i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9

Available online at www.sciencedirect.com

ScienceDirect

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

Numerical simulation of the hybrid filtration

combustion of biomass

Mario Toledo*, Carlos Rosales, Claudio Silvestre, Sebastia

n Caro
Department of Mechanical Engineering, Universidad T
ecnica Federico Santa Marı́a, Av. Espan
~ a 1680, Valparaiso,
Chile

article info abstract

Article history: Hybrid filtration combustion waves in a porous media were numerically analyzed for
Received 1 June 2016 biomass and methane-air mixtures. Temperature and products of the combustion waves
Received in revised form were studied in the range of equivalence ratio from 0.2 to 1.1. A model, based on a two-
11 August 2016 temperature approximation and combustion chemistry is developed to analyze tempera-
Accepted 17 September 2016 ture profiles and species of the phenomena. The numerical solutions predictions for
Available online xxx temperatures and chemical products are in good agreement with experimental data for
some biomasses contrasted. It was found that with ultra-lean methane mixtures, the
Keywords: combustion temperature presents essentially higher values than stoichiometric mixtures.
Filtration combustion Experimental validation suggests that the chemistry of the process is very sensitive to CO
Biomass formation. For low values of the biomass fraction in the porous medium, it is found that
Hydrogen homogenous reactions are dominant whereas for high values (over 40%) the heterogeneous
Syngas reactions are more influent. Thus, for ultra-lean methane-air mixtures, hydrogen pro-
duction is enhanced when the biomass fraction in the medium is increased.
© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

During the last years, hybrid porous media reactors have

Introduction
Strict environmental standards and the extensive use of fossil
fuels as primary energy sources have imposed stringent re-
quirements over combustion technologies. This scenario
challenges technology developers and scientists to seek the
promotion and improvement of techniques that efficiently
use the available resources and, moreover, the use of renew-
able energy sources. In this context, biomass has gained
attention, providing an alternative to conventional fossil
fuels, reducing petroleum dependence while decreasing the
net CO2 emissions coming from carbon based fuels. Under
these circumstances, hybrid filtration combustion is a tech-
nology that offers interesting advantages and an efficient
energy conversion from solid fuels of this nature [1,2].
been developed aiming to partially oxidize fossil fuels to
produce reducing gases. The gases produced are mainly
composed by hydrogen and carbon monoxide, among other
products of gasification. This gaseous mixture is produced
from carbon rich reactants exposed to the high temperatures
of filtration combustion. Hybrid filtration combustion com-
bines porous media combustion processes and gasification of
solid fuels by replacing a fraction of the inert solid volume
with a solid fuel. Porous media combustion is a well-known
technology that has been extensively studied during recent
years [3e10]. It is defined as a combustion wave that travels
through a medium composed by an inert solid and the inter-
stitial spaces (porosity) that percolate through the solid
occupied space. In the case of hybrid filtration combustion, a
combustion wave is produced by a flow that can contain hot

* Corresponding author.
E-mail address: mario.toledo@usm.cl (M. Toledo).
http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
0360-3199/© 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
2 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9

