You are on page 1of 8

international journal of hydrogen energy 33 (2008) 7419–7426

Available at www.sciencedirect.com

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

Theoretical analysis of the direct decomposition of


methane gas in a laminar stagnation-point flow:
CO2-free production of hydrogen

O. Bautistaa, F. Méndezb,*, C. Treviñoc


a
Sección de Estudios de Posgrado e Investigación-IPN, México, D.F., 02550, Mexico
b
Facultad de Ingenierı́a, UNAM, México, D.F., 04510, Mexico
c
Facultad de Ciencias, UNAM, 04510, México, D.F., Mexico

article info abstract

Article history: In this work, a theoretical analysis is developed to predict the decomposition temperature
Received 20 September 2007 of methane gas, CH4, in a planar stagnation-point flow over a catalytic carbon surface.
Received in revised form Hydrogen is produced (without CO2 as a byproduct) by means of a heterogeneous reaction
10 June 2008 mechanism, which is modeled with five heterogeneous reactions, including adsorption and
Accepted 26 September 2008 desorption reactions. The mass species, momentum, and energy conservation equations
Available online 12 November 2008 for the gas phase are solved, taking into account that the temperature of decomposition is
characterized by the Damköhler number. Therefore, the critical temperature conditions for
Keywords: the catalytic thermal decomposition are found by using a high activation energy analysis
Thermal decomposition for the desorption kinetics of the adsorbed hydrogen component, HðsÞ. Specifically, the
Methane numerical estimations show that, for increasing values of the velocity gradient associated
Hydrogen production with the stagnation flow, the temperature of decomposition grows, depending on the
Endothermic reaction surface coverages of the product species.
Surface coverage ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights
reserved.

1. Introduction requires high temperatures (1500–2000 K) in order to obtain


reasonable quantities of hydrogen. An alternative method for
Today, it is well known that hydrogen fuel, the combustion of hydrogen production is the direct thermal decomposition of
which produces only water vapor, could be a feasible energy methane (TDM), which produces no further byproducts except
resource for the future. The direct decomposition of some valuable black carbon and low endothermicity (compared to
hydrocarbons to generate hydrogen by using carbon catalysts MSR). The TDM process decomposes natural gas (NG) in
is a viable alternative to the conventional steam reforming a high-temperature solar chemical reactor [1,2]. When feeding
processes. Methane is a preferred source of hydrogen because the reactor with NG, the overall reaction is equivalent to
of its high ratio of hydrogen to carbon and its abundant CH4 /2H2 þ Csolid . This process results in two products: a H2-
supply. The traditional hydrogen production methods include rich gas fuel and high-value carbon black material (CB). The
methane steam reforming (MSR) and partial oxidation of TDM occurs at temperatures above 700  C and in the absence
methane (POM) [1], both of which are accompanied by of oxygen.
production of the greenhouse gas CO2, which needs to be Since the pioneering works of Muradov [3,4] and Suelves
reduced. However, non-catalytic methane decomposition et al. [5] identifying the carbon-based catalytic decomposition

* Corresponding author. Tel.: þ52 55 56228103; fax: þ52 55 56228106.


E-mail address: fmendez@servidor.unam.mx (F. Méndez).
0360-3199/$ – see front matter ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2008.09.060
7420 international journal of hydrogen energy 33 (2008) 7419–7426

