Two-section reactor model for autothermal

reforming of methane to synthesis gas

P. Maarten Biesheuvel and Gert Jan Kramer

Shell Global Solutions International B.V., Badhuisweg 3, 1031 CM, Amsterdam, The Netherlands.

Abstract
A one-dimensional and stationary reactor model is presented to describe the catalytic conversion of
a gaseous hydrocarbon fuel with air and steam to synthesis gas by autothermal reforming (ATR) and
catalytic partial oxidation (CPO). The model defines two subsequent sections in the reactor, namely an
upstream oxidation section, and a downstream reforming section. In the oxidation section all of the
oxygen is converted, with partial conversion of the fuel. An empirical fuel utilization ratio is used to
quantify which part of the fuel is converted in the oxidation section as function of the relative flows of
air and steam. In the oxidation section, the gas temperature rapidly increases toward the
toptemperature at the intersection with the reforming section. In this section the temperature
decreases while the fuel is further converted with water and CO2 as oxidant. For methane as fuel,
simulation results are presented and compared with experiments. For multicomponent fuels such as
natural gas and naphtha, it is described how the two-section model can be applied.

Introduction
Hydrogen is manufactured industrially from hydrocarbon fuels either by steam reforming or by
gasification (Rostrup-Nielsen, 1993). In steam reforming, the fuel (in most cases, natural gas) reacts
with steam toward synthesis gas over a Ni-based heterogeneous catalyst. Due to the endothermicity of
the steam reforming reaction, heat must be supplied to the reactor through the reactor tube walls.
Gasification is an autothermal process in which the fuel is mixed with pure oxygen and non-
catalytically converted to synthesis gas. An example of its industrial use is the Shell Gasification
Process (Elvers et al., 1989).
The recent surge in demand for on-spot, small-scale, cheap and simple hydrogen production
technologies based on hydrocarbon fuels is driven by the interest in fuel cells for electricity generation
(Docter and Lamm, 1999; Ahmed and Krumpelt, 2001; Moon et al., 2001). Light hydrocarbon fuels,
such as natural gas, are envisioned for stationary applications, e.g., at the scale of a single household
(1-5 kWe, kW electricity), while liquid fuels (gasoline, naphtha) have potential for automotive
applications (25-75 kWe; Service, 1999). To obtain a gas mixture that can be fed to a fuel cell, the
hydrocarbon fuel is first converted to synthesis gas in a reformer. When a proton exchange membrane
(PEM) fuel cell is used, the synthesis gas must be further processed, first in a water-gas-shift reactor,
where CO is converted to CO2 and additional hydrogen is obtained, after which remaining traces of
CO are removed (e.g., by selective oxidation) before the mixture is finally fed to the anode side of the
fuel cell.

1
 For these applications, down-scaled versions of the gasification and steam reforming processes are
impractical. First of all, downscaling is not very well possible (especially for gasification) and otherwise
the resulting processes are too bulky, expensive and difficult to operate. However, an air-based,
autothermal catalytic fixed bed technology in which the hydrocarbon fuel is mixed with air and steam
may result in a small reformer that is safe and easy to operate and automate. Two technologies are
available, derived from their industrial analogues. Autothermal reforming (ATR), based on mixing air in
with the steam/fuel mixture, is an autothermal adaptation of steam reforming. In the gasification
process, it is possible to replace the oxygen with air and use a catalyst instead of an underoxidized
burner; this is catalytic partial oxidation (CPO). Catalytic partial oxidation (CPO) has –within Shell–
originally been developed as a catalytic alternative for the industrial SGP process. Originally intended
for operation with pure oxygen, it has been modified for use with air and steam to suit the needs of fuel
processing for fuel cells.

Ttop

Oxidation Tout
section
Reforming section
T0 top

*
0 Reactor length, L 1

Figure 1. Gas phase temperature in CPO/ATR of methane. The measured catalyst temperature at the
upstream side is assumed equal to the gasphase temperature at the intersection of the oxidation and
reforming section (0, “top”).

While the two technologies have a distinctly different origin, we are of the opinion that the two
technologies have all but merged though CPO typically operates at low to zero steam-to-carbon ratio’s
(e.g., S/C<1) while ATR operates at a higher steam load (S/C>1). In ATR and CPO several similar
catalyst systems have been used and tested, such as based on supported precious metals. In both
processes the aim is to achieve the thermodynamic equilibrium composition which is determined by
feed conditions (composition, temperature), pressure and heat loss. These conditions are optimized to
obtain a maximum syngas yield in hydrogen and carbon monoxide (in the relevant yield figure, CO is
added to H2, because of its conversion in the subsequent water-gas-shift reactor), as well as to
minimize slip of hydrocarbons (especially methane levels can be significant at thermodynamic
equilibrium, dependent on exit temperature and amount of steam).

2
 Comparing CPO with ATR, in CPO temperatures in the catalyst bed tend to be higher and syngas
yield somewhat lower. However, premixing and preheating is more straightforward (certainly for
S/C=0) and reaction rates tend to be higher (higher space velocity possible). In any case, the choice
for the optimal air/fuel and steam/fuel ratio as well as inlet temperature and operating pressure must
be made at the level of the entire fuel processor system.
It must be stressed that both in CPO and ATR the practical objective is to bring the mixture to
thermodynamic equilibrium. While there is no a-priori reason why a catalytic process could not be
more selective towards syngas than what is dictated by thermodynamics, the fact that CO and
hydrogen are so much more reactive with oxygen than the feed hydrocarbon molecules makes this
goal elusive. This reactivity bias will in fact cause an ‘overshoot’ of the reaction: excess amounts of
CO2 and water are produced at the “top” of the reactor (see Figure 1), after which thermodynamic
equilibrium is approached via reforming reactions. Thus, if thermodynamic equilibrium is not reached,
there is both an excess amount of CO2 and water as well as an unnecessary slip of methane and
higher hydrocarbons, both leading to a sub-optimal syngas yield.

100

H2-selectivity (%)

95
Conversion, Selectivity

CO-selectivity (%)
90 Conversion
Methane (%)

85

80
0.1 1 10 100 1000
Pressure P (kPa)

Figure 2. Thermodynamic equilibrium in adiabatic operation of air-based CPO of methane (no steam,
o
O2/C=0.5, air: 21 mol% O2, 79% N2), Tinlet=300 C. Methane conversion and selectivities to CO and H2
as function of pressure.

