Vous êtes sur la page 1sur 11

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/265338956

Investigation of coil outlet temperature effect


on the performance of naphtha cracking
furnace
ARTICLE in CHEMICAL ENGINEERING RESEARCH AND DESIGN AUGUST 2014
Impact Factor: 2.28 DOI: 10.1016/j.cherd.2014.08.010

READS

375

5 AUTHORS, INCLUDING:
Marziyeh Hamidi

M. R. Rahimpour

Australian National University

Shiraz University

8 PUBLICATIONS 25 CITATIONS

331 PUBLICATIONS 3,178 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,


letting you access and read them immediately.

SEE PROFILE

Available from: M. R. Rahimpour


Retrieved on: 16 March 2016

CHERD-1674; No. of Pages 10

ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

Contents lists available at ScienceDirect

Chemical Engineering Research and Design


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

Investigation of coil outlet temperature effect on


the performance of naphtha cracking furnace
Kazem Barazandeh a , Ourmazd Dehghani b , Marziyeh Hamidi c ,
Elham Aryafard b , Mohammad Reza Rahimpour b,
a

Department of Chemical Engineering, Amirkabir University of Technology, No. 424, Hafez Ave., Tehran, Iran
Department of Chemical Engineering, School of Chemical and Petroleum Engineering, Shiraz University,
Shiraz 71345, Iran
c Petroleum Engineering Department, Islamic Azad University, Science and Research Branch, Fars, Iran
b

a b s t r a c t
In this study, liquid furnace has been investigated to nd out the reason of over-cracking which results in excessive tar
and coke formation so that the furnace run length is decreased. For this purpose, a one-dimensional mathematical
model is used to analyze the furnace temperature prole. In addition, inlet feedstocks and outlet gas products of
furnace have been studied. The results obtained from the simulation demonstrate differences between modeled coil
outlet temperature and cracking severity with those of experimental data. For overcoming this problem, furnace
temperature were sufciently reduced with the purpose of reaching constant cracking severity at 0.5. Correction of
the coil outlet temperature led to signicant increase in cracking severity, propylene production, furnace capacity
and run length while gasoline production and number of decoke cycles decreased.
2014 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

Keywords: Naphtha; Thermal cracking; Mathematical modeling; Over-cracking; Furnace run length

1.

Introduction

Cracking is a process whereby heavy hydrocarbons are broken


down into light and useful hydrocarbons such as parafn and
olen. Different types of cracking applied in petroleum and
petrochemical industries are:

Steam cracking: In steam cracking, a gaseous or liquid


hydrocarbon feedstock like naphtha, LPG or ethane is
thermally cracked through the use of steam without the
presence of oxygen. It is the principal industrial method for
producing the lighter olens such as ethylene and propylene.
Thermal cracking or pyrolysis: This type is the cracking of
heavy residues under severe thermal conditions. Thermal
cracking is currently used to upgrade very heavy fractions
and produce valuable light fractions and feedstocks for the
petrochemical industry.

Catalytic cracking: this process involves the presence of acid


catalysts (usually solid acids such as silica-alumina and zeolites) (Xu et al., 2011; Choudhary and Mulla, 2001).
Hydrocracking: Hydrocracking takes place on catalyst in
the presence of elevated partial pressure of hydrogen gas.
According to the reaction conditions (temperature, pressure
and catalyst activity), products of this process vary from
ethane and LPG to heavier hydrocarbons which are all saturated hydrocarbons.
In this case, steam thermal cracking of naphtha has been
studied. Steam cracking of naphtha is applied to olens
production with three main products including ethylene,
propylene and butadiene. Propylene has been the fastest growing of the three main olens, mainly because of the strong
demand for polypropylene. acrylonitrile (ACN), oxo alcohols,
propylene oxide, cumene and acrylic acid are other important
derivatives of propylene (Xu et al., 2012). Due to the increase

Corresponding author. Tel.: +98 711 2303071; fax: +98 711 6287294.
E-mail address: rahimpor@shirazu.ac.ir (M.R. Rahimpour).
http://dx.doi.org/10.1016/j.cherd.2014.08.010
0263-8762/ 2014 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