Nomenclature a dimensionless parameter


b dimensionless parameter
a velocity gradient, adsorption reaction
D nondimensional Damköhler number
cp specific heat
DTD characteristic Damköhler number to define the
d desorption reaction
decomposition temperature
Di molecular diffusion coefficient of the species i
3 nondimensional activation energy parameter,
f dimensionless stream function
RTN =E5d
h thickness of the flat plate
g nondimensional parameter
kj reaction rates
G surface molar concentration
m mass of the species i
4 nondimensional temperature, Eq. (38)
p partial pressure of the species i
F nondimensional temperature, Eq. (26)
Pr Prandtl number of the methane gas, nrcp =l
h nondimensional transverse coordinate for the gas
qr heat for the endothermic reaction on the surface
phase
of the flat plate
l thermal conductivity
Sj sticking probability or accommodation coefficient
m dynamic viscosity
t physical time
n kinematics coefficient of viscosity
T temperature
r density
TN temperature of the methane gas far from the flat
qCH3 surface coverage of the component CH3
plate
qCH2 surface coverage of the component CH2
Tw temperature of the flat plate
qCH surface coverage of the component CH
u, v longitudinal and transverse velocities
qH surface coverage of the component H
Wi molecular weight of the species i
qC surface coverage of the component C
W molecular weight of the mixture
qV surface coverage of empty sites
x, y Cartesian coordinates
s dimensionless time
Yi concentration of the species i
u surface reaction rate
yi nondimensional concentration of the species i
Zw rate of collisions Subscripts
w conditions at the flat plate
Greek letters
N denotes conditions far from the plate

of methane for hydrogen production, a significant number of kind of theoretical analysis of TDM of the CH4 gas on a catalytic
experimental and quasi-analytical studies of TDM using surface in a laminar stagnation flow has not been performed.
various carbon catalysts in packed-bed or fluidized-bed reac- Therefore, we develop in this work a theoretical model to
tors have been published. However, most of these authors predict the decomposition temperature of methane gas in
[6–9] have only reported experimental data for the thermal a stagnation-point flow configuration. In general, there are
decomposition of methane using different types of reactors. important differences between combustion and TDM, such as
Recently, these works have been extended to consider special the presence of reactions involving oxygenated species in
effects, such as the influence of the primary particle size of the combustion and the absence of these reactions in TDM.
carbon catalyst, that can improve the thermocatalytic Nevertheless, there are some common reactions and
decomposition of the methane in a fluidized-bed reactor [10], processes; thus, we adopt some steps of the kinetic scheme
the kinetics of methane decomposition in a fixed bed reactor investigated by Deutschmann et al. [16], retaining those
[11], and the catalyst deactivation [12]. The activity, activation reactions that permit only the decomposition of methane and
energy and reaction order of different particle sizes of carbon deposition of black carbon on the catalysts.
[13] were analyzed to improve the overall hydrogen produc-
tion process. Similar processes of hydrogen production using
carbonaceous catalysts were reported on recently [14,15]. At 2. Theoretical formulation
the same time, new theoretical methods and predictions in
the specialized literature are missing. Only a few theoretical 2.1. Gas-phase governing equations
and fundamental studies have been conducted. Therefore, in
the present work, we study theoretically the conversion of In Fig. 1, we show the typical model, the coordinate system
methane gas to obtain hydrogen gas without CO2, using and a sketch of the laminar stagnation-point flow configura-
a simple analytical model that takes into account the tion. Gaseous methane with a concentration denoted by Y CH4 N
following fact: in order to accelerate the TDM under laboratory flows with a velocity gradient a and temperature TN ,
conditions, it is necessary to elevate NG’s temperature to perpendicular to a catalytic flat plate of finite thickness h. We
a level suitable for decomposition (700  C). However, we assume that the material of the flat plate is composed of
should emphasize that the present theoretical predictions a carbon catalyst. The plate temperature is maintained at
must be completed and compared with experimental results. a uniform value of Tw , slightly greater than TN , in order to
From a fundamental point of view and to our knowledge, this induce the thermal decomposition of the methane. In
international journal of hydrogen energy 33 (2008) 7419–7426 7421

