Beruflich Dokumente
Kultur Dokumente
journal homepage:www.elsevier.com/locate/cherd
A B S T R A C T
Naphtha reforming is one of the most important processes in refineries in which high value-added reformate for gasoline pool and
aromatics such as benzene, toluene, and xylene are produced. It is necessary to establish new naphtha reforming units and develop the
traditional units to increase the efficiency of the processes. In this study, according to the recent progresses in the naphtha reforming
technology, mathematical modeling of this process in continuous catalyst regeneration mode of operation is accomplished in two
dimensions (radial and axial) by consider-ing cross flow pattern. In addition, a new catalyst deactivation model has been proposed and a
new reaction network model based on 32 pseudo-components with 84 reactions is investigated. Then, this model has been validated by
comparing with industrial data, and its results have acceptable agreement.
© 2013 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Keywords: Catalytic naphtha reforming process; Continuous catalytic regeneration; Two dimensional mathematical modeling; Reaction
network; Catalyst deactivation model.
∗
Corresponding author at: Dept. of Chemical Engineering, School of Chemical and Petroleum Engineering, Shiraz University, Shiraz 71345, Iran. Tel.: +98 711
2303071; fax: +98 711 6287294.
E-mail addresses: rahimpor@shirazu.ac.ir, mrahimpour@ucdavis.edu (M.R. Rahimpour). Received 26 April
2013; Received in revised form 11 September 2013; Accepted 14 December 2013
0263-8762/$ – see front matter © 2013 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cherd.2013.12.012
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1705
m2 Ri inner diameter, m
C
concentration, kmol m−3 Ro outer diameter, m
C −3 2 −1
ACP alkyl-cyclopentane concentration, kmol m Sa specific surface, m g
C
CA coke weight fraction on acidic function of cata- T temperature, K
−1 U
lyst, kg kg overall heat transfer coefficient, W m−2 K−1
C
CM coke weight fraction on metallic function of cat- Uj internal energy of component j
−1 −1
alyst, kg kg ur radial velocity, m s
C −3
j0 inlet concentration of component j, kmol m yj mole fraction of component (j)
Cp specific heat capacity at constant pressure,
−1 −1
kJ kmol K Greek letters
Cv specific heat capacity at constant volume, ε void fraction of catalyst bed
−1 −1 −1 −1
kJ kmol K viscosity, kg m s
C −3 v
T total concentration, kmol m ij stoichiometric coefficient of component j in
dp particle diameter, m reaction i
2 −1 b −3
De effective diffusivity, m s reactor bulk density, kg m
−1
Ec coke formation activation energy, J mol s sphericity,
F H −1
molar flow rate, kmol h−1 heat of reaction, kJ mol
−1
Fj molar flow rate of component j, kmol h z control volume length, m
Ft −1 r control volume thickness in radial direction, m
total molar flow rate to the reactor, kmol h
H 3 −1
enthalpy, J mol −1
˛A constant of deactivation, m kmol
−1 ˛ 3 −1
Hj enthalpy of component j, J mol CA constant of deactivation, m kmol
k thermal conductivity, W m K
−1 −1 ˛ 3 −1
CM constant of deactivation, m kmol
k
CA constant of deactivation equation for acidic
−1 n 1.5 −1.5
function, kg kg kPa1 m kmol ˛M
k
CM constant of deactivation equation for metallic Subscript
−1 n 1.5 −1.5
function, kg kg kPa1 m kmol
k −1 −1 Pi number, 3.14156
eff effective thermal conductivity, W m k 3 −1
k constant of deactivation, m kmol
in reaction rate constant for reaction
K equilibrium constant for reaction
in
L length of reactor, m e exit condition
mnumber of reactions
−1 i numerator for reaction
Mj molecular weight of component j, kg kmol
j numerator for component
M −1
mean molecular weight in the flow, kg kmol nnumber of carbon atom
nnumber of components
Ttotal
n constant of deactivation equation
1
n2 constant of deactivation equation Abbreviations
nA acidic function activity power number A aromatics
n acidic function activity power number
CA ACH alkyl-cyclohexane
n metallic function activity power number
M ACP alkyl-cyclopentane
n
CM metallic function activity power number CCR continuous catalyst regeneration reformer
−2 −1
Nj molar flux of component j, kmol m h F scaling factor
P total pressure, kPa HChydrocarbon
P partial pressure of aromatic with n carbon, kPa
An ◦
P IBP initial boiling pint, C
ACHn partial pressure of alkyl-cyclohexane with n
IP iso-paraffin
carbon, kPa NP normal-paraffin
P partial pressure of alkyl-cyclopentane with n
ACPn RON research octane number
carbon, kPa Pt platinum
P partial pressure of hydrogen, kPa
H2 Sn tin
P partial pressure of iso-paraffin with n carbon,
IPn
kPa
P partial pressure of normal-paraffin with n car-
NPn
bon, kPa using catalyst with less stability but higher selectivity and yield, requiring
r radius, m lower recycle ratio, economic design, produc-ing more reformate with
−1 −1 higher aromatic content and steady production of hydrogen with constant
ri rate of ith reaction, kmol kg cat h
purity (93% compared to 80% in the semi regenerative process) (Aitani,
2005a,b; Antos and Aitani, 1995; Hongjun et al., 2010; Edgar, 1983;
Lee et al., 1997). In addition, the deposited coke in the semi
regenerative
1706 CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727
reformer can be varied from a few weight percent on the first reactor to 20
wt% coke in the last reactor. In the cyclic regen-erative reformers, catalyst
regeneration is usually based on economics performance for one or more
of the reactors in the reforming process loop. While in the CCR process 2. Kinetic model
that operates with a platinum/tin catalyst, the deposited coke on the cata-
lyst is usually restricted to less than 7 wt% in the last reactor (Antos and An effective kinetic model of naphtha reforming must repre-sent all the
Aitani, 1995). Thus, according to these advantages of CCR process the major types of reactions in the reforming process. A kinetic model
capacity of the CCR reforming units has increased quickly in the recent inherently satisfies both heat and material balances; because it takes into
years. Generally, the naphtha reforming unit consists of four reactors, account the stoichiometry of the reactions. As discussed, the naphtha
catalyst regeneration unit, distillation column which separates products feedstock consists of several hundred components and each of them takes
into light hydrocarbons and reformates, and due to endothermic reac-tions part in various reactions. Thus, presenting a detailed kinetics-based model
in the reforming process, inlet stream to each reactor is preheated via a considering all components and reactions is infeasi-ble. In past decades,
furnace. many attempts have been made (Marin et al., 1983; Froment, 1987;
Ramage et al., 1987; Jorge and Eduardo, 2000; Hu et al., 2003; Weifeng et
al., 2006) to propose a perfect kinetic model for naphtha reforming
Accomplished researches regard naphtha reforming pro-cess could be process by con-sidering most kinetic lumps and reactions. Padmavathi and
classified in three categories. The first category consists of the Chaudhuri (1997) proposed appropriate model for naphtha reforming
accomplished studies on kinetic models of cat-alytic naphtha reforming. kinetic, but in their model, some of the main lumps and reactions have
To reduce the complexity of the reaction system, similar chemical been ignored. Whiles in the reforming pro-cess, aromatics are the main
components are lumped together and represented as a pseudo component, products, 8-carbon aromatics were not subdivided in details. In addition,
and the reactions among the pseudo components are considered (Hongjun some of the main reactions such as dehydrocyclization of paraffins to
et al., 2010). Primary studies about lumped kinetic have been done by aromat-ics, isomerization and transalkylation of aromatics have been
Smith (1959), Krane et al. (1959) Kmak (1972), Ramage et al. (1980), ignored. In the proposed model, 8-carbon aromatics are subdi-vided to
Wolff and Kramarz (1979) and Jenkins and Stephens (1980). To satisfy four lumps (ethylbenzene, para-xylene, meta-xylene and ortho-xylene) and
various demands on the reforming process, different pseudo components the variations of each of them are inves-tigated. Also ignored reactions in
have been considered in the proposed network kinetic models. Marin et al. Padmavathi and Chaudhuri (1997) study, are characterized in Tables 1–9.
