02638762/02/$10.00+0.

00
# Institution of Chemical Engineers
Trans IChemE, Vol 80, Part A, May 2002

CATALYST SHAPE AS A DESIGN

PARAMETEROPTIMUM SHAPE FOR
METHANE-STEAM REFORMING CATALYST
J. S. SOLTAN MOHAMMADZADEH 1 and A. ZAMANIYAN2
1

Department of Chemical Engineering, Sahand University of Technology, Tabriz, IR Iran.

2
Gas research division, Research Institute of Petroleum Industry, Tehran, Iran.

igh catalyst activity plays an important role in the safe and ef cient operation of
methane-steam reforming reactors. Since transport resistance limits the effectiveness
factor in the reactor to values much less than unity, increasing the catalyst surface area
increases its activity. Using a mathematical model of a terrace wall reactor, the effect of catalyst
shape on performance of the reactor has been studied. Considering different parameters of
catalyst effectiveness factor, reaction gas temperature, reactor pressure drop and mechanical
strength, it has been shown that a multichannel cylinder is a good choice among different
alternative catalyst pellet shapes. Compared to a standard spherical shape, a multichannel cylinder gives 10% less total pressure drop and its effectiveness factor is almost three times higher.
Keywords: terrace wall furnace; methane-steam reforming; catalyst shape; catalyst design.

INTRODUCTION

(ii) maintain low tube wall temperature to ensure a long

operation life;
(iii) cause a low and constant pressure drop.

The Steam reforming process is still the most economical

route for the production of hydrogen and synthesis gas from
natural gas. It generates synthesis gas for ammonia and
methanol production as well as hydrogen, oxo gas and town
gas. After desulphurization, natural gas feed is mixed with
steam and is fed to the reforming furnace where chemical
reactions take place in the presence of a nickel containing
catalyst. A mixture of CO, CO2, H2, remaining CH4 and
equilibrium steam leaves the reformer furnace. Four different furnace types are commonly used today (radiant wall,
terrace wall, down- ring and up- ring). The reformer reactors are high performance furnaces, which contain 40 to 400
tubes. The internal tube diameter is in the range 70 to
160 mm, with a tube wall thickness of 10 to 20 mm. The
heated length is 6 to 12 m depending on the furnace type.
The tubes are made from high alloy nickel chromium steel
(e.g. HK40: Cr 25%, Ni 20%, Co 4% or IN519: Cr 24%, Ni
24%, Nb 1.5%, Co 3%). The tubes are supported outside the
furnace chamber either from the oor or the ceiling1. The
reliability of the tubes is an important factor, because failures can result in long down-periods for re-tubing, and
hence, loss of production. Reactions take place at temperatures from 500 to 950 C and pressures between 1500 and
3000 kPa. Reactor tubes are lled with nickel containing
catalyst pellets. Ni-Al 2O3 catalyst properties are dictated by
severe operating conditions of high temperature and high
steam partial pressure (close to 3000 kPa). In the reforming
reaction, the catalyst should:

To meet these requirements, the catalyst must have

suf cient activity, resistance to carbon formation, mechanical strength and suitable shape2. Catalyst activity, its
resistance to coke formation and its mechanical properties
depend mainly on the particular catalyst formulation and
preparation method. Transport resistances limit the catalyst
effectiveness factor to values much less than unity, so only a
thin exposed layer of the catalyst pellet takes part in the
reaction. Although, for a given formulation decreasing
catalyst pellet size can increase the catalyst exposed area
per unit reactor volume, the catalyst particle size cannot be
reduced freely due to the excessive pressure drop. With a
given formulation, pellet size and preparation method,
catalyst pellet shape is an important factor in achieving
maximum activity for minimum pressure drop.
In this work, the effect of catalyst pellet shape on the
performance of the catalyst and steam reforming reactor
have been studied. An industrial terrace wall steam-reforming reactor has been considered. Terrace wall arrangement is
one of the most common arrangements for methane-steam
reforming reactor furnaces. The heat transfer model of
terrace wall furnace has been coupled with a chemical
reaction model to predict the behaviour of an industrial
steam-reforming reactor3. Assuming a xed mass of an
industrial catalyst with different physical shapes, performance of the reactor has been studied.