CHERD-1674; No. of Pages 10

ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

in demand for propylene, alternative ways for production of


propylene are available such as uid catalytic cracking (FCC)
(Wang et al., 2008), methanol-to-olens (MTO) and methanolto-propylene (MTP) (Mei et al., 2008). However, nowadays
propylene is mainly produced as a co-product of ethylene via
steam cracking of naphtha (Xu et al., 2012). Steam cracking
is an endothermic and energy intensive process which takes
place in pyrolysis furnace (Choudhary and Mulla, 2001).
The furnaces are the most important and expensive pieces
of equipment in the unit and they are considered as the heart
of the ethylene plant. Each furnace is composed of convection
and radiation sections and reactions take place in the second
section of pyrolysis coils (Maciel Filho and Sugaya, 2001). The
pyrolysis yield pattern is strongly affected by the hydrocarbon
partial pressure, the temperature prole, and the residence
time. Moreover, moderation of radiant heat ux ensures the
coke formation rate is kept to a minimum. By carefully optimizing these parameters, the best yield pattern are obtained
while retaining a long continuous on-stream time between
successive decoke operations.
For developing the cracking severity parameter, some criteria exist such as: the coil outlet temperature, the extent of feed
gasication, product yield ratio and the rate of decomposition
of model components (Shu and Ross, 1982). In commercial
units, only easily accessible exit yields and parameters can be
used as a severity factor (Ghashghaee and Karimzadeh, 2011).
In this study, the propylene/ethylene mass ratio (P/E) has been
proposed as the best index to calculate the severity for optimizing the furnace performance since measuring the amount
of propylene and ethylene with chromatography analyzing of
furnace efuent gas is so easy.
In furnaces, temperature is one of the most important process variable since it inuences not only the pyrolysis reaction,
but also determines the mechanical limits for furnace design.
At a xed residence time, higher temperatures will increase
the depth of cracking and ethylene yield, up to a point beyond
which over-cracking and product degradation occur, leading
to excessive carbon and polymer formation.
Typical run length (the period between two successive
decoking periods) of cracking furnaces is 3090 days depending on the feed, operating conditions and the reactor types
(Shokrollahi Yancheshmeh et al., 2013). During this run length
and endothermic cracking process, a large amount of coke
deposits on the inner walls of the tubular cracking reactor (Choudhary and Mulla, 2001). Coke formation leads to
serious problems, such as reduction in olen selectivity as
a consequence of increased pressure drop in the cracking
coil, elimination of one or more products at the end of the
period (Edwin and Balchen, 2001), decrease in wall heat transfer in reactor and at last it reduces the life of reactor tubes
(Choudhary and Mulla, 2001). When the skin temperature of
cracking coil reaches its maximum allowable value, decoking must be carried out via a mixture of air and steam
(Shokrollahi Yancheshmeh et al., 2013; Edwin and Balchen,
2001; Heynderickx et al., 2006; Jianxin et al., 2012). The furnace
must rst be taken out of production in decoking procedure
and then a ow of steam or a steam/air mixture is passed
through the furnace coils. It converts the hard solid carbon
layer to carbon monoxide and carbon dioxide. Once this reaction is complete, the furnace can be returned to service.
The main effects of thermal cracking plant and its optimization have been investigated in various cases. Basu and
Kunzru (1992) reported the effect of temperature, residence
time and weight ratio of steam to naphtha on the plant.

Table 1 Design feed of liquid furnace.


Feedstock

Design ow rate (kg/h)

Mass fraction

Light end
Rafnate
LPG
C5 cut
C5 +
Propane recycle
C4 cut recycle

21,645
11,995
2620
1736
2873
937
8390

0.4312
0.2390
0.0522
0.0346
0.0572
0.0187
0.1671

With increasing temperature, yields of methane and ethylene


increased whereas the yield of propylene passes through
maxima. Bajus et al. (1980) has studied the effect of residence
time, the temperature, the reactor material, the ratio of
steam to hydrocarbon and the sulfur compounds presence on
products yields and coke formation. Pinter et al. (2004) have
indicated that the optimal yields of ethylene and propylene
can be achieved at intermediate pyrolysis temperatures and
residence times.

