hemicellulose Lig-C Lig-O Lig- H TANN TGL
41.0 25.7 16.60 2.14 9.75 3.36 1.44

Slow pyrolysis
Product distribution of pyrolysis and gasification have been categorized as char, gases (non-
condensable volatiles) and liquid or condensable vapors (bio-oil). The study has been
described in two twofold: i) to predict the slow pyrolysis of wood biomass at low sweep gas
flow rate (100 mL/min) thus enabling secondary tar reactions, and ii) to predict the product
distribution of steam gasification of the primary pyrolysis products, as illustrated in Fig.1.
Since secondary pyrolysis becomes more dominant at temperature above 600 oC (Sun, Yu et
al. 2011), the model assumed a primary pyrolysis temperature of 500 oC, and solid residence
time of 8 minutes. The secondary reactions have been modeled in the second reactor, and the
vapor residence time (40 seconds) estimated from the reactor volume and sweep gas flow
(100 ml/min) was used. During thermal cracking, the steam flow rate is considered to be zero.

Fig.1 Process flowscheme in AspenPlus®

The primary pyrolysis liquid products consisted of aromatic and aliphatic compounds,
aldehydes, alcohols, levoglucosan, phenols, other hydrocarbon compounds and water
(Faravelli, Frassoldati et al. 2013). Although not experimentally validated, the model was
able to predict the yield of light hydrocarbons such as ethane and ethylene , and heavier
hydrocarbons including benzene, toluene , 1,3-cyclopentadiene , furan , phenols and others
have been reported to part of the bio-oil or tar composition from biomass pyrolysis (Abou
Rjeily, Cazier et al. 2023).
Gas composition

Fig.2 Comparison of the volumetric composition of H2, CO and CH4 from model with experimental results.

The experimental results obtain in this study were in close agreement with the previous
studies in the literature (Tao, Dong et al. 2015). The mean absolute error between the model
and experimental results was 21.53%. The model predicts slightly higher H 2 composition
(26.5 %) compared to 21 % obtained from the experiment at 750 oC. Lower concentration of
CO (41%) was predicted by the model, compared to 48 % observed in the experiment. The
CO composition predicted by the model is within range with that predicted by Tao et al. (Tao,
Dong et al. 2015). However, the model predicts low concentration of methane at all
temperatures, when compared to the experimental data in this study and that of Tao et al.
(Tao, Dong et al. 2015).
Char yield
As shown in Fig.3, the char yield predicted by the model was in close match with
experimental data. However, the char yield obtained from experimental data was slightly
lower than that predicted by the model. The mean absolute error of 14.1 % was obtained.
The difference in the amounts of char could be attributed to the losses during the removal of
the char sample from the reactor.
 Fig.3 Comparison of the char yield (wt.%) from model with experimental results.

Effect of temperature on product distribution

For the non-cracking process, the primary pyrolysis yields at 500 oC consisted of 12.27 wt.%
gas, 48.15 wt.% tar/bio-oil, 12.81 wt. % water, and 26.7 wt. % char. At the conditions of
thermal cracking (750 to 930 oC), the char yield decreased as the gas yield increased, as
shown in Table 3. However, there was only a slight change in the tar yield. Perhaps, at the
model conditions, the increase in gas yield was mainly as a result of the heterogeneous
reaction of the pyrolytic products with char (Ranzi, Debiagi et al. 2017b). At higher
temperatures, further investigation revealed that vapor residence time is the controlling factor
during thermal cracking. Detailed analysis of the effect of vapor residence time has been
discussed in the subsequent subsection.
Table 3: Effect of temperature on the gas, tar and char yield.

wt.%
o
750 C 850 oC 930 oC
Gas 57.73 64.10 65.52
Tar 5.69 5.81 5.85
Char 22.33 15.74 14.24
Water 14.20 14.34 14.38

The total product composition after thermal cracking have been shown in Table 4. The AspenPlus
model was able to predict the polymerization, dehydration and other reactions that occur during the
thermal cracking of bio-oil components (Tao, Dong et al. 2015).

Table 4: Total product composition of the pyrolysis product

mg/g_biomass
o
750 C 850 oC 930 oC
H2 14.512 17.687 18.765
CO 305.937 353.706 367.656
CO2 140.670 135.993 134.282
CH4 21.384 29.505 30.179
C2H4 88.635 94.715 94.671
 Methanol 34.241 34.489 34.571
Formaldehyde 0.011 0.003 0.002
Acetaldehyde 12.194 12.414 12.496
Hydroxymethylfurfural 0.011 0.004 0.002
Benzene 1.408 2.139 2.227
Toluene 0.000 0.000 0.000
Napthalene 6.160 9.361 9.745
Cyclopentadiene 0.001 0.011 0.048
H2O 142.454 143.467 143.835
Char 223.300 157.300 14240.000

Effect of vapor residence time

As the temperature was raised from 750 to 930 oC, the gas yield increased at the expense of
the thermal cracking of bio-oil. The char yield did not any significant decrease in the range
of 750 to 930 oC, this observation agrees with the previous studies in the literature (Abou
Rjeily, Cazier et al. 2023).