(i) allow possible nearly full conversion of the hydrocarbon feed and a close approach to equilibrium for
the methane steam reforming reaction at the reformer
exit;

REACTOR MATHEMATICAL MODEL

Figure 1 shows a schematic diagram of the terrace wall
reactor. Rows of vertical reaction tubes packed with Ni
383

384

SOLTAN MOHAMMADZADEH and ZAMANIYAN

Figure 2. Schematic diagram of a section of furnace wall and burners.

Figure 1. Schematic diagram of a terrace wall methane-steam reforming

reactor.

based reforming catalyst are installed in the furnace. Rows of

burners in contact with the furnace walls supply the heat
required for the endothermic reaction. Reactor feed consisting mainly of methane and steam enters the reactor tubes
from the top. To model heat transfer in the furnace; furnace
wall, tube skin, process gas and combustion gases are divided
into different zones, and in each zone heat transfer governing
equations are set-up. Establishing the energy balances for the
four types of heat transfer areas, a set of simultaneous
nonlinear equations are obtained. Solving the set of nonlinear
equations by iterative methods, temperature distribution on
the furnace walls, tubes skin, reaction gas inside the tubes
and combustion gas in the furnace are obtained4,5.
To simplify the complex and coupled phenomena of heat
transfer and reaction into a mathematical model, a number
of simplifying assumptions are made. These assumptions
provide a tractable model without sacri cing accuracy. The
major assumptions are:
(1) due to the large number of burners, high momentum of
the combustion gases and good mixing in the furnace,
the temperature and chemical composition of the
combustion gases in the furnace is uniform;
(2) combustion gases, furnace wall and reaction tubes are
grey bodies;
(3) elements of the furnace wall and reaction tubes are
isothermal.
Considering the above generally acceptable assumptions, the
governing equations of different heat transfer zones are
established.
Figure 2 shows a burner and the associated furnace wall.
With all the heat transfer components into the element (m,n)
an energy balance dictates that (see Figure 3):
1 0
1
Conduction
Convection
@ from neighbouringA @ from combustion A
elements
gases
0
1 0
1
Radiation from all
Heat lost
A0
@ the elements of tube A @
to the
and combustion gases
surroundings

Figure 3. An element of furnace wall.

That is:
qm1;n qm;n1 qm1;n qm;n1

qconv qR qg1 qlost 0

qg1

J1 Eb1
R1

!
IT
Eb2;n
Eb1 nX
J1

AD
R1
R R3;n
n1 2;n

The conduction, convection and heat loss terms were

derived from basic heat transfer equations. Convective
heat transfer coef cients inside and outside the furnace
were predicted using empirical correlations6. The radiation
terms were derived by an electric analogy and the shape
factors were determined using the Cartesian coordinates of
all the furnace wall and tube skin elements.
Figure 4 shows the electric analogy of an element of the
furnace wall, for which:
qR

er 1
4

Dx Dy eg sTg4 eg sTm;n
2

nX
IT
1
1
AD

R1 n1 R2;n R3;n

Trans IChemE, Vol 80, Part A, May 2002

CATALYST SHAPE AS A DESIGN PARAMETER

385

modes of heat transfer into the tube wall element, an energy

balance can be set up8.
0
1 0
1
Radiation from all
Convection from
B
C B
C
@ elements of furnace A @ combustion gas to A
wall and flame
tube elements
0
1 0
1
Heat transfer from
Radiation from
B
C B
C
@ tube skin to process A @ combustion A 0
fluid and catalyst
gas

11

Figure 4. Electric analogy of an element of furnace wall.

That is:

IT is the number of elements of the tubes. The resistances

and emissive powers are7:
1e
R1
A1 er

R2

1
A1 F12 1 eg

R3

1 ew
A2 ew

4
Eb1 sTm;n

Eb2 sTw4

Using correlations for ame height, actual dimension of

the burner ame has been taken into account. A correlation
for the ratio of ame visible height to nozzle diameter has
been given as7:

Tf
Mg 0:5
L 5:3

C
1 C
D
C
ATn
Mn

Trans IChemE, Vol 80, Part A, May 2002

12

qin U Ain Tpg Tw N

13

dti
d
1
1
ln te
U 2l
dti
a

14

a is the convective heat transfer coef cient in the catalyst

bed which is determined using correlations developed by Xu
and Froment4.
Similar to the furnace wall elements an electric analogy
approach was used for radiation terms. Figure 6 shows a
schematic diagram of an element of reaction medium inside
the reactor tubes. An energy balance of heat transfer in the
element is:
0
1 0
1
Energy
Energy
B
C B
C
@ transported A @ transferred from A
by fluid
tube skin
0
1
Heat
B
C
@ generated by the A 0
chemical reactions

