HYDROGEN PRODUCTION FROM METHANE REFORMING:

THERMODYNAMIC ASSESSMENT

ADILSON J. ASSIS (1)

CARLA E. HORI
CÍCERO N. ÁVILA-NETO
TATIANA V. FRANCO
(1)
School of Chemical Engineering – Federal University of Uberlândia – Av.João Naves de Ávila,
2121 – Santa Mônica – CEP: 38408-100 – Uberlândia, MG, Brazil

RESUMO
As composições no equilíbrio químico das reações de reforma a vapor, seca e oxidativa do
metano são calculadas usando dois métodos distintos: (1) avaliação das constantes de equilíbrio;
(2) multiplicadores de Lagrange. Ambos os métodos resultam em sistemas de equações
algébricas não lineares, resolvidos numericamente no “software” livre Scilab © INRIA-ENPC,
através da função “fsolve”. A contribuição deste trabalho é realizar uma análise termodinâmica
comparativa para estas reações, determinando a influência das principais variáveis operacionais
sobre o equilíbrio químico. Para a validação dos resultados, empregam-se dois procedimentos: (1)
comparação com dados experimentais e/ou de simulação, específicos de equilíbrio químico; (2)
cálculo da taxa de reação intrínseca, que nas condições de equilíbrio deve ser nula. São
estudados os efeitos da temperatura, pressão, razão inicial H2O/CH4 (reforma a vapor), razão
inicial CH4:CO2:N2 (reforma seca) e razão inicial O2/CH4 (reforma oxidativa) sobre os produtos das
reações.

ABSTRACT
The main contributions of this study are to conduct a comparative thermodynamic analysis of
methane reforming reactions and to asses the influence of key operational variables on chemical
equilibrium using an in-house code, developed in the open-source software Scilab © INRIA-ENPC
(www.scilab.org). Equilibrium compositions are calculated by two distinct methods: (1) evaluation of
equilibrium constants; (2) Lagrange multipliers. Both methods result in systems of non-linear
algebraic equations, solved numerically using the Scilab function “fsolve”. Comparison between
experimental and simulated equilibrium data, published in the literature, was used to validate the
simulated results. Effects of temperature, pressure, initial H2O/CH4 ratio (steam reforming), initial
CH4:CO2:N2 ratio (dry reforming) and initial O2/CH4 ratio (partial oxidation) on the reaction products
were evaluated.

KEYWORDS
Methane reforming, hydrogen, chemical equilibrium, scilab software.

1. INTRODUCTION
Hydrogen can be produced using several routes. At the present moment, the main route is the
catalytic reforming of methane, which includes steam-methane reforming (SMR), partial oxidation
(POX), autothermal reforming (ATR) and dry-methane reforming (DMR).

Steam-Methane Reforming produces a H2/CO ratio equals to three, which is high, when compared
to other reforming processes [1]. Xu and Froment (1989) demonstrated that besides steam-

1
Correspondência deverá ser enviada a Adilson José de Assis:
Tel.: + 55 (34) 3239-4292, Ramal: 233; e-mail: ajassis@ufu.br

1
methane reforming (Equation 1), there is another important reaction which takes place in these
systems: the water-gas shift reaction (Equation 2). The sum of both reactions results in Equation 3
[2].

°
CH 4 + H 2 O → CO + 3H 2 ∆H 298 = 206.0kJ / mol (1)

°
CO + H 2O → CO2 + H 2 ∆H 298 = −41.0kJ / mol (2)

°
CH 4 + 2 H 2 O → CO2 + 4 H 2 ∆H 298 = 165.0kJ / mol (3)

Nevertheless, steam reforming has the disadvantage of intensive energy requirements due to the
overall endothermic nature. In order to reduce energy costs, partial or stoichiometric oxidation
(Equation 4) of methane has been investigated as an alternative [3].

°
CH 4 + 1 2 O2 → CO + 2 H 2 ∆H 298 = −35.7 kJ / mol (4)

The autothermal reforming process combines the endothermic reforming with the exothermic
oxidation. In the reaction process, it is assumed that the heat required for the endothermic
reforming can be supplied by the heat released from the oxidation reaction [4].