A second aspect of thermodynamic equilibrium is that for operation at low air and steam flows
(relative to the fuel), such as done almost by definition in studies of “pure” CPO of methane (studies in
which the O2/CH4 ratio is set to 0.5 and no steam is added, based on CH4 + 0.5 O2
CO + 2 H2),
CH4-conversion and CO- and H2-selectivities are clearly below 100% unless operation is at a very high
(inlet) temperature or a very low pressure (Dissanayake et al., 1991; Wang and Ruckenstein, 1999).
Results of thermodynamic calculations are shown in Figure 2 for adiabatic CPO of methane without
steam and with air (O2/CH4=0.5). In this case, only by decreasing the pressure to very low values, is it
possible to obtain full methane conversion and 100% selectivity toward H2 and CO. Thus, incomplete
methane conversion and non-selective behavior at practical conditions (T, P) is not necessarily due to

3
a catalyst with poor selectivity/activity, but in many cases a direct consequence of the thermodynamic
equilibrium (Prettre, 1946; Chang and Heinemann, 1993; Bodke et al., 1998). In practice this is well-
known and to optimize syngas yield and keep methane slip below a certain threshold, more air (and/or
water) are added than the “pure CPO” reaction suggests (e.g., Kramer et al., 2001).
Thermodynamic equilibrium models required for the above calculations (both for CPO and ATR) are
readily available or are rather easily set up, e.g., based on (part of) the equations used in this paper
(Docter and Lamm, 1999). However, for reactor design and catalyst development, non-equilibrium
reactor modeling is required.
This paper will focus on the development of a reactor model that describes the approach to
thermodynamic equilibrium, thus reflecting the situation in practical reactors. Obviously, it is required
that the model is thermodynamically consistent at each reactor coordinate (at each location). By this
we mean that the equations that are used at each location to describe the conversion of the off-
equilibrium gas mixture only contain local parameters (local temperature, gas composition) but
correctly predict the thermodynamic equilibrium composition and temperature when given enough
residence time or space. To be practical, the model must contain a limited number of fit parameters
(reaction orders, activation energy, kinetic constants, etc.) Furthermore, the high upstream
temperature (top-temperature) that is observed in CPO/ATR (Prettre et al., 1946; Papp et al., 1996;
Heitnes Hofstad, 1997; Wang and Ruckenstein, 1999) must be an integral part of the model. Finally,
atomic (mass) balances and the enthalpy balance must be explicitly accounted for.
Several relevant modeling efforts for CPO without steam (and some also for CPO/ATR with steam)
can be found in literature (Hickman and Schmidt, 1993a,b; De Groote and Froment, 1995, 1996;
Deutschmann and Schmidt, 1998; Arena et al., 2000; De Smet et al., 2000, 2001), some based on
effective rate equations, others on detailed descriptions of transport to the catalyst surface and
reactions at the surface. Though several of these models contain some of the elements that are
discussed above, we could not find a model that we could confidently use for our objectives.
Based on the above objectives, we set up a reactor model for synthesis gas formation in ATR and
CPO processes. Our focus here is on methane, but it is explained how the model can be adapted to
multicomponent fuels such as natural gas and naphtha. A one-dimensional and stationary (time-
independent) model is set up that does not describe mass transfer to the catalyst surface or reaction
at the catalyst surface in any detail but is based on effective rate equations expressed in gas phase
properties (such as Eq. [10]) which makes it more straightforward to arrive at a thermodynamically
consistent reactor model. For the existing models in literature that contain detailed descriptions of
surface reactions in the form of reaction networks it is only possible to be thermodynamically
consistent when enthalpy and entropy of each of the possible surface species are considered, when
the enthalpy balance is expanded to include all of these surface species and when each adsorption,
surface reaction and desorption step contains a driving force-term (such as the term within brackets in
Eq. [9]) based on equilibrium of that specific step (the equilibrium constant in the driving force-term, Ki,
is temperature-dependent, at least for each adsorption and desorption step).
In the model we only consider convective transport of mass and heat in the axial coordinate, which is
possible because in CPO and ATR (both autothermal processes; in contrast to steam reforming) heat

4
transfer in a perpendicular direction is negligible (only due to heat loss). Dispersive transport of mass
and heat (in axial direction) are also neglected because of the high gas (space) velocities. We use a
stationary model because it needs to be practical and we aim at describing stationary laboratory scale
experiments and using the validated reactor model for design purposes. Instationary aspects are
important during start-up, shut-down and load-change, but these periods are short in CPO and ATR
that operate at a high space velocity.
The model divides the reactor into two subsequent sections. The first (upstream) section is the
(partial) oxidation section where all oxygen is converted with part of the fuel and the highest “top”
temperature is reached. This section is followed by the reforming section in which the remaining fuel is
further converted with water and CO2 as only possible oxidants to arrive at thermodynamic equilibrium,
see Figure 1 (Zhu et al., 2001).
It is well-known that oxygen conversion (in the oxidation section) is very fast. Based on a kinetic
analysis Zhu et al. (2001) estimate the oxidation section to be a few percent of the total reactor, while
Kramer et al. (2001) estimate the oxidation section to comprise some ten percent of the total reactor
volume, based on mass-transfer limited transport in the oxidation section. In the present model we do
not assign reactor space to the oxidation section and focus on the reforming section.
At the end of the oxidation section (top), the crude syngas contains all components (CO, CO2, H2
and H2O) as well as part of the fuel, but no oxygen. The fraction of the fuel that is converted in the
oxidation section is the “fuel utilization ratio” (UT). A high fuel utilization ratio (for a given amount of air
and steam) means that large amounts of CO and H2 are formed in the oxidation section, resulting in
low top-temperatures; a low value for UT results in high top-temperatures and large amounts of CO2
and H2O formed in the oxidation section. In the latter case, more fuel is unconverted in the oxidation
section and must be converted in the reforming section. For both reasons, a high value of UT is
favourable for CPO/ATR performance. Note that the fuel utilization ratio is equal to one over the
oxygen usage ratio in Kramer et al. (2001).
The fuel utilization ratio depends on the relative amounts of air and steam premixed with the fuel,
and obviously depends on the rates by which the oxygen and the fuel are transported to the surface,
adsorb and react. When the rate of transport and adsorption of the fuel (methane) is increased
(relative to that of oxygen), the fuel utilization ratio increases (Kramer et al., 2001).
The fuel utilization ratio is calculated from the catalyst temperature at the upstream side of the
catalyst bed using an optical pyrometer and applying the appropriate mass and enthalpy balances (see
Results and Discussion). When deriving the fuel utilization ratio from the measured temperature, it is
assumed that at the end of the oxidation section the temperatures of gas and catalyst are equal (the
“top-temperature”) and that they coincide with the measured upstream catalyst temperature (see
Figure 1). Our experimental observations indicate that this assumption is valid. Such a constant
catalyst temperature throughout the oxidation section can be explained when the oxidation reaction is
effectively heat- and mass-transfer limited, when (under CPO/ATR conditions) the Chilton-Colburn
analogy holds and when the Lewis number (the dimensionless ratio between mass and heat diffusivity)
is equal to unity. This case is argued further in Kramer et al. (2001). The assumption that the gas
phase has the same temperature as the catalyst at the intersection of oxidation and reforming section