15

That is:

10

As a boundary condition, temperature of the elements

adjacent to the ame is assumed equal to the adiabatic
ame temperature.
As in Figure 5, an element of the reaction tube wall with a
length of DZ has been considered. Considering different

Figure 5. Schematic diagram of an element of reaction tube.

qR qconv qin qg2 0

T To
UAin Tw i
rg Cp us As Ti To
2
rB As DzSDHi ri Zi 0

16

Figure 6. Schematic diagram of an element of reaction medium inside the

tube.

386

SOLTAN MOHAMMADZADEH and ZAMANIYAN

Energy balance for the element of combustion gases is in the

form of:
0

Net heat transfer

B
B by radiation
B
B
@ with all the

Heat transfer

C B
C
C B by convection C
CB
C
C B
C
A @ with all the A

surfaces
surfaces
0
1 0
1
Energy released
Net energy
B
C B
C
B from combustion C B transfer C
CB
C0
B
B
C B
C
@ of fuel in the
A @ with the A
burners

17

gas

Methane-steam reforming reactions occur inside the reaction tubes. Small amount of heavier hydrocarbons are
assumed to be converted to methane at the entrance to the
reformer by hydrocracking reactions, and their methane
equivalent is considered as the hydrocarbon feed to the
reformer. The main reactions occurring in the reactor are
methane-steam reforming (I), water gas shift (II) and
methanation (III) reactions.
CH4 H2 O $ CO 3H2
CO H2 O $ CO2 H2
CH4 2H2 O $ CO2 4H2

I
II
III

18

Form and constants of the intrinsic reaction rates are

taken from the results of Xu and Froment9.
r1

k1
PH2:52

PCH4 PH2 O

PH3 PCO

K1

1KCO PCO KH2 PH2 KCH4 PCH4 KH2 O PH2 O =PH2 2

k
r2 2
PH2

P P
PCO PH2 O H2K CO2
2

1KCO PCO KH2 PH2 KCH4 PCH4 KH2 O PH2 O =PH2 2

k3
r3 3:5
PH2

P4H PCO2
2
2
PCH4 PH2 O K
3

1KCO PCO KH2 PH2 KCH4 PCH4 KH2 O PH2 O =PH2 2

19

Rate of formation of each component is:

rCO r1 r2

rCO2 r2 r3
rCH4 r1 r3

20

A one-dimensional heterogeneous model has been used for

chemical reaction in the reaction tubes. Since only two of

the three reactions in equation (18) are independent, two

continuity equations for CH4 and CO2 are:
dxCH4 Ar rB ZCH4 rCH4

dz
FCH
4
21
dxCO2 Ar rB ZCO2 rCO2

dz
FCH
4
Transport resistances for each reaction have been taken into
account by introducing effectiveness factor:
r
ri Ps rs dV
V
0

Zi
22
ri Pss rs
To nd the pro le of partial pressure of each component
in the catalyst pellet, planar geometry has been used in
solution of continuity equations for CH4 and CO2. Since the
gradients of partial pressure of all components are limited to
a very thin layer near the surface, using planar geometry is
justi ed. This is con rmed by checking the values of
effectiveness factors for the three reactions in equation (18).
Parallel cross-linked pore model with uncorrelated pore
size distribution and orientation has been used in evaluation
of effective diffusivity of components in the catalyst pellet4.
Pressure drop in the reactor is found using a momentum
equation. Changes in physical properties of uids have
also been taken into account1012.
CATALYST SHAPE
In the steam-reforming reaction, transport resistances
have a large effect on the catalyst performance, leading to
effectiveness factors much less than unity. Since only a thin
layer of the catalyst is involved in the reaction, only the
exposed surface of the catalyst pellet plays an important role
in its effectiveness. Catalysts with desired mechanical and
hydrodynamic characteristics and high exposed surface area
are desirable for this reaction. To compare the different
shapes of catalyst pellet, a reference spherical shape has
been considered. A number of catalyst particle shapes with
the following properties have been considered:
ease of manufacturing with conventional techniques;
good mechanical strength;
high external surface area.