Fig.xx Effect of vapor residence time and temperature on the gas yield
 Fig.xx Effect of vapor residence time and temperature on the tar cracking

Fig.xx Effect of vapor residence time and temperature on the char yield

In less than 1 sec, the gas yield increases by over 112 % via the thermal cracking of the bio-oil. The
char yield increased by less than 1 % within the 1 second of the vapor residence time. As shown in
Fig.X, the reaction becomes even more faster at higher temperatures. The reactions become even
more faster at higher This indicates that most of the gases produced during thermal cracking
originates from thermal cracking of bio-oil. Although char is produced during thermal cracking, the
yield is not significant enough to increase the overall yield of char.
The results in this study justified the use of very low vapor residence time ( < 10 s) at higher cracking
temperatures. For instance , Tao et al. (Tao, Dong et al. 2015) studied the thermal/catalytic cracking
of tar at vapor residence times of around 11 s for 500 and 9s at 700 C .

Steam gasification

Gas and tar yield

The gasification products assumed to consist of mainly aromatic compounds such as benzene,
naphthalene and cyclopentadiene.

Total product distribution

wt.%
o
750 C 850 oC 930 oC
Gas 76.4 82.85 88.84
 Tar 4.79 4.8 4.8
Char 11.3 7.8 3.54
Water 7.5 4.54 2.7

mg/g_biomass
750 oC 850 oC 930 oC
H2 36.238 42.700 47.936
CO 471.348 485.303 513.399
CO2 184.673 220.304 241.287
CH4 2.418 2.490 2.404
C2H4 67.712 75.189 79.763
Methanol 28.828 29.066 29.139
Formaldehyde 0.144 0.047 0.022
acetaldehyde 10.275 10.464 10.533
Hydroxymethylfurfural 0.152 0.050 0.024
Benzene 0.365 0.579 0.820
Toluene 0.000 0.000 0.000
Napthalene 1.599 2.533 3.589
Cyclopentadiene 0.000 0.003 0.018
H2O 75.102 45.414 27.859
Char 113.000 78.000 50.000

The presence of aromatic compounds indicates that steam has only a little influence on the
conversion of the aromatics (Li and Suzuki 2009). Under the given reaction conditions,
neither the soot primarily formed nor the organic cracking products such as methane are
completely converted by steam to carbon monoxide and hydrogen, even at the highest
temperature. Similar observations have been reported in the literature (Li and Suzuki 2009).

Effect of solid residence time

Fig.x shows the variation of temperature with the H2: CO ratio at different residence times. With a
residence time of 20 minutes, the maximum H2/CO ratio (1.5) is obtained at a temperature of 675 C,
while that at 5mins residence time, the maximum H2/CO ratio of 1.42 is obtained at 765 C. It is clear
that the H2/CO ratio increases as the residence time is increased, and the temperature of maximum
yield reduces with increase in residence time. Possibly this observation was caused by the heat
accumulation during the slow pyrolysis at higher residence times.