Sciences, using the superadiabatic regime of filtration com-

Nomenclature bustion to process combustible waste in two stages. The first
!u Filtration velocity
stage considers a hybrid filtration combustion reactor where a
steam-air flow is filtered through the waste and porous ma-
q Porosity
terial, gasifying the solid fuel, and a second stage burns the
yj Mole fraction of specie j
produced gases in a steam generator; reporting efficiencies of
rg Molar density of the gas phase
!
wj Diffusion velocity
the gasification process of up to 95% and the capacity to pro-
cess materials with ash content of up to 90% and up to 60% of
Dj Diffusion coefficient of specie j
humidity [21].
lg Thermal conductivity of the gas phase
On the other hand, some works presenting mathematical
d Characteristic size of the solid particle
models based on mass and energy equilibrium have been
L Length of the reactor/domain
presented by Salgansky et al. [14,15,22,23], Glazov and
h''j Stoichiometric coefficient of product j
Polianczyk [24] and Toledo and Rosales [25]. More in detail,
h'j Stoichiometric coefficient of reactant j
Salgansky et al. [14], presented a study of superadiabatic re-
ls Thermal conductivity of the solid phase
gimes, theoretically describing a steam-air wave filtered
rf Forward reaction rate
ðHÞ through a porous medium composed by carbon and an inert
rg Reaction rate of the homogeneous reactions
solid, with good agreement with experimental results. The
rb Backwards reaction rate
numerical model considered was one-dimensional and un-
Wj Molecular weight of specie j
steady, having two temperatures in an adiabatic reactor. Using
WS Molecular weight of the solid
the same model, Kislov et al. [15] presented a numerical -and
Ci Heat capacity of the inert solid
experimental-study of the gasification of a coal with air and
f Equivalence ratio
CO2, comparing the results with coal gasified by steam and air
hg Enthalpy of the gas
!v Velocity of the combustion wave
in a porous medium finding that an equal volumetric replace-
ment of steam by CO2 in the gas flow produced a significant
Tg Temperature of the gas phase
decrease of the temperature of combustion and that the calo-
Ts Temperature of the solid phase
rific values of the gaseous products were similar in both cases.
Cpj Specific heat of specie j
Salgansky et al. [26] modeled the filtration combustion of a
ε Heat fraction retained in the solid particle
 pyrolyzing solid fuel in a porous medium at a steady-state
hfj Enthalpy of formation of specie j
 through a theoretical analysis and numerical techniques,
hfS Enthalpy of formation of the solid
using a chemical scheme that included the formation of a coke
x Heat transfer coefficient
residual and oxidation reactions; results showed that at a low
ng Kinetic viscosity of the gaseous phase
content of porous material, pyrolysis occurs in a zone distant
Cf Heat capacity of the fuel
from the combustion zone, generating a pyrolysis front that
LHVj Lower heating value of specie j
moves as the combustion wave propagates in the domain.
Sj Summation of source terms
Amelin et al. [23] studied the dependence of temperature and
QRg Standard heat of reaction
rate of filtration combustion of carbon on the main operational
t Simulation time
parameters of the process through a model, being able to reveal
Sub-index the critical conditions to produce a stationary regime, finding
g Gas that in such systems, the combustion wave velocity is deter-
s Solid mined by the rate of oxidant supply. Glazov and Polianczyk [24]
f Solid fuel present a model for the steady-state filtration of endothermic
i Inert solid oxidizers through a carbon and inert solid mixture. The
gaseous phase contained steam and/or carbon dioxide. Pre-
dictions of the simulations showed qualitative agreement with
air, water-steam and/or a gaseous fuel-air mixture. This wave experimental data. Rabinovich et al. [27] presented the nu-
propagates reforming the solid fuel contained in the porous merical results of the modeling of the combustion wave sta-
medium within a wide range of thermal power with high ef- bility of the filtration combustion wave of a solid fuel, finding
ficiency, high energy concentration per unit of volume [11] that the propagation of a plane combustion wave is unstable
and stable combustion over a wide range of equivalence ra- under certain conditions, presenting an inclination that is
tios [12]. Experimental hybrid filtration combustion has been found to be related to the dimensionless gas flow rate and
studied for syngas production using several fuels, clearly width of the reactor, determining these conditions.
showing that this technology represents a strong and feasible Toledo and Rosales [25] presented a theoretical model for
alternative for syngas production from solid and gaseous hybrid filtration combustion of a solid fuel which has not been
fuels, also glimpsing new opportunities for the development tested. Therefore, such model is applied in this work for the
of new reactors [13]. Reports using carbon [14,15], coal [2,12], analysis of hybrid filtration combustion of biomass. In this
wood [1,16], eucalyptus nitens, pinus radiata, oat and wheat way, the study provides additional characterization of this
cane [17], avocado seeds [18], polyethylene [19] and algae [20], new tool as well as it contributes to the description and un-
among other materials, are available. An interesting applica- derstanding of the phenomena. This can lead to further de-
tion of the technology has been developed by the Institute of velopments and optimization of technologies that use this
Problems of Chemical Physics of the Russian Academy of combustion technique.

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9
Hybrid filtration combustion model
The process to be modeled consists on the filtration of a
methane-air mixture through a porous medium composed by
a mixture of alumina spheres and biomass. Fig. 1 shows a
schematic of the process. While a methane-air mixture is
filtered through the medium, a combustion wave travels
through it following the heat transfer mechanisms raised up
by the chemistry of the phenomena.
The governing equations presented by Toledo and Rosales
[25] considered a general three-dimensional and non-steady
analysis of this scheme. The model involves the coupling of
chemical kinetics and transport mechanisms by diffusion and
convection by means of energy balance equations for the solid
and gaseous phases, as well as mass conservation equations
for the gas phase and for the solid fuel. First, the general
equations of the model are summarily presented, which are
later specialized for a one-dimensional case.
The main assumptions considered in the formulation of
the model are: (1) the process is isobaric, (2) the propagation
velocity of the combustion wave in the medium is lower than
the filtration velocity, (3) all the biomass is consumed and (4)
the solid product of pyrolysis is composed essentially by car-
bon. The model does not assume thermal equilibrium be-
tween the solid and gas phases, and heat exchange between
them is taken into account.
Under these considerations, the process is modeled by the
following equations (full derivation and detailed description
for expressions to obtain the parameters can be found in
Toledo and Rosales [25], nomenclature is equivalent):
Conservation of mass for gaseous species
v