For comparison, in addition to the reference sphere, a

cylinder, a single channel cylinder (ring), a multi-channel
cylinder and a multi-channel cube have also been considered
(refer to Figure 7).
Since, using catalyst particles with the same volume but
different shapes changes the void fraction of the catalyst bed
and length of the reaction tube, a xed mass of catalyst has
been taken as a basis.
A summary of the main characteristics of the reaction
tubes packed with a xed mass of catalysts of different
shapes has been given in Table 1. In Table 1, the shape
factor fs is de ned as:

surface of sphere
fs
surface of particle of same volume
Void fraction of the bed for each shape is estimated based on
the catalyst pellet shape factor13.
Trans IChemE, Vol 80, Part A, May 2002

CATALYST SHAPE AS A DESIGN PARAMETER

387

Figure 7. Various shapes of the catalyst pellets.

NUMERICAL SOLUTION
After setting up the governing equations of heat transfer
and chemical reactions in the reactor furnace, the physical
domain is discretized into a set of grids for numerical
calculations. To reduce the number of equations, on the
furnace wall, grids were generated on an area enclosing one
burner. The whole length of a reaction tube has been
discretized. The nite difference method was used to
convert the rst order heat conduction differential equations.
An iterative numerical method of solution has been
adopted for solving the set of simultaneous nonlinear
equations. To initiate the calculations, seed values of
temperature were assumed for the combustion gas in the
furnace and grids on the furnace wall, tube skin and
the process gas inside the reaction tubes. The initial composition of the reaction mixture and pressure distribution

inside the tubes and physical properties were obtained.

A system of M nonlinear equations and M unknowns was
set up.

Number of combustion

wall elements
gas elements

Number of
Number of

tube elements
process gas elements
Number of

The Newton method was used to solve the set of equations to obtain new values for the temperatures and compositions. To obtain a reliable solution with a reasonable
computation time, stability and convergence of the solution
were checked by changing the grids size and initial values.

Table 1. Main characteristics of the catalyst pellet shapes.

Shape
Sphere
Cylinder
Single channel cylinder
Multi-channel cylinder
Multi-channel cube

Dimension
D
HD
HD
D 2di
HD
D 4di
L 6di

Shape factor,
fs

Void fraction,
e

Tube length,
L (m)

1.0
0.8736
0.5769

0.37
0.40
0.60

7.50
7.85
11.77

0.4555

0.65

13.46

0.4051

0.70

15.70

For case studies, the numerical value of H and D were 17 mm.

Trans IChemE, Vol 80, Part A, May 2002

388

SOLTAN MOHAMMADZADEH and ZAMANIYAN

Table 2. Characteristics of the feed, catalyst and the reformer reactor.

Characteristics of the feed:
Temperature, K
Pressure, bar
Flow rate of gas per tube,
Nm3 hr 1
Flow rate of steam per
tube, Nm3 hr 1
Rate of equivalent CH4
feed, kmol hr1
Feed molar ratio
Characteristics of the catalyst:
Catalyst type
Mass of catalyst, kg
Shape
Dimensions, mm
Density, kg m3
Thickness of active
layer, mm
Thermal
conductivity, w m 1 K1

811.2
21.5
53.9
405.1
3.1
H2O=CH4 5.48 CO2=CH4 0.16
H2=CH4 0.14 N2=CH4 0.01
Ni=MgO Al2O3
18630
single channel cylinder (ring)
L(17) ID(6) OD(17)
2355.5
2

Performance of the model was veri ed by simulation of

the performance of a terrace wall steam-reforming reactor in
one of the Iranian re neries. The main characteristics of the
feed, catalyst and the furnace are summarized in Table 2.
In Table 3 results of the simulation run are compared with
the measured values of temperature and composition in
the reactor. Good agreement between the measured and
calculated values is an indication of the good performance
of the model. Consistent results of the parametric studies
and sensitivity analysis of the model proved its reliable
performance3.
RESULTS AND DISCUSSION
Data on tube wall temperature pro le is one of the most
important factors in design and safe operation of a steamreforming reactor. This data is useful in rating and determining the suitable material of construction for reaction tubes at
the design phase. During the operation, information on the
maximum skin temperature, hot spot locations and points
with a large thermal gradient are crucial for a safe operation.
After a change in operating parameters or before implementing modi cation in reactor design or con guration it is
important to evaluate their effect on reactor tube maximum
temperature. Figure 8 shows the skin temperature pro le of
a reaction tube and the reacting medium inside. Figure 8