Finally, dry-methane reforming (Equation 5) produces a synthesis-gas with a H2/CO ratio near one,
which can be used for adjusting H2/CO ratio in steam reforming, suitable for Fischer–Tropsch
reactions and methanol production [5].

°
CH 4 + CO2 ↔ 2CO + 2 H 2 ∆H 298 = 247.4kJ / mol (5)

Although the thermodynamics equilibrium is widely discussed in the literature, it has a scarce
systematization and the available programs are usually proprietary, except the CANTERA’s
software. Therefore, the main objectives of this study are to conduct a comparative thermodynamic
analysis for methane reforming reactions and to asses the influence of key operational variables on
chemical equilibrium.

2. OBJECTIVES
The objective of this work is the calculation of the equilibrium compositions of the methane
reforming reactions by two distinct methods, Lagrange multipliers and the evaluation of equilibrium
constants, using the open-source software Scilab.

3. METHODOLOGY
Equilibrium compositions are calculated by two distinct methods: (1) evaluation of equilibrium
constants; (2) Lagrange multipliers. Both methods result in systems of non-linear algebraic
equations, solved numerically using the open-source software Scilab (www.scilab.org).

3.1 EVALUATION OF EQUILIBRIUM CONSTANTS METHOD (EEC)

An independent multi-reaction system was organized and each reaction is associated with a
reaction coordinate (ξ) and with a distinct equilibrium constant (Kj). The equilibrium constant for
each reaction is described by Smith et al. (2000) [6]. In this work, the fugacity coefficient for each
species in the mixture, φˆi , is calculated using the virial’s equation considering that there are
interactions between the gases (Equations 6). The cross virial’s coefficient, used in the Equation 7,
follow the correlation proposed by Tsonopoulos and Reidman (1990) [7].

2
 P  1 
ln φˆi =  Bii + M ijk  (6)
RT  2 

 
M ijk = ∑∑ yk y j  2δ ki − δ kj  (7)
k j  
This method is interesting to evaluate when a determined operational condition (T, P, composition)
favours a certain reaction, through the analysis of the value of the equilibrium constant of this
reaction compared to another one present in the system [6].

3.2 LAGRANGE MULTIPLIERS METHOD (LM)

In agreement with Smith et al. 2000, the system composition in the equilibrium can be found
solving N equilibrium equations, w mass balance equations and a restriction equation, represented
by Equations 8-10, respectively:

 P
∆G fi + RT ln  yiφˆi  + ∑ λk aik = 0 (8)
 P0  k

( i = 1, 2,..., N )

∑n a
i
i ik = Ak ( k = 1, 2,..., w) (9)

∑y
i
i =1 ( i = 1, 2,..., N ) (10)

Where ∆G fi is the standard Gibbs function of the formation of species i, R is the universal gas
constant, yi is the molar fraction of species i in the mixture, φˆ is the fugacity coefficient of species i in
i

the mixture, λk is the Lagrange’s multipliers oh each element k, aik is the number of atoms of each k
element of molecule i and Ak is the total number of atomic mass of the k element. The advantage of
this method is to be independent of the reaction system because the choice of a group of molecules
is equivalent to choice a system of independent reactions between the species [6].

3.3 SIMULATION METHOD

In both cases, the variation of standard enthalpy and Gibbs free energy of the reaction system
(∆H°j e ∆G°j, respectively) are functions of the temperature and must be corrected with the heat
capacities of the gas, Cp. The input parameters of the simulator are: molar fraction of the reactants,
temperature and pressure in the reactor input and operational temperature and pressure. The
resolution of the equations is made through implemented computational codes in the free platform
of calculation Scilab © INRIA-ENPC. The numeric calculation uses a function “fsolve”, which one
utilizes a modification of the hybrid Powell method with a tolerance of 10-10.

