i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

Available at www.sciencedirect.com

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

A novel integrated, thermally coupled fluidized bed

configuration for catalytic naphtha reforming to enhance
aromatic and hydrogen productions in refineries

M.R. Rahimpour*, R. Vakili, E. Pourazadi, D. Iranshahi, K. Paymooni

Chemical Engineering Department, School of Chemical and Petroleum Engineering, Shiraz University, Shiraz 71345, Iran

article info abstract

Article history: In the recent years, refineries have focused on developing new ways to gain more from
Received 17 August 2010 their asset utilization owing to increasing demand for high octane gasoline. In this regard,
Received in revised form a thermally coupled fluidized bed naphtha reactor (TCFBNR) is proposed in this study. The
23 November 2010 first and the second reactors of a conventional catalytic naphtha reactor configuration (CR)
Accepted 27 November 2010 are substituted by thermally coupled fluidized bed reactors. In this novel configuration,
Available online 31 December 2010 naphtha reforming reactions which are highly endothermic are coupled with the
exothermic hydrogenation of nitrobenzene to aniline. Some drawbacks of CR such as
Keywords: pressure drop, internal mass transfer limitation and radial gradient of concentration and
Naphtha reforming temperature are successfully solved in this novel configuration. In addition to some
Thermally coupled reactor mentioned advantages of this novel configuration, TCFBNR configuration enhances the
Aromatics enhancement aromatic production rate about 20.54% and 7.13% higher than CR and TCNR, respectively.
Hydrogen production Also, the TCFBNR is capable to enhance hydrogen production rate in the shell side, the
Fluidized bed reactor aniline flow rate in the tube section and simultaneously improves the thermal behavior of
Nitrobenzene hydrogenation endothermic side and reduces the undesirable temperature drop. The modeling results of
TCFBNR is compared with the results of CR and thermally coupled fixed-bed naphtha
reactor (TCNR). These studies provide a good initial insight for some modifications and
revamping of the old facilities with more efficient ones.
ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.

1. Introduction is a better one to be revamped and improved owing to its

Naphtha and reformate are complex mixtures of paraffins,
naphthenes and aromatics in the C5eC12 range [1]. Catalytic
naphtha reforming is extensively practiced in the petroleum-
refining to produce high octane gasoline and hydrogen for
frequently use in the hydro-treatment processes to improve the
feedstock [1,2]. The process of hydrogenation, dehydrogenation
and isomerization have all benefitted from the catalyst, reactor
and feed treatment technologies invented for catalytic reform-
ing processes [1]. Additionally, catalytic naphtha reforming unit
considerable effect on overall refinery profits [3]. The above
descriptions have motivated a variety of research on the cata-
lytic naphtha reforming such as catalyst preparation and coking
[4e9], reaction kinetics [10e15], optimization [16e21] and
improvement of reaction condition [22e26].
According to the previous studies, two problems have been
found in the catalytic reforming process which should be
initially addressed by researchers and then by process licen-
sors. The first problem is a carbon deposition on the catalyst
surfaces which causes catalyst deactivation and catalytic bed

* Corresponding author. Tel.: þ98 711 2303071; fax: þ98 711 6287294.
E-mail address: rahimpor@shirazu.ac.ir (M.R. Rahimpour).
0360-3199/$ e see front matter ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2010.11.112
2980 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

Fig. 1 e A simplified process for conventional catalytic naphtha reforming (CR).

clogging [27] while the second one is the increasing demand Innovations in catalytic reactors, which constitute the heart
for high octane gasoline [1]. These problems would be of process technologies, are often the preferred starting point.
addressed by simultaneous application of fluidization and In this way, multifunctional auto-thermal reactor is a novel
process intensification concepts (using thermally coupled concept in process intensification. At present, a promising
fluidized bed reactors). field of using multifunctional auto-thermal reactors is the
coupling of endothermic and exothermic reactions. In this
1.1. Fluidized bed reactor type of reactors, an exothermic reaction is used as the heat
producing source to drive the endothermic reaction(s) [38,39].
The application of fluidized bed reactors has gained wide
interest in the chemical and petroleum industries [28]. The
main advantages of fluidized bed reactors are: Table 1 e Specifications of conventional naphtha reactor,
feed, product and catalyst of plant for fresh catalyst.
 Negligible pressure drop. Parameter Numerical value Unit
 More effective use of catalyst owing to the possibility of
3
applying smaller catalyst particles and vigorous gasesolid Naphtha feedstock 30.41  10 kg/h
Reformate 24.66  103 kg/h
contact [29,30] (very small catalyst particles cannot be
H2/HC mole ratio 4.73 e
applied in fixed-bed reactors owing to plugging and high LHSV 1.25 h1
pressure drop). Mole percent of hydrogen in recycle 69.5 e
 More compact design [31]. Diameter and length of 1st reactor 1.25, 6.29 m
 The possibility of using inexpensive metal alloys for reactor Diameter and length of 2nd reactor 1.67, 7.13 m
vessels (due to lower operating temperature) [32]. Diameter and length of 3rd reactor 1.98, 7.89 m
 More effective temperature controlling and prevention from Distillation fraction of naphtha feed and reformate
catalyst destruction owing to rapid mixing of solids in TBP Naphtha feed Reformate
fluidized bed reactors [30]. ( C) ( C)
 Continuous or periodic catalyst replacement [30,32]. IBP 106 44
10% 113 73
30% 119 105
A large number of theoretical and experimental studies 50% 125 123
have been recently performed on the fluidized bed reactor 70% 133 136
configurations [33e36]. 90% 144 153
FBP 173 181
1.2. Process intensification Typical properties of catalyst
dp 1.2 mm
Process intensification (PI) is currently one of the most Pt 0.3 wt%
significant trends in chemical engineering and process tech- Re 0.3 wt%
nology. It is attracting more and more attention of the sa 220 m2/g
rB 0.3 kg/L
research world [37]. It is the strategy of reducing environ-
3 0.36 
mental emissions, energy and materials consumption.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1 2981

Fig. 2 e A schematic concept for thermally coupled fluidized bed naphtha reactor (TCFBNR).