8.6

Characteristics of the furnace:

Number of tubes
Number of burners
Tube ID, m
Tube OD, m
Total tube length, m
Heated tube length, m
H1, m
H2, m
H3, m
y,
Centre to centre tube distance, m
Tube centre to burner centre distance, m
Rectangular burner size, m m

MODEL VERIFICATION

138
96
0.124
0.153
13.31
11.74
4.95
4.71
1.56
84.35
0.22
1.25
0.32 0.56

Table 3. Comparison of the model prediction with plant data.

Parameter
Reactor outlet temperature, K
Reactor outlet pressure, kPa
Maximum tube skin temperature, K
Product ow rate, kgmol hr1
H2
CO
CO2
N2
CH4
H2O

Plant

Model

% Error

1117.2
1925
1227.2

1135.8
1928
1210.6

1.7
0.2
1.4

10.58
1.60
1.75
0.02
0.26
12.97

10.66
1.66
1.82
0.02
0.25
12.48

0.8
3.8
4.0
0
3.8
3.8

Figure 8. Pro le of tube skin temperature and reaction medium temperature.

Figure 9. Rate of the three reactions in equation (18) as a function of the

reaction tube length.

Figure 10. Effectiveness factor for the three reactions in equation (18) as a
function of the reaction tube length.

Trans IChemE, Vol 80, Part A, May 2002

CATALYST SHAPE AS A DESIGN PARAMETER

389

Figure 11. Effectiveness factor of steam-reforming reaction for different

catalyst pellet shapes.

Figure 14. Conversion of CH4 at the reactor outlet for different catalyst
pellet shapes.

shows that there is a relatively constant temperature difference between the tube wall and the reaction uid. At the
length of about 6 m there is a distinct change in the uniform
pattern mainly due to the second row of burners at this
location.
Figure 9 shows the rate of the three reactions in equation
(18), and the net rate of formation of CH4, CO and CO2. In
the entrance zone of the reaction tube (i.e. Z < 0.7 m), the
endothermic effect of reaction I and III lowers the reaction
gas temperature, leading to a decrease in the net rate of
reaction I and III. In this region the rate of exothermic
reaction II increases. Beyond this point, the rate of reaction I
and III increases, although this increasing trend is attened

near the end of reactor. Although the temperature of reaction

gas increases along the reaction tube, the gas composition
approaches equilibrium, leading to decreasing reaction rate.
The rate of reaction II shows a sign change at the reactor
length of about 4 m. This indicates a change in the direction
of this reaction. Due to the near-equilibrium chemical
composition and low reaction tube temperature, there is a
sharply decreasing trend of reaction rate near the end of the
reaction tubes.
Figure 10 shows the effectiveness factor for the three
reactions in equation (18) as a function of the reaction tube
length. Generally, the effectiveness factors of reaction I and
III are less than 0.03 in the reactor. The effectiveness factors
of both reaction I and III increase slightly near the entrance
of the reaction tubes, then there is a slightly decreasing
trend. The effectiveness factor of reaction II experiences a
discontinuity and sign change at about 4 m of the tube
length, due to the changing direction of the reaction.
A higher effectiveness factor translates to a more ef cient
use of a mass of catalyst. The Effectiveness factor of
reaction I has been plotted for the ve catalyst pellet
shapes in Figure 11. The general trends are the same for
all shapes. The sphere has the lowest effectiveness factor
while the multi-channel cube has the highest. Figure 11
shows that the effectiveness factor of the sphere and cylinder
fall into one category, while, single channel cylinder, multichannel cylinder and multi-channel cube, fall into another
category with a higher effectiveness factor.
Pressure drop is a key parameter in the steam-reforming
reactor. On one hand, low pressure drop is a desirable
characteristic of a particular catalyst shape, because of a
lower mechanical energy loss. On the other hand, high
pressure drop indicates a high turbulence and a high heat
transfer coef cient. A high heat transfer coef cient, leads to
a high reaction rate, a more active catalyst, lower reaction
tube wall temperature and longer operation life of the tubes.
Figure 12 shows the pressure distribution along the reaction
tube for different shapes. Sphere and cylinder fall into the
high pressure drop category, while the single and multichannel cylinder and multi-channel cube fall into the low
pressure drop category.
In Figure 13 the pro le of the temperature of the reaction
gas inside the tubes have been compared for different shapes
along the reaction tube. The general trend shows a slight
decrease in temperature, due to the cooling caused by the
endothermic reactions. The temperature then increases
monotonically, because of the decreased endothermic reac-

Figure 12. Pressure pro le along the reaction tube for different catalyst
pellet shapes.