(1983) and Froment (1987) published the reac-tion network for the whole
range of naphtha reforming. In addition, by considering this fact that the
reaction pathways derived for each carbon number fraction are almost
identi-cal to each other, Ramage et al. (1987) used the same reaction
network within each carbon number group. Jorge and Eduardo (2000) Proper kinetic modeling of reforming reactions is the key to effective
proposed a 24 lumps model in order to model the kinetic of catalytic representation of the process over a wide operating range. The first effect
naphtha reforming reactions. Hu et al. (2003) pro-posed kinetic model of the kinetics of the reactions on the product properties is the effect of
with 17-lumpes while Weifeng et al. (2006) proposed kinetic model with operating variables on the octane number (Taskar and Riggs, 1997).
18-lumpes. The second class of studies have been focused on improving Therefore, by consid-ering important kinetics in the reforming process and
the operation and selectivity of the catalyst and reduce the probability of all of the accomplished studies, in this study, a new reaction network is
coke formation on the catalyst surface by adding some metals such as Sn, proposed.
In, and Ge to the catalyst. The literature sur-vey about these concepts is
discussed extensively in previous work (Iranshahi et al., 2012). The third Paraffins, naphthenes and aromatics are the different hydrocarbon
category consists of the performed studies on presenting and modeling groups in the reforming feedstock. In the pre-sented model, the naphtha
novel config-urations with better performance. Some studies in this field feedstock has been subdivided to naphthenes (alkylcyclohexanes: ACH,
were carried out by Taskar and Riggs (1997), Juarez et al. (2001), and alkylcyclopentanes: ACP), paraffins (normal paraffins: NP, isoparaffins:
Weifeng et al. (2007). Rahimpour (2009), proposed a novel fluidized-bed IP), and aro-matics (A) lumps with carbon numbers ranging from C 6 to C9
membrane reactor for naphtha reforming in the presence of catalyst +
, this kinetic model consists of 32 pseudo components with 84 reactions.
deactivation to enhance hydrogen pro-duction. Also Mostafazadeh and
As shown in Fig. 2, in this model, all of the main reactions are included,
Rahimpour (2009) assessed the membrane insertion in naphtha reforming
and the defined pseudo components are related by major reforming
catalytic bed reactions. In addition, rate con-stants in the proposed model are obtained
by optimization, so that deviation between output results of model and
com-mercial data is minimized. In the following, main reactions in the
reforming process are discussed, and the reactions, rate
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1707
K RT − ) T (kPa)
(1)ACHn ↔ An + 3H2 r1n = k1n ACHn 1n (kmol kgcat K1n = exp A −
−1 ) a E A B
C6 H (kJ (mol H2 ) R
68.73 18.75 19.50 59.90 3
C7 24.80 × 10
208.47 20.70 19.50 60.23 3
C8 25.08 × 10
a 64.50 17.89 19.50 60.37 3
for An = MX 23.27 × 10
a 65.10 19.15 19.50 60.32 3
for An = OX 23.49 × 10
a 64.74 18.66 19.50 60.13 3
for An = PX 23.36 × 10
a 68.70 18.71 19.50 60.40 3
for An = EB 24.78 × 10
+ 66.05 20.38 19.50 61.05 3
C9 21.33 × 10
a Ignored reactions in Padmavathi and Chaudhuri (1997) study.
constants, characteristics and the calculated heat of reactions are listed. catalyst (Antos and Aitani, 1995; Taskar and Riggs, 1997). The reaction is
endothermic and produces a sharp drop in tem-perature. The needed
reaction constants for calculating the reaction rate constants (k1n ) and
↔ =
P P
H K
NP
=
exp E
(kmol kg h kPa ) K
=
exp −
B
(kPa)
n+ 2
n 2n 2n ACHn 2 2n 2n RT cat 2n T
R
−43.64 24.37 33.11 −9.12 3
C6 −5.24 × 10
−32.85 29.10 33.11 −8.95 3
C7 −3.95 × 10
−32.55 27.81 33.11 −8.34 3
C8 −3.91 × 10
+ −28.00 29.76 33.11 −8.12 3
C9 −3.37 × 10
P
−1 −1 −2 −1
H (kJ mol−1 )
=
a −
E A B
(3)ACH H IP r k P P IPn
k exp a E (kmol kg h kPa ) K exp A B (kPa)
n+ ↔ n = 3n H K = − cat 3n −
2 3n ACHn 2 3n 3n RT T
R
C6 −52.00 26.36 33.11 −14.35 3
−6.25 × 10
−39.30 25.75 33.11 −13.65 3
C7 −4.73 × 10
C8 −49.40 29.00 33.11 −13.17 3
−5.94 × 10
+ −49.00 29.76 33.11 −12.39 3
C9 −5.89 × 10
P P −1 −1
H (kJ mol− ) a
R A B
−1
(4)NP ACP H r k P ACPn
K
H2
k exp a− E (kmol kg h kPa ) K exp A B (kPa)
n↔ n+ 2 4n = 4n NPn − 4n 4n = RT cat 4n = −T
1 E
r kn P P P
kn a E −1 −1 −1 A B
(5)IPn ACPn H n ACPn H2 exp (kmol kg h kPa ) K exp (kPa)
↔ + 2 5 = 5 K 5 = − cat 5n =
IPn − 5n RT −T
1 E
−
H (kJ mol ) a R A B
a 76.67 29.83 33.11 16.18 3
C5 9.22 × 10
70.60 30.87 33.11 16.03 3
C6 8.50 × 10
C7 67.20 32.95 33.11 15.47 3
8.08 × 10
78.60 34.19 33.11 16.37 3
C8 9.45 × 10
+ 80.00 32.96 33.11 16.02 3
C9 9.62 × 10
a Ignored reactions in Padmavathi and Chaudhuri (1997) study.
1708 CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727
−
= =
H(kJ (mol H2 )−1 ) a E A B
n ↔ n +2 6n
=
6n NPn
−
K6n 6n RT cat 6n −T
R
a 66.50 16.87 18.86 35.73 3
C6 7.99 × 10
a 63.20 17.17 18.86 35.21 3
C7 7.60 × 10
a 3
C8 6.79 × 10
for An = MX 56.51 17.32 18.86 34.13
56.95 17.89 18.86 34.21 3
for An = OX 6.85 × 10
56.70 17.01 18.86 34.16 3
for An = PX 6.81 × 10
59.66 16.96 18.86 34.63 3
for An = EB 7.17 × 10
+a 59.29 18.90 18.86 33.56 3
C9 6.36 × 10
(7)IP ↔ A + 4H r k (P P P 4 ) k exp a E −1 −1 −1 K exp A B 4
An H 2 (kmol kg h kPa ) (kPa)
n n
H(kJ (mol H2 )−1 ) a − E A B
2 7n = 7n IPn − K
7n 7n = RT cat 7n = −T
R
a 68.18 16.44 18.86 36.01 3
C6 8.20 × 10
a 64.83 16.90 18.86 35.47 3
C7 7.79 × 10
a 3
C8 7.30 × 10
for An = MX 60.73 16.27 18.86 34.81
61.17 16.78 18.86 34.87 3
for An = OX 7.35 × 10
60.90 16.04 18.86 34.83 3
for An = PX 7.32 × 10
63.88 15.93 18.86 35.23 3
for An = EB 7.68 × 10
+a 60.77 17.66 18.86 34.81 3
C9 7.3 × 10
a Ignored reactions in Padmavathi and Chaudhuri (1997) study.