In the last years, auto-thermal reactors have been converted
to important topic as many articles are in the literature
[40e43]. In order to enhance the aromatic production rate,
thermal coupling of the endothermic reaction (naphtha
reforming) and exothermic hydrogenation of nitrobenzene to
aniline in TCFBNR configuration are proposed in this study.
2. Process description
2.1. Conventional naphtha reforming (CR)
A schematic process diagram of conventional catalytic
naphtha reforming is shown in Fig. 1. Since naphtha
reforming reactions are highly endothermic, temperature and
consequently the reaction rate decrease along the reactor.
Therefore, three adiabatic reactors are utilized and inter stage
furnaces are assisted before each reactor. In order to produce
high octane gasoline from straight run naphtha, a bimetallic
platinumerhenium catalyst is used. The H2/HC ratio is one of
the restricted parameters in the controlling unit of the
naphtha reforming process. It is adjusted in accordance with
the inlet amount of feed to the first reactor. In industry, it is
recommended to maintain H2/HC in the range of 4e6 [44]. If
the H2/HC reduces, the catalysts will be subjected to coking
and a rapid deactivation. It imposes a huge burden on cata-
lysts and jeopardizes the catalyst life. Subsequently, the fresh
naphtha feed is sprayed into a recycled stream containing

Table 2 e Operating conditions for hydrogenation reaction of nitrobenzene to aniline.

Parameter Numerical value Unit

Inlet temperature 780 K

Inlet pressure 1.1  105 Pa
Catalyst density 1400 kg/m3
Diameter of catalyst particle 1.2  103 m
Void fraction 0.46 

1st reactor
Available cross-sectional area of the tube side for flow 3.31 m2
Nitrobenzene molar flow rate 7.32 mol/s
Hydrogen molar flow rate 29.5 mol/s
Aniline molar flow rate 0.72 mol/s
Steam molar flow rate 72.46 mol/s

2nd reactor
Available cross-sectional area of the tube side for flow 3.31 m2
Nitrobenzene molar flow rate 7.32 mol/s
Hydrogen molar flow rate 29.5 mol/s
Aniline molar flow rate 0.72 mol/s
Steam molar flow rate 72.46 mol/s
2982 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

60e90% hydrogen (by mole) and then heated and enters the
Table 3 e Rate constants and heat of reactions for
top of the first reactor. The outlet stream from the third naphtha reforming.
reactor enters a flash separator at desired condition and
k ¼ AexpðB  ðE=1:8TÞÞ A B E DH298K
separated into the gaseous and liquid product. The liquid
product is called reformate which mainly contains aromatics. kf1 9.87 23.21 36,350 71,038.06
The operating conditions and catalyst specification for kf2 9.87 35.98 58,550 36,953.33
kf3 1 42.97 63,800 51,939.31
conventional system are presented in Table 1.
kf4 1 42.97 63,800 56,597.54
Ke1 1.04  103 46.15 46,045 
Ke2 9.87 7.12 8000 
2.2. Thermally coupled fluidized bed naphtha reactor
(TCFBNR)


Fig. 2 shows a conceptual process diagram for TCFBNR kf 2  
r2 ¼ ke2 pn ph  pp (6)
configuration. In TCFBNR configuration, three fluidized bed Ke2
reactors are fabricated meanwhile hydrogenation reaction of
nitrobenzene to aniline is coupled with naphtha reaction in

kf 3
the first and the second reactors. In the first and the second r3 ¼ pn (7)
pt
reactors of coupled configuration, the endothermic naphtha
reactions take place in the shell side (the endothermic side)

kf 4
while the exothermic hydrogenation reaction of nitrobenzene r4 ¼ pp (8)
pt
provides the required heat of naphtha reactions in the tube
side (the exothermic side). The hydrogenation reaction of where kf i and Kei are forward rate constant and equilibrium
nitrobenzene to aniline takes place on the commercial constant, respectively. The reaction rate constants (kf i ), equi-
spherical PdeAl2O3 (1.1 wt %) in the exothermic side. The librium constants (Kei ), activation energies (Ei) and standard
specific properties and operating conditions of the ex- heat of reactions (DH298K) are tabulated in Table 3 [45].
othermic side is presented in Table 2. The applied molar flow
rates in the exothermic side are similar to ones used by Abo- 3.2. The exothermic side
Ghander et al. [43].
Over the last 145 years, aniline has become one of the hundred
most important building blocks in chemistry. Aniline is
known to be used in more than 300 different end products
such as isocyanates, rubber processing chemicals, dyes and
3. Reaction scheme and kinetic expressions
pigments, agricultural chemicals and pharmaceuticals [46].
Nitrobenzene is used as a raw material for aniline production
3.1. The endothermic side
by almost entire world producers, using fixed-bed or fluidized
bed vapor phase reactors [46,47].
Considering the proposed model by Smith [10], four dominant
The hydrogenation of nitrobenzene to aniline in the tube
reactions are used to describe the catalytic reforming process,
side can be expressed by:
the catalytic reforming system is significantly simplified. The
reactions are as follows:
Dehydrogenation of naphthenes to aromatics: C6 H5 NO2 þ 3H2 /C6 H5 NH2 þ 2H2 O (9)

Naphthenes(CnH2n) 4 Aromatics(CnH2n6) þ 3H2 (1) The rate expression provided by Klemm et al. [48] is as
follows:
Dehydrocyclization of paraffins to naphthenes:
k0 KNB KH2 PNB P0:5
H2
r¼ 2 (10)
Naphthenes(CnH2n) þ H2 4 Paraffins(CnH2nþ2) (2)
1 þ KNB PNB þ KH2 P0:5H2

Hydro-cracking of naphthenes to lower hydrocarbons: The reaction rate constant (k0 ), equilibrium constants (Ki), the
activation energy (E ) and standard heat of reaction (DH298K)
Naphthenes(CnH2n) þ n/3H2 / Lighter ends (C1eC5) (3) are presented in Table 4.