5
(at the top, see Figure 1) is reasonable given the high heat and mass transfer rates, but can also be
underpinned by the fact that in the oxidation section the gas is colder than the catalyst (the exothermic
reactions take place on the catalyst surface while the gas is heated up by hot molecules coming off the
surface), while in the (endothermic) reforming section the catalyst should be (somewhat) cooler than
the gas. At the intersection of the exothermic and the endothermic regions the temperature of the
catalyst and the gas phase should be the same.
In the model it is assumed that the water-gas-shift (WGS) equilibrium is established at each location
in the reforming section (thus also at the top where the reforming section starts). That the WGS-
equilibrium is established is quite often reported in work on steam reforming and dry reforming
(Bodrov et al., 1964; Soliman and Adris, 1992; Papp et al., 1996; Hou and Hughes, 1997). Also, in
experiments with partial fuel conversion (but complete oxygen consumption) we find that the syngas
from the reformer is at or close to the WGS-equilibrium. It must be noted that assuming that the WGS-
equilibrium is established is not a critical assumption in the current model because of the following
reasons: the reaction enthalpy of the WGS-reaction is not very large (small influence on enthalpy
balance) and the reaction rate equation that is used for the fuel in the reforming section, as well as the
syngas yield-number, are independent of whether the WGS-reaction is at equilibrium or not (for the
syngas yield, CO and H2 are added together).
In the Theory section the model for CPO and ATR is explained in detail for methane as fuel, and
more schematically for natural gas and naphtha. In the Results and Discussion section, simulation
results are presented that are representative of our experience for CPO and ATR of methane.

Theory

Inputs
The two-section model describes the catalytic conversion of methane mixed with air and steam. The
inlet molar flows, φCH4,0, φair,0, φwater,0 [mol/s] are known as well as the inlet temperature, T0 [K] and the
system pressure, P [Pa] (absolute pressures are used throughout). It is assumed here that the fraction
of O2 in air is χ (namely, χ=0.2095), the remainder being N2, thus φO2,0=χ⋅φair,0 and φN2,0=(1-χ)⋅φair,0. The
humidity of air is neglected as well as Ar, CO2, etc. The oxygen-to-carbon ratio is α=φO2,0/φCH4,0 and the
steam-to-carbon ratio is γ=φH2O,0/φCH4,0. Simulations are based on a fuel flow of φCH4,0=10 mmol/s which
corresponds to 8 kWh (heating value) and ~3 kWe produced in the fuel cell, which is typical of the
scale of a single household.
The reactor model consists of a description of the oxidation section in which all oxygen is converted
as well as part of the methane. This conversion is considered to be extremely fast, thus the catalyst
volume required negligible. The exit of the oxidation section is called the “top” having a certain gas
composition and a top-temperature. The top is the virtual intersection of the oxidation section and the
reforming section. In the model, the entire catalyst volume is assigned to the reforming section.
Heat loss can be implemented in both sections. For the oxidation section it must be an algebraic
expression (as function of top-temperature, surface area, etc.) or a constant number, Qox [W]. Heat

6
loss in the reforming section is a function of the local temperature and is implemented in the
(differential) enthalpy balance, Eq. [12].

Simple relations
Some simple relations that will be used without further reference are summarized here. Molar flows

φi [mol/s] are related to the total flow, φtot by φ i = x i ⋅ φtot , with xi the molar fraction [mol/mol]. Thus,

φtot = ∑ φ i and ∑x i = 1 . At the inlet, i runs over max. 4 species (CH4, O2, N2 and possibly H2O)
i i

and over 6 species at the top and in the reforming section (CH4, CO, CO2, H2, H2O and N2). Partial

pressures pi [Pa] are related to the total pressure P by ∑p

i
i = P and pi = x i ⋅ P .

Thermodynamics
The relevant thermodynamics that is included in the oxidation section and reforming section is
summarized in a general form. Enthalpy, Hi [J/mol], and entropy, Si [J/(mol⋅K)], of a certain
component, i, are a function of the local temperature, T, see Appendix A.
The two relevant reactions are the water-gas-shift (WGS) reaction,
CO + H2O ←→ CO2 + H2 ,
and the methane-steam-reforming (MSR) reaction,
CH4 + H2O ←→ CO + 3 H2 .
For each reaction, the equilibrium constant, Ki, is given by
νi
 1    pi 
K i = exp  −
  ∑ν H i i −T ∑ ν i Si   =
 ∏  *
P 
[1]
 RT  i i  i

with νi the stoichiometric constants in the reaction and P =101.3 kPa. For the two reactions the result
*

is

 1
KWGS = exp  − ( ( HH 2 + HCO 2 − HCO − HH 2O ) − T (SH 2 + SCO 2 − SCO − SH 2O ))  =
 RT 
, [2]
pH 2 ⋅ pCO 2 xH 2 ⋅ xCO 2 φH 2 ⋅ φCO 2
= = =
pH 2O ⋅ pCO xH 2O ⋅ xCO φH 2O ⋅ φCO

 1
K MSR = exp  − ( ( 3 ⋅ HH 2 + HCO − HCH 4 − HH 2O ) − T ( 3 ⋅ SH 2 + SCO − SCH 4 − SH 2O ))  =
 RT 
2 2 2
. [3]
⋅ pCO  1 
pH3 2 ⋅ xCO  P 
xH3 2 ⋅ φCO 1  P φ3H 2
= = =
pH 2O ⋅ pCH 4  P *  xH 2O ⋅ xCH 4  P *  2 
φH 2O ⋅ φCH 4 φtot * 
P 
Atomic balances are based on the fact that at stationary conditions, and for a one-dimensional
reactor model, at each location, the mole flow of each atom (C, H, O, N) is constant (thus equal to the
inlet, 0, conditions). Thus, at each location, j, the balances for C, H, O and N are