qrg yj þ V$ rg !
u yj  V$ qrg Dj Vyj ¼ Sj for j ¼ 1;…;NSP
vt
Eq. (1) presents the mass conservation (in molar terms) for
the gaseous species. There is one equation of this form for each
one of the NSP species present in the gaseous phase, being their
mole fractions yj the dependent variables. The first term on the
left hand side (LHS) represents the molar variation of species j
through time. The second term on the LHS is the advection of
the species j by the filtration velocity, and the third term rep-
resents the diffusion of the species j. Sj is the net source term,
which considers the formation and consumption of the
different species by the homogeneous and heterogeneous re-
actions that take place in the reactor, and also the contribution
of the volatiles released from the solid fuel. Expressions to
Fig. 1 e Schematic of the hybrid filtration combustion
process.
Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
3
obtain these sources can be found in Ref. [25]. Basically, the
data required to use those expressions are the Arrhenius pa-
rameters for the homogeneous and heterogeneous reactions.
In this particular case, we take the global homogeneous reac-
tion presented in Table 1, and the abridged heterogeneous
reaction mechanism listed in Table 2. Also, the volatilization
process must be defined. Therefore, to model the process of
volatilization through thermal decomposition and pyrolysis, a
two first-order irreversible reaction scheme is considered,
which can be written -according to Di Blasi [28]- generically as:
biomass/ð1  m1 ÞS1 þ m1 V1 (2.a)
biomass/ð1  m2 ÞS2 þ m2 V2 (2.b)
Here, the thermal decomposition of the biomass material
can be described through the release of two characteristic
volatile gas mixtures, namely V1 and V2, each one with a
prescribed chemical composition. These volatiles are consid-
ered to contain H2, H2O, CO, CO2, and CH4, as derived from the
gross reactions presented by Salgansky et al. [14]. The release
of V1 and V2 leaves corresponding solid residues S1 and S2
which can further react heterogeneously with the surround-
ing gas. For this application we assume that S1 and S2 are
composed only by carbon. The constants m1 and m2 indicate the
mass fraction of the fuel that is transformed into volatiles by
their respective reaction. Both reactions proceed simulta-
neously according to kinetic parameters given in Table 3.
Conservation of mass for solid fuel
vh i h i
rf ð1  qÞ þ V$ rf ð1  qÞ!
v j ¼ Bf (3)
vt
Eq. (3) expresses the mass conservation principle for the
(1) fuel in the solid phase. The dependent variable rf is the partial
density of fuel in the solid phase. That is the mass of solid fuel
per unit of volume occupied by solids (the rest of such volume
is occupied by the inert solid, being in this case the alumina
spheres). The second term on the LHS arises due to the
propagation of the combustion wave, while Bf is the rate of
solid fuel consumption by volatilization and heterogeneous
reaction. This term is computed by adding up all the rates of
solid fuel consumption given by the reactions in Eq. (2). Spe-
cific details can be found in Ref. [25].
Conservation of energy for the gas phase
v

  
qrg hg þ V$ rg hg !
u ¼ V$ qlg VTg þ HV;v þ HV;HE þ HV;Q (4)
vt
The conservation of energy of the gas phase is described by
Eq. (4). The molar enthalpy for the gas is given by Eq. (5), which
directly links Eq. (4) with the mass conservation equations for
the species, and where hj are the total enthalpies for those
species (computed as functions of the gas temperature).
Table 1 e Homogeneous reactions.
No. Reaction Reaction constant
1 CH4 þ 2O2 / CO2 þ 2H2O 2.6$108exp(15643.8/T) [23]
4 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9