Hydro-cracking of paraffins to lower hydrocarbons:

Table 4 e Rate constants and heat of reaction for

Paraffins(CnH2nþ2) þ (n  3)/3H2 / Lighter ends (C1eC5) (4) hydrogenation of nitrobenzene.
k ¼ A0 expðB0  ðE0 =1:8TÞÞ A0 B0 E0 DH298K
The following rate expressions are described for the above
0
reactions: k 0.186 0 10 443,000

 KH2 4.427  103 0 0 
kf 1  
r1 ¼ ke1 pn  pa p3h (5) KNB 1.51  105 0 0 
Ke1
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1 2983

Fig. 3 e An axial element of thermally coupled fluidized bed naphtha reactor (TCFBNR).

Table 5 e Mass & energy balance equations.

Definitions Equation

TCFBNR
Mass balance for bubble phase   b
dj dFi;j XXX
dj KCtj abj yie;j  yib;j  þ dj gj rs rbi;j;k ¼ 0 (11)
Ac dz j k i

Mass balance for emulsion phase   1  dj dFi;j 

e
 XXX
dj Kbei;j Ctj abj yib;j  yie;j  þ 1  dj re hj rei;j;k ¼ 0 (12)
Ac dz j k i
 
Energy balance for bubble & emulsion phases
Cpj d Ftj Tj   XX   XX  
¼ 1  dj re hj rej;k DHf ;k þ gj dj rB hj rbj;k DHf ;k
Ac dz j k j k
pDi
 UShell ðTTube  TShell Þ (13)
Ac
Boundary conditions z ¼ 0; Fi;j ¼ F0;j ; Tbj ¼ Tej ¼ T0j (14)

TCNR
Mass balance 1 dFi;j
¼ hj rb ri;j;k (15)
Ac dz
Energy balance
 
1 d Ftj Tj pDi XX  
Cp ¼ UShell ðTTube  TShell Þ þ rb hj rj;k DHf ;k (16)
Ac j dz Ac j k

Boundary conditions z ¼ 0; Fi;j ¼ F0;j ; Ti ¼ T0i (17)

Auxiliary correlations  2  2 Ref.

C3 =T C5 =T
Component heat capacity Cp ¼ C1 þ C2 þ C4 [49]
sinhðC3 =TÞ coshðC5 =TÞ
C1 TC2
Viscosity of reaction mixtures m¼ [49]
1 þ ðC3 =TÞ þ ðC4 =T2 Þ
 0:3  2 0:3
hc Dt r mg
Heat transfer coefficient Nu ¼ ¼ 420 s Ret ðPrg Þ0:3 [50]
kg rg gr2s d3p

dP 150m ð1  3Þ2 Q 1:75r ð1  3Þ Q 2

Ergun equation ¼ þ [51]
dz f2s d2p 33 Ac f s dp 33 A2c
2984 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

 Bubble and emulsion phases in TCFBNR are assumed to be

Table 6 e Empirical correlations for the hydrodynamic
parameters of fluidized bed reactor. isotherm.

Parameter Equation Ref.

The related mass and energy balances plus the boundary
 2 
1:75 dp rg umf 150ð1  3mf Þ dp rg umf conditions for TCNR and TCFBNR are presented in Table 5. In
Superficial þ ¼ Ar [52]
33mf 4s m 33mf 4s m
Eqs. (13) and (16), the positive and negative signs are used for
velocity at
minimum endothermic and exothermic sides, respectively. The pressure
fluidization drops in the CR and TCNR configurations are calculated based
d3p rg ðrp  rg Þg
Archimedes Ar ¼ [52] on the Ergun equation [51]. Furthermore, other useful corre-
m2
number
lations for heat transfer and physical properties are consid-
Bubble db ¼ dbm  ðdbm  dbo Þexpð0:3z=DÞ [53]

0:4 ered to solve the set of developed models.
diameter dbm ¼ 0:65 p4 D2 ðu0  umf Þ
dbo ¼ 0:376ðu0  umf Þ 2 The reported empirical correlations of the hydrodynamic

1=2
Mass transfer Kbe ¼ umf
3 þ ð4Dmf 3mf ub =pdb Þ [54] parameters for TCFBNR in Table 6 are extracted from the
coefficient literatures [52e54].
(bubble-
emulsion
phase)
pffiffiffiffiffiffiffiffi
Bubble rising ub ¼ u  um þ 0:711 gdb [53] 5. Numerical solution
velocity
Specific surface ab ¼ 6d=db [52] The developed model is composed of the set of ordinary
area for differential equations (ODEs) related to mass and energy
bubble
conservative rules as well as non-linear algebraic equations of
Volume d ¼ ðu  umf Þ=ub [52]
fraction of
the kinetic model, auxiliary and hydrodynamic correlations.
bubble phase Backward finite difference approximation is applied to solve the
to overall bed set of ODEs. Therefore, the ODEs are changed into a set of non-
Density for re ¼ rp ð1  3mf Þ [52] linear algebraic equations. The length of each reactor is divided
emulsion to 100 separate segments and the GausseNewton method is
phase
used to solve the obtained set of non-linear algebraic equations
in each segment for both sides simultaneously. This procedure
is repeated for all nodes in the reactor so that the results of each
4. Mathematical modeling node are used as the inlet conditions for the following node.