Stream of CH4 2.2. Heterogeneous reaction model

T Y U =ax V =-ay The heterogeneous reaction mechanism proposed by the

8
8
thermal decomposition of methane follows the basic ideas
suggested by the simplified model of Reinke et al. [17].
Furthermore, the reaction steps and specific values of the rate
parameters for the present model were taken directly from
the reaction mechanism developed by Deutschmann et al. [16]
and are shown in Table 1.
The kinetic model is represented by five heterogeneous
reactions. The reactions finished with a and d represent
y
adsorption and desorption, respectively. Here, Ct(s) denotes
a free site on the surface of the catalytic plate. All surface
x
reactions are assumed to be of the Langmuir–Hinshelwood
Externally heated catalytic plate type. The adsorption kinetics are given by a sticking proba-
bility, Sj, or accommodation coefficient, which represents the
portion of the collisions with the surface that successfully
Fig. 1 – Schematic of the physical model, showing the leads to adsorption. The rate of collisions, Zw, can be
stagnation-point flow configuration. computed using the classical kinetic theory, with
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Zw ¼ p= 2pmkT, where p and m are the partial pressure and
the mass of the species involved and k is the Boltzmann
constant (k ¼ 1.38  1023 J/K). However, the kinetic desorption
addition, we assume that this process only occurs on the
is well represented by an Arrhenius law with high activation
surface of the catalytic material. Therefore, homogeneous
energy for the adsorbed species, and the corresponding
reactions for the gas phase are not taken into account, and the
concentrations can be represented by the surface coverage qi
stagnation-point boundary layer governing equations for the
defined by the ratio of the number of sites occupied by surface
gas phase (assumed to be frozen) are the following:
species i to the total number of available sites. The fractional
vðruÞ vðrvÞ coverage of each species is determined by writing differential
þ ¼ 0; (1)
vx vy balances on all surface species. For the stationary case, the
differential equations then become a set of coupled algebraic
 
vu vu v vu equations. For the adsorbed species, the quasi-steady gov-
ru þ rv ¼ m þ rN mN a; (2)
vx vy vy vy erning equations are then given by

    dqCH3
vT vT v vT ¼ k1a q2V  k2 qCH3 qV ¼ 0; (7)
rcp u þ v ¼ l ; (3) dt
vx vy vy vy
and dqCH2
  ¼ k2 qCH3 qV  k3 qCH2 qV ¼ 0; (8)
vY i vY i v vY i dt
ru þ rv ¼ rDi ; (4)
vx vy vy vy
dqCH
for i ¼ CH4 and the reaction products. Here, u and v are the ¼ k3 qCH2 qV  k4 qCH qV ¼ 0; (9)
dt
longitudinal and transverse components of the velocity field,
respectively, while x and y represent the longitudinal and dqH
¼ 2k1a q2V þ k5a q2V  k5d q2H ¼ 0; (10)
transverse axes to the flat plate. In addition, T and Y i are the dt
temperature and the concentration of the species i. The dqC
¼ k1a q2V ; (11)
density, viscosity, specific heat at constant pressure, and dt
and
thermal conductivity of the gas phase are given by r, m, cp and
l, respectively. Di is the molecular diffusion coefficient of the qV þ qCH3 þ qCH2 þ qCH þ qC þ qH ¼ 1: (12)
species i. The associated boundary conditions are the
following: Here, qV denotes the surface coverage of empty or vacant sites.
In addition, we assume by Eq. (11) that the carbon deposits
vYCH4 uWCH4 vT uqr
u¼v¼  ¼  ¼ 0 for y ¼ 0; (5)
vy rDCH4 vy l

u  ax ¼ T  TN ¼ Y i  Y iN ¼ 0 for y/N: (6) Table 1 – Heterogeneous reaction model


No. Reaction S A (mol cms) E (kJ/mol)
Here, Wi is the molecular weight of the species i; qr is the heat
a
necessary for the endothermic surface reaction per mole of 1 CH4 þ 2CtðsÞ/CH3 þ H 0.01 *** 201
methane, considering the overall reaction CH4 /CðsolidÞ þ 2H2 ; 2 CH3 þ 1CtðsÞ/CH2 þ H *** 3.7E21 20
u is the surface reaction rate given in units of moles of 3 CH2 þ 1CtðsÞ/CH þ H *** 3.7E21 20
4 CHðsÞ þ 1CtðsÞ/CðsÞ þ HðsÞ *** 3.7E21 20
methane consumed per unit time and unit surface area of the
5a,d H2 þ 2CtðsÞ%H þ H 0.046 3.7E21 77.8
catalytic plate.
7422 international journal of hydrogen energy 33 (2008) 7419–7426