7
φCH 4,0 = φCH 4, j + φCO, j + φCO 2, j
φCH 4,0 ⋅ ( 4 + 2 ⋅ γ ) = 4 ⋅ φCH 4, j + 2 ⋅ ( φH 2, j + φH 2O, j )
. [4]
φCH 4,0 ⋅ ( 2 ⋅ α + γ ) = φCO, j + 2 ⋅ φCO 2, j + φH 2O, j
φN 2,0 = φN 2, j

Oxidation Section
In the oxidation section, kinetics are not considered but a set of algebraic equations describes the
conditions at the exit of the oxidation section (the top) which is the intersection with the downstream
reforming section. The top-temperature, Ttop, and the individual gas flows (φCH4,top, φCO,top, φCO2,top,
φH2,top, φH2O,top, φN2,top) are unknown. At the end of the oxidation section oxygen has already vanished.
Seven equations are required. We have the WGS-equilibrium, Eq. [2], the four (atomic) mass
balances, Eq. [4], and the enthalpy balance

∑H
i
i ,T0 ⋅ φi ,0 = ∑H i
i ,Ttop ⋅ φi ,top + Qox [5]

with Qox the heat loss in the oxidation section [W], which can be made an explicit function of T0, Ttop,
surface area, etc.
Finally, we introduce the fuel-utilization ratio, UT, which describes which part of the fuel is converted
in the oxidation section. UT is a function of process conditions, catalyst formulation and structure. For
a specific catalyst we have derived the empirical function
2
UT = -1.6492⋅α +2.664⋅α-0.2348-0.033⋅γ [6]
from experiments with methane at T0~400 C and P=170 kPa in a wide range of values for α and γ
o

(being respectively the oxygen-to-carbon and steam-to-carbon ratio). In these experiments the top-
temperature was measured and the above model used to derive the appropriate UT-value. Eq. [6] is a
very accurate description of these data.
Using Eq. [6], φCH4,top is given by
φCH 4,top = (1 − UT ) ⋅ φCH 4,0 . [7]

Note that this set of equations does not explicitly describe the selectivities to H2 and CO (relative to
the formation of H2O and CO2). However, selectivities can be directly calculated from the oxidation-
model (based on Eq. [6], atom and enthalpy balances and the WGS-equilibrium). In general, (for a
given α and γ), the higher UT, the more selective the oxidation section and the lower the top-
temperature.

Reforming section
Transport and reaction in the reforming section are described by a plug-flow model. The four atomic
balances are used as well as the enthalpy balance and the WGS-equilibrium. One differential mass
balance is required to obtain a one-dimensional, stationary, plug-flow model for the reforming section.
Ideally, this is a mass balance in one of the rate-limiting reactants, typically in the fuel component (but
not necessarily so). Indeed, we will use a differential mass balance that describes the conversion of
methane. Based on literature (Akers and Camp, 1955; Bodrov et al., 1964; Ridler and Twigg, 1989;
Rostrup-Nielsen, 1993; Steghuis et al., 1998) and our own experience, a power law rate equation is

8
proposed for the reforming of methane that is first order in the partial pressure of the fuel component,
methane. A first order rate equation in the fuel is compatible with the idea that mass transfer,
adsorption or dissociation of the fuel is rate limiting.
At stationary conditions, a differential mass balance over a “slice” in the reforming section results for
component i in
dφ i
= Ri ⋅ S . [8]
dz
with S the cross-sectional surface area [m ], z the reactor coordinate [m], φi the mole flow of i [mol/s]
2

and Ri the formation rate of i [mol/(m ⋅s)].

3

For methane conversion, the following effective rate equation is proposed based on a first order
dependence on the partial pressure of methane and an empirical term (in between brackets) that is
based on local partial pressures and (via KMSR) on the local temperature, which describes how closely
thermodynamic equilibrium has been approached. The result is
 pH3 2 ⋅ pCO 1 
RCH 4 = −a ⋅ k 0 ⋅ e − EA / RT pCH4  1 −  [9]
 K MSR ⋅ pCH 4 ⋅ pH 2O P * 2 
2 3
in which a is the specific surface area a (m of active phase per m of reactor volume) which is a
function of the amount of active phase and details of catalyst preparation, form and structure; k0 is the
pre-exponential kinetic number, EA the activation energy [J/mol] of the reaction and R the gas constant
[J/(mol⋅K)]. The activation energy of reforming reactions is often reported to be in the 100
kJ/mol-range: Hou and Hughes (1997) suggest EA=100 kJ/mol (based on their Fig. 4) and EA=95
kJ/mol (Fig. 12), while Steghuis et al. (1998) measure EA=105-114 kJ/mol. KMSR is a direct function of
the local temperature and is given by the first equality in Eq. [3].
Because the WGS-equilibrium is assumed to be established throughout the reforming section, the
driving force-term (between brackets) based on the MSR-reaction, can be rewritten to a form based
on the dry reforming (DR) equilibrium (CH4+CO2 <--> 2⋅CO+2⋅H2) or to a form based on any of the
other possible reactions of methane with water and CO2. All give exactly the same result.
Eq. [9] is implemented in Eq. [8] while partial pressures are recalculated to flows, φi. Furthermore, z
3
and S are replaced by the catalyst volume, V [m ] and a dimensionless reactor coordinate, 0<L*<1,
z⋅S=L*⋅V. Finally, the group a⋅k0⋅V is replaced by λ [mol/(Pa⋅s)], resulting in

d φCH 4  pH3 2 ⋅ pCO 1 

= −λ e − E A / RT
p  1 −  →
dL* CH 4  K MSR ⋅ pCH 4 ⋅ pH 2O P * 2
 

d φCH 4 − E A / RT −1
 φH3 2 ⋅ φCO  P  1
2

= −λ e P φ φ  1 − . [10]
 K MSR ⋅ φCH 4 ⋅ φH 2O  P *  φtot 
CH 4 tot
dL* 
2

Note that Eq. [11] does not directly specify whether methane reacts with water or CO2, nor does it
specify the products; it only describes the conversion rate of methane. However, in conjunction with
the WGS-equilibrium and the atom balances, Eqs. [4], the conversion rates of all other components
*
are fixed at each location, L .