4. RESULTS

4.1 STEAM-METHANE REFORMING

Accordingly to Seo et al. (2002), in the equilibrium, the species of the steam-methane reforming are
CH4, H2O, CO, CO2 and H2, C(solid) and the radicals H, O, OH, HO2, HCO, CH e CH2. However,
the simulations indicate that the radicals concentrations are neglected when compared to the
concentrations of the other products. Therefore, they will be disrespected in this work.

3
Applying the Denbigh method [8] with the products considered by Seo et al. (2002), one concludes
that only Equations 1 and 2 are necessary for the calculation of the thermodynamic balance
through the analysis of the balance constants.

The H2 yield and the H2/CO selectivity are defined by Fogler (1999) [9] as:

N HOut2
YH 2 = (11)
In
N CH 4
− N CH
Out
4

N HOut2
S H 2 / CO = Out
(12)
N CO

As the reactor temperature is raised from 600 to 800 °C, CH4 conversion increases from 44 to 90
%. In agreement with the Figure 1 (b), for temperatures around 800 °C, the raise of initial H2O/CH4
ratio from one to two increases H2 yield in 4.67%. For temperatures around 430 °C, this yield is
practically the same for all conditions.

M H2
o (a) (b) P (bar) H2O/CH4
H2 1 1 1
l 2 10 1
e 4 3 1 2
y 4 10 2
f i
e 2
r
a H2O l
CO d 3
c
(%)
t CH4
i CO2 1
o
n
Temperature (°C) Temperature (°C)
Figure 1. (a) Equilibrium composition of the SMR products (P = 1 bar and H2O/CH4 = 1); (b) H2
yield for SMR, pressure and H2O/CH4 sensitivity.

4.2 DRY-METHANE REFORMING

Dry-methane reforming analysis is carried out considering the same species as in the case of
steam-methane reforming, with the addition of N2 as an inert. As the reactor temperature is raised
from 600 to 800 °C, CH4 conversion increases from 43 to 91%. Figure 2 (b) shows the H2 yield
simulation for the DMR with CH4:CO2:N2 sensitivity.

M
(a) CO (b)
o
H2 H2
l
e
y 3
f i
2
r e
a l
c N2 d 1
(%)
CH4:CO2:N2
t CO2
CH4 1 1:1:0,5
i H2O 2 1:1:4
o 3 1:1:8
n
Temperature (°C) Temperature (°C)
Figure 2. (a) Equilibrium composition of the DMR products (P = 1 atm and CH4:CO2:N2 = 2:2:1); (b)
H2 yield for DMR, CH4:CO2:N2 sensitivity.

4
4.3 PARTIAL OXIDATION OF METHANE
For partial oxidation, the following species are considered: CH4, H2O, CO, CO2, H2 and O2. Despite
of Zhu et al. (2001) [10] consider the formation of others hydrocarbons such as C2H6, C2H4, C2H2,
CH3OH, HCHO and HCOOH, they have a very small yield (<10-7%) and, for this reason, they are
not consider in this work.

M T = 1000 °C
o
(a) (b)
1 O2/CH4
CH4 H2
l 1 0.5
e H2 2 1.0
H2O y 3 1.5
f i
r e 2
a CO l
CO2
c d
(%)
t
i 3
o
n
Feed rate O2/CH4 Temperature (°C)
Figure 3. (a) Equilibrium composition of the POX products (P = 1 atm); (b) H2 yield for POX, O2/CH4
sensitivity.

Figure 3 (a) shows the simulated results of the POX equilibrium composition, and it’s also possible
to note that, at 1000°C, CH4 is completely consumed when O2/CH4= 0.5.

The H2 yield is strongly influenced by the temperature and by the O2/CH4 ratio in the feed. As it is
shown in Figure 3 (b), the highest values are achieved for higher values of temperature and smaller
initial O2/CH4 ratios.

The software ScanIt was used to obtain the data of Table 1-3 because all the authors showed the
results by graphics.