onto the carbon catalyst. All reaction rates in Eqs. (7)–(11) are 100
in s1 units. In this sense, the reaction rates given in Table 1 10-1
can be transformed to the appropriate units by taking
10-2
SCH4 pY CH4 w W SH2 pY H2 w W 10-3
k1a ¼ 3=2
pffiffiffiffiffiffiffiffiffiffiffiffi; k5a ¼ pffiffiffiffiffiffiffiffiffiffiffiffi (13) θC
GWCH 4
2pRT GWH3=22
2pRT 10-4
10-5 θH
10-6 θV
kr ¼ GAr expð  Er =RTÞ for r ¼ 2; 3; 4; 5d; (14) τ =1e-15
θ 10-7 θCH3
where G is the surface molar concentration in mol/cm2 and 10-8 θCH2
corresponds to the surface site density (w1015 sites/cm2) 10-9 θCH
divided by the Avogadro number, Av ¼ 6.02283  1023 10-10
molecules/cm2, and R is the universal gas constant. Y i are the 10-11
concentrations close to the catalytic surface and are to be 10-12
obtained after solving the coupled gas equations with the 10-13
governing equations for the surface coverage of the adsorbed 10-14
species. 400 600 800 1000 1200 1400 1600 1800
From Eqs. (7) to (12), we find that the surface coverage of T
product species is related to qV , as well as to the temperature,
Fig. 2 – Surface coverage for the species involved in the
through the relationships
TDM.
k1a
qCH3 ¼ qV ; (15)
k2

k1a
qCH2 ¼ qV ; (16) coverage for each product species as a function of the
k3
temperature for a given time, Eqs. (22)–(25).

k1a
qCH ¼ qV ; (17) 2.3. Gas-phase nondimensional governing equations
k4

qH ¼ bqV ; (18) For the gas phase, a stream function jðx; yÞ is introduced to
satisfy the mass-conservation equation (Eq. (1)): ru ¼ vj=vy
and rv ¼ vj=vx. We also define the following nondimen-
dqC
¼ k1a q2V ; (19) sional variables:
ds
rffiffiffiffiffiffiffiffiffiffiffiffiffi Z y
where s represents the nondimensional time given by j a
f ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi; h ¼ rðx; y0 Þdy0 ;
x rN mN a r N mN 0
t
s¼ ; (20)
ða þ bÞ2 =k1a cp WCH4 ðT  TN Þ
F¼ ; YCH4 ¼ Y CH4 : (26)
with a ¼ 1 þ k1a ð1=k2 þ 1=k3 þ 1=k4 Þw1, b ¼ ½ð2k1a þ k5a Þ=k5d 1=2 . qr
Recognizing that k1a =k2  1, k1a =k3  1 and k1a =k4  1, the The resulting nondimensional governing equations now
conservation relationship (12) can be rewritten, in a first take the form
approximation, as 3 2  2
d f d f rN df
þ f þ  ¼ 0; (27)
qC þ ða þ bÞqV ¼ 1; (21) dh3 dh2 r dh

and the solution of Eqs. (15)–(19) allows us to write qCH3 , qCH2 ,


2
qCH , qH and qC as functions of the nondimensional time s, the d Y dY
þ ScCH4 f ¼ 0; (28)
parameters a, b and the reaction rates in the next form: dh2 dh
s and
qC ¼ ; (22)
1þs 2
d F dF
þ Pr f ¼ 0; (29)
dh2 dh
b
qH ¼ ; (23)
ð1 þ bÞð1 þ sÞ where Pr is the Prandtl number of the gaseous
    mixture,Pr ¼ mcp =l and Sci is the Schmidt number of the
k1a 1 k1a 1
qCH ¼ ; qCH2 ¼ ; (24) species, Sci ¼ m=rDi . The nondimensional boundary condi-
k4 ð1 þ bÞð1 þ sÞ k3 ð1 þ bÞð1 þ sÞ
  tions are then given by
k1a 1 1 df dYi dF
qCH3 ¼ ; and qV ¼ : (25) f¼ ¼  G Lei ¼ þ G ¼ 0 at h ¼ 0; (30)
k2 ð1 þ bÞð1 þ sÞ ð1 þ bÞð1 þ sÞ dh dh dh
Substituting Eq. (21) for qV in Eq. (19), we obtain
df
dqC =ds ¼ k1a ð1  qC Þ2 , which can be integrated to obtain Eq.  1 ¼ F ¼ Yi  YiN ¼ 0 for h/N; (31)
dh
(22). The other Eqs. (23) and (24), are algebraically derived. In
Fig. 2, we show the corresponding values of the surface where
international journal of hydrogen energy 33 (2008) 7419–7426 7423