Table 4 – Rate constants and heat of reactions in isomerization of naphthenes and paraffins. (8)ACP n ↔ ACHn
r =k
8n 8n
B
A− (−)
T
−1 a E A B
H (kJ mol ) R
−17.10 15.11 23.81 −5.20 3
C6 −2.06 × 10
−27.90 24.73 23.81 −6.63 3
C7 −3.36 × 10
−29.20 25.78 23.81 −6.83 3
C8 −3.51 × 10
+ −31.00 26.13 23.81 −7.16 3
C9 −3.73 × 10
(9)NP IP r k (P P ) k exp a E −1 −1 −1 K exp A B
( )
IPn (kmol kg h kPa )
K a − =
n↔ n − −
H (kJ mol−1 ) E A B
9n
=
9n NPn 9n 9n
=
RT cat 9n −T
R
a −9.23 25.08 26 0.80 3
C4 −1.11 × 10
a −6.94 24.89 26 1.17 3
C5 −0.83 × 10
−9.86 24.70 26 0.69 3
C6 −1.19 × 10
C7 −6.45 24.15 26 1.27 3
−0.76 × 10
−13.95 25.54 26 0.04 3
C8 −1.68 × 10
+ −21.00 24.48 26 −1.09 3
C9 −2.53 × 10
a Ignored reactions in Padmavathi and Chaudhuri (1997) study.
cat
B
10n 10n −
K10n 10n
=
RT 10n −T
H mol
−1
)
E
R
a OX PX −10.50 25.21 21.24 12.13 3
n=1 −1.26 × 10
a OX MX −17.60 24.50 21.24 11.00 3
n=2 −2.11 × 10
a MX EB 12.60 24.53 21.24 15.81 3
n=3 1.50 × 10
a OX EB 10.84 25.17 21.24 15.54 3
n=4 1.30 × 10
a EB PX −11.89 25.84 21.24 11.91 3
n=5 −1.43 × 10
a MX PX 0.71 24.47 21.24 13.7 3
n=6 −0.085 × 10
a Ignored reactions in Padmavathi and Chaudhuri (1997) study.
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1709
these reactions are the most difficult reaction to promote (Matar and the number of carbon atoms (Matar and Hatch, 2000). The reactions are
Hatch, 2000). The most reactions of dehydrocy-clization are unaffected by pressure, whiles from a kinetic viewpoint, the equilibrium
depends mainly on the temper-ature. It should be mentioned that
endothermic and promoted by high temperature, low
isomerization reactions are promoted by the acidic function (Antos and
pressure and by both metallic and acidic functions of the Aitani, 1995). The isomerization of alkylcyclopentanes into alkylcy-
catalyst. The needed reaction constants for calculating clohexanes involves a ring rearrangement. In the subsequent stage, the
the reaction rate constants k k k k k and k7n) produced alkylcyclohexanes undergo dehydrogena-tion reaction and
( 2n, 3n, 4n, 5n, 6n
and equilibrium constants K K K K K and K produce aromatics. Thus, isomerization of naphthenes is considered as a
( 2n, 3n, 4n, 5n, 6n 7n )
of dehydrocyclizationreactionsarepresented in desirable reaction. Aro-matics with 8-carbon atoms also undergo the
Tables 2 and 3. isomerization reactions, the isomerization between P-xylene, M-xylene
and O-xylene occur rapidly, while isomerization between ethylbenzene
and xylene isomer occurs slowly (Weifeng et al., 2006). Calculated
2.3. Isomerization
constants for isomerization reac-tions of naphthenes, paraffins and
aromatics are shown in Tables 4 and 5.
Three types of isomerization reactions are investigated in the reforming
process, including isomerization of paraffins, naph-thenes and aromatics.
Paraffins isomerization rearranges the molecule with essentially no
change in the volume. These reactions are fast, slightly exothermic and do
not affect
n−3
5 r =k P
NPn
2 H2 → 2 (3P1+ P2 5
n 1 P P P n−3 n P
P
+ + E 1 1
(13) for = 5 IPn
IP H
5+ 2→2 (IP4 + 3 2 1 ) IPn + 3 H2 → 15 i
i=1
− −
r13n = k13n k13n = exp a − (kmol kgcat h )
Pt RT
−1 E
H(kJ (mol H2 ) ) a
R
a −40.30 36.15 34.61
C4
a −45.30 41.57 34.61
C5
C6 −37.40 40.53 34.61
C7 −41.85 41.57 34.61
C8 −36.00 41.51 34.61
C9+ −35.65 42.93 34.61
H(kJ (mol H2 )
−1
) a E
R
C6 −45.45 42.15 34.61
ACPn +3 H2 → 15 Pi
i=1
−1 ) a E
H(kJ (mol H2 ) R
a −68.74 40.23 34.61
C5
C6 −54.00 41.55 34.61
C7 −52.71 43.75 34.61
C8 −51.95 43.65 34.61
C9+ −50.40 44.15 34.61
Transalkylation is the reaction between two similar or dis-similar Breaking of carbon bonds in the reforming is called hydroc-racking.
molecules involving transfer of an alkyl group. In the alkyl-transfer Bond’s breaking can occur at any position along the hydrocarbon chain.
reactions, shift of alkyl groups from one ring to the other, depends on the This reaction by removing the low octane number paraffins and
number of alkyl groups on the aromatic ring/s, the type of alkyl naphthenes from reformate and pro-ducing lighter hydrocarbons, such as
substituent/s and the chain length (Roldán et al., 2004). In the proposed methane, ethane and propane, helps to improve the octane in the products
model, transalkylation between aromatics is considered. Two toluene rings (Matar and Hatch, 2000). Hydrocracking is an irreversible reaction and is
can disproportionate to produce one benzene ring and one Para xylene favored at high temperature and high pressure, also is cat-alyzed by the
ring, in addition, two Ortho-xylene rings disproportionate to produce one acidic or metallic function of the catalyst (Antos and Aitani, 1995; Matar
+
toluene and higher aro-matics (A9 ). These reactions are promoted by and Hatch, 2000). Reaction constants for hydrocracking of paraffins and
the catalyst metallic function, and with increase in the reaction tempera- naphthenes are shown in Tables 7 and 8, respectively.
ture, the conversion increases. It should be mentioned that transalkylation
occurs mainly in very severe conditions of pressure (Roldán et al., 2004).
Calculated constants for reac-tion rate in the transalkylation reactions are
presented in Table 6. 2.6. Hydrodealkylation
R
C7 1 n −41.81 7.64 17.92
C8 −42.32
for An+1 = PX 1 n 5.57 17.92
for An+1 = EBa 2 n−1 −20.02 5.55 17.92
+ −52.57
C9
for Ak = OX 1 n 8.91 17.92
for Ak = T
a 2 n−1 −30.80 5.57 17.92
hydrogen. Like hydrocracking, this reaction is favored by high recycle hydrogen gas and, after heating to the desired reac-tion
temperature and high pressure also it is an irreversible reaction, and temperature (798 K), it is sent to a series of reactors. The specification of
promoted by the metallic function of the cat-alyst. hydrodealkylation is conventional process such as reactors, feed, product and catalyst are
the slowest of all the reforming reactions. This reaction is desirable shown in Table 10.
because it changes the relative equilibrium distribution of the aromatics in In the CCR process, spherical catalysts are used in the reac-tors to
favor of benzene and octane number (Taskar and Riggs, 1997; Matar and facilitate catalyst circulation by gravity. Coked catalyst is withdrawn from
Hatch, 2000). The needed hydrodealkylation constants to calculate the
the last reactor and sent to the regenera-tion unit, then, regeneration of the
reaction rate constants (k16n ) is presented in Table 9. catalyst is performed as the catalyst moves down in the regeneration unit
on a con-tinuous basis. Regenerated catalyst is added to the top of the first
reactor to keep the efficiency of process at the certain level, and catalyst
regeneration leads to reduction in operat-ing cost, whiles, poor catalyst
will lead to low product yields and increase in operating expense. As
3. Process description shown in Fig. 4, cata-lyst flows vertically in the reactor; and the feed
flows radially through the catalyst bed (reaction side) of the reactor, thus
The Continuous Catalyst Regeneration (CCR) process is a part of process cross flow pattern exist in the reactors.
used in the petroleum and petrochemical indus-tries. With the advent of
the continuous catalytic regeneration process, significant progress was
observed. Also, catalytic reformers and in particular, continuous catalyst
regeneration, have indispensable role in the refiners’ plans due to current
environment regulations for reducing air pollutants induced by sulfur of 4. Catalyst deactivation model
gasoline (Aitani, 2005a,b).