2.

Process description

The cracking plant is one of the plants composing the 10th


olen complex receiving its feed from other complexes. Liquid
furnaces are designed to crack liquid feedstock and recycled
propane and butane.
For supplying the liquid furnaces feed, the C5 + feed, the C5
cut, rafnate and light ends are mixed and sent to the liquid
surge feed drum. The mixture is pump out and then mixed
with the LPG feedstock. The propane and C4 cut recycles are
mixed to the other liquid feeds at the discharge of the pump.
The liquid feedstock is then preheated before being sent to the
furnaces. Feedstocks specications are given in Table 1.
The main sections of a liquid furnace are described below.

2.1.

Convection section

Since the radiant coil operates at high temperature, the residual part of the red duty is recovered in the furnace convection
section. At each furnace, the feed and dilution steam are
heated by ue gas in their respective convection banks prior
to mixing. The purpose of these convection banks is to heat
the feed/steam mixture to the required radiant coil inlet temperature.

2.2.

Radiant section

As the feed/steam mixture enters the radiant coil, it is


rapidly heated to a temperature at which pyrolysis is initiated.
Thermal cracking is sustained over the radiant coil by the continuous supply of heat from the radiant gases surrounding the
coil. The cracked gas exits the radiant section, owing upward
through the roof of the rebox.

2.3.

Temperature quench system

In order to stop the reaction at the radiant coil outlet, the pyrolysis product is rapidly cooled, or quenched, in the transfer line
exchanger (TLE) by several hundred degrees in a time interval
of milliseconds.

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

ARTICLE IN PRESS

CHERD-1674; No. of Pages 10

chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

Table 2 Feedstock composition.

Table 4 Coil properties.

Feedstock

Mass fraction

Parameter

Value

Light end
Rafnate
LPG
C5 cut
C5 +
Propane recycle
C4 cut recycle

0.70
0
0
0.03
0.03
0.1
0.14

Radiant coil type


Number of inlet parallel tubes
Number of outlet tubes
Inlet tube length (m)
Outlet tube length (m)
Inlet tube diameter (m)
Outlet tube diameter (m)
Inlet tube wall thickness (m)
Outlet tube wall thickness (m)
Maximum inlet temperature ( C)
Maximum outlet temperature ( C)
Coil inlet temperature ( C)
Coil outlet temperature ( C)
Coil inlet pressure (bara)
Coil outlet pressure (bara)
TLE inlet pressure (bara)
Feed ow rate per coil (kg/h)
Steam ow rate per coil (kg/h)
Steam feed ratio

GK5
2
1
12.068
13.028
0.048
0.068
0.006
0.0065
950
1110
600
865
2.6
2.05
2.0
781.25
390.625
0.5

3.
Complications associated with the
furnace
The liquid furnace problems are:
(1) The liquid furnace feed is not the same as design ow
rate, therefore, furnaces operating conditions should be
regulated according to the received feed so that the outlet
products be in the optimal condition.
(2) When the furnace is adjusted at its full capacity, the fuel
gas pressure guard restricts the fuel gas ow to prevent
the emergency furnace shutdown. Once the pressure of
gas fuel in furnace exceeds its allowable limit (2.2 bar),
emergency shutdown alarm system of furnace will be
activated and furnace efuent temperature will decline
below 865 C. For overcoming this problem, according to
the feed specications, the furnace should operates at its
90% capacity to control coil outlet temperature at 854 C.
(3) All furnaces have been designed to achieve a run length
of at least 60 days, but average of their run length never
exceeds 45 days. This problem leads to reduction in
operating time and repetition of furnace decoking which
increases usage of utilities and expenditure more time
compared with ordinary operation.
To overcome these problems, inlet feed and outlet gas
product of furnace have been investigated. During this study,
received feed composition was nearly constant. The composition of feed and its analysis are presented in Tables 2 and 3,
respectively.
The properties of liquid furnace coil are provided in Table 4.
The sample of liquid furnace outlet gas was analyzed
with Renery Gas Analyzer (RGA) with detection accuracy of
0.01 mol%. The analysis of gaseous samples has been done

Table 3 Analysis of furnace feed components.