Table 2 e Heterogeneous reactions. Numerical solution and experimental contrast

No. Reaction Reaction constant [24]
3 2C þ O2 / 2CO 2.013$100exp(9.6/T)
The model is solved using an implicit finite difference scheme,
4 C þ O2 / CO2 9.18$101exp(8200/T) considering a forward first order scheme for the time related
5 C þ CO2 / 2CO 1.24$106exp(29600/T) terms, a backward scheme for the convective terms and a
6 C þ H2O / CO þ H2 5.8$102exp(16500/T) second order central difference scheme for the diffusion
terms. The discretization produces a linear system of equa-
tions that can be solved using a tridiagonal matrix algorithm
through a Fortran 90 code.
Table 3 e Parameters for the volatilization reactions.
The reactor has a length of 0.35 m (the station x ¼ 0 is
Reaction Pre-exponential Activation located at the point where the gaseous reactants are fed). It is
[s1] energy [K]
filled with alumina spheres (0.0056 m in diameter) that create
biomass / (1  m1)S1 þ m1V1 2$105 104,500 a medium with a porosity of 40%. Initial temperature of the
biomass / (1  m2)S2 þ m2V2 1.3$107 167,200 domain is 300 K. The thermal conductivity of the solid and gas
phases are 1.016 and 0.022 W/mK respectively. The densities
X
NSP
of the solid fuel and the alumina spheres are 630 and 3987 kg/
hg ¼ yj hj (5)
j¼1
m3. The same diffusion coefficient (0.00005 m2/s) was
considered for all the gaseous species. The thermal properties
In Eq. (4), the HV,v term represents the enthalpy carried into of the gaseous species were obtained from Turns [29].
the gas phase by the volatiles released from the solid fuel. The simulations are started by imposing (at t ¼ 0 s) a
HV,HE corresponds to the net heat source for the gas due to the temperature of 1200 K for the solid phase over an intermediate
heterogeneous reactions. The last term, HV,Q, is the heat interval 0.136 m < x < 0.225 m of the domain in order to
transfer to the gas by convection from the solid phase surface generate the ignition. The discretization of the domain con-
in contact with the gas (the solid phase surface has a tem- sists on a fixed and uniform grid with 60 nodes. The time-step
perature Ts which is determined by Eq. (5)). The heat released that allows convergence of the numerical solution is 0.0001 s.
by the homogeneous reactions is implicit in Eq. (4), since the The fraction ε, that represents the heat of heterogeneous re-
terms hj already include the enthalpies of formation. actions stored in the solid particles was 30%. The equivalence
ratio is expressed considering only the gaseous reactants fed
Conservation of energy for the solid phase to the reactor.
Fig. 2a shows the numerical results obtained for tempera-
v

 

 
ture and concentrations with f ¼ 1.1 and 20 min of simulation.
ð1  qÞ rf Cf þ ri Ci Ts þ V$ ð1  qÞ rf Cf þ ri Ci ! v Ts
vt It can be noted that the gas temperature is higher than the
ε X
Nhe
  solid temperature (at 0.2 m in the axial position), which in-
¼ V$ðð1  qÞls V Ts Þ  rðHÞ QRg  xF Ts  Tg (6)
WS g¼1 g dicates that the main energy release is produced by the
combustion of the gaseous reactants coming from the feed
The conservation of energy of the solid phase is presented
and the volatilization of the solid fuel. The simulation results
in Eq. (6), where the first term in the right hand side represents
evince that the highest degree of chemical conversions takes
the heat conduction through the solid, while the second term
place at the position of the combustion wave, showing a steep
is the fraction of the heat released by the heterogeneous re-
decrease of the gaseous reactants concentrations and an in-
actions that remains in the solid phase. The last term is the
crease for the final products in the gaseous mixture. In
heat transfer by convection from the solid phase to the gas.
particular, it is observed that methane is not completely
Note that Ts represents the temperature of the solid phase
consumed, which can be related to the release of methane
without distinction between inert and fuel particles. Because
through the gasification of the biomass and the partial
of the thorough mix and contact between such particles, we
oxidation conditions.
have assumed that they are in local thermal equilibrium, so
A comparison of the numerical results (temperature of the
that the thermal state of the solid medium can be character-
solid phase) with experimental data shows very good agree-
ized by only one local temperature Ts.
ment (Fig. 2b). The experimental data are taken from the study
Having presented the general governing equations of the
presented by Araus et al. [16]. It can be observed that the
model, we concentrate now in the analysis for a one-
trends in temperature profiles exhibit similar behavior. The
dimensional case.
comparison is based on the temperature of the solid phase
The boundary and initial conditions considered are:
given that the thermocouples in the experimental setup were
x¼0: Ts ¼ Tg ¼ To ; yj ¼ y0j ; rf ¼ r0f located inside a ceramic rod, and the overall temperature
measurement were estimated to be close to the temperature
dTs dTg dyj drf
x¼L: ¼ ¼ 0; ¼ 0; ¼0 (7) of the solid phase. Also the simulation provide a good pre-
dx dx dx dx
diction of traveling velocity equal to 0.00472 cm/s (upstream
t¼0: Ts ¼ Tg ¼ To ; yj ¼ y0j ; rf ¼ r0f wave propagation) in comparison with experimental data of
Some physical properties that are temperature dependent, 0,00495 cm/s.
such as thermal conductivity, specific heat, are considered as Fig. 3 shows a comparison of the simulation results with
constant. the experimental data provided by Caro et al. [17] for f ¼ 1.1