As shown in Fig. 3, an element is considered in the axial

direction of TCFBNR to develop the mass and energy balances. 6. Results and discussion
The following assumptions are made in the modeling of CR,
TCNR and TCFBNR configurations: 6.1. Model validation

 The gas mixture is considered to be an ideal gas (CR, TCNR
and TCFBNR).
 The catalytic bed of TCFBNR is assumed to be composed of
bubble and emulsion phases.
 Reactions occur in the emulsion phase (TCFBNR).
 Axial diffusion of mass and heat are negligible compared to
convection term (CR, TCNR and TCFBNR).
 The case is investigated under steady-state condition (CR,
TCNR and TCFBNR).
In order to investigate the accuracy of the developed model,
the modeling results of conventional naphtha reactor (CR) are
validated with the plant data under steady sate condition in
Table 7. The achieved daily results from plant are also
provided and compared with the model predictions in Table 8
to show the capability of proposed model. Analyses of
components (paraffin, naphthene and aromatic) are per-
formed by PONA test. The PONA test is a GC apparatus which
operates with Helium as a carrier gas. A simplified conceptual

Table 7 e Comparison between model prediction and plant data for fresh catalyst.
Reactor no. Inlet temperature (K) Inlet pressure (kPa) Catalyst distribution (wt %) Input feedstock (Mole %)

1 777 3703 20 Paraffin 49.3

2 777 3537 30 Naphthene 36.0
3 775 3401 50 Aromatic 14.7

Outlet temperature (K) Aromatic in reformate (Mole %)

Plant CR TCNR TCFBNR Plant CR TCNR TCFBNR

1 722 726.1 759.8 761.1  35.1 44.3 45.7

2 753 751.7 798.7 802.5  47.5 61.0 65.5
3 770 770.8 797.3 799.3 57.7 56.0 73.7 86.2
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1 2985

every month based on our data sheets from the domestic

Table 8 e Comparison between predicted production rate
and plant data. refinery [1,44]. The aromatic is tested especially by ASTM 2159
equivalent to UOP 273 method.
Time Naphtha feed Plant Model Devi %
The performances of TCNR and TCFBNR configurations are
(day) (ton/h) (kmol/h) (kmol/h) ðModelPlant
Plant  100Þ
investigated and compared with the CR performance in the
0 30.41 225.9 221.58 1.91 following figures.
34 30.41 224.25 222.64 0.71
62 31.00 229.65 228.18 0.64
97 30.78 229.65 227.09 1.11 6.2. The endothermic side
125 31.22 229.65 231.19 0.67
160 31.22 229.65 231.72 0.90 6.2.1. Paraffin and naphthene
188 28.55 211.6 210.16 0.68
The reactants molar flow rates in CR, TCNR and TCFBNR are
223 30.33 222.75 225.19 1.09
compared in Fig. 5(a) and (b). The paraffin consumption rate
243 31.22 233.05 232.74 0.13
298 30.67 228.65 228.75 0.04 increases considerably in the coupled configurations in
321 30.76 227.64 229.69 0.90 comparison with CR. The difference between the paraffin
398 42.35 317.30 326.18 2.79 molar flow rate profiles in TCNR and TCFBNR is owing to the
425 42.32 317.94 326.12 2.57 efficient heat transfer in the fluidized bed reactor (Fig. 5(a)).
461 42.32 317.94 326.38 2.65 The naphthene molar flow rate is depicted in Fig. 5(b). The
490 42.32 317.94 326.56 2.71
thermal effect of coupled reactors, especially the fluidized bed
524 42.32 313.09 326.77 4.36
configuration, changes the equilibrium conditions and shifts
567 42.54 317.94 328.83 3.42
610 42.54 313.9 329.06 4.82 the reactions to more production and almost complete
717 37.86 286.15 290.70 1.62 conversion (see Fig. 5(b)).
771 38.51 282.10 296.40 5.06
a 140

schematic for related gas chromatogram is described by Fig. 4.

120
As illustrated, the system is composed of three parts: Split
Paraffin molar flow rate (kmole/hr)

injector, Temperature programmed oven and finally the

ionization detector. The analysis time is around one hour and 100
a half. The components are identified according to the peaks
which are created by the GC. In order to detect all the indi-
80
vidual compounds, more complex temperature program and
also time are required. For the process engineer it is often
sufficient to know the PONA group concentrations in order to 60
CR
verify the feedstock or product qualities, and a least time-
TCNR
consuming GC method is chosen. This test is usually taken TCFBNR
40

20
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Length of reactor (Dimensionless)

b 100
90
TCFBNR
TCNR
Naphthene molar flow rate (kmole/hr)

80
CR
70

60

50

6
40
4
30
2
20 0
0.8 0.9 1
10

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Length of reactor (Dimensionless)

Fig. 5 e Molar flow rates for (a) paraffine and (b) naphthene
Fig. 4 e A schematic of PONA test (illustration of GC vs. the dimensionless length for three configurations of CR,
apparatus). TCNR and TCFBNR.
2986 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

2500

2450
a 180
2400

Total molar flow rate (kmole/hr)

160
2350
Aromatic molar flow rate (kmole/hr)

140
2300

120
2250

100 2200 TCFBNR

TCNR
CR
80 2150

TCFBNR 2100
60
TCNR
CR 2050
40 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Length of reactor (Dimensionless)

20
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Fig. 7 e Variation of total molar flow rate in the CR, TCNR
Length of reactor (Dimensionless) and TCFBNR.

b 1550
1500 6.2.2. Aromatics, hydrogen and light ends
Hydrogen molar flow rate (kmole/hr)

As shown in Fig. 6(a), higher aromatic production rate and as

1520 a consequence high octane gasoline can be achieved in TCFBNR
1450
1510 configuration. The aromatic production rate in coupled
configurations is remarkably higher than the one in CR. The
1500
1400 aromatic production rate in the outlet of the second reactor in
1490
0.8 0.9 1 the coupled reactors is obviously higher than the one at the
1350
outlet of the third reactor in CR. On the other hand, as previ-
ously mentioned in Table 1, the catalyst load of the third reactor
equals the sum of the catalyst loads of the first and the second
CR
CTR
1300
TCNR
data2 reactors. Therefore, an enormous reduction in the capital and
TCFBNR
data3 operational costs (Pt metal in the catalyst) is possible by elim-
1250 inating the third reactor in TCNR and TCFBNR setups.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 On the other hand, stricter legislation on sulfur content in
Length of reactor (Dimensionless)
the fuels, and the processing of heavier crude oils in the hydro-
c 800 cracking units cause an incremental demand for hydrogen in