Sci WCH4 Pr u very small compared with the others. The chemical reaction is
Lei ¼ ; G ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (32)
Pr rN mN a then governed by the adsorption of methane as well the
adsorption–desorption kinetics of hydrogen. From Eq. (34), we
and the first parameter represents the Lewis number. The obtain an equation for qV :
reaction rate, u, which can be obtained from Eqs. (7) to (9),  
represents the consumption of the methane gas at the surface of the qV ¼ K5  K25 þ O K35 ; for K5 ¼ 1=b2 /0; (35)
flat plate. By definition, reaction 1a of Table 1 for the methane gas is and
dissociative, and the frequency for the consumption of this species is  
qH ¼ 1  K5 þ K25 þ O K35 : (36)
given by k1a q2V . Therefore, the product of k1a q2V G represents the
number of moles of methane gas which are consumed at the surface Threfore, the reaction rate can be approximated by
of the catalytic plate. Substituting Eq. (25) for qV in the above rela-
k1a GK5
tionship, we obtain u¼ ; (37)
ð1 þ sÞ2
k1a G
u ¼ k1a q2V G ¼ $ (33) where we use the leading-order term in qV . Due to the fact that
ð1 þ bÞ2 ð1 þ sÞ2
k5d has a large activation energy (desorption of HðsÞ), the
The above equation is readily derived. surface reaction rate is strongly dependent on temperature.
The surface chemical reaction produces vacant sites that are
2.4. Asymptotic solution rapidly occupied by fresh adsorbed species. Additional free
sites are obtained when the newly generated adsorbed
From Eq. (10), we obtain the relationship between qV and qH product of the surface reaction is rapidly desorbed. This
given by process finally leads to a thermal runaway, which character-
qV ¼ K5 qH ; (34) izes the catalytic thermal decomposition process.
In Eq. (37), it is shown that the reaction rate depends on
where K5 is a function mainly of the temperature and given by the adsorption kinetics of methane and on the desorption
K5 ¼ 1=b. Fig. 3 shows the values of the heterogeneous reaction kinetics of hydrogen, which depends strongly on tempera-
rates kj as functions of the inverse of the temperature, 1000/T. ture. To generate the thermal decomposition conditions, it is
The slowest rate of all corresponds to the desorption reaction enough to consider a temperature increase due to the
of the adsorbed atomic hydrogen, k5d. It is important to note energy absorbed in the endothermic reaction of the order
that, in Fig. 3, the cases of k2, k3 and k4 are indistinguishable. Tw  TN wRTN =E5d . Therefore; it is convenient to define
However, the rate plays an essential role in determining the a new variable of order unity for the nondimensional
thermal decomposition condition. Fig. 2 shows the evolution temperature as
of the surface coverage of the surface species as a function of
E5d qr
temperature (Eqs. (20)–(25)), assuming that the reactant 4¼ Fw1: (38)
cp WCH4 RT2N
consumption is negligible, that is, Y iw ¼ Y iN . The surface
coverage of HðsÞ is almost unity for temperature T  1000 K. With this new variable, the nondimensional governing equa-
The surface coverages of H and V are the most important, tions take the form
neglecting all other contributions of Eq. (12), that is, qH þ qV x1. 3 2  2
d f d f df
In addition, in Fig. 2, the distributions of the transition species 3
þ f 2 þ 1 þ 34  ¼ 0; (39)
dh dh dh
CH, CH2 and CH3 follow a similar behavior. However, for
2
values of qH w1, we obtain from Eq. (34) that qV wK5 . In the d yi dyi
þ Sci f ¼ 0; (40)
same figure, it can be noted that the surface coverage of C is dh2 dh
2
d 4 d4
þ Pr f ¼ 0; (41)
1016 dh2 dh