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9 5

Fig. 2 e (a) Temperature profile and gaseous concentrations for the gaseous and solid phases at f ¼ 0.7 and a filtration
velocity of 0.15 m/s. (b) Temperature of the solid phase for numerical results and experimental measurements from Araus
et al. [16] for f ¼ 0.7 and a filtration velocity of 0.15 m/s.

and a filtration velocity of 0.261 m/s. The plot shows that the of the products can be observed in Fig. 4a, for the original
simulation provides a good prediction for CO2, H2 and CH4, scheme, as the equivalence ratio changes except in CO and
with respect to average values for the shown biomass results. CH4. Thus, the methane proportion fed to the reactor does not
Nevertheless, CO presents significant differences. Moreover, it highly influence the production of H2. It is observed that an
can be noted that the numerical results for CO and CO2 have a increment of f decreases CO production from 28.9% to 21.6%.
similar tendency considering the differences in the produc- Higher concentrations of methane are found as f increases,
tion with the cases of oat and wheat cane. It is observed that which is related to the enrichment of the mixture and non-
the numerical hydrogen production shows similar results reacting methane. On the other hand, H2O and CO2 present
than olive. On the other hand, methane production prediction slight changes within the studied range. Also, H2 production is
matches pine results. In general, the model provides a good favored by using a low f. Moreover, an increment of f pro-
prediction of the main chemical species produced for selected duces a decrease on CO, which suggests that the carbon-
biomass fuels. oxygen partial oxidation reaction is predominant in the
Since the main differences are observed in CO production, scheme. This hypothesis is based on the concentrations
the influence of every reaction related to CO production is shown in Fig. 4b, where that reaction has been removed from
studied. It was found that the incomplete oxidation reaction the scheme. In that case, the CO concentration decreased to
of carbon is the main source of CO in the reaction mechanism.
For this reason, reaction 1 (Table 2) was removed from the
scheme of heterogeneous reactions. Therefore, heteroge-
neous reactions considered are: C þ O2 ¼> CO2; C þ CO2
¼>2CO; C þ H2O ¼> CO þ H2. Using this new scheme, CO
production decreases from 21.4 to 4.24%, which is in good
agreement with the experimental results contrasted. This
change also affects the rest of products, having an increase up
to 21.3% of CO2 and 2.59% of H2; CH4 was almost unaltered,
with a concentration of 8.67% in the computations. Under
these circumstances, the removal of the reaction 1 allowed an
improvement on the numerical results.

Variation of operational parameters

Equivalence ratio

Fig. 4 shows the results for the concentrations predicted by the

simulations of the hybrid filtration combustion of biomass for
different equivalence ratio in the range 0.2  f  1.1. This has
been done for the original scheme of chemical reactions
contained in Table 2 and also for the modified scheme which
has Reaction 1 removed. Slight changes in the concentrations
Fig. 3 e Comparison of the numerical results and the
experimental results of Caro et al. [17] for the product
concentrations at f ¼ 1.1 and a filtration velocity of
0.261 m/s.

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
6 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9

60 1000 60 1100

1050
50 950 50
Ts
Tg
1000
CO Ts
40 CO 900 40
CO2 Tg

Concentration, %
950
Concentration, %

CO2

Temperature, K
H2O

Temperature, K
H2O H2
H2 CH4
30 CH4 850 30 900

850
20 800 20

800
10 750
10
750

0 700
0 700
0.2 0.4 0.6 0.8 1.0 1.2
0.2 0.4 0.6 0.8 1.0 1.2

Equivalence Ratio Equivalence Ratio

(a) (b)
38

36

34
EROI, %

32

30

28
0.2 0.4 0.6 0.8 1.0 1.2

Equivalence Ratio

(c)
Fig. 4 e Product concentrations and temperature of the solid and gas phases as a function of the equivalence ratio. (a) results
using the original heterogeneous reaction scheme from Table 2. (b) results with the modified scheme. (c) EROI calculated as a
function of the equivalence ratio with the results of the modified scheme. Biomass/alumina 50/50 v/v. Filtration velocity of
0.15 m/s.