Coke deposition on the catalyst is an undesirable by-product in the
The CCR process consists of regeneration unit and adia-batic cross- reforming process, and the amount of deposited coke depends on feed
flow reactors that are arranged in a conventional side-by-side pattern. As quality, catalyst state and transit time of catalyst from the reactors (Taskar
shown in Fig. 3, the first reactor is always smaller than the other reactors and Riggs, 1997). Catalyst deactivation is primarily a function of coke
and the last reactor is always the largest. In this process, each catalyst deposition on the catalyst. Controlling coke deposition is equivalent to
particle moves continuously from the first reactor toward the last reac-tor. control-ling catalyst deactivation. In the CCR process, the catalyst
Then, catalyst particles are sent to the regeneration unit to restore. Also, circulation rate can be adjusted to maintain a specified weight percent
hydrotreated naphtha is combined with the carbon on the catalyst (Antos and Aitani, 1995).
1712 CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727
The main characteristics of a catalyst are: activity, selec-tivity and functions directly affect the selectivity of reactions (Lieske et al., 1987).
stability (Barbier, 1986). The activity expresses the catalyst ability to Catalysts used in this processes are bimetallic Pt-Sn and support is high
increase the rate of the reactions. The selec-tivity expresses the catalyst purity alumina (acidic function). The metal site is provided by platinum
ability to favor desirable reactions rather than others. The stability and metal promoter is tin. The reaction in this process starts on the metal
characterizes the change of the catalyst performance with time (i.e. sites and continues on the acid sites (Carter et al., 1982; Lieske et al.,
activity and selec-tivity) while operating conditions and feed are stable. 1987).
The catalyst affects reaction rates through its two different func-tions:
metallic and acidic, which promote different type of reactions (Carter et The metallic function is mainly responsible for dehydro-genation,
al., 1982). Coking on the metallic and acidic transalkylation and hydrodealkylation activity and these reactions are
practically unaffected by the presence or
Table 10 – Specifications of reactors, feed, product and catalyst of plant for fresh catalyst.
Parameter Numerical value Unit
Naphtha feed stock 233637.01 kg/h
Reformate 216488.63 kg/h
H2 /HC mole ratio 2.193 −
Mole percent of hydrogen in recycle 0.83 −
1st reactor 2nd reactor 3rd reactor 4th reactor
Inlet temperature (K) 798 798 798 798
Inlet pressure (kPa) 595 550 505 460
Inner and outer Diameter (m) 1.25, 2.19 1.25, 2.35 1.30, 2.53 1.3, 2.89
Length (m) 8.50 10.35 12.1 15.39
Catalyst distribution (wt %) 12 18 25 45
ASTM D86 ◦
Naphtha feed ( C)
IBP 81
5% 91.2
10% 93.2
20% 96.9
30% 101.1
40% 105.7
50% 111.4
60% 117.6
70% 124.5
80% 132.7
90% 143.1
95% 150.5
FBP 159
Typical properties of catalyst
dp 1.8 mm
Pt 0.3 wt%
Sn 0.3 wt%
220 2
sa m /g
680 3
B kg/m
ε 0.36 −
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1713
Table 11 – Catalyst functions on the main reactions in the proposed kinetic model (metallic (M) and acidic (A) functions).
Reactions Agent Reactions Agent
ACHn ↔ An + 3H2 M r9n NPn ↔ IPn A
r1n
r2n ACHn + H2 ↔ NPn M +A r10n R↔P A
r3n ACHn + H2 ↔ IPn M +A r11n 2R ↔ P + C M
5
(n−3) 5=
15
absence of acid sites, while the acidic support provides activity for the and published literatures. According to previous studies (Figoli et al.,
isomerization reactions. Two of the naphtha reform-ing reactions 1983a,b; Mieville, 1991; Bishara et al., 1984), the rate of deactivation
(hydrocracking and dehydrocyclization) can also be catalyzed by the depends on operating conditions such as: temperature, pressure and
metallic and acidic function (Antos and Aitani, 1995; Barbier, 1986; hydrogen over hydrocarbon mole ratio, also according to models proposed
Lieske et al., 1987). Table 11 shows the catalyst functions on the main by Liu et al. (1997, 2002) and Tailleur and Davila (2008) the rate of
reactions in the proposed kinetic model. deactivation depends on alkylcyclopentane concentrations. Thus, the rate
of coke formation on the fresh catalyst could be written as follows:
This section proposes a model for catalyst deactiva-tion rate that is
developed on the basis of pilot-plant data
exp(−Ec /RT) ×
o (17)
r ∝
C0.5
C n n2 ACP
P 1 (H2 /HC)
o
where rC , Ec , CACP , P and (H2 /HC)are the rate of coke formation on
the fresh catalyst, coke formation activation energy, alkyl-cyclopentane
concentration, pressure and hydrogen over hydrocarbon mole ratio,
respectively. In addition, the rate of reaction over a spent catalyst is
calculated by multiplying the catalyst activity on the dominant function
and rate of reaction over fresh catalyst.
o (18)
ri = a i × ri
where ri represents the ith reaction. The catalyst activity (ai ) is a function
of deposited coke weight fraction, which could be written as:
˛ ai i ni = 1 1
i
− = × n If
dCCi If n i =/ 1
ai = exp(−˛i × CCi )
ai =
(1/(ni 1))
dai
(1 +(ni − 1)˛i × CC ) −
i (19)
r =a ×r o , (20)
Ci Ci C i
0
where rCi and rC i are coke formation rate of spent catalyst on dominant
function and rate of coke formation on domi-nant function of fresh
catalyst, respectively. Similar to catalyst activity in the reaction rate,
Fig. 4 – Structure of reforming reactor in CCR process. catalyst activity in coke formation
1714 CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727
daCi n
5. Mathematical modeling Fig. 5 – (a) Control volume for modeling with the lngth of dz (b) top view
of control volume.
To develop a model for CCR process, heat and material balance equations
6. Numerical solution
are made over a small element of the reactor with the length of z and the
thickness of r. Then these equations are combined with the deactivation
The obtained mathematical model in the previous section is a set of partial
model, Ergun equation, also thermodynamic and kinetic relations as
differential equations (PDEs). These equations are integrated through each
auxiliary corre-lations for predicting the behavior of reforming process. In
reactor to describe the molar flow rate of reformate and physical
the previous studies, that have been done for modeling of CCR process
properties along the length and radius of the reactors. By considering these
(Hongjun et al., 2010; Weifeng et al., 2007; Mahdavian et al., 2010),
complex equa-tions, a numerical method must be employed. In this study,
variations of parameters are assumed in one dimension. While, in this
finite difference method with explicit solution of all the dif-ferential
study, the mathematical modeling of the reactor is done in two dimensions
equations in the mathematical model is employed. In this method, the
(radial and axial), and its results had better compliance with the actual
reactor is divided into a series of nodes and for each node derivatives in
data. A differ-ential volume element of the reactor on which the energy
differential equations become difference equations that are functions of the
and mass balance equations has been applied, is shown in Fig. 5a, also top
surrounding cells (Nauman, 2001). Then, boundary conditions of the pro-
view of this control volume is shown in Fig. 5b. The obtained results of
cess are used for initial nodes of the reactor and the result of each node is
the mathematical model are described in Table 13. Also, details of the
mathematical modeling for this process are described in Appendixes B– used as boundary conditions for the follow-ing nodes and this procedure is
D.Basic assumptions repeated for all nodes in the reactor (Nauman, 2001).