Figure 13. Temperature pro le of the reaction gas inside the reaction tubes.

Trans IChemE, Vol 80, Part A, May 2002

390

SOLTAN MOHAMMADZADEH and ZAMANIYAN

tions rate, decreased heat absorption and nearly constant

heat ux. Figure 13 shows that at any length of the reaction
tube, the gas temperature is lower for multi-channel cube,
multi-channel cylinder and single channel cylinder, while
the temperature for sphere and cylinder is higher. For
example at a length of 5 m, the difference in temperature
for multi-channel cube and sphere is about 60 K. More
active catalysts lead to lower gas temperature and lower
tube skin temperature which increases reaction tube operating life span.
For a given mass of catalyst but with different pellet
shapes, the conversion of CH4 at the reactor outlet is shown
in Figure 14. Spherical and cylindrical pellet shapes give a
conversion of about 70%, while the highest conversion of
about 85% is attainable with the multi-channel cube and a
slightly lower conversion (ca. 83%) is obtainable with multichannel cylinder. Single channel cylinder leads to a conversion of about 80%.
Considering the effect of catalyst shape on its pressure
drop, heat transfer and kinetic characteristics, it seems
that the multi-channel cube is the most desirable followed
by the multi-channel cylinder, single channel cylinder,
cylinder and nally sphere. In addition to these parameters,
a desirable catalyst should be easy and practical to
manufacture, handle and operate. It should also have good
mechanical strength and attrition resistance. Considering
these factors, the multi-channel cube has sharp edges,
which can break and produce nes and particles, plugging
the catalyst voids, leading to excessive pressure drop
and uneven ow distribution. These can cause hot spots,
hot bands and hot tubes. Multi-channel cylinder and single
channel cylinder seem to be the most suitable shapes.
Comparing these two shapes, it seems that they have similar
tendencies for attrition and ne production, while the
mutichannel cylinder can have higher mechanical strength
and a lower tendency for breakage and particle formation, in
addition to its better effectiveness and temperature and
pressure distribution.
CONCLUSION
A mathematical model for simulation of a catalytic
terrace wall methane-steam reformer has been developed.
The reliability of the complete model has been veri ed by
comparing its predictions with actual plant data. The effect
of catalyst pellet shape on performance of the reactor has
been studied. Among different catalyst pellet shapes of
sphere, cylinder, single channel cylinder, multi-channel
cylinder and multi-channel cube, the multi-channel cube
has the highest effectiveness, lowest pressure drop and
lowest reaction gas temperature. Although the multi-channel
cylinder is the second choice in terms of these parameters, it
has better loading characteristics and higher attrition resistance, and hence, is the preferred shape for catalyst pellets.

NOMENCLATURE
A
Ain
As
A1
A2
Cp

molar ratio of fuel to combustion products with theoretical air

inner surface area of the tube element, m2
cross sectional area of the tube, m2
surface area of the furnace wall element, m2
outer surface area of the tube element, m2
heat capacity of process gas, kJ kg1 K1

C
D
dte
dti
Eb1
Eb2
F12
DHi
J1
k1
K1
L
Mn
Ms
N
Pi
qconv
qg 1
qg 2
qin
qlost
qm1,n
qR
R1
R2
R3
Ri
Tf
Tg
Ti
Tm,n
Tn
To
Tpg
Tw
U
us
Dx
Dy
Dz