TCFBNR
750 TCNR
Light ends molar flow rate (kmole/hr)

8.5
CR

8 CR
700 TCNR
7.5 TCFBNR
H2/HC molar ratio

650 7

6.5

600
6

5.5
550
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Length of reactor (Dimensionless) 5

Fig. 6 e Products’ molar flow rates of (a) aromatic, (b) 4.5

hydrogen and (c) the light ends along the reactor for CR, 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Length of reactor (Dimensionless)
TCNR and TCFBNR.
Fig. 8 e H2/HC ratio through the reactor for CR, TCNR and
TCFBNR.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1 2987

a 30
b 30

TCFBNR
25 25

Nitrobenzene molar folw rate (Kmol/hr)

TCNR

Aniline molar flow rate (Kmol/hr) 20 20

15 15

10 10
TCFBNR
TCNR

5 5

0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Length of reactor (Dimensionless) Length of reactor (Dimensionless)

c 320 d 110
TCNR
100
TCFBNR
310

Hydrogen molar flow rate (Kmol/hr)

90
H O molar flow rate (Kmol/hr)

300 80

70
290
60

280 50

TCNR
TCFBNR 40
270
30

260 20
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Length of reactor (Dimensionless) Length of reactor (Dimensionless)

Fig. 9 e Changes of (a) aniline, (b) nitrobenzene, (c) water and (d) hydrogen in the exothermic side for TCNR and TCFBNR.

refineries and petrochemical sites which force refiners to the probability of deactivation by coking especially in the third
improve their process efficiency. According to Fig. 6(b), TCFBNR reactor which has the highest temperature profile. However,
can be considered as a proper candidate for hydrogen produc- higher H2/HC molar ratios (i.e., higher hydrogen production in
tion in refineries. The hydrogen production rate in TCFBNR is the reaction side) shifts naphthenes dehydrogenation reac-
18 kmol/h more than the one in CR. The small figure illustrates tion to the reactant side in accordance with the Le Chatelier’s
the superiority of TCFBNR configuration to the other configu- principle which reduces the quality of the reformate. There-
rations owing to achieving higher hydrogen production rate at fore, a hydrogen perm-selective membrane layer can be
the end of the third reactor as a result of applying fluidized bed assisted in TCFBNR to adjust the H2/HC molar ratio at a proper
concept. The off gas production rate is depicted in Fig. 6(c). The value.
off gases are used as a source of LPG and maybe these gases can
be used as a heating source (combustion reaction) for naphtha 6.3. The exothermic side
reactors through coupling.
6.3.1. Molar behavior
6.2.3. Total molar flow rate change in the endothermic side The exothermic molar flow rates are investigated through
The variation of total molar flow rate for TCNR and TCFBNR is Fig. 9(a)e(d). Adding steam to the feed line eliminates the coke
about 16.6% and 18.7%, respectively according to Fig. 7. In the formation on the catalyst, reduces the temperature along the
coupled configurations, the heat capacity, viscosity, density hydrogenation side and maintains the temperature low by
and other physical properties are considered variable in both diluting the reacting mixture on the hydrogenation side [43].
shell and tube sides in order to increase the accuracy of the Since the hydrogenation reaction of nitrobenzene is an irre-
predicted results which are achieved by the developed versible reaction, feed distribution lines are in parallel. A
model. reduction in the inlet feed affects the reaction rates and
consequently decreases the reaction rates. 94.1% and 96.6% of
6.2.4. H2/HC molar ratio nitrobenzene is converted in the first and the second reactors
H2/HC molar ratio as one of the key parameters in naphtha of TCFBNR, respectively. The produced hydrogen in the
reforming process is shown in Fig. 8. Higher H2/HC molar ratio endothermic side can be utilized as feed in the exothermic
along the reactors enhances the catalyst lifetime and reduces side (e.g. using membrane concept in the coupled reactors).
2988 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

810
a 0.08 0.08
Bubble phase ...........
Emulsion phase 800

Temperature of endothermic side (K)

790
0.06 0.06

780
Aromatic mole fraction

P araffin mole fraction

770
0.04 0.04
760

750

0.02 0.02 CR
740
TCNR
TCFBNR
730

0 0 720
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Length of reactor (Dimensionless) Length of reactor (Dimensionless)

b 0.08 0.08 Fig. 11 e Thermal profiles of CR, TCNR and TCFBNR in the
endothermic side.

0.06 0.06
Nitrobenzene mole fraction

Aniline mole fraction

reforming is highly endothermic, the tube side temperature

Bubble phase .......... decreases abruptly in each reactor of CR. In this regards, using
0.04 Emulsion phase 0.04
a thermal heat source leads to lower temperature drop and
shifts the reactions. Fig. 12(a) and (b) justifies the thermal
behavior of TCFBNR. Fig. 12(a) reveals that the transferred
0.02 0.02 heat from the solid wall is notably less than the consumed
heat in the endothermic side. Thus, temperature decreases
gradually along the first reactor. On the other hand, the tube
side of the second reactor charges a huge thermal burden on
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 the shell side (Fig. 12(b)) and increases the temperature of
Length of reactor (Dimensionless) shell side.
Fig. 10 e Variation of (a) aromatic and paraffine mole
fractions in the bubble & emulsion phases and (b) 6.5.2. The exothermic side
nitrobenzene and aniline mole fractions in the bubble & Fig. 13 demonstrates the axial temperature profile in the
emulsion phases for TCFBNR. exothermic side of coupled configurations. There is a peak in
the temperature profile of the first reactor owing to the
generated heat in the exothermic side and the amount of
transferred heat (see Fig. 12(a)). The higher conversion in the
exothermic side increases the temperature in the TCFBNR.
6.4. Bubble and emulsion phases