Component

Mass fraction

Propane
Vinyl acetylene
n-Butane
i-Butane
22-Mbutane
tr2-Butene
1-Butene
i-Butene
2M-1-butene
n-Pentane
i-Pentane
2-Mpentane
3-Mpentane
Cyclopentene
n-Hexane

0.11
0.005
0.285
0.02
0.01
0.006
0.05
0.014
0.01
0.21
0.205
0.025
0.015
0.02
0.015

according to ASTM D2504 standard for hydrogen and ASTM


D6159 standard for other components.

4.

Kinetic model

The kinetic model could be sorted into three kinds: molecular reaction kinetic model, empirical model and mechanism
model of free radical reaction (Geng et al., 2012).
Among different models for naphtha cracking, a set of 22
molecular reactions have been proposed in current work. 20
reactions of them are postulated by Gao et al. (2009) in addition
to 8 radical reactions and 7 coke formation reactions are based
on Zou et al. (1993) model.

First order reactions

Naphtha 0.5H2 + 0.76CH4 + 1.16 C2 H4 + 0.13 C2 H6


+ 0.38 C3 H6 + 0.09 C3 H8 + 0.008 C4 H10 + 0.245 C4 H8
+ 0.113 C4 H6 + 0.08 C4 H4

(1)

2C2 H6 C3 H8 + CH4

(2)

C3 H8 C2 H4 + CH4

(3)

C4 H10 C3 H6 + CH4

(4)

C4 H10 2C2 H4 + H2

(5)

C4 H10 C2 H4 + C2 H6

(6)

C4 H8 C4 H6 + H2

(7)

vC3 H5 C2 H2 + CH3

(8)

Reversible reactions
C2 H6 C2 H4 + H2
C3 H6 C2 H2 + CH4

(9)
(10)

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

CHERD-1674; No. of Pages 10

ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

C4 H10 C4 H8 + H2

(11)

First order reactions

C3 H8 C3 H6 + H2

(12)

r1 = k1

C3 H4

C3 H4

(13)
r2 = k2

1-C4 H8 2-C4 H8
H

(14)

(15)

H + 2-C4 H8 2-C4 H9

(16)

H + 1-C4 H8 i-C4 H9

(17)

+ 1-C4 H8 1-C4 H9

r3 = k3

r4 = k4

r5 = k5

1-C4 H9 C2 H4 + C2 H5

(18)

C2 H5 + C2 H5 n-C4 H10

(19)

H + C3 H4 vC3 H5

(20)

H + C3 H4

aC3 H5

r6 = k6

r7 = k7

(21)
r8 = k8

F

(22)

C3 H8 + C2 H4 C2 H6 + C3 H6

(23)

2C3 H6 3C2 H4

(24)
(25)

C2 H2 + C2 H4 C4 H6

(26)

C2 H4 + C2 H6 C3 H6 + CH4

(27)
(28)

C3 H6 + C4 H6 C7 H8 + 2H2

(29)

C4 H8 + C4 H6 C8 H10 + 2H2
C4 H6 + C4 H6 C8 H8 + 2H2

C2 H4 Coke

C6 H6 Coke

(34)

C7 H8 Coke

(36)

C4 H6 Coke

(37)

Ft

(42)

RT

F

 P 

C4 H10

Ft

(43)

RT

F

 P 

C4 H10

Ft

RT

C4 H8

 P 

F

(44)

Ft

(45)

RT

F

 P 

vC3 H5

Ft

(46)

RT

Reversible reactions
r9 = k9

F

r11 = k11

RT

F

C4 H10

C4 H10

Ft

 P 

FC3 H4

Ft

RT

1-C4 H8

 P 

F

Ft

RT

F1-C4 H8 FH

RT

F2-C4 H8 FH

 P 2 
t

Ft2

RT

F1-C4 H8 FH

 P 2 
t

Ft2

F

1C4 H9

Ft

RT

 P 
t

RT

FC2 H5 FC2 H5

FH FC3 H4
Ft

 P 2 
RT

 P 
RT

2-C4 H8

 P 

1-C4 H9

 P 
t

Ft

RT

F

2-C4 H9

 P 
t

Ft

RT

Fi-C4 H9

 P 
t

RT

Ft

FC2 H4 FC2 H5

kR20

 P 2 
t

Ft2

kR19

(52)