9
 Finally, the differential enthalpy balance is given by

d ∑H φ i i
i
dz
(
= −U ⋅ PR ⋅ T − T * . ) [11]

Here, an engineering heat transfer relation is used for heat loss through the wall of the reactor. U is
an effective heat transfer coefficient [W/(m ⋅K)], T* a representative temperature outside the reactor,
2

PR the perimeter [m]. Multiplying each side by V/S and replacing U⋅PR⋅V/S by κ [W/K] results in

d ∑H φ i i
i
dL * (
= −κ ⋅ T − T * . ) [12]

At each location, that is at each L*-value, the two differential equations, Eqs. [10] and [12], are
solved as well as the four atom balances, Eqs. [4], the WGS-equilibrium, Eq. [2], and the first equality
in Eq. [3]. Because the temperature, T, changes as function of L*, all terms Hi, Si, KWGS and KMSR
change with L*, as well as all φi’s and φtot; P, κ, λ and T* are constant throughout.

Coke formation
From separate thermodynamic calculations one can derive the following inequality that predicts
below what critical temperature, Tcritical [K], coke formation is thermodynamically expected,

2 pCO 2 * xCO 2 P * φCO 2 P*

Tcritical=3.1368⋅(ln(A)) -45.791⋅ln(A)+973.51 with A = P = = φ tot
. [13]
2
pCO 2
xCO P φCO
2
P

After having determined the thermodynamic equilibrium gas phase composition assuming zero coke
formation, Eq. [13] can be used to check whether that assumption was correct. Coke formation is not
predicted in the relevant cases for ATR/CPO we considered, with the equilibrium temperature of the
synthesis gas typically ~100 K above that required for coke formation. This is in alignment with our
experiments in which coke formation in the reactor is not observed under normal operating conditions.

Natural gas
The model for ATR/CPO of methane can be extended to describe the conversion of natural gas - if
methane makes up the majority of the hydrocarbons in the natural gas, which is generally the case.
The two-section model for methane can be used as long as we assume that all hydrocarbons except
for CH4 are consumed in the oxidation step (Akers and Camp, 1955). The parameters α and γ
required in Eq. [6] are best defined relative to the total inlet hydrocarbon flow, φC,0: α=φO2,0/φC,0 and
γ=φH2O,0/φC,0 with φC,O given by a summation over all hydrocarbons

φC,0 = ∑C i
i ⋅ φCi ,0 [14]

with Ci the C-number of component i (e.g., C=2 for C2H6). Further, Eq. [7] must be replaced by
φCH 4,top = (1 − UT ) ⋅ φC ,0 . [15]

10
Naphtha and other multicomponent fuels
Naphtha is a complex mixture of components mainly in the C5-C12 range. Instead of dealing with
each component separately, it is possible to simplify the analysis by proper averaging. For example, a
fuel mixture of 25 mol% toluene, 25% methylcyclohexane, 25% n-heptane and 25% iso-octane can be
summarized by the formula C7.25H14 with the enthalpy function H = -6⋅10 ⋅T +0.2638⋅T +15.249⋅T-
-5 3 2

153627 (J/mol) correct within ~1% in the relevant temperature range 298<T [K]<1400.
A reformer model for naphtha can be derived from the methane two-section model. For the oxidation
section, a utilization-function for naphtha must be used which is not necessarily the same as Eq. [6],
as well as an additional equation for the formation of methane and other hydrocarbons (though
methane is the most relevant one; the formation rates of higher hydrocarbons are negligible with ppm-
or ppb-levels in the equilibrium mixture). Methane formation in the oxidation section can be set to zero,
φCH4,top=0, or an equilibrium can be used, such as the MSR-equilibrium or the dry reforming equilibrium
(DR).
In the reforming section a differential mass balance for naphtha conversion (e.g., first order in
naphtha) replaces Eq. [11]. A driving force-term, such as used in Eq. [10] is not required, as at
thermodynamic equilibrium naphtha-conversion is complete. For methane formation, the MSR- or DR-
equilibrium can be used or a reaction rate equation, e.g., based on the methanation mechanism (with
the rate dependent on the partial pressures of CO and H2). We find that the DR-equilibrium is in many
cases established (at the exit of the reactor) and use the DR-equilibrium in the reforming section.

Results and Discussion

In the first part of Results and Discussion, experimental results of the conversion of methane in
ATR/CPO are compared with predictions of the two-section model. In the second part somewhat other
parameter settings are used and simulation results are presented for top-temperature and methane
conversion behavior.
Two remarks must be made with respect to methane conversion, ζ, and utilization, UT. First,
whereas ζ is the “actual” conversion measured at the end of the reactor, thus at the end of the
reforming section, the fuel utilization ratio, UT, is actually equal to the methane conversion at the end
of the oxidation section (0<UT<1). The difference between UT and ζ is a measure of the fuel
conversion in the reforming section.
Second, in CPO/ATR of methane, the methane conversion, ζ, is directly related to the syngas yield,
Y, (mole flow of produced CO+H2 divided by the mole flow of CH4 fed to the reactor). This relation
follows from combination of the atom balances for C, H and O, and is given by
Y = 4⋅ζ - 2⋅α [16]
in which ζ is the methane conversion (0<ζ<1) and α the (molecular) oxygen-to-carbon ratio. Thus,
methane conversion results can be recalculated directly to syngas yield numbers. Eq. [16] can be used
to calculate the actual syngas yield (based on ζ), the syngas yield after the oxidation section (based on
UT) and the maximum attainable syngas yield (based on the conversion at thermodynamic
equilibrium).

11
Comparison of two-section model with experimental results
Experimental data for top-temperature and methane conversion are presented for a system
o
pressure of 170 kPa and an inlet temperature of T0~400 C in Figure 3 and Figure 4 for different
values of the oxygen-to-carbon ratio, α, and the steam-to-carbon ratio, γ. Each experiment is analyzed
individually and a different heat loss is calculated for each experiment by the difference of the total
enthalpy leaving and entering the reactor (inlet temperature measured using a thermocouple; exit
temperature by assuming the WGS-equilibrium; exit composition based on GC-analysis). In the
analysis the heat loss is completely assigned to the oxidation section; therefore Qox of Eq. [5] can be
measured. In the analysis of the data (Figures 3 and 4) heat loss in the reforming section is set to zero
(U in Eq. [12] set to zero). Due to the fact that the heat loss is different for each experiment, the lines
for toptemperature (Figure 3), conversion and thermodynamic equilibrium (Figure 4) are somewhat
jagged. The heat loss was in all cases small compared to the total energy content (in combustion) of
the entire gas flow.
From each experimental value of top-temperature (symbols in Figure 3) the utilization of methane
(UT) can be calculated (circles in Figure 4) using the oxidation model that is based on atom balances,
the enthalpy balance and the WGS-equilibrium (i.e., given these constraints, only for one value of UT
is the measured top-temperature, at the exit of the oxidation section, obtained). The top-temperature
decreases when the amount of air is reduced or the amount of steam is increased. The empirical
utilization function, Eq. [6], is derived from the derived utilization data and describes these data quite
well, see the lower solid lines in Figure 4. Using Eq. [6] and the other elements of the oxidation model
the top-temperature can be back-calculated. The result is plotted as the solid lines in Figure 3. For the
highest oxygen-to-carbon ratio, α=0.59, the prediction is accurate; for the two lower oxygen-to-carbon
ratio’s the prediction is less accurate but is certainly showing the right trends.