1014 with the boundary conditions


1012 df dyi D expð4w Þ
f¼ ¼  ;
1010 dh dh ð1 þ sÞ2
d4 D expð4w Þ
108 ¼ þ ¼ 0 at h ¼ 0; (42)
kj (1/sec)

dh ð1 þ sÞ2
106 k1a
df
k2  1 ¼ 4 ¼ yi  yiN ¼ 0 for h/N;
104 dh
2 k3
10
k4 where
0
10
k5a Yi qr
10-2 yi ¼ (43)
k5d 3 Lei cp WCH4 TN
10-4
and
10-6
0.5 1.0 1.5 2.0 2.5 qr Pr 2
D¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffik1a GK5N ð1 þ sÞ ; (44)
1000/T 3cp TN rN mN a

Fig. 3 – Rate constants, ki (sL1), as a function of the wall where K5N is K5 computed with T ¼ TN. For large activation
temperature (1000/T ). energy of the desorption reaction for the methane,
7424 international journal of hydrogen energy 33 (2008) 7419–7426

3 ¼ RTN =E5d /0, the temperature variation close to the 1190


thermal decomposition condition is very small; thus, in the
first approximation, both adsorption reactions can be 1180
assumed to be temperature independent. The Damköhler
number for the surface reactions is expressed as D. With the
1170
limit 3/0, the solution to Eqs. (39)–(42) can be found else-
where [18]:

T∞ (K)
 1160
dyi    D expð4w Þ
¼ 0:57 Sci2=5 yiN  yiw ¼ ; (45)
dh h¼0 ð1 þ sÞ2
1150
 a = 16seg-1, Γ = 1.66x(10)-8 mol/cm2
d4 D expð4w Þ
¼ 0:57 Pr2=5 4w ¼  : (46)
dhh¼0 ð1 þ sÞ2 1140
The first equation takes into account reactant consumption.
From Eq. (46) we can obtain the critical conditions for the 1130
thermal catalytic decomposition of methane as 3 4 5 6 7 8 9
2=5 γ
0:57 Pr
DTD ¼ ð1 þ sÞ2 : (47)
expð1Þ Fig. 5 – Decomposition temperature as a function of the
From Eq. (45), the reactant concentration at the wall, at conversion parameter g.
thermal decomposition conditions, is given by
3=5
3 Lei cp WCH4 TN
Yiw ¼ YiN  : (48) we choose an average value for it, and it is given by Pr ¼ 0.72,
qr
taken directly from Ref. [19]. For the density rN , we assume an
In the above relationship, it is very important to note that, in
ideal-gas behavior for the methane gas. For the viscosity mN
the first approximation, the methane consumption can be
and specific heat cp of the methane gas, we use the following
neglected to obtain the critical endothermic conditions for the
correlations:mN ¼ 1:34  105 ðTN =293Þ0:87 and cp ¼ 19:89þ
temperature decomposition. Therefore, the mass diffusion in
5:024  102 TN þ 1:269  105 T2N  11:01  109 T3N , taken from
the boundary layer is important after reaching decomposition
Refs. [20] and [21], respectively. The condition for the TDM
conditions.
given by Eq. (47), along with the definition of the Damköhler
number, Eq. (44), represents the parametric dependence
on the TDM. The Damköhler number for this condition is
3. Results given by