o
o (24)ri = aA × ri
(23)ri = aM × ri
(26)− n
(25)− n daA
daM
dC
CM = ˛M × aM
M
dC
CA
= ˛A × aA
A
(27)n /1 a = − × 1 (28)n /1 a = − × 1
nM = 1 aM exp( ˛M CC M ) nA = 1 aA exp( ˛A C CA )
M= A=
M = A =
da n da n
(31) CM ˛ a CM (32) CA ˛ a C
− = CM × CM − × − = CA × CA
A − ×
dCCM dCCA
(33)n C /1 aC = 1 (34)n C /1 aC = 1
o k
P n1(H 2 /HC) n2 ×
ACP
(1 + (nCM − 1)˛CM × CCM )
C
0.5
M
r CA =
o k
P n1 (H2 / HC) n 2 ACP
(1 + (nCA − 1)˛CA × CCA )
C
0.5
×
A
8. Results and discussion Results of modeling involve the following main issues in the process:
Table 13 – Mass and energy balances and auxiliary relations for CCR process.
Mass balance
1 ∂ 1 m j = 1, 2, ..., n
(37)D (A ∂Cj ) ∂ (A u C ) a r ε ∂Cj
ej r b i=1 ij
= ∂t i = 1, 2, ..., m
j
Ar ∂r ∂r − rm r + ×
Ar ∂r i i
eff
× j=1 j=1
Ar − i × − j=1 ∂r = j−
i = 1, 2, ..., m
∂rm
=− ( H × a × r )
(40) ∂r ur CT CP i i i i = 1, 2, ..., m
∂T b
i=1
Velocity distribution
n m j = 1, 2, ..., n
(41) ∂ur ur ur ∂CT b a r
∂r − × ∂r + j=1 i=1 × ij i i = 1, 2, ..., m
=−r CT Ct i
Collector (mass balance)
(42)Fj |z − Fj |z+dz + Fje = 0 j = 1, 2, ..., n
∂F
j
(43) ∂z = 2 Ri ure Cje j = 1, 2, ..., n
Collector (energy balance)
| n
∂T T ∂CP T j=1 1
z − FT CP T|z+dz +
(44)FT CP T Fje Cpje (Te − T) = 0
(45) = − − (2 Ri × ure × CTe ) + × 2 Ri × ure × CTe × Cpe × (Te − T)
∂z CP ∂z
FT FT CP
results and commercial values. Thus, the proposed model predicts the observed during the passage of a stream through one reactor to another
reactor and catalyst performance and quality of reformate very well. Some due to piping and instruments such as a furnace.
variables such as feedstock properties, temperature and pressure of In order to consider the variation of temperature in cat-alytic naphtha
reaction, space veloc-ity, and hydrogen over hydrocarbon molar ratio, reforming, the temperature profile of the plant is depicted in Fig. 6b. In
affect the enhancement of main products yield and reactor performance this 3-D plot, temperature vs. dimen-sionless radial and axial coordinates
(Pisyorius, 1985). Since the operating conditions (pressure and of reactor is illustrated. Since the major reforming reactions are highly
temperature) of process side play a significant role in the per-formance of endothermic, temperature decreases by proceeding of the reactions, and
the catalyst as well as the yield and quality of reformate, these variables then production rate of reactions decrease. Consequently, to maintain the
have been investigated firstly. reaction temperature at suitable level, commer-cial catalytic reformers are
designed with multiple reactors and intermediate furnaces. In the first
reactor, naphthenes concentration are high and dominant reaction in this
8.1. Investigation of operating conditions during the process reac-tor (dehydrogenation) is very fast and endothermic, thus, it causes a
very sharp temperature drop in the radial direction in the first reactor,
which is demonstrated in Fig. 6b. Due to the remarkable decrease in the
In this section, variation of operating conditions (pressure and temperature of first reac-tor, the outlet stream of the first reactor is
temperature) in CCR process is investigated. One of the advan-tages of preheated via a furnace to the same entrance temperature of the first reac-
the CCR process is the negligible pressure drop in the reactors, which tor and then is sent to the second reactor. In the second
allows the use of smaller catalyst particles with higher efficiency. As
shown in Fig. 6a, the variations of pressure inside the reactors are slight,
and sudden drops are
Table 15 – Comparison between predicted temperature, pressure and molecular weight with plant data.
Reactor number Outlet temperature Outlet pressure Outlet molecular weight
reaction rates through its two different functions: metallic and acidic
(Carter et al., 1982), which coke can be formed on the both of them. This
coke causes a decrease in the activity of the catalyst.
Fig. 7 – Coke concentration distribution in (a) metallic function, (b) acidic function and (c) total coke. (d) comparison coke content in various length
of reactors.
a1 a
Z=0 300
Molar flow rate along inner annulus of reactor
Z=L/4
molar flowrate(kmol/h)
250 Mean molar flow rate along radius of reactor
0.95
catalyst
Mean
Z=3/4*L 200
ofthe
Z=L
0.9
activity
0.85 150
Metallicfunction
0.8
Alkylcyclohexane
100
0.75 50
0.7 0
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
Radius of reactor (Dimensionless) Annulus Length(Radius) of reactor (Dimensionless)
b
b1 180
0.95 Z=0 160 Molar flow rate along inner annulus of reactor
rate(kmol/h)
catalyst
flow
100
of the
0.8
Acidic functionactivity
0.75
molar
80
0.7
eAlkylcyclopentan
60
0.65
40
0.6
20
0.55
0.5 0
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
Radius of reactor (Dimensionless) Annulus Length(Radius) of reactor (Dimensionless)
Fig. 8 – Catalyst activity in (a) metallic function and (b) Fig. 10 – Variations of molar flow rate of (a) alkyl-cyclohexane
acidic function. and (b) alkyl-cyclopentane in the process.
carbon atoms such as ethylbenzene, and ortho-, meta-, and para-xylene are reactions, such as r1691-1 and r169-2 ), molar flow rate decreases.
depicted in Fig. 12a, and aromatics with six and seven carbon atoms, also
Aromatics molar flow rate along the length of collector is shown by solid
+
heavy aromatics (A9 ) are shown in Fig. 12b. Aromatics molar flow rate lines in Fig. 12a and b. Produced aromatics in the reactor collect in the
+
except heavy aromat-ics (A9 ), during the process have a similar pattern. annulus, and then, are sent to the next reactor. In the next reactor,
As can be seen in Fig. 12a and b, aromatics are produced along the radius aromatics production contin-ues. Thus, as shown in Fig. 12a and b by
of reactor, but production rate in the first two reactors is more than the last solid lines, molar flow rate of aromatics increases during the process.
two reactors, because dehydrogenation as a fast reaction occurs mostly in
the first two reactors that pro-duces aromatic components. Also, heavy Hydrogen, as a main byproduct in the reforming pro-cess, has a dual
+
aromatics (A9 ) have similar conditions, but in the last reactor due to role in the naphtha reforming as a product and reactant. It acts as a
consumption of heavy aromatics in the slow reactions (hydrodealkylation product in the dehydrogena-tion and dehydrocyclization reactions and as a
reactant in the hydrodealkylation and hydrocracking reactions. But total
hydrogen during the process increases because its production rate is more
than the consumption rate. As shown in Fig. 13, hydrogen molar flow rate
increases during the process but production rate decreases, because, as
previously discussed, hydrodealkylation and hydrocracking reactions, as
slow reac-tions, take place in the last two reactors. Other products in the
naphtha reforming process are light ends that have a main role in the
refineries. Light ends are produced in the hydroc-racking reactions that
occur in all reactors. But, owing to these reactions take place mostly in the
last two reactors, produc-tion rate of light ends increases in these reactors,
as shown in Fig. 14.