RT

F

(51)

Ft

kR17

(50)

Ft

RT

FC3 H4

F

kR18

RT

kR16

(49)

 P 2 

Ft 2

kR15

 P 2 

Ft2

kR14

Ft2

kR13

 P 2 

RT

FC3 H6 FH2

(48)

 P 2 

Ft 2

kR12

RT

RT

FC4 H8 FH2

 P 

 P 2 

Ft 2

(47)

RT

FC2 H2 FCH4

kR11

RT

F

r17 = k17

kR10

Ft 2

 P 

Ft

r16 = k16

r20 = k20

Ft

r18 = k18

 P 

C 3 H6

 P 2 

FC2 H4 FH2

kR9

RT

F

r13 = k13

r14 = k14

Ft

 P 

C2 H6

(38)

For above reactions, rate equations are given by following


relations:

 P 

C4 H10

r19 = k19
C8 H10 Coke

(41)

RT

F

(35)

C8 H8 Coke

Ft

(32)
(33)

 P 

C3 H8

r15 = k15

C3 H6 Coke

(40)

RT

F


Coke formation reactions

Ft

(30)
(31)

 P 

C2 H6

r12 = k12

C2 H4 + C4 H6 C6 H6 + 2H2

(39)

RT

F

r10 = k10
C3 H6 + C2 H6 C4 H8 + CH4

Ft

Second order reactions


aC3 H5 + C3 H6 C2 H5 + C4 H6

 P 

C6.5 H14

RT

F

n-C4 H10

 P 
t

Ft

RT

vC3 H5

 P 

F

Ft

RT

(53)

(54)

(55)

(56)

(57)

(58)

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

ARTICLE IN PRESS

CHERD-1674; No. of Pages 10

chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx


r21 = k21

 P 2 

FH FC3 H4

Ft2

RT

kR21

F

aC3 H5

Ft

 P 
t

RT

Table 5 Arrhenius parameters for naphtha thermal


cracking.

(59)

Rate coefcient

Second order reactions


r22 = k22

Ft2


r23 = k23

r24 = k24

FC3 H8 FC2 H4

FC3 H6 FC3 H6

FC2 H6 FC3 H6

r27 = k27

FC2 H2 FC2 H4

FC2 H6 FC2 H6

FC2 H4 FC4 H6

FC3 H6 FC4 H6

 P 2 
t

 P 2 
t

 P 2 
t

 P 2 
t

RT

FC4 H8 FC4 H6
Ft2

 P 2 

RT

Ft2

r30 = k30

RT

Ft2

r29 = k29

 P 2 

RT

Ft2

r28 = k28

RT

Ft2

 P 2 

RT

Ft2

r26 = k26

RT

Ft2

r25 = k25

 P 2 
RT

Ft2

r31 = k31

FaC3 H5 FC3 H6

 P 2 
t

RT

FC4 H6 FC4 H6
Ft2

 P 2 
t

RT

k1
k2
k3
k4
k5
k6
k7
k8
k9
kR9
k10
kR10
k11
kR11
k12
kR12
k13
kR13
k14
kR14
k15
kR15
k16
kR16
k17
kR17
k18
kR18
k19
kR19
k20
kR20
k21
kR21
k22
k23
k25
k26
k27
k28
k29
k30
k31
k32
k33
k34
k35
k36
k37
k38

(60)

(61)

(62)

(63)

(64)

(65)

(66)

(67)

(68)

(69)

Coke formation reactions


r32 = k32

r33 = k33

r34 = k34

r35 = k35

r36 = k36

r37 = k37

r38 = k38


F

C2 H4 Pt

1.34 
(70)

Ft RT


F

C3 H6 Pt

1.34 
(71)

Ft RT


F

C6 H6 Pt

1.34 
(72)

Ft RT


F

C7 H8 Pt

1.37 

C8 H8 Pt

C4 H6 Pt

(74)

1.37 
(75)

Ft RT


F

C8 H10 Pt

Ft RT

5.