5. DISCUSSION
Table 1 shows the comparison between the simulated results and literature data for the steam-
methane reforming. Lutz et al. (2003) [11] used the software CHEMKIN with the program for
chemical equilibrium calculations Stanjan, with a ratio H2O/CH4 = 2 and P = 10 atm. The average
relative error between the results of this work (EEC method) and those one of Lutz et al. (2003) is
7.21% and the average relative error for the LM method is 7.68%. The same analysis is done for
the results reported by Seo et al. (2002), but they use the software Aspen PlusTM for the equilibrium
compositions with a ratio H2O/CH4 = 1 and P = 1 bar. This time, the average relative error between
the simulated results of this work (EEC method) and those of Seo et al. (2002) is 24.85%, while for
the LM method the average relative error is 24.89%. The discrepancy of the errors between the two
author’s data can be a function of the consideration of carbon solid in the reforming products by
Seo et al. (2002). It’s important to note that both softwares used by the authors, use the LM method
for the calculation of the equilibrium composition.

5
 Table 1. Comparison between simulates and data results for SMR equilibrium composition.
500 °C 600 °C 700 °C 800 °C 900 °C 1000 °C
Operational
Compounds Simulation Simulation Simulation Simulation Simulation Simulation
conditions Author Author Author Author Author Author
LM EEC LM EEC LM EEC LM EEC LM EEC LM EEC
CH4 0.267 0.260 0.260 0.210 0.202 0.203 0.133 0.126 0.126 0.058 0.054 0.050 0.015 0.013 0.015 0.003 0.003 0.004
H2O 0.537 0.524 0.524 0.434 0.420 0.421 0.322 0.309 0.314 0.230 0.222 0.222 0.185 0.181 0.184 0.175 0.173 0.176
CO 0.002 0.002 0.004 0.015 0.016 0.015 0.055 0.056 0.061 0.114 0.115 0.115 0.154 0.154 0.153 0.169 0.168 0.168
CO2 0.037 0.041 0.038 0.059 0.063 0.061 0.065 0.068 0.065 0.051 0.053 0.050 0.036 0.038 0.038 0.029 0.030 0.027

Lutz et al. (2003)

H2O/CH4 = 2
P = 10 atm
H2 0.157 0.173 0.174 0.282 0.299 0.300 0.425 0.441 0.434 0.547 0.556 0.563 0.610 0.614 0.61 0.624 0.626 0.625
CH4 0.320 0.317  0.194 0.192 0.137 0.079 0.078 0.055 0.027 0.026 0.018 0.009 0.009 0.004 0.004 0.004 0.000
H2O 0.249 0.245  0.132 0.128 0.151 0.054 0.053 0.064 0.019 0.019 0.018 0.007 0.007 0.004 0.003 0.003 0.000
CO 0.019 0.019  0.090 0.091 0.059 0.185 0.185 0.155 0.229 0.229 0.224 0.243 0.243 0.242 0.247 0.247 0.246
CO2 0.071 0.072  0.062 0.063 0.046 0.025 0.025 0.023 0.007 0.007 0.004 0.002 0.002 0.000 0.001 0.001 0.000

Seo et al. (2002)

H2O/CH4 = 1
P = 1 bar
H2 0.341 0.347  0.522 0.526 0.525 0.657 0.659 0.652 0.718 0.719 0.716 0.729 0.739 0.739 0.745 0.745 0.748

Table 2. Comparison between simulates and data results for DMR equilibrium conversion.
400 °C 500 °C 600 °C 700 °C 800 °C
Operational conditions Conv. (%) Simulation Simulation Simulation Simulation Simulation
Author Author Author Author Author
LM EEC LM EEC LM EEC LM EEC LM EEC
Akpan et al. (2007)
CH4:CO2:N2 = 2:2:1 CH4 3.96 3.95 3.38 16.02 16.02 15.8 43.01 43.10 45.1 74.43 74.59 77.8 91.24 91.35 93.2
P = 1 atm