values closer to those experimentally reported. As this suggest that the high temperatures of ultra-lean equivalence
modified scheme has less oxygen consuming reactions, a ratios promote the formation of H2.
higher rate of production of complete combustion species can From the variation of f, it is of interest to provide an idea of
be found. For this reason, higher values of CO2 and H2O are the performance and technical feasibility of the reactor.
obtained. Therefore, for such purpose, the energy return on investment
In both cases (Fig. 4), an increase in f implies a decrease of (EROI) was calculated from the simulations considering the
the temperature of the solid and gas phases. It is noted that in procedure presented by Ripoll et al. [20] for the hybrid filtra-
every result, the temperature of the gas phase is higher than tion combustion. The EROI is considered by comparing the
the solid phase, which is expected as the assumption of energy input in terms of the heating values of the gas and solid
absence of thermal equilibrium was considered. The tem- fuels inserted with the heating values of the output. Further-
peratures obtained with the original scheme (Fig. 4a) are lower more, the EROI is calculated with Eq. (8) using the data from
than the temperatures obtained with the modified chemical Fig. 4b, assuming that constant products are obtained during
scheme (Fig. 4b), which is related to the removal of the reac- simulation time t and the products considered were H2, CO
tion of partial oxidation of carbon, favoring heat release by the and CH4 as in Ref. [20], with LHV values for gases (in kJ/m3) and
complete combustion reactions of carbon and CH4. Results biomass (in kJ/kg) found in literature [17,29].

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9 7

P
products 5
LHVi $yi
Energy of the product i
EROI ¼ ¼ (8)
Energy of the reactants LHVCH ;0 $yCH ;0 þ rf $0:5$L$LHV
!
f
4 4
u $t 4 0.2
0.4
EROI variation as a function of f results are shown in 0.7

Hydrogen Concentration, %
1.0
Fig. 4c, suggesting that an increase of f, and therefore, a
change of the composition of the gasifying agent, provides a 3

higher energy conversion performance. An exponential in-

crease of EROI is observed, stepping from 28.2% at f ¼ 0.2 up to
37.5% at f ¼ 1.1. These results are in a similar order of 2
magnitude to the calculations of Ref. [20], which indicates that
Fig. 4c provides a qualitative description of the performance of
the technology, showing its capacity to transform energy from 1
biomass. Ruoppolo et al. [30] report that a fluidized bed gasifier
presents an energy conversion of 46% with air as a gasifying
agent in an equivalence ratio of 0.3 for the solid fuel, which 0
suggests that the hybrid filtration combustion technology 10 20 30 40 50 60

presents an alternative to compete with well-developed Biomass Fraction in the Porous Medium, %
technologies.
Fig. 6 e Hydrogen production as a function of the biomass
fraction in the porous medium for different equivalence
Fractions of biomass in porous medium
ratios. Filtration velocity of 0.15 m/s.

Fig. 5 shows the results obtained of the maximum tempera-

ture of the solid phase as a function of the biomass volumetric
As H2 is one of the main compounds of syngas, Fig. 6 shows
fraction (from 12.5% to 62.5%) in the porous medium for
its production varying the presence of biomass in the porous
f ¼ 0.2, 0.4, 0.7 and 1.1. The temperature rises as the biomass
medium for several equivalence ratios. Results indicate that
presence in the medium increases for f  0.7. With f ¼ 0.2 and
H2 production is severely influenced by the biomass presence,
0.4, the temperature presents a minimum at 40% and 30%,
rather than the equivalence ratio. For a low content of
respectively. Nevertheless, temperatures for f ¼ 0.2 are the
biomass in the reactor, H2 is independent of the equivalence
highest temperatures for the equivalence ratios considered
ratio, but, if the biomass content increases, the equivalence
and the results show that the temperature decreases by an
ratio gains importance. With an increase of the biomass
increase of f. The observed minimum can be related to change
content, the higher H2 values are obtained with the lowest
in the predominance of the reaction mechanisms, passing
equivalence ratio, suggesting that the heterogeneous re-
from homogeneous to heterogeneous as f is increased.
actions play a fundamental role in the process, regardless of
the increase of the CH4 feed.
1100

Conclusions
0.2
1050 0.4
0.7 The process of gasification/pyrolysis of biomass in a porous
1.0
media burner was numerically studied, considering the
1000
Solid Temperature, K

filtration of a methane-air mixture. Simulations were per-

formed for 0.2  f  1.1 and a filtration velocity of 0.15 m/s,
varying the biomass content on the porous medium.
950
The governing equations considered [25] are thermally
validated considering the absence of thermal equilibrium
900 among the solid and gas phases and the simple reaction
mechanism proposed, having good agreement with experi-
mental data and the predictions of the combustion simulations.
850 The results of the removal of the partial oxidation reaction
of carbon and oxygen (that produces CO2) to the chemical
mechanism considered, provide a better prediction than the
800
results using the complete proposed scheme for CO, H2 and
10 20 30 40 50 60
CH4, in detriment of CO2. The variation of the number of het-
Biomass Fraction in the Pororus Medium, % erogeneous reactions considered decreases precision on the
maximum temperature predicted, increasing the temperature
Fig. 5 e Maximum temperature of the solid phase as a of the solid phase compared with experimental data. The
function of the biomass fraction in the porous medium for experimental value was 937 K for wheat (for a filtration ve-
different equivalence ratios. Filtration velocity of 0.15 m/s. locity of 0.261 m/s and f ¼ 1.1), while the use of the original and