1.37 

Ft RT


F

6.565E+11
3.750E+12
4.692E+10
7.000E+12
7.000E+14
4.099E+12
1.000E+10
1.585E+12
7.943E+13
7.943E+08
5.012E+13
7.943E+08
1.637E+12
1.780E+04
5.888E+10
9.030E+02
3.162E+13
1.585E+13
2.239E+12
1.950E+12
2.512E+10
1.288E+13
1.000E+11
1.000E+14
1.000E+11
1.000E+14
1.514E+14
1.995E+08
1.995E+09
5.012E+16
3.467E+10
1.412E+12
3.467E+10
8.912E+12
6.026E+07
2.536E+10
1.000E+11
1.026E+09a
7.083E+10a
8.385E+06
9.740E+05
5.610E+11
1.510E+06
5.000E+14
2.770E+09
5.610E+18
1.390E+12
1.390E+12
1.390E+12
1.390E+12

E (J/mol)
2.20E+05
2.73E+05
2.12E+05
2.49E+05
2.95E+05
2.56E+05
2.09E+05
1.57E+05
2.97E+05
1.59E+05
2.93E+05
1.59E+05
2.61E+05
1.35E+05
2.15E+05
9.34E+04
2.64E+05
2.57E+05
2.51E+05
2.59E+05
8.37E+03
1.63E+05
1.05E+04
1.72E+05
8.37E+03
1.59E+05
1.30E+05
3.18E+04
0
3.47E+05
8.37E+03
1.57E+05
8.37E+03
2.47E+05
9.00E+04
2.47E+05
2.51E+05
1.73E+05
2.53E+05
1.44E+05
1.49E+05
2.42E+05
1.24E+05
2.24E+05
1.16E+05
2.74E+05
1.41E+05
1.41E+05
1.41E+05
1.14E+05

Units: m3 /(mol s).

(73)

Ft RT


F

A (s1 )

1.37 
(76)

Kinetic parameters of above reactions such as Arrhenius


parameters are provided in Table 5.

Mathematical modeling

In this study, for solving the material, energy and momentum


balances, a one dimensional mathematical model has been
developed as the Reynolds number in the tubular reactors is
high enough (>105 ) (Van Geem et al., 2007a, 2007b) following
assumptions are considered for this purpose:

A plug ow regime is assumed,


The gas mixture is an ideal gas,
Radial concentration gradients are insignicant,
Hydrodynamic and thermal entrance region effects are negligible,

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

ARTICLE IN PRESS

CHERD-1674; No. of Pages 10

chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

Axial diffusion of mass and heat are neglected,


A quasi-steady state condition has been considered.

875
850

dj
dz

Mj
G

rri S(i.j)

(77)


4
dT =
G Cp
rri Hi + Uo (Te T)
dZ
D

Temperature (C)

825

With the above assumptions, the governing equations were


obtained as follows:

775
COT at measuring point

750
725
700
675

(78)

650
625

600

U0 (Te T) = C1 (T4 Te4 ) + C2 (Tg4 Te4 ) + U(Tg Te)

(79)

d
G2
2F G2
dp

= 2

dz
gD
 g dz

(80)

Eqs. (77)(80) are material balance, energy balance in tube


side, energy balance and momentum balance for component
j, respectively.
The friction factor for the straight parts of the cracking coil
and the tube bend are considered respectively as follows:
f = 0.092

Re0.2
D

(81)

f = 0.092

Re0.2

+
D
D

(82)

tube bend parameter () is expressed as follows:


=

0.7 + 0.35


90



0.051 + 0.19

D
Rb


(83)

The rate of coke formation can be calculated from the following equation:
Mc rc
C
= (D 2tc )
t
4c

6.

(84)

Model solution

For solving a set of ordinary differential equations (ODE)


including mass, energy and momentum balance equations
which are coupled with non-linear algebraic equations, forward nite difference method is applied to the system of
ordinary differential equations. Solving of these equations has
been done in MATLAB programming environment.