Table 3. Comparison between simulates and data results for POX equilibrium conversion.
O2/CH4 = 0.5 O2/CH4 = 1.0 O2/CH4 = 1.5
Operational conditions Compounds Simulation Simulation Simulation
Author Author Author
LM LM LM
CH4 0.214 0.156 0.060 0.023 0.005 0.000
H2O 0.107 0.078 0.237 0.198 0.409 0.401
Zhu et al. (2001)
T = 600 °C CO 0.155 0.200 0.118 0.148 0.062 0.063
P = 1 atm
CO2 0.107 0.078 0.195 0.175 0.270 0.268
H2 0.417 0.488 0.390 0.456 0.254 0.268

6
Table 2 shows the validation of the simulated results by comparison with CH4 equilibrium
conversion data of Akpan et al. (2007) [12]. The average relative error between the results of this
work (EEC method) and those one of Akpan et al. (2007) is 6.32% and the average relative error
for the LM method is 6.40%.

Table 3 shows the validation of the results by comparison with equilibrium composition data of Zhu
et al. (2001). The average relative error between the results of this work (LM method) and those of
Zhu et al. (2001) is 28.7%.

6. CONCLUSION
Thermodynamics analysis was carried out for the steam-methane reforming, dry-methane
reforming and partial oxidation of the methane. Literature’s data were compared with the results of
Lagrange’s multipliers method and the Equilibrium constants evaluation method. The analyses
enabled to validate the simulated results of this work and to verify the best operational conditions
for each methane reforming reaction. The developed computer codes in Scilab are available upon
request.

7. ACKNOWLEDGEMENTS
This work has been supported by Brazilian funding agencies, CAPES and CNPq (Grant n.
475934/2006-7).

8. REFERENCES
[1] Seo, Y.-S., Shirley, S. T., Kolaczkowski, S. T. Evaluation of thermodynamically favourable
operating conditions for production of hydrogen in three different reforming technologies.
Journal of Power Sources; v. 108; p. 213-225; 2002.

[2] XU, J., FROMENT, G. F. Methane Steam Reforming, Methanation and Water-Gas Shift: I.
Intrinsic Kinetics. AlChE Journal; 1989.

[3] Corbo, P., Migliardini, F. Hydrogen production by catalytic partial oxidation of methane
and propane on Ni and Pt catalysts. International Journal of Hydrogen Energy; v. 32; p. 55-66;
2007.

[4] Li, B, Maruyama, K., Nurunnabi, M., Kunimori, K., Tomishige, K. Temperature profiles of
alumina-supported noble metal catalysts in autothermal reforming of methane. Applied
Catalysis A; v. 275; p. 157-172; 2004.

[5] O’Connor, A. M., Schuurman, Y., Ross J. R. H., Mirodatos, C. Transient studies of carbon
dioxide reforming of methane over Pt/ZrO2 and Pt/Al2O3. Catalysis Today, v. 115, p. 191-198;
2006.

[6] Smith, J. M., Van Ness, H. C., Abbott, M. M. Introdução à termodinâmica da engenharia
química. LTC Rio de Janeiro; 5 ed; 2000.

[7] Tsonopoulos, C., Heidman, J. L. From the virial to the cubic equation of state. Fluid Phase
Equilibria; v. 57; p. 261-276; 1990.

[8] Denbigh, K. G. The principles of chemical equilibrium. Cambridge University Press; 4. ed;
1981.

[9] FOGLER, H. S. Elementos de Engenharia das Reações Químicas. 3.ed. Rio de Janeiro,
LTC, 2002.

7
[10] Zhu, J., Zhang, D., King, K. D. Reforming of CH4 by partial oxidation: thermodynamic and
kinetic analyses. Fuel; v. 80; p. 899-905; 2001.

[11] Lutz, A. E., Bradshaw, R. W., Keller, J. O., Witmer, D. E. Thermodynamic analysis of
hydrogen production by steam reforming. International Journal of Hydrogen Energy; v. 28;
p.159-167; 2003.

[12] Akpan, E., Sun, Y., Kumar, P., Ibrahim, H., Aboudheir, A., Idem, R. Kinetics, experimental
and reactor modelling studies of the carbon dioxide reforming of methane (CDRM) over a
new Ni/CeO2–ZrO2 catalyst in a packed bed tubular reactor. Chemical Engineering Science; v.
62; p. 4012-4024; 2007.