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
8 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9
modified schemes predicted 963 K and 1024 K, respectively.
This analysis suggests that the removal of the reaction does
not totally improve the numerical results and also simplifies
the proposed model, despite the importance of that reaction in
the gasification/pyrolysis physical-chemical description.
Identifying this, the chemistry of the process requires a
detailed attention for further applications and improvement of
the model, with special emphasis on the CO formation rates. It
must be noted that the model can predict a thermal descrip-
tion of the process as long as one of the partial oxidation re-
actions remains in the scheme, to provide the energy required
for the gasification reactions, which implies that the com-
bustion reactions highly influence the gasification/pyrolysis of
the solid and the reactions of the volatiles released.
On the other hand, the variation of the equivalence ratio for
a 50/50 v/v presence of biomass in the porous medium indicate
that this is not a relevant parameter to obtain significant var-
iations in the H2 production, having that the higher H2 presence
was obtained in the leanest equivalence ratio tested. Never-
theless, the equivalence ratio showed a significant impact on
the EROI, suggesting that the energy conversion performance
of the reactor increases as a richer mixture is filtered.
In addition, the variation of the biomass fraction in the
porous media highly influenced H2 production, where an in-
crease in the biomass fraction implied a significant increase in
H2 production. With f ¼ 0.2, an increase of H2 from 0.13% to
4.17% is obtained as the biomass fraction increased from
12.5% to 62.5%. These results remark the predominance of the
biomass fraction over equivalence ratio for H2 production.
Thus, heterogeneous reactions take a more important role
than homogeneous reactions in simulation results.
Considering the strong assumptions made for the
modeling of the process and the absence of a pyrolysis model
for every specific kind of biomass contrasted, the simulations
provide a suitable convergence and shows good agreement
with the expected results for the temperatures and product
concentration. This study contributes handing over a useful
tool to describe the phenomena for further developments that
aim to use the important potential of biomass to produce H2
and energy using hybrid filtration combustion. Further studies
of the technology shall develop or adapt a more complex
chemical mechanism for each biomass or solid fuel consid-
ered to improve the predictions of the species produced.
Acknowledgments
The authors wish to acknowledge the support of the CON-
ICYT-Chile (FONDECYT 1121188 and FONDAP 15110019/SERC-
Chile).
references
[1] Toledo M, Vergara E, Saveliev AV. Syngas production in hybrid
filtration combustion. Int J Hydrogen Energy 2011;36:3907e12.
[2] Toledo M, Utria K, Gonza
lez F, Zún
~ iga J, Saveliev A. Hybrid
filtration combustion of natural gas and coal. Int J Hydrogen
Energy 2012;37:6942e8.
Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120
[3] Abdul Mujeebu M, Abdullah MZ, Abu Bakar MZ,
Mohamad AA, Muhad RMN, Abdullah MK. Combustion in
porous media and its applications e a comprehensive
survey. J Environ Manag 2009;90:2287e312.
[4] Bingue JP, Saveliev AV, Fridman AA, Kennedy LA. Hydrogen
production in ultra-rich filtration combustion of methane
and hydrogen sulfide. Int J Hydrogen Energy 2002;27:643e9.
[5] Toledo M, Utria K, Saveliev AV. Ultrarich filtration
combustion of ethane. Energy Fuels 2014;28:1536e40.
[6] Toledo M, Bubnovich V, Saveliev A, Kennedy L. Hydrogen
production in ultrarich combustion of hydrocarbon fuels in
porous media. Int J Hydrogen Energy 2009;34:1818e27. http://
dx.doi.org/10.1016/j.ijhydene.2008.12.001.
[7] Drayton MK, Saveliev AV, Kennedy LA, Fridman AA, Yao-
En L. Syngas production using superadiabatic combustion of
ultra-rich methane-air mixtures. Twenty-Seventh Symp
Combust Vol One 1998;27:1361e7.
[8] Kennedy LA, Bingue JP, Saveliev AV, Fridman AA, Foutko SI.
Chemical structures of methane-air filtration combustion
waves for fuel-lean and fuel-rich conditions. Proc Combust
Inst 2000;28:1431e8.
[9] Smith CH, Pineda DI, Zak CD, Ellzey JL. Conversion of jet fuel
and butanol to syngas by filtration combustion. Int J
Hydrogen Energy 2013;38:879e89. http://dx.doi.org/10.1016/
j.ijhydene.2012.10.102.
[10] Toledo M, Gracia F, Caro S, Go
mez J, Jovicic V. Hydrocarbons
conversion to syngas in inert porous media combustion. Int J
Hydrogen Energy 2016;41:5857e64.
[11] Durst F, Pickena € cker K, Trimis D. Porenbrenner e kompakte,
emissionsarme Brenner mit grossem
Leistungsmodulationsbereich. Gas Wasserfach Gas Erdgas
1997;138:116e23.
[12] Toledo M, Araus K, Vasconcelo D. Syngas production from
coal in presence of steam using filtration combustion. Int J
Hydrogen Energy 2015;40:6340e5.
[13] Abdul Mujeebu M. Hydrogen and syngas production by
superadiabatic combustion e a review. Appl Energy
2016;173:210e24.
[14] Salgansky EA, Kislov VM, Glazov SV, Zholudev AF,
Manelis GB. Filtration combustion of a carbon-inert material
system in the regime with superadiabatic heating. Combust
Explos Shock Waves 2008;44:273e80.
[15] Kislov VM, Glazov SV, Salgansky EA, Kolesnikova YY,
Salganskaya MV. Coal gasification by a mixture of air and
carbon dioxide in the filtration combustion Mode. Combust
Explos Shock Waves 2016;52:320e5.
[16] Araus K, Reyes F, Toledo M. Syngas production from wood
pellet using filtration combustion of lean natural gas-air
mixtures. Int J Hydrogen Energy 2014;39:7819e25.
[17] Caro S, Torres D, Toledo M. Syngas production from residual
biomass of forestry and cereal plantations using hybrid
filtration combustion. Int J Hydrogen Energy
2015;40:2568e77.
[18] Domı́nguez MP, Araus K, Bonert P, Sa
nchez F, San Miguel G,
Toledo M. The avocado and its waste: an approach of fuel
potential/application. In: Environment, energy and climate
change II: energies from new resources and the climate
change. Springer International Publishing; 2016. p. 199e223.
http://dx.doi.org/10.1007/698_2014_291.
[19] Gentillon P, Toledo M. Hydrogen and syngas production from
propane and polyethylene. Int J Hydrogen Energy
2013;38:9223e8.
[20] Ripoll N, Silvestre C, Paredes E, Toledo M. Hydrogen
production from algae biomass in rich natural gas-air
filtration combustion. Int J Hydrogen Energy. http://dx.doi.
org/10.1016/j.ijhydene.2016.03.082
[21] Aldoshin SM. Achievements and innovation prospects of
chemical science. Her Russ Acad Sci 2012;82:143e57.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 6 ) 1 e9 9