The mass and energy balance equations for TCFBNR configu- 6.6. Influence of tube numbers
ration are developed for the bubble and emulsion phases. The
bubble phase contains a minor portion of catalyst particles One of the main issues in the exothermic section is the
and the reaction mostly occurs in the emulsion phase. maximum allowable temperature through the catalytic zone.
Consequently, mole fractions of reactants in the emulsion A high temperature may cause a destructive catalyst deacti-
phase are lower than the ones in the bubble phase. However, vation and a decrease in the production rate.
this will be reversed for mole fractions of products. The mole The heat transfer area increases well with increasing the
fractions of aromatic and paraffin in the endothermic side as number of tubes which affect the temperature profile in both
well as nitrobenzene and aniline mole fractions in the sides. The effect of the tubes number in each reactor on the
exothermic side are illustrated in Fig. 10(a) and (b). temperature profile of the exothermic side is shown in Fig. 14.
The optimum number of tubes plays an important role and
6.5. Thermal behavior considered as an objective function during the optimization
step which improves the performance of process in addition to
6.5.1. The endothermic side saving of operational and capital costs. Therefore, 1000 tubes
Fig. 11 illustrates the temperature profile in the endothermic are considered in this study to have a desirable temperature
side of various reactor configurations. Since naphtha profile in the exothermic side.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1 2989

1150
a 45
Generated heat in exothemic side
40 Transferred heat from tube wall N=1
1100
Consumed heat in endothermic side N=100
35 N=500
1050 N=1000
30 18

Temperature (K)
16 1000
Heat (W)

25

0 0.01 0.02 0.03

20 950
3
15 2.5
900
2
10

0.23 0.24 0.25 850

5

0 800
0 0.05 0.1 0.15 0.2 0.25 0.3
Length of reactor (Dimensionless)

750
0 0.1 0.2 0.3 0.4 0.5 0.6
b 25 Transferred heat from solid wall Length of reactor (Dimensionless)
Consumed heat in endothermic side
Generated heat in exothermic side
Fig. 14 e Effect of tube’s number on the developed hot spot
20
in the exothermic side.

15
Heat (W)

7. Conclusion
10 2

1.5 A remedy for increasing demand of hydrogen and high octane

1 gasoline is investigated in this study. A thermally coupled
5 0.54 0.56 0.58
fluidized bed reactor (TCFBNR) has been proposed in this
study for catalytic naphtha reforming process. In this novel
configuration, the exothermic hydrogenation of nitrobenzene
0
0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 is coupled with naphtha reactions and provides the required
Length of reactor (Dimensionless)
heat for naphtha reforming process. In TCFBNR, the advan-
Fig. 12 e A comparison of generated heat in the exothermic tages of fluidization concept including low pressure drop,
side, consumed heat of endothermic side and the high heat transfer and etc. are added to the coupled configu-
transferred heat from the tubes side to the shell side for (a) ration and identifies more proper candidate for naphtha
1st and (b) the 2nd reactors. reforming process. The thermal behavior of endothermic side
is improved and the aromatic content of reformate increases
considerably in TCFBNR in comparison with CR and TCNR.
Furthermore, a considerable increase (about 18 kmol/h) is
observed in the hydrogen production in TCFBNR. The nitro-
805 benzene conversion reaches 94.1% and 96.6% at the outlet of
the first and the second reactors, respectively. This novel
800
configuration has the capability to provide a relatively pure
hydrogen feedstock for the exothermic side. Moreover, the
Temperature of exothermic side (K)

795
pressure drop problem is solved in TCFBNR owing to applying
790 fluidized bed concept. Commercial and economic feasibility
of the proposed configuration must be evaluated in order to
785
consider commercialization of the process. The experiments
780 can be carried out in conjunction with the mathematical
modeling in order to improve the predicted results of the
775
TCFBNR
model and become as a powerful tool for designing new
TCNR plants.
770

765

760 Acknowledgment
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Length of reactor (Dimensionless)
The authors gratefully acknowledge the cooperation of
Fig. 13 e Temperature profile of exothermic side (for 1000 Mr. Hamid Rahmanifard during the preparation of this
tubes) for TCNR and TCFBNR. manuscript.
2990 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1