a a
900 300
(kmol/h)
Molar flow rate along inner annulus of reactor Molar flow rate along inner annulus of reactor
700
rate
O-xylene
500
200
400
150
100
Normal paraffin
300
200
100 50
Eight
Ethylbenzene
0
0
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
Annulus Length(Radius) of reactor (Dimensionless) Annulus Length(Radius) of reactor (Dimensionless)
b b
500
1000 Molar flow rate along inner annulus of reactor
900 Molar flow rate along inner annulus of reactor 450 Mean molar flow rate along radius of reactor
Mean molar flow rate along radius of reactor
800
600
350
700 300
500
250 Benzene
200
molar
400
Aromatic
150
Iso-paraffin
200
300 100
100
50
0
0
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
Annulus Length(Radius) of reactor (Dimensionless)
Annulus Length(Radius) of reactor (Dimensionless)
Fig. 11 – Variations of mloar flow rate of (a) normal-paraffin and (b) iso- Fig. 12 – Molar flow rate of (a) aromatics with eight carbon atoms and
paraffin in the process. (b) Toluene, Benzene and heavy aromatics with A9+.
8000
6000
components. Thus, during the process, the trend of viscosity obeys
temperature conditions. As shown in Fig. 15b, viscos-ity decreases along
the radius of reactors due to temperature reduction, and an ascendant trend
can be observed during the process, because the outlet stream of reactors
Hydrogen
4000
is heated by the furnace. One of the other significant physical properties in 2000
1800
Molar flow rate along inner annulus of reactor
1600 Mean molar flow rate along radius of reactor
(kmol/h)
1400
1200
Light end molar flow rate
1000
800
600
400
200
130
120
100
1.75
1.65
1.6
9. Conclusions
30
superficial velocity, and hydrogen to hydrocarbon molar ratio (Antos and
Molecular
25
Aitani, 1995) have effect on the per-formance of the catalytic naphtha
reforming process. In this work, variations of these parameters along the
radial and axial directions of reactors, also the variations of proposed
pseudo components in the reforming process have been investigated.
Acceptable agreement has been obtained between commer-cial values and
simulated results of the molar flow rate of pseudo components, coke
content of catalysts, operating con-ditions and physical properties which
20 indicate the acceptable prediction ability of the proposed kinetic model.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Radius of reactor (Dimensionless)
Fig. 15 – Physical properties in the process (a) heat capacity (b) viscosity
(c) molecular weight.
1722 CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727
Appendix A. Catalyst deactivation model Then, Eq. (A9) is replaced in Eq. (A6):
uz ∂CC =r
CM
(A10)
M
Reaction rate of pseudo-components based on activity of cat-alyst and
∂z
reaction rate on the fresh catalyst can be obtained by:
By rearranging above equation, we have:
∂C = r (A11)
CM CM
o (A1)
ri = a i × ri
∂z uz
Then, by considering catalyst function for each reaction,
Thus, the rate of reaction that is catalyzed by metallic function could be Thus, the coke weight fraction on the metallic function (CCM ),
written as: depends on the rate of coke formation on the metal-lic function of spent
catalyst (rCM ) and catalyst velocity (uz ). Therefore to solve Eq. (A11),
o (A2)
ri = aM × ri the rate of coke formation must be calculated:
∂F (A6) (CCM ) could be calculated. With same procedure, the rate of reactions that
= rCM × ( B × 2 × × r × r)
CM are catalyzed by acidic site could be achieved, the results shown in Table
∂z 12. In addition, required constants to calculate catalyst deactivation rate
and activity are presented in Table A.1.
where FCM and rCM , are mass flow rate of coke and rate of coke
formation on the metallic function of catalyst. In addition, FCM can be
defined as: Appendix B. Developing of governing
. (A7) equations
F =C ×m
CM CM Cat.
B.1. Mass balance
Also, mass flow rate of catalyst is defined by:
(A9)
FCM = CCM × ( b × 2 × × r × r × uz ) Input − Output + Consumption = Accumulation (B1
)
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1723
kg coke
The molar flux comprises two terms, ones due to the bulk motion and
nA 1 –
the other one for diffusion of component j as fol-lows:
˛A 14.5 kg
cat
kg ∂Cj (B9)
coke
nC A 1 – Nj = −Dej ∂r + Cj ur
10.18 kg or
˛CA cat
0.5 ∂r
kgcat (kmol)
1.386 n 1.5
kCA kgcoke ×(kPa) 1 m
0.5 By substituting the above equation in Eq. (B8), we have:
kgcat (kmol)
J
Ec 4055
1 ∂ ∂Cj 1 ∂ m
mol
n1 0.94 – D (Ar )−
ej Ar ∂r ∂r Ar ∂r (Ar ur Cj ) + bai
i=1
n2 1.33 –
× ij ri = ε∂C j = 1, 2, ..., n (B11)
j
∂t i = 1, 2, ..., m
where nj is mole of component j in control volume and Nj is molar flux of − 1 ∂ (A u C ) = −ur Cj ∂A ∂Cj ∂ur
r r j r
component j, by writing the Taylor’s exten-sion of second term in left
Ar ∂r Ar ∂r − ur ∂r − Cj ∂r
hand of mass balance equation, is obtained:
ur Cj ∂Cj ∂ur (B13)
= − r − u ∂r − C ∂r
m r j
∂(Nj × Ar ) ∂nj
− (B3)
· dr + b ai × vij ri × Ar × dr =
∂r i=1 ∂t
By substituting recent equation in Eq. (B12):
ur Cj ∂Cj
m j = 1, 2, ..., n
Then, by dividing the above equation over the volume of the ∂ur
differential element: −
r − ur ∂r − Cj ∂r + b 1ai × ij ri = 0
i i = 1, 2, ..., m
m =
1 ∂(Nj × Ar ) 1 ∂nj (B14)
− (B4)
+ b ai × vij ri =
Ar ∂r i=1
Ar × dr ∂t
Consequently, final equation of mass balance in the CCR pro-cess obtains:
m j = 1, 2, ..., n
To simplify the above equation, by considering concentra-tion of ∂Cj Cj Cj ∂ur b
= + eff − b × ij + ×
within the by conduction by conduction Ar ∂r ∂r n i j ∂t
n ∂Hj n ∂Uj ∂Cj
n ∂U j=1 ∂C
system by massto the system i=1
the system surroundings j j j
− N = ε C + ε (U − H ) (B22)
(B16) j=1 j ∂r j=1 j ∂t j=1 j j ∂t
n
∂Hj = ∂Hj × ∂T = CP × ∂T
∂( j
∂r ∂T ∂r ∂r