E5d
All results plotted in Figs. (2)–(4) and the additional Figs. (5)–(7) qr E5d Pr G2 A5d exp  RT 0:57 Pr2=5
ð1 þ sÞ2 :
N
DTD ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼
given in this section were obtained with the following values cp TN RTN rN mN a ð2 þ gÞ expð1Þ
for the parameters: qr ¼ 89,750 J/mol, E5d ¼ 201 kJ/mol. The (49)
above data were taken directly from Muradov et al. [10].
where the dimensionless methane-conversion parameter is
Although the Prandtl number is dependent on temperature,
defined as

1200
10

1150

1100
Γ = 1.66Ε−7, γ = 5
Γ = 1.66Ε−8, γ = 5 1
1050
T∞ (K)

Γ = 1.66Ε−7, γ = 8
mH2 /mC

Γ = 1.66Ε−8, γ = 8
1000

950

0.1
900

850
0 5 10 15 20 25 30
a (1/s) 0.1 1 10

Fig. 4 – Decomposition temperature as a function of the


gradient of velocity for the stagnation-point flow for two Fig. 6 – Ratio of the hydrogen to the carbon produced in the
different values of the parameter G and of the parameter g. process of TDM.
international journal of hydrogen energy 33 (2008) 7419–7426 7425

3=2
SH2 YH2w WCH4 temperature profiles for the gaseous phase, taking into
g¼ 3=2
; (50) account that, in the first approximation, the critical condition
SCH4 YCH4w WH2
to find the decomposition temperature based on the methane
and represents a measure of the conversion process for the consumption is negligible; therefore, the methane concen-
methane gas. Taking into account that the parameters SH2 , tration remains uniform.
WH2 , SCH4 and WCH4 are fixed, the ratio given by YH2w =YCH4w
determines the conversion rate. Therefore, the temperature of
decomposition for the methane gas can be obtained from Eq. 4. Conclusions
(49) for different values of the parameters. In order to obtain
this temperature, we note that it is highly dependent on the We have developed a theoretical model to predict the critical
activation energy of reaction 5d (Table 1) and the ratio of temperature of decomposition of methane into H2 and valu-
the adsorption rates of the components H2 and CH4 given by able carbon. The above formulation reduces the problem to
the dimensionless parameter g. Fig. 4 shows the decomposi- the solution of a set of non-linear equations. Since the solu-
tion temperature, TN , as a function of the velocity gradient of tion procedure does not require integration of any differential
the stagnation-point flow, a, for two different values of the equations, the set of non-linear equations that emerges can be
parameter G and two different values of the parameter g. It is solved with a standard numerical routine, thereby reducing
clear from this figure that the analysis predicts correctly the considerably the required computation time. The formulation
trends, both quantitatively and qualitatively, according to is also advantageous for investigation of surface kinetics. As
the technical literature [1,2,6–8]. Similarly, in Fig. 5 we present a concluding remark, it should be emphasized that our
the decomposition temperature as a function of the conver- formulations are applicable only when the stream gas
sion parameter g, with a ¼ 16 s1 and G ¼1.66  108 mol/m2. temperature on the catalytic surface is sufficiently high and
From this figure, we can see that, for increasing values of g, near to the TDM.
the decomposition temperature is also increased, a state that
qualitatively corresponds to the typical experimental data
reported previously. For instance, Lee et al. [2] obtained
experimentally that, during the decomposition process, the
Acknowledgments
methane conversion always increases for larger values of the
O. Bautista acknowledges DGAPA of UNAM for supporting this
decomposition temperature. Therefore, our theoretical
work.
predictions are in accordance with the above experimental
results at least qualitatively, recognizing that both approaches
use different physical models. Finally, Fig. 6 shows the ratio of
mH2 =mC as a function of the nondimensional time, where mH2
and mC are the mass of hydrogen and carbon produced in the Appendix A
process, showing qualitatively that the hydrogen produced Supplemental material
decreases as the time increases, causing a decreasing effi-
ciency for the catalytic plate, because the carbon produced in Supplementary information for this manuscript can be
the reaction is deposited on the surface of the catalytic plate. downloaded at doi: 10.1016/j.ijhydene.2008.09.060.
Finally, in Fig. 7, we show the corresponding velocity and