1050

1000
Top-temperature ( C)

α=0.59
o

950

900 α=0.52

850 α=0.42

800
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Steam-to-Carbon ratio, γ

Figure 3. Top-temperature in ATR/CPO of methane. Different steam-to-carbon ratio’s, γ, and oxygen-

to-carbon ratio’s, α. T0~400 C, P=150 kPa, varying heat loss. Lines are predictions of the oxidation
o

model.

12
 The data for methane conversion, ζ (squares) are presented in Figure 4 together with the predicted
conversion at thermodynamic equilibrium (upper solid lines). By adjusting λ and EA (given a suitable
form of the rate equation, Eq. [10]), the two-section reactor model (dashed lines) can be fitted to the
methane conversion data.
For each oxygen-to-carbon ratio, α, the difference between the utilization, UT, and the actual
conversion, ζ (the difference representing the conversion in the reforming section), first increases with
γ (starting from γ=0) but above a certain γ-value (0.5-0.75) becomes constant: UT and ζ decrease
simultaneously as if the conversion in the reforming section becomes independent of γ. The constant
difference of ζ and UT increases with α, namely from ζ-UT=0.17 at α=0.42 to 0.20 at α=0.52 and 0.23
at α=0.59. Though the gap between UT and ζ remains fairly constant for a given α and for enough
steam (above γ~0.5-0.75), the gap between UT and the thermodynamically attainable conversion
increases with increasing amounts of steam, γ.
For each value of α (α=0.42 in Figure 4a, α=0.52 in Figure 4b and α=0.59 in Figure 4c),
thermodynamic equilibrium conversion (upper solid line) was reached at the end of the reforming
section for zero steam addition, γ=0. With increasing amounts of steam, the actual conversion, ζ, first
increases, then levels off and finally starts to decrease. This while the thermodynamically attainable
conversion continuously increases with γ. As a consequence, the gap between the actual conversion,
ζ, and that what is thermodynamically attainable increases with γ. Interestingly, the gap between the
actual conversion, ζ, and thermodynamic equilibrium is about the same for each value of the oxygen-
to-carbon ratio, α, though both values strongly increase with α.
That thermodynamic equilibrium is only reached at low to zero steam loads is due to the decreasing
utilization in the oxidation section and the lower reaction rates in the reforming section when α is
increased. The lower reaction rate is due to the lower temperatures throughout the reforming section,
and the higher total gas flow rate, φtot, resulting in a lower methane partial pressure. Both effects are
due to the presence of the additional steam. Thus, the thermodynamic advantage of adding more
steam (if available at a low cost at T0) can only be attained with more catalyst (or at a lower space
velocity).
As discussed, the gap between the actual, observed, conversion, and the potential,
thermodynamically attainable, conversion does not depend much on the oxygen-to-carbon ratio, α.
This shows that several opposing mechanisms occur simultaneously: though reactor temperatures are
much higher at the high air-load (high value for α), the methane partial pressure is decreased due to
the additional nitrogen, while the actual conversion required in the reforming section (to keep the gap
constant) increases with α (from ~0.17 at α=0.42 to ~0.23 at α=59).

13
 1.0
Oxygen-to-Carbon ratio, α=0.42

0.9

Utilization and Conversion

0.8
Thermodynamic Equilibrium

0.7
Conversion

0.6
Utilization

0.5
0 0.25 0.5 0.75 1 1.25 1.5
Steam-to-Carbon ratio, γ

1.0
Oxygen-to-Carbon ratio, α=0.52
Thermodynamic Equilibrium
0.9
Utilization and Conversion

Conversion
0.8

0.7

Utilization
0.6

0.5
0 0.25 0.5 0.75 1 1.25 1.5
Steam-to-Carbon ratio, γ

1.0 Thermodynamic
Equilibrium
Conversion
0.9 Oxygen-to-Carbon ratio, α=0.59
Utilization and Conversion

0.8

0.7 Utilization

0.6

0.5
0 0.25 0.5 0.75 1 1.25 1.5
Steam-to-Carbon ratio, γ

Figure 4. Methane utilization and conversion. The fuel utilization ratio, UT, equals the methane
conversion at the end of the oxidation section and is derived from top-temperature data (circles). The
lower solid line is the empirical fit, Eq. [6]. The two-section reactor model is used for the methane
conversion, ζ, at the end of the reforming section (dashed line) and is compared with conversion data
o
(squares). Upper solid line is thermodynamic equilibrium. T0~400 C, P=170 kPa, EA=160 kJ/mol,
λ=17.26 mol/(Pa⋅s), φCH4,0=5.3 mmol/s, κ=0 W/K, varying heat loss Qox).

14
Further Simulations
For other parameter settings, simulation results are presented in Figure 5 - Figure 7. Figure 5
focuses on the oxidation section and shows the top-temperature and methane utilization as function of
the oxygen-to-carbon ratio, α. The top-temperature is plotted on the left-hand y-axis and increases
with more oxygen and less steam. The utilization is plotted on the right-hand y-axis (dashed lines) and
as in Figure 4 increases with more oxygen and less steam. Again, as in Figure 4 the maximum,
thermodynamically attainable, conversion (solid lines) increases with more steam and oxygen. For the
pressure of 150 kPa, only for sufficient amounts of oxygen (in this case, for α>0.6) is the
thermodynamic methane conversion close to 100%. For a lower α, the thermodynamic, maximum, fuel
conversion is significantly reduced and most significantly without steam (γ=0).
Figure 5 indicates once more that with more steam the utilization (conversion in oxidation section)
decreases while the attainable (maximum) conversion increases: more fuel must be converted in the
reforming section.
Simulation results of the development of temperature and methane conversion through the
reforming section are presented in Figure 6. Methane conversion is directly coupled to syngas yield
using Eq. [16]. The top-temperature is higher without steam which is one reason why conversion is
already complete at L*~0.3. For γ=1.6 methane is converted along the entire reactor coordinate, with
thermodynamic equilibrium not yet reached at L*=1.
Comparing γ=0 and γ=1.6 shows that for γ=1.6 more catalyst is required to arrive at thermodynamic
equilibrium, but that the methane conversion and syngas yield associated with equilibrium are higher
(+1.5% and +1%, respectively) than at zero steam load.