j Number of side (1, exothermic side; 2, endothermic

Nomenclature side)
k Reaction number index
f Feed conditions
Ac cross section area of each tube, (m2) in inlet conditions
Ar Archimedes number, (e) p Particle
As Lateral area of each tube, (m2) g Gas phase
ab specific surface area of bubble, (m2 m3) s At catalyst surface
av specific surface area of catalyst pellet, (m2 m3) b Bubble phase
cP specific heat of the gas at constant pressure, e Emulsion phase
(J mol1 K1) mf Minimum fluidization
ct total concentration, (mol m3)
Di inner diameter, (M) Abbreviations
dp particle diameter, (M) CR Conventional catalytic naphtha reforming
db bubble diameter, (M) TCNR Thermally coupled fixed-bed naphtha reactor
Fi Molar flow rate of component i in the fluid phase, TCFBNR Thermally coupled fluidized bed naphtha reactor
(mol mol1)
Ft total molar flow per tube, (mole s1) references
Fe molar flow in emulsion side, (mole s1)
Fb molar flow in bubble side, (mole s1)
Kbei mass transfer coefficient for component i in fluidized [1] Antos GJ, Aitani AM. In: Catalytic naphtha reforming. 2nd ed.
bed, (m s1) New York: Marcel Dekker Inc.; 1995.
[2] Li J, Tan Y, Liao L. Modeling and optimization of a semi-
kg Thermal conductivity of gas phase, (W m1 K1)
regenerative catalytic naphtha reformer. Proceedings of the
L Reactor length, (M)
2005 IEEE conference on control applications, Toronto,
ri reaction rate of component i, (mol kg1 s1) Canada; August 28e31, 2005.
rbi reaction rate of component i in bubble phase, [3] Hu Y, Xu W, Su H, Chu J. A dynamic model for naphtha
(mol kg1 s1) catalytic reformers. Proceedings of the 2004 IEEE
T bulk gas phase temperature, (K) International Conference on Control Applications, Taipei,
Tshell temperature of shell side, (K) Taiwan; September 2e4, 2004.
[4] Carvalho LS, Pieck CL, Rangel MC, Figoli NS, Grau JM.
Ttube temperature of tube side, (K)
Trimetallic naphtha reforming catalysts. I. Properties of the
Ushell overall heat transfer coefficient between exothermic
metal function and influence of the order of addition of the
and endothermic streams, (W m2 K1) metal precursors on PteReeSn/Al2O3eCl. Appl Catal A:
U superficial velocity of fluid phase, (m s1) General 2004;269:91e103.
ub velocity of rise of bubbles, (m s1) [5] Carvalho LS, Pieck CL, Rangel MC, Figoli NS, Grau JM.
ue linear velocity of emulsion phase, (m s1) Trimetallic naphtha reforming catalysts II. Properties of the
yib mole fraction of component i in the bubble phase, acid function and influence of the order of addition of the
metallic precursors on PteReeSn/Al2O3eCl. Appl Catal A:
(mol mol1)
General 2004;269:105e16.
yie mole fraction of component i in the emulsion phase, [6] Gonzalez-Marcos MP, Inarra B, Guil JM, Gutierrez-Ortiz MA.
(mol mol1) Development of an industrial characterisation method for
t time, (s) naphtha reforming bimetallic PteSn/Al2O3 catalysts through
z axial reactor coordinate, (M) n-heptane reforming test reactions. Catal Today 2005;
107e108:685e92.
Greek letters [7] Boutzeloit M, Benitez VM, Mazzieri VA, Especel C, Epron F.
DHf,i enthalpy of formation of component, i (J mol1) Effect of the method of addition of Ge on the catalytic
3B void fraction of catalytic bed, (e) properties of PteRe/Al2O3 and PteIr/Al2O3 naphtha reforming
3mf void fraction of catalytic bed at minimum catalysts. Catal Commun 2006;7:627e32.
[8] Borgna A, Garetto TF, Apesteguia CR, Moraweck B. Formation of
fluidization, (e)
bimetallic alloys in naphtha reforming Pt-Ge/Al2O3 catalysts:
r density of fluid phase, (kg m3)
an EXAFS study. Appl Catal A: General 1999;182:189e97.
rB density of catalytic bed, (kg m3) [9] Jess A, Hein O, Kern C. Deactivation and decoking of a naphtha
rs density of catalyst, (kg m3) reforming catalyst. Stud Surf Sci Catal 1999;126:81e8.
h catalyst effectiveness factor, (e) [10] Smith RB. Kinetic analysis of naphtha reforming with
mg viscosity of fluid phase, (kg m1 s1) platinum catalyst. Chem Eng Prog 1959;55:76e80.
d Bubble phase volume as a fraction of total bed [11] Weifeng H, Hongye S, Yongyou H, Jian C. Lumped kinetics
model and its on-line application to commercial catalytic
volume, (e)
naphtha reforming process. J Chem Ind Eng 2006;57:1605e11.
g Volume fraction of catalyst occupied by solid particle [12] Boyas RS, Froment GF. Fundamental kinetic modeling of
in bubble, (e) catalytic reformer. Ind Eng Chem Res 2009;48:1107e19.
m Viscosity of fluid phase, (kg m1 s1) [13] Krane HG, Groh AB, Schuhnan BL, Sinfeh JH. Reactions in
4s Sphericity of catalyst, (e) catalytic reforming of naphthas. Paper presented in Fifth
World Petroleum Congress; 1960.
Superscripts and subscripts [14] Kmak WS. A kinetic simulation model of the powerformmg
i Component i process. AIChE National Meeting; 1972.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 9 7 9 e2 9 9 1
[15] Juarez JA, Macias EV. Kinetic modeling of naphtha catalytic
reforming reactions. Energy Fuels 2000;14:1032e7.
[16] Stijepovic MZ, Linke P, Kijevcanin M. Optimization approach
for continuous catalytic regenerative reformer process.
Energy Fuels 2010;24:1908e16.
[17] Weifeng H, Hongye S, Shengjing M, Jian C. Multiobjective
optimization of the industrial naphtha catalytic reforming
process. Chin J Chem Eng 2007;15:75e80.
[18] Taskar UM. Modeling and optimization of a catalytic
naphtha reformer. Doctor of philosophy thesies; 1996.
[19] Rahimpour MR, Iranshahi D, Bahmanpour AM. Dynamic
optimization of a multi-stage spherical, radial flow reactor
for the naphtha reforming process in the presence of catalyst