references
1.0

[1] Steinberg M. Fossil fuel decarbonization technology for


0.8 mitigating global warming. Int J Hydrogen Energy 1999;24:
771–7.
[2] Lee K, Han GY, Yoon KJ. Thermocatalytic hydrogen
production from the methane in a fluidized bed with
0.6
activated carbon catalyst. Catal Today 2004;93–95:81–6.
ϕ
[3] Muradov N. CO2-free production of hydrogen by catalytic
pyrolysis of hydrocarbons. Energy Fuels 1998;12:41–8.
0.4 df/dη [4] Muradov N. Catalysis of methane decomposition over
elemental carbon. Catal Commun 2001;2:89–94.
[5] Moliner R, Suelves I, Lazaro M, Moreno O. Thermocatalytic
decomposition of methane over activated carbons: influence
0.2
of textural properties and surface chemistry. Int J Hydrogen
Energy 2005;30:293–300.
[6] Abanades S, Flamant G. Production of hydrogen by thermal
0.0 methane splitting in a nozzle-type laboratory-scale solar
0 1 2 3 4 reactor. Int J Hydrogen Energy 2005;30:843–53.
η [7] Hirsch D, Steinfeld A. Radiative transfer in a solar chemical
reactor for the co-production of hydrogen and carbon by
Fig. 7 – Velocity and temperature profiles in the gaseous thermal decomposition of methane. Chem Eng Sci 2004;59:
region. 5771–8.
7426 international journal of hydrogen energy 33 (2008) 7419–7426

[8] Dunker A, Ortmann JP. Kinetic modeling of hydrogen [14] Suelves I, Lázaro MJ, Moliner R, Pinilla JL, Cubero H.
production by thermal decomposition of methane. Int J Hydrogen production by methane decarbonization:
Hydrogen Energy 2006;31:1989–98. carbonaceous catalysts. Int J Hydrogen Energy 2007;
[9] Abanades S, Flamant G. Solar hydrogen production from the 32:3320–6.
thermal splitting of methane in a high temperature solar [15] Pinilla JL, Suelves I, Lázaro MJ, Moliner R. Kinetic study of
chemical reactor. Sol Energy 2006;80:1321–32. the thermal decomposition of methane using carbonaceous
[10] Muradov N, Chen Z, Smith F. Fossil hydrogen with reduced catalysts. Chem Eng J 2008;138:301–6.
CO2 emission: modeling thermocatalytic decomposition of [16] Deutschmann O, Schmidt R, Behrendt F, Warnatz J. Twenty-
methane in a fluidized bed of carbon particles. Int J Hydrogen sixth symposium (international) on combustion, The
Energy 2005;30:1149–58. Combustion Institute; 1996. p. 1747.
[11] Bai Z, Chen H, Li W, Li B. Hydrogen production by methane [17] Reinke M, Mantzaras J, Shaeren R. High-pressure catalytic
decomposition over coal char. Int J Hydrogen Energy 2006;31: combustion of methane over platinum: in situ experiments
899–905. and detailed numerical predictions. Combust Flame 2004;
[12] Pinilla JL, Suelves I, Utrilla R, Gàlvez ME, Làzaro MJ, Moliner R. 136:217–40.
Hydrogen production by thermo-catalytic decomposition of [18] Treviño C. Gas-phase ignition of premixed fuel by catalytic
methane: regeneration of active carbons using CO2. J Power bodies in stagnation flow. Combust Sci Technol 1983;30:213.
Sources 2007;169:103–9. [19] Kays WM, Crawford ME, Weigand B. MP for convective heat
[13] Ryu BH, Lee SY, Lee DH, Han GY, Lee T-J, Yoo KJ. Catalytic and mass transfer. Bernhard Weigand: McGraw-Hill; 2004.
characteristic of various rubber-reinforced carbon blacks in [20] White FW. Fluid mechanics. McGraw-Hill; 2003.
decomposition of methane for hydrogen production. Catal [21] Kyle BG. Chemical and process thermodynamics. Englewood
Today 2007;123:303–9. Cliffs: Prentice-Hall; 1984.

You might also like