[22] Salgansky EA, Fursov VP, Glazov SV, Salganskaya MV,
Manelis GB. Model of vapor-air gasification of a solid fuel in a
filtration Mode. Combust Explos Shock Waves 2006;42:55e62.
[23] Amelin II, Salgansky EA, Volkova NN, Zholudev AF,
Alekseev AP, Polianczyk EV, et al. Parametric domain of the
stationary filtration combustion wave in the charge with a
low carbon content. Russ Chem Bull 2011;60:1150e7.
[24] Glazov SV, Polianczyk EV. Filtration combustion of carbon in
the presence of endothermic oxidizers. Combust Explos
Shock Waves 2015;51:540e8.
[25] Toledo M, Rosales C. Hybrid filtration combustion. In: Minic D,
editor. Hydrogen energy e challenges perspect. InTech; 2012.
[26] Salgansky EA, Polianczyk EV, Manelis GB. Modeling filtration
combustion of a pyrolyzing solid fuel. Combust Explos Shock
Waves 2013;49:38e52.
[27] Rabinovich OS, Malinouski AI, Kislov VM, Salgansky EA.
Effect of thermo-hydrodynamic instability on structure
and characteristics of filtration combustion wave of solid
fuel. Combust Theory Model 2016;20:877e93.
[28] Di Blasi C, Branca C, Galgano A, Meier D, Brodzinskil D,
Malmros O. Supercritical gasification of waste water
from up draft wood gasifiers. Biomass Bioenergy
2007;31:802e11.
[29] Turns SR. An introduction to combustion: concepts
and applications. 2nd ed. McGraw-Hill Higher Education;
2000.
[30] Ruoppolo G, Ammendola P, Chirone R, Miccio F. H2-rich
syngas production by fluidized bed gasification of biomass
and plastic fuel. Waste Manag 2012;32:724e32.

Please cite this article in press as: Toledo M, et al., Numerical simulation of the hybrid filtration combustion of biomass, International
Journal of Hydrogen Energy (2016), http://dx.doi.org/10.1016/j.ijhydene.2016.09.120