1200 1.4
Top-temperature ( C)

γ=0 γ=1.6 1.3

o

1100
1.2

1000 1.1

1
900
Utilization, Conversion

0.9
γ=1.6 γ=0
800
γ=0 0.8

0.7
700 γ=1.6
0.6
600
0.5

500 0.4
0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7
Oxygen-to-Carbon ratio, α

Figure 5. Oxidation Section in ATR/CPO of methane. Variation of oxygen-to-carbon ratio, α, and

steam-to-carbon ratio, γ. Top-temperature on left y-axis. Thermodynamic conversion (solid lines on
o
right y-axis) as well as the fuel utilization (dashed lines). T0=400 C, Qox=0 W, P=150 kPa.

15
 1100

1050

1000

Temperature ( C)
o 950

900

850
γ=0
800

γ=1.6
750
0 0.2 0.4 0.6 0.8 1
Dimensionless reactor length, L *

100
γ=0

95 γ=1.6
Methane conversion (%)

90

85

80

75

70
0 0.2 0.4 0.6 0.8 1
Dimensionless reactor length, L *

Figure 6. Temperature profile (a) and methane conversion (b) as function of reactor coordinate, L*, for
two steam-to-carbon ratio’s, γ. The circle shows the equilibrium value for γ=1.6 (α=0.57, P=150 kPa,
φCH4,0= 10 mmol/s, T0=400 C, λ=0.10 mol/(Pa⋅s), EA=100 kJ/mol, Qox=0 W, κ=0 W/K).
o

In case of heat loss through the reactor wall, κ>0, the situation becomes more complicated and it is
possible that for a low γ, resulting in fast reactions and low catalyst volumes, the optimum L
*

corresponds to a higher methane conversion (thus, a higher syngas yield). Figure 7 shows that for γ=0
the maximum methane conversion is found at L*~0.14 and is 95.7%, whereas for γ=1.6 the maximum
methane conversion is 91.5% at L*~0.63. This behavior is opposite to what is predicted for
thermodynamic equilibrium and without heat loss, in which case methane conversion is higher for
γ=1.6 than for γ=0, see Figure 6. In the decreasing portions of the conversion curves (Figure 7)
thermodynamic equilibrium has already been rather well-established but due to heat loss the
temperature decreases and methane is formed again (methanation).

16
 100

95 γ=0
γ=1.6
Methane conversion (%) 90

85

80

75

70
0 0.2 0.4 0.6 0.8 1
Dimensionless reactor length, L *

Figure 7. Methane conversion with heat loss in the reforming section, as function of γ and L*.
(κ=1 W/K, T*=298.15 K, all other data see Figure 6).

Conclusions
A two-section reactor model describes several key features of fuel conversion in autothermal
reforming (CPO and ATR) based on air and steam, such as the high upstream temperature as well as
the decreasing temperature throughout the reforming section. The model can be fitted to data to
obtain an optimum value for the kinetic constant, activation energy and reaction order. For a given
amount of catalyst and with increasing steam loads, the temperature in the reactor decreases while
the methane conversion (and yield of hydrogen and CO) first increases and then decreases, while the
conversion that is thermodynamically feasible continuously increases. Thus, the thermodynamic
advantage of adding more steam can only be attained at a lower space velocity. The model can be
extended to describe multicomponent fuels such as natural gas and naphtha.

Notation
2 3
a Specific surface area of catalyst m /m
3
ci Concentration of component i mol/m
EA Activation energy of reaction i J/mol
Hi Enthalpy of component i J/mol
2
k0 Pre-exponential kinetic number mol/(m .Pa.s)
Ki Equilibrium constant of reaction i (either WGS or MSR)
*
L Dimensionless axial reactor coordinate
pi Partial pressure of component i Pa
P Absolute pressure Pa
* 5
P Reference pressure (=1.013⋅10 ) Pa

17
PR Perimeter of the reactor m
Qox Heat loss in oxidation section W
R Gas constant (=8.3144) J/(mol.K)
3
Ri Reaction rate of reaction i mol/(m .s)
2
S Cross-sectional surface area of reactor m
Si Entropy of component i J/(mol.K)
t Time s
T Temperature K
T0 Inlet temperature K
*
T Temperature outside of reactor K
UT Fuel utilization ratio
2
U Effective heat transfer coefficient W/(m .K)
3
V Reactor volume m
xi mole fraction of component i mol/mol
z Reactor coordinate in axial direction m

Greek symbols
α Oxygen-to-carbon ratio of inlet flow (=φO2,0/φCH4,0)
γ Steam-to-carbon ratio of inlet flow (=φH2O,0/φCH4,0)
ζ Methane conversion (at the end of reforming section)
κ =U⋅π⋅V/S W/K
λ =a⋅k0.V mol/(Pa.s)
νi Stoichiometric constant of reactant i
φi Mole flows of component i mol/s
φi,0 Inlet mole flows of component i mol/s
φtot Total mole flow mol/s
φi,top Mole flow of component i at the intersection of oxidation and reforming section mol/s
χ Fraction of oxygen in air (=0.2095)

Literature Cited
Ahmed, S., and M. Krumpelt, “Hydrogen from hydrocarbon fuels for fuel cells,” Int. J. hydrogen energy,
26 (2001) 291.
Akers, W.W., and D.P. Camp, “Kinetics of the methane-steam reaction,” AIChE J., 1, 471 (1955).
Arena, F., F. Frusteri, and A. Parmaliana, “Kinetics of the partial oxidation of methane to formaldehyde
on silica catalyst,” AIChE J., 46, 2285 (2000).
Bodke, A.S., S.S. Bharadwaj, and L.D. Schmidt, “The effect of ceramic supports on partial oxidation of
hydrocarbons over noble metal coated monoliths,” J. Catalysis, 179, 138 (1998).
Bodrov, N.M., L.O. Apel’baum, and M.I. Temkin, “Kinetics of the reaction of methane with steam on
the surface of nickel,” Kinet. Catal., 5, 614 (1964).