Conclusion

References
Abdelouahed, L., O. Authier, G. Mauviel, J. P. Corriou, G. Verdier and A. Dufour (2012). "Detailed
Modeling of Biomass Gasification in Dual Fluidized Bed Reactors under Aspen Plus." Energy & Fuels
26(6): 3840-3855.
Abou Rjeily, M., F. Cazier, C. Gennequin and J. H. Randrianalisoa (2023). "Detailed Analysis of Gas,
Char and Bio-oil Products of Oak Wood Pyrolysis at Different Operating Conditions." Waste and
Biomass Valorization 14(1): 325-343.
Ahmed, A. M. A., A. Salmiaton, T. S. Y. Choong and W. A. K. G. Wan Azlina (2015). "Review of kinetic
and equilibrium concepts for biomass tar modeling by using Aspen Plus." Renewable and Sustainable
Energy Reviews 52: 1623-1644.
Bong, C. P. C., Y. V. Fan, C. T. Lee and J. J. Klemeš (2022). "Process analysis and optimisation for a
sustainable circular economy." Cleaner Engineering and Technology 11: 100578.
Buragohain, B., S. Chakma, P. Kumar and P. Mahanta (2013). "Comparative Evaluation of Kinetic,
Equilibrium and Semi-Equilibrium Models for Biomass Gasification." International Journal of Energy
and Environment (IJEE) 4: 581-614.
Debiagi, P. E. A., C. Pecchi, G. Gentile, A. Frassoldati, A. Cuoci, T. Faravelli and E. Ranzi (2015).
"Extractives Extend the Applicability of Multistep Kinetic Scheme of Biomass Pyrolysis." Energy &
Fuels 29(10): 6544-6555.
Faravelli, T., A. Frassoldati, E. B. Hemings and E. Ranzi (2013). Multistep Kinetic Model of Biomass
Pyrolysis.
Faravelli, T., A. Frassoldati, G. Migliavacca and E. Ranzi (2010). "Detailed kinetic modeling of the
thermal degradation of lignins." Biomass and Bioenergy 34(3): 290-301.
Gorensek, M. B., R. Shukre and C.-C. Chen (2019). "Development of a Thermophysical Properties
Model for Flowsheet Simulation of Biomass Pyrolysis Processes." ACS Sustainable Chemistry &
Engineering 7(9): 9017-9027.
Grønli, M. (1996). Theoretical and experimental study of the thermal degradation of biomass,
Doctoral Thesis, The Norwegian University of Science and Technology.
Janajreh, I., I. Adeyemi, S. S. Raza and C. Ghenai (2021). "A review of recent developments and future
prospects in gasification systems and their modeling." Renewable and Sustainable Energy Reviews
138: 110505.
Jarvik, O., M. Sulg, P. Cascante Cirici, M. Eldermann, A. Konist, J. Gušča and A. Siirde (2020). "Co-
Pyrolysis and Co-Gasification of Biomass and Oil Shale." Environmental and Climate Technologies 24:
624 - 637.
Kaushal, P. and R. Tyagi (2017). "Advanced simulation of biomass gasification in a fluidized bed
reactor using ASPEN PLUS." Renewable Energy 101: 629-636.
Konist, A., A. Valtsev, L. Loo, T. Pihu, M. Liira and K. Kirsimäe (2015). "Influence of oxy-fuel
combustion of Ca-rich oil shale fuel on carbonate stability and ash composition." Fuel 139: 671-677.
Li, C. and K. Suzuki (2009). "Tar property, analysis, reforming mechanism and model for biomass
gasification—An overview." Renewable and Sustainable Energy Reviews 13(3): 594-604.
Martínez-Lera, S. and J. Pallarés Ranz (2016). "On the development of a wood gasification modelling
approach with special emphasis on primary devolatilization and tar formation and destruction
phenomena." Energy 113: 643-652.
Morf, P., P. Hasler and T. Nussbaumer (2002). "Mechanisms and kinetics of homogeneous secondary
reactions of tar from continuous pyrolysis of wood chips." Fuel 81(7): 843-853.
Mutlu, Ö. and T. Zeng (2020). "Challenges and Opportunities of Modeling Biomass Gasification in
Aspen Plus: A Review." Chemical Engineering & Technology 43: 1674-1689.
Peters, J. (2015). Pyrolysis for biofuels or biochar? A thermodynamic, environmental and economic
assessment.
Ranzi, E., P. E. A. Debiagi and A. Frassoldati (2017a). "Mathematical Modeling of Fast Biomass
Pyrolysis and Bio-Oil Formation. Note I: Kinetic Mechanism of Biomass Pyrolysis." ACS Sustainable
Chemistry & Engineering 5(4): 2867-2881.
Ranzi, E., P. E. A. Debiagi and A. Frassoldati (2017b). "Mathematical Modeling of Fast Biomass
Pyrolysis and Bio-Oil Formation. Note II: Secondary Gas-Phase Reactions and Bio-Oil Formation." ACS
Sustainable Chemistry & Engineering 5(4): 2882-2896.
Sierra Jimenez, V., C. M. Ceballos Marín and F. Chejne Janna (2021). "Simulation of thermochemical
processes in Aspen Plus as a tool for biorefinery analysis." CT&F - Ciencia, Tecnología y Futuro 11(2):
27-38.
Sun, Q., S. Yu, F. Wang and J. Wang (2011). "Decomposition and gasification of pyrolysis volatiles
from pine wood through a bed of hot char." Fuel 90(3): 1041-1048.
Tao, J., C. Dong, Q. Lu, H. Liao, X. Du, Y. Yang and E. Dahlquist (2015). "Catalytic Cracking of Biomass
High-Temperature Pyrolysis Tar Using NiO/AC Catalysts." International Journal of Green Energy 12(8):
773-779.
Visconti, A., M. Miccio and D. Juchelková (2015). "An aspen plus® tool for simulation of lignocellulosic
biomass pyrolysis via equilibrium and ranking of the main process variables." International Journal of
Mathematical Models and Methods in Applied Sciences 9: 71-86.
Zhang, Z. and S. Pang (2017). "Experimental investigation of biomass devolatilization in steam
gasification in a dual fluidised bed gasifier." Fuel 188: 628-635.