7.

Simulated COT

800

Results and discussion

Model validation was carried out by comparing the mathematical modeling results with the design data in the
manufactures documents over 195 operating days. Table 6
compares modeling results and design data. As can be seen,
they were in good agreement, while comparison between
experimental data and model results indicated that they were
not the same. For eliminating this error, sampling was done
with more accuracy, but new data conrmed the differences
between experimental and modeling results again.
For matching the modeling results to cracked gas analysis,
furnace temperature was sufciently reduced with the purpose of reaching constant cracking severity at 0.5. Furnace
efuent temperature declined from 865 C to 850 C as a consequence of reduction in furnace temperature.

10

12

14

16

18

20

22

24

26

28

30

Coil length (m)

Fig. 1 COT correction effect on thermal cracking severity,


from October 9, 2012 to April 21, 2013.
On the other hand, investigating the coil properties and
location of coil outlet temperature (COT) sensor, indicated
that COT of rebox roof was 865 C. Pyrolysis continues in
the transfer-line volume. Since the reaction is endothermic,
and there is no heat input over the transfer-line volume, the
reaction consumes a small amount of sensible heat and the
temperature drops to 15 C along this line. Therefore, to control
the coil outlet temperature at 865 C, the temperature measured by sensor must be set at 850 C.
Fig. 1 illustrates the temperature prole along the coil (from
entrance to 25.1 m) and reduction of temperature in transfer
line (from 25.1 m to 29.2 m) where adiabatic cracking takes
place.
Fig. 2 shows the position of temperature measuring at modeling and industrial furnace.
On January 17, 2013, according to the mismatch between
the modeling and industrial temperature results, outlet temperatures of three furnaces operating at 90% capacity were
reduced so that coil outlet temperature at furnace roof does
not exceed design temperature. This alteration indicated different conclusions that will be mentioned in this section.

7.1.

Effect of coil outlet temperature correction

Correction of the coil outlet temperature led to signicant


increase in propylene production and decrease in gasoline
production, this effect was clearly observed in experimental
studies.
Figs. 3 and 4 illustrate the comparison of experimental
data and modeling results, from October 9, 2012 to April 21,
2013 before and after COT correction. As can be seen, propylene yield increases after correction, but there is no specic
change in the amount of ethylene. It should be mentioned
that in addition to the enhancement in propylene selectivity,
which is one of the valuable products in olen units; methane
production that has the least value among furnace products,
was decreased. Overall, liquid furnaces protability increased
signicantly.
In liquid furnaces, design value for propylene/ethylene
mass ratio dened as thermal cracking severity is 0.5, while
Calculation of this parameter for experimental data indicated
that propylene/ethylene mass ratio was less than 0.5 before
COT correction (see Fig. 5). Therefore, it had deviation from
design and model values which was indicative of furnace over
cracking.
Fig. 6 illustrates the propylene and gasoline productions of the unit when three furnaces were running at

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

ARTICLE IN PRESS

CHERD-1674; No. of Pages 10

chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

Table 6 Comparison between modeling results and design data.


Component

Modeling results (wt.% dry)

Hydrogen
Methane
Acetylene
Ethylene
Ethane
Methyl-acetylene
Propadiene
Propylene
Propane
Vinyl-acetylene
Butadiene
Butane (sum)
Butane (sum)
Total C5 -C9 s
Total C10 +
Carbon oxide
Carbon dioxide

Design data (wt%. dry)

1.0856
19.526
0.78517
32.721
3.4519
0.7691
0.49128
16.676
0.85702
0.095277
4.4897
5.0575
2.3919
10.106
1.3736
0.11502
0.008071

Relative error %
1.4579
4.4171
12.1641
1.4428
0.3459
9.8714
22.82
1.3254
28.5817
5.8633
1.9716
3.6373
5.0833
2.8269
8.1575
4.5636
19.29

1.07
18.7
0.7
33.2
3.44
0.7
0.4
16.9
1.2
0.09
4.58
4.88
2.52
10.4
1.27
0.11
0.01

Fig. 2 Temperature prole along a coil.