18
Chang, Y.-F., and H. Heinemann, “Partial oxidation of methane to syngas over Co/MgO catalysts. Is it
low temperature?,” Catalysis Letters, 21, 215 (1993).
Daubert, T.E., and R.P. Danner, Data Compilation Tables of properties of pure compounds, AIChE,
New York (1985).
De Groote, A., and G.F. Froment, “Reactor modeling and simulations in synthesis gas production,”
Reviews in Chem. Eng., 11, 145 (1995).
De Groote, A., and G.F. Froment, “Simulation of the catalytic partial oxidation of methane to synthesis
gas,” Applied Catalysis A: General, 138, 245 (1996).
De Smet, C.R.H., M.H.J.M. de Croon, R.J. Berger, G.B. Marin, and J.C. Schouten, “Kinetics for the
partial oxidation of methane on a Pt gauze at low conversions,” AIChE J., 46, 1837 (2000).
De Smet, C.R.H., M.H.J.M. de Croon, R.J. Berger, G.B. Marin, and J.C. Schouten, “Design of
adiabatic fixed bed reactors for the partial oxidation of methane to synthesis gas. Application to
production of methanol and hydrogen-for-fuel-cells,” Chem. Eng. Sci., 56, 4849 (2001).
Deutschmann, O., and L.D. Schmidt, “Modeling the partial oxidation of methane in a short-contact-
time reactor,” AIChE J., 44, 2465 (1998).
Dissanayake, D., M.P. Rosynek, K.C.C. Kharas, and J.H. Lunsford, “Partial oxidation of methane to
carbon monoxide and hydrogen over a Ni/Al2O3 catalyst,” J. Catalysis, 132, 117 (1991).
Docter, A., and A. Lamm, “Gasoline fuel cell systems,” J. Power Sources, 84, 194 (1999).
th
B.Elvers et al. (ed.). Ullmann’s Encyclopedia of Industrial Chemistry, Volume A12, p. 207, 5 Ed.,
VCH, Weinheim (1989).
Heitnes Hofstad, K., T. Sperle, O.A. Rokstad, and A. Holmen, “Partial oxidation of methane to
synthesis gas over a Pt/10% Rh gauze,” Catalysis Letters, 45, 97 (1997).
Hickman, D.A., and L.D. Schmidt, ”Production of syngas by direct catalytic oxidation of methane,”
Science, 259, 343 (1993a).
Hickman, D.A., and L.D. Schmidt, “Steps in CH4 oxidation on Pt and Rh surfaces: high temperature
reactor simulations, “ AIChE J., 39, 1164 (1993b).
Hou, K., and R. Hughes, “The kinetics of methane steam reforming over a Ni/α-Al2O catalyst,” Chem.
Eng. J., 82, 311 (2001).
Kramer, G.J., W. Wieldraaijer, P.M. Biesheuvel, and H.P.C.E. Kuipers, “The determining factor for
catalyst selectivity in Shell’s catalytic partial oxidation process,” ACS Fuel Chemistry Division
Preprints, 46, 659 (2001).
Moon, D.J., K. Sreekumar, S.D. Lee, B.G. Lee, and H.S. Kim, “Studies on gasoline fuel processor
system for fuel-cell powered vehicles application,” Applied catalysis A: General, 215, 1 (2001).
Papp, H., P. Schuler, and Q. Zhuang, “CO2 reforming and partial oxidation of methane,” Topics in
Catalysis, 3, 299 (1996).
Prettre, M., Ch. Eichner, and M. Perrin, “The catalytic oxidation of methane to carbon monoxide and
hydrogen,” Trans. Faraday Soc., 43, 335 (1946).
Ridler, D.E., and M.V. Twigg, “Steam Reforming.” Ch. 5 in Catalyst Handbook. M.V. Twigg (Ed.),
Manson Publishing, London (1989).
Rostrup-Nielsen, J.R., ”Production of synthesis gas,” Catalysis Today, 18, 305 (1993).

19
Service, R.F., “Bringing fuel cells down to earth,” Science, 285, 682 (1999).
Soliman, M.A., A.M. Adris, A.S. Al-Ubaid, and S.S.E.H. El-Nashaie, “Intrinsic kinetics of nickel/calcium
aluminate catalyst for methane steam reforming,” J. Chem. Tech. Biotechnol., 55, 131 (1992).
Steghuis, A.G., J.G. van Ommen, and J.A. Lercher, “On the reaction mechanism for methane partial
oxidation over yttria/zirconia,” Catalysis Today, 46, 91 (1998).
Wang, H.Y., and E. Ruckenstein, “Catalytic partial oxidation of methane to synthesis gas over γ-Al2O3-
supported rhodium catalysts,” Catalysis Letters, 59, 121 (1999).
Zhu, J., D. Zhang, and K.D. King, “Reforming of CH4 by partial oxidation: thermodynamic and kinetic
analyses,” Fuel, 80, 899 (2001).

Appendix A
Enthalpy, Hi, and entropy, Si, are expressed as (Daubert and Danner, 1985):

 c c   e e 
H i = H 0 ,i + ai (T − T0 ) + bi c i  coth i − coth i  − d i ei  tanh i − tanh i  ,
 T T0   T T0 

     
  sinh  c i      cosh  ei   
T c c c c   T    − d  e i tanh ei − ei tanh e i − ln  T  
S i = S 0,i + ai ln + bi  i coth i − i coth i − ln  i  
T0 T T T0 T0  sinh  c i    T T T0 T0  cosh ei   
   T      T  
   0      0  
with T0=298.15 K and coth=1/tanh. The values for ai-ei are summarized in Table A.

Table A. Thermodynamic data.

H0,i (kJ/mol) S0,i (J/mol/K) ai (J/mol/K) bi (J/mol/K) ci (K) di (J/mol/K) ei (K)

CH4 -74.5 186.27 33.298 79.933 2086.9 41.602 991.96

O2 0 205.04 29.103 10.04 2526.5 9.356 1153.8

CO -111 197.54 29.108 8.773 3085.1 8.4553 1538.2

CO2 -394 213.69 29.37 34.54 -1428 26.4 588

H2 0 130.57 27.617 9.56 2466 3.76 567.6

H 2O -242 188.72 33.363 26.79 2610.5 8.896 1169

N2 0 191.609 29.105 8.6149 1701.6 0.10347 909.79

20