N AH = (B17)
− n j r j = i 1, 2, ..., m n n ∂T
nj U j ) ∂Hj
j = 1, 2, ..., n
∂t
=1
j
j 1
+
= N )
r dr
j =((Jj + Cj ur ) × CPj ×
j=1 ∂r ∂r
j=1
Uj is internal energy component j in the system and Hj is enthalpy of n n
∂T ∂T
component j, by considering the Taylor’s exten-sion, we have: =
CC )
∂r j=1 ((Jj + Cj ur ) × CPj ) ur ( j Pj ∂r
j=1
(K A )− n ∂r
∂r Ar n n n
Ar ∂r eff r
Cj
∂( N j A r Hj )
1 ∂ ∂T 1 j=1
n n ( CC ) C ( ( )C ) C yC C C
1 ∂Uj 1 ∂nj j Pj = T j=1 CT Pj = T j Pj = T P
j=1 j=1
= n + U (B18)
Ar × dr j ∂t Ar × dr j=1 j ∂t (Uj − Hj ) = −Pvj = −RT
j=1
n n
∂Cj ∂Cj ∂CT
Substituting Eq. (B5) in the above equation obtains:
j=1 (Uj − Hj ) ∂t = −RT ∂t = −RT ∂t
1 ∂ ∂T 1 n n j=1
∂(Nj Ar ) ∂Hj
(Keff Ar )− ∂Uj ∂Uj ∂T ∂T
Ar ∂r ∂r Ar Hj ∂r −Nj ∂r
∂T × ∂t = CVj × ∂t
=
j=1 j=1
∂t
n n
∂Uj ∂Cj C C R
(B19) Vj = Pj −
= εCj ∂t + εUj ∂t
j=1 j=1
CV = CP − R
The term ∂(Nj Ar ) is replaced from Eq. (B8):
Also:
∂r
(K A ) Hj − b j n n ∂T ∂T n n ∂T
∂r + n ai × vij ri + ε ∂t
Ar ∂r eff r ∂Uj
1 ∂ ∂T m ∂C
C C = CC ( CC )
n j=1 i=1
Cj = j Vj × j Vj = j Vj
j=1 ∂t ∂t ∂t j=1 ∂t
j=1 j=1
n ∂Hj n (B20)
− Nj ∂r
j=1
∂Uj ∂Cj
=ε +ε n n n
Cj ∂t Uj ∂t Cj
j=1 j=1 (C C ) C ( ( )C ) C yC C C
j Vj = T j=1 CT Vj = T j Vj = T V
(B23)
(B24)
(B25)
(B26)
(B27)
(B28)
(B29)
(B30)
(B31)
(B32)
(B33)
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1725
n n
In the above equations, CV and CP are the average heat capacities of ∂Fj
mixture at constant volume and pressure, respec-tively. Thus, by applying (C5)
j=1 ∂z = 2 Ri ure(Cje )
these auxiliary equations, we have: j=1
1 ∂ ∂T m
∂z
(Keff Ar
Ar ∂r ∂r ) − b ( Hi × ai × ri )
i=1 Also, energy balance in the Z coordinates for a control volume is:
∂T ∂T ∂CT (B34)
− ur CT CP ∂r = εCT CV ∂t − εRT ∂t
n
And then,
FT CP T|z − FT CP T|z+dz + Fje Cpje (Te − T) = 0 (C7)
j=1
1 ∂ ∂T m
M n (C9)
Due to steady state conditions and low amount of conduc-tive heat
transfer in comparison with convective heat transfer, therefore, heat
M j Fj
transfer simplify to: j=1
=
F
∂T
m T
n
−ur CT CP ∂r − b( Hi × ai × ri ) = 0 (B36)
Fj Cpj
i=1
j=1
Cp = (C10)
Consequently, final equation for energy balance in the CCR process FT
obtains: n n C n
je
∂T m
b
=− ( Hi × ai × ri ) (B37) (C C ) C (( )C ) C (y C ) C C
∂r ur CT CP je Pje = Te C
Te Pje = Te je Pje = Te Pe
i=1 j=1 j=1 j=1
(C11)
Appendix C. Collector equations
To represent the governing equations in the collector, mass and energy By substituting the above equation in governing energy bal-ance:
balances have been applied for a control volume with length of dz and
2
cross section area of Ri , thus, the mass balance for control element in Z
n
direction is:
FT CP T|z − FT CP T|z+dz +Cje ure 2 Ri × dz × Cpje (Te − T) = 0 j=1
Fj |z − Fj |z+dz + Fje = 0 (C1) (C12)
Fje = Cje ure Are = Cje ure 2 Ri × dz (C2) ∂(FT CP T) = 2 Ri × ure × CTe × Cpe × (Te − T) (C13)
∂z
where Cje and ure are concentration of component j and radial velocity in
inner radius of reactor. By substituting recent equa-tion in Eq. (C1), final Now, by applying algebraic definitions Eq. (C13) simplifies:
equation for mass balance in the collector obtains:
Then, to calculate total molar flow rate FT, the summa-tion of recent 1 ∂T + 1 ∂C + 1 ∂F = 1 ×2 R
P T i
equation for all pseudo-components must be considered:
n n T ∂z
∂Fj CP ∂z FT ∂z F T CP T
= (2 Ri ure Cje ) (C4) × ure × CTe × Cpe × (Te − T) (C15)
j=1 ∂z
j=1
1726 CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727
And by rearranging Eq. (C15) we have: deactivation and on product yield and quality during naphtha catalytic
reforming. Appl. Catal. 13, 113–125.
∂T = − T ∂C T ∂FT 1 Carter, J.L., Mcvinker, G.B., Weissman, W., Kmak, M.S., Sinfelt, J.H., 1982.
P
∂z C ∂z − F ∂z +
FT C P
Bimetallic catalysis; application in catalytic reforming. Appl. Catal. 3, 327–
P T 346.
× 2 Ri × ure × CTe × Cpe × (Te − T) (C16) Edgar, M.D., 1983. Catalytic reforming of naphtha in petroleum refineries.
In: Applied Industrial Catalysis. Academic Press, New York, pp. 1–123.
Then by substituting Eq. (C6) in the recent equation we have: Figoli, N.S., Beltramini, J.N., Barra, A.F., Martinelli, E.E., Sad, M.R., Parera,
J.M., 1983a. Influence of total pressure and hydrogen: hydrocarbon ratio
∂T T ∂CP T 1 on coke formation over naphtha-reforming catalyst. In: Coke Formation on
∂z = − C ∂z − (2 Ri × ure × CTe ) + × 2 Ri Metal Surfaces ACS Symp. Ser., vol. 202, pp. 239–252.
P FT FT C P
× ure × CTe × Cpe × (Te − T) (C17) Figoli, N.S., Beltramini, J.N., Marinelli, E.E., Sad, M.R., Parera, J.M., 1983b.
Operational conditions and coke formation on Pt-Al2 O3 reforming
catalyst. Appl. Catal. 5, 19–32.
Appendix D. Velocity distribution Fogler, H.S., 1992. Elements of Chemical Reaction Engineering, 2nd ed.
Prentice-Hall, Englewood Cliffs, NJ.
Froment, G.B., 1987. The kinetic of complex catalytic reactions.
For velocity distribution by using mass balance in the CCR process
Chem. Eng. Sci. 42, 1073–1087.
Hongjun, Z., Mingliang, S., Huixin, N., Zeji, L., Hongbo, J., 2010. Modeling
∂Cj C C ∂ur b m j = 1, 2, ..., n and simulation of moving bed reactor for catalytic naphtha reforming.
j j
Petrol. Sci. Technol. 28, 667–676.
=− − + i a× r i = 1, 2, ..., m (D1)
∂r r Hu, Y.Y., Su, H.Y., Chu, J., 2003. Modeling and simulation of
ur ∂r ur
= 1 i ij i
commercial catalytic reformers. Chin. J. Chem. Eng. 17, 418–424.
Then the summation of recent equation for all pseudo-components must Iranshahi, D., Bahmanpour, A.M., Pourazadi, E., Rahimpour, M.R., 2010a.
be calculated Mathematical modeling of a multi-stage naphtha reforming process using
novel thermally coupled recuperative reactors to enhance aromatic
n n n n m
∂Cj Cj 1 ∂ur b production. Int. J. Hydrogen Energy 35, 10984–10993.