deactivation using differential evolution (DE) method. Int J
Hydrogen Energy 2010;35:7498e511.
[20] Weifeng H, Hongye S, Yongyou H, Jian C. Modeling,
simulation and optimization of a whole industrial catalytic
naphtha reforming process on Aspen Plus platform. Chin J
Chem Eng 2006;14:584e91.
[21] Tailleur RG, Davila Y. Optimal hydrogen production through
revamping a naphtha-reforming unit: catalyst deactivation.
Energy Fuels 2008;22:2892e901.
[22] Rahimpour MR. Enhancement of hydrogen production in
a novel fluidized-bed membrane reactor for naphtha
reforming. Int J Hydrogen Energy 2009;34:2235e51.
[23] Iranshahi D, Rahimpour MR, Asgari A. A novel dynamic
radial-flow, spherical-bed reactor concept for naphtha
reforming in the presence of catalyst deactivation. Int J
Hydrogen Energy 2010;35:6261e75.
[24] Kolesnikov IM, Zuber VI, Svarovskaya NA, Kolesnikov SI.
Reforming of naphtha cut in a fluidized bed of catalysts.
Chem Technol Fuels Oils 2008;44:133e8.
[25] Iranshahi D, Pourazadi E, Paymooni K, Bahmanpour AM,
Rahimpour AM, Shariati A. Modeling of an axial flow,
spherical packed-bed reactor for naphtha reforming process
in the presence of the catalyst deactivation. Int J Hydrogen
Energy 2010;35:12784e99.
[26] Rahimpour MR, Iranshahi D, Pourazadi E, Bahmanpour AM.
A comparative study on a novel combination of spherical
and membrane tubular reactors of the catalytic naphtha
reforming process. Int J Hydrogen Energy 2011;36(1):
505e17.
[27] Gibilaro LG. Fluidization-dynamic: the formulation and
applications of a predictive theory for the fluidized state.
Boston: Butterworth-Heinemann; 2001.
[28] Deshmukh SARK. Membrane assisted fluidized beds reactor
experimental demonstration for partial oxidation of
methanol. Doctor’s degree at the University of Twente; May
27, 2004.
[29] Deshmukh SARK, Heinrich S, Mörl L, Annaland MVS,
Kuipers JAM. Membrane assisted fluidized bed reactors:
potentials and hurdles. Chem Eng Sci 2007;62:416e36.
[30] Levenspiel O. In: Chemical reaction engineering. 3rd ed. New
York: John Wiley & Sons; 1999.
[31] Abashar MEE. Coupling of steam and dry reforming of
methane in catalytic fluidized bed membrane reactors. Int J
Hydrogen Energy 2004;29:799e808.
[32] Grace JR, Elnashaie SSEH, Lim CJ. Hydrogen production in
fluidized beds with in-situ membranes. Int J Chem React Eng
2005;3:A41.
[33] Rahimpour MR, Bayat M, Rahmani F. Dynamic simulation of
a cascade fluidized-bed membrane reactor in the presence of
2991
long-term catalyst deactivation for methanol synthesis.
Chem Eng Sci 2010;65:4239e49.
[34] Tabis B. Methanol synthesis in a fluidized-bed reactor
coupled with an external heat exchanger: the effect of
feedback deformation. Chem Eng J 2001;83:191e200.
[35] Ye G, Xie D, Qiao W, Grace JR, Lim CJ. Modeling of fluidized
bed membrane reactors for hydrogen production from steam
methane reforming with Aspen Plus. Int J Hydrogen Energy
2009;34:4755e62.
[36] Gallucci F, Sintannaland MV, Kuipers JAM. Theoretical
comparison of packed bed and fluidized bed membrane
reactors for methane reforming. Int J Hydrogen Energy 2010;
35:7142e50.
[37] Ftankiewicz F, Moulijn JF. Re-engineering the chemical
processing plant: process intensification. New York: Marcel
Dekker Inc.; 2004.
[38] Zanfir M, Gavriilidis A. Modelling of a catalytic plate reactor
for dehydrogenation-combustion coupling. Chem Eng Sci
2001;56:2671e83.
[39] Dautzenberg FM, Mukherjee M. Process intensification using
multifunctional reactors. Chem Eng Sci 2001;56:251e67.
[40] Patel KL, Sunol AK. Modeling and simulation of methane
steam reforming in a thermally couples membrane reactor.
Int J Hydrogen Energy 2007;32:2344e58.
[41] Kirillov VA, Fadeev SI, Kuzin NA, Shigarov AB. Modeling of
a heat-coupled catalytic reactor with co-current oxidation
and conversion flows. Chem Eng J 2007;134:131e7.
[42] Khademi KH, Jahanmiri A, Rahimpour MR. A novel
configuration for hydrogen production from coupling of
methanol and benzene synthesis in a hydrogen perm-
selective membrane reactor. Int J Hydrogen Energy 2009;34:
5091e107.
[43] Abo-Ghander NS, Grace JR, Elnashaie SSEH, Lim CJ. Modeling
of a novel membrane reactor to integrate dehydrogenation of
ethyl benzene to styrene with hydrogenation of
nitrobenzene to aniline. Chem Eng Sci 2008;63:1817e26.
[44] Operating data of catalytic reformer unit, domestic refinery;
2005.
[45] Rase RF. Chemical reactor design for process plants, vol. 2.
John Wiley & Sons Inc.; 1977.
[46] Ullmann’s encyclopedia of industrial chemistry. 6th ed., vol.
3. Germany: Wiley-Vch; 2003. 161e175.
[47] Diao S, Qian W, Luo G, Wei F, Wang Y. Gaseous catalytic
hydrogenation of nitrobenzene to aniline in a two-stage
fluidized bed reactor. Appl Catal A: General 2005;286:30e5.
[48] Klemm E, Amon B, Redlingshok HF, Dieterich E, Emig G.
Deactivation kinetics in the hydrogenation of nitrobenzene
to aniline on the basis of a coke formation kinetics
investigations in an isothermal catalytic wall reactor. Chem
Eng Sci 2001;56:1347e53.
[49] Perry RH, Green DW, Maloney JO. In: Perry’s chemical
engineers’ handbook. 7th ed. McGraw-Hill; 1997.
[50] Yang WC. Fluidization, solid handling, and processing:
industrial applications. New Jersey: Westwood; 1998.
[51] Fogler HS. In: Elements of chemical reaction engineering.
2nd ed. Englewood Cliffs, NJ: Prentice-Hall; 1992.
[52] Kunii D, Levenspiel O. Fluidization engineering. USA:
ButterwortheHeinemenn; 1991.
[53] Mori S, Wen CY. Estimation of bubble diameter in gaseous
fluidized beds. AIChE J 1975;21:109e15.
[54] Davidson JF, Harisson D. Fluidized particles. New York:
Cambridge University Press; 1963.