90% capacity, from January 13, 2013 to January 31, 2013.
As seen, when coil outlet temperature was corrected,
propylene production increased and gasoline production
decreased.

Each furnace has been designed to achieve a run length


of 60 days, when operates under conditions specied in the
basis of design. Coke deposition growing in the coil during
furnace operation plays as a thermal insulation and leads to
35

26
C5+
Methane
Modeled Methane @865C
Modeled C5+ @865C

24

30

Ethylene
Propylene
Modeled Ethylene@865C
Modeled Propylene@865C

20
18

COT=865C

2013/1/17

COT=850C

16

Yield (Wt %)

Yield (Wt %)

22
25
COT=865C

2013/1/17

COT=850C

20

14
15

12
10
(2012/10/9)20 37 42 55 58 64 70 77 82 100102111133140145146170175 (2013/4/21)

Time (day)

Fig. 3 Comparison between experimental data and


modeling results during 195 days from October 9, 2012 to
April 21, 2013 for methane and C5 + .

10

(2012/10/9) 20 37 42 55 58 64 70 77 82 100102111133140145146170175 (2013/4/21)

Tima (day)

Fig. 4 Comparison between experimental data and


modeling results during 195 days from October 9, 2012 to
April 21, 2013 for ethylene and propylene.

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

ARTICLE IN PRESS

CHERD-1674; No. of Pages 10

chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

1050

0.54

Tube skin temperature (c )

0.52

Severity

0.5
2013/1/17

0.48

COT=865C

COT=850C

0.46
0.44
0.42
0.4
(2012/10/9) 20 37 42 55 58 64 70 77 82 100102111133140145146170175 (2013/4/21)

COT= 865c
COT= 855c
COT= 845c

1000

950

900

850

800

750

Time (day)

10

12

14

Fig. 5 Severity as a function of time before and after COT


correction.

20

22

24

26

1080
Average tube skin temperature
Maximum tube skin temperature

Tube skin temperature (c)

620
600

Product (ton/day)

18

Fig. 8 Tube skin temperature along coil length for


different COTs.

640

580
560
540

2013/1/17

COT=865C

COT=850C

520

1070
1060
1050
1040
2013/1/17

1030
COT=850C

COT=865C

1020

500
480

16

Coil length (m)

Propylene Product
Gasoline Product

1010

(2013/1/13) 3

10 11 12 13 14 15 16 17

(2013/1/31)

(2013/1/9) 3

Fig. 6 Effect of temperature correction on propylene and


gasoline production from January 13, 2013 to January 31,
2013.
reduction in the heat absorbed by process gas. Fig. 7 indicates
the coking rate along a coil and represents that coking rate
has its maximum value when coil outlet temperature is set at
865 C compared to 855 C and 845 C.
Coil skin temperature increases during cracking operation
because of the coke layer thickness. The maximum allowable
operating tube skin temperature is 1110 C. The operation of
the furnace has to be conducted in such a way that this temperature does not exceed, if not, it must eventually lead to
furnace shutdown. Fig. 8 illustrates the tube skin temperature prole along coil length for three different adjusted COT
values.

9 10 11 12 13 14 15 16 17 18 19 (2013/1/29)

Time (day)

Time (day)

Fig. 9 Effect of reduction in furnace outlet temperature on


tube skin temperature, from January 9, 2013 to January 29,
2013.
Tube skin temperature is measured with a pyrometer and
recorded every day. Fig. 9 indicates the maximum and average temperature for the tube skin. As can be seen in this
gure, maximum temperature for the tube skin reduces more
than 10 C after temperature correction. Reduction in tube
skin temperature decreases the tube carburization so the run
length will increase. On the other hand, as this temperature is
away from maximum allowable tube temperature, the lifetime
of furnaces enhances signicantly.
Fig. 10 illustrates the effect of COT correction on the furnace run length for four different operating cycles. As can be
seen, at fourth column that the furnace was running with the
70

3
COT=865 c
COT=855 c
COT=845 c

50
Run Length(day)

Coking rate(mm/month