= − − Cj + ur ai × ij ri(D2)
j=1 ∂r j=1 r ur ∂r Iranshahi, D., Pourazadi, E., Paymooni, K., Bahmanpour, A.M., Rahimpour,
j=1 j=1 i=1
M.R., Shariati, A., 2010b. Modeling of an axial flow, spherical packed bed
n m
∂CT CT CT ∂ur reactor for naphtha reforming process in the presence of the catalyst
deactivation. Int. J. Hydrogen Energy 35, 12784–12799.
b
=− − + a× r (D3)
∂r r ur ∂r ur i ij i
j=1 i=1 Iranshahi, D., Pourazadi, E., Paymooni, K., Rahimpour, M.R., 2012. A novel
ur n m dynamic membrane reactor concept with radial-flow pattern for reacting
∂CT ur ∂ur b
material and axial-flow pattern for sweeping gas in catalytic naphtha
× =− (D4)
− + C
ai × ij ri reformers. AICHE J. 58, 1230–1247.
CT ∂r r ∂r T j=1
i=1
Iranshahi, D., Rahimpour, M.R., Asgari, A., 2010c. A novel dynamic radial
By rearranging Eq. (D4) velocity distribution equation can be obtained: flow, spherical-bed reactor concept for naphtha reforming in the
presence of catalyst deactivation. Int. J. Hydrogen Energy 35, 6261–
6275.
n m
∂ur ur ur ∂CT b
Jenkins, J.H., Stephens, T.W., 1980. Kinetics of catalytic reforming.
(D7) Hydrocarb. Process. 11, 163–167.
=− − × + ai × ij ri
∂r r CT ∂r Ct Jorge, A.J., Eduardo, V.M., 2000. Kinetic modeling of naphtha catalytic
j=1 i=1
reforming reactions. Energy Fuels 14, 1032–1037.
To find ∂CT in the recent equation, we can use ideal gas low:
Juarez, J.A., Macias, E.V., Garcia, L.D., Arredondo, E.G., 2001. Modeling
∂r
P (D8) and simulation of four catalytic reactors in series for naphtha
CT = reforming. Energy Fuels 15, 887–893.
RT Kmak, W.S., 1972. A Kinetic simulation model of the
powerformmg process. AIChE Meeting, Houston, TX.
Then by doing algebraic operations on the above relations, we have: Krane, H.G., Groh, A.B., Schuhnan, B.L., Sinfeh, J.H., 1959. Reactions in
catalytic reforming of naphthas. In: 5th World Petroleum Congress,
New York, May 30–June 5.
∂CT = 1 ∂P − P ∂T (D9) Lee, J.W., Ko, Y.C., Jung, Y.K., Lee, K.S., Yoon, E.S., 1997. A Modeling and
∂r RT ∂r simulation study on a naphtha reforming unit with a catalyst circulation and
RT ∂r
2
regeneration system. Comput. Chem. Eng. 21, 1105–1110.
References
Lieske, H., Sarkany, A., Volter, J., 1987. Hydrocarbon adsorbtion and coke
formation on Pt/Al2 O3 and Pt-Sn/Al2 O3 catalysts. Appl. Catal. 30, 69–
Aitani, A.M., 2005a. Catalytic naphtha reforming. In: 80.
Encyclopedia of Chemical Processing., Liu, K., Fung, S.C., Ho, T.C., Rumschitzki, D.S., 1997. Kinetics of catalyst
http://dx.doi.org/10.1081/E-ECHP-120039766. coking in heptane reforming over Pt-Re/Al2 O3 . Ind. Eng. Chem. Res.
Aitani, A.M., 2005b. Catalytic Naphtha Reforming. Science and 36, 3264–3274.
Technology. Marcel Dekker, New York. Liu, K., Fung, S.C., Ho, T.C., Rumschitzki, D.S., 2002. Heptane reforming
Antos, G.A., Aitani, A.M., 1995. Catalytic naphtha reforming, 2nd ed. Marcel over Pt-Re/Al2 O3 : reaction network, kinetics, and apparent selective
Dekker, New York. catalyst deactivation. J. Catal. 206, 188–201.
Barbier, J., 1986. Deactivation of reforming catalysts by coking—a review. Mahdavian, M., Shohreh, F., Fazeliz, A., 2010. Modeling and
Appl. Catal. 23, 225–243. simulation of industrial continuous naphtha catalytic
Bishara, A., Stanislaus, A., Hussain, S., 1984. Effect of feed
composition and operating conditions on catalyst
CHEMICAL ENGINEERING RESEARCH AND DESIGN 9 2 ( 2 0 1 4 ) 1704–1727 1727
reformer accompanied with delumping the naphtha feed. Int. Ramage, M.P., Graziani, K.R., Schipper, P.H., Krambeck, F.J., Choi, B.C.,
J. Chem. React. Eng. 8. 1987. A review of mobil’s industrial process modeling philosophy. Adv.
Marin, G.B., Froment, G.B., Lerou, J.J., Backer, W.D., 1983. Chem. Eng. 13, 193–266.
Simulation of a catalytic naphtha reforming unit W. Eur. Fed. Roldán, R., Romero, F.J., Jiménez1, C., Borau, V., Marinas, J.M., 2004.
Chem. Eng. 2, 1–7. Transformation of mixtures of benzene and xylenes into toluene by
Matar, S., Hatch, L.F., 2000. Chemistry of Petrochemical Processes, 2nd ed. transalkylation on zeolites. Appl. Catal. A 266, 203–210.
Gulf Publishing Company, Texas.
Mieville, R.L., 1991. Coking kinetics of reforming. Stud. Surf. Sci. Smith, R.B., 1959. Kinetic analysis of naphtha reforming with platinum
Catal. 68, 151–159. catalyst. Chem. Eng. Prog. 55, 76–80.
Mostafazadeh, A.K., Rahimpour, M.R., 2009. A membrane catalytic bed Tailleur, R.G., Davila, Y., 2008. Optimal hydrogen production through
concept for naphtha reforming in the presence of catalyst deactivation. revamping a naphtha-reforming unit: catalyst deactivation. Energy
Chem. Eng. Process. 48, Fuels 22, 2892–2901.
683–694. Taskar, U., Riggs, J.B., 1997. Modeling and optimization of a
Nauman, E.B., 2001. Chemical Reactor Design, Optimization, and Scaleup. semiregenerative catalytic naphtha reformer. AIChE J. 43, 740–753.
McGraw-Hill, New York.
Padmavathi, G., Chaudhuri, K.K., 1997. Modeling and simulation of Weifeng, H., Hongye, S., Yongyou, H., Jian, C., 2006. Modeling, simulation
commercial catalytic naphtha reformers. Can. J. Chem. Eng. 75, 930–937. and optimization of a whole industrial catalytic naphtha reforming
process on Aspen Plus platform. Chin. J. Chem. Eng. 14, 584–591.
Pisyorius, J.T., 1985. Analysis improves catalytic reformer Weifeng, W., Hongye, S., Shengjing, M., Jian, C., 2007. Multi objective
troubleshooting. Oil Gas J. 23, 147–151. optimization of the industrial naphtha catalytic reforming process.
Rahimpour, M.R., 2009. Enhancement of hydrogen production in a novel Chin. J. Chem. Eng. 15, 75–80.
fluidized-bed membrane reactor for naphtha reforming. Int. J. Hydrogen Wolff, A., Kramarz, J., 1979. Kinetic models of catalytic reforming.
Energy 34, 2235–2251. Chem. Technol. Fuels Oils 15, 870–877.
Rahimpour, M.R., Iranshahi, D., Pourazadi, E., Paymooni, K., Bahmanpour,
A.M., 2011. The aromatic enhancement in the axial-flow spherical packed-
Further reading
bed membrane naphtha reformers in the presence of catalyst deactivation.
AIChE J. 57, 3182–3198.
Perry, R.H., Green, D.W., Maloney, J.O., 1999. Perry’s Chemical
Ramage, M.P., Graziani, K.P., Krambeck, F.J., 1980. Development of mobil’s Engineers’ Handbook, 7th ed. McGraw-Hill, New York.
kinetic reforming model. Chem. Eng. Sci. 35,
41–48.