molar fuel to air ratio

burner nozzle diameter, m
tube external diameter, m
tube internal diameter, m
black emmisive power of the furnace wall element, W m 2
black emmisive power of the tube wall element, W m2
shape factor of the furnace wall element with respect to tube
element,
enthalpy of reaction i, kJ kgmol 1
radiosity of furnace wall element, W m2
reaction rate constant
reaction equilibrium constant
visible ame height, m
molecular weight of fuel, kg kgmole1
molecular weight of surrounding gasl, kg kgmole 1
number of tubes
partial pressure, kPa
convection heat transfer between gas and surface, W
radiation heat transfer between combustion gas and furnace wall
element, W
radiation heat transfer between combustion gas and tube wall
element, W
heat transfer from tube wall to process uid, W
heat lost from furnace wall element to surroundings, W
conduction heat transfer between element (m 1, n) and element
(m,n) of the furnace wall, W
radiation heat transfer between furnace elements and tube wall
elements, W
surface resistance of furnace wall element, m2
space resistance between furnace element and tube element,
m 2
surface resistance of tube element, m2
rate of reaction i, kgmol kgcat 1 h 1
adiabatic ame temperature, K
temperature of combustion gases, K
temperature of process uid entering a tube element, K
temperature of furnace wall element, K
fuel temperature, K
temperature of process uid leaving a tube element, K
temperature of process uid, K
temperature of external surface of tube wall element, K
overall heat transfer coef cient, kJ m 2 h 1 K 1
super cial velocity of process uid, m h 1
length of furnace wall element in x direction, m
length of furnace wall element in y direction, m
length of furnace wall element in z direction, m

Greek letters
a
convective heat transfer coef cient in catalytic
kJ m2 h1 K1
eg
emissivity of combustion gas
er
emissivity of furnace wall
ew
emissivity of tube gas
Zi
effectiveness factor of reaction i
l
conductivity of tube, kJ m1 h 1 K 1
rb
density of catalytic bed, kgcat m3
bed
rg
density of gas, kg m3
s
Stephan-Boltzman constant, 5.669 E 8 W m2 K 4

bed,

REFERENCES
1. Adris, A. M., Pruden, B. B., Lim, C. J., and Grace, J. R., 1996, On
reported attempts to radically improve the performance of the steam
methane reforming reactor, Can J Chem Eng, 74: 177186.
2. Rostrup-Nielsen, J. R., Christiansen, L. J. and Bak Hansen, J. H., 1988,
Activity of steam reforming catalysts: Role and assessment, Appl Cat,
43(287): 287303.
3. Soltan Mohammadzadeh, J. S. and Zamaniyan, A., 2000, Simulation of
terrace wall methane-steam reforming reactors, Iran J Sci Tech.
4. Xu, J. and Froment, G. F., 1989, Steam reforming, methanation and
water-gas shift reaction: IIDiffusion limitation and reactor simulation,
AIChE J, 35(1): 97103.
5. Rao, M. V. R., Plehiers, P. M. and Froment, G. F., 1988, The coupled
simulation of heat transfer and reaction in a pyrolysis furnace, Chem
Eng Sci, 43(6): 12231229.
6. Holman, J. P., 1992, Heat Transfer (Mc-Graw Hill Inc, New York, USA).

Trans IChemE, Vol 80, Part A, May 2002

CATALYST SHAPE AS A DESIGN PARAMETER

7. Kudo, K., Taniguchi, H. and Guo, K., 1992, Heat transfer simulation in
a furnace for steam reformer, Heat Trans Japanese Res, 20(8): 750764.
8. Hottle, H. C. and Saro m, A. F., 1967, Radiative Transfer (Mc-Graw
Hill Inc, New York, USA).
9. Xu, J. and Froment, G. F., 1989, Methane steam reforming, methanation
and water-gas shift: I-Intrinsic kinetics, AIChE J, 35(1): 8896.
10. Fogler, H. S., 1992, Elements of Chemical Reaction Engineering
(Prentice-Hall International Inc, Toronto, Canada).
11. Reid, R. C., Prausnitz, J. M. and Poling, B. E., 1988, The Properties of
Gases and Liquids (McGraw-Hill Inc, Toronto, Canada).
12. Bird, R. B., Stewart, W. E., and Lightfoot, E. N., 1966, Transport
Phenomena (John Wiley and Sons Inc, London, UK).
13. Kunii, D. and Levenspiel, O., 1991, Fluidization Engineering (Butterworth-Heinemann, New York, USA).

Trans IChemE, Vol 80, Part A, May 2002

391

ADDRESS
Correspondence concerning this paper should be addressed to Professor
J. S. Soltan Mohammadzadeh, Department of Chemical Engineering,
Sahand University of Technology, Tabriz, 51335-1996, I.R. Iran.
E-mail: sultanm@hotmail.com
The manuscript was received 2 August 2001 and accepted for publication after revision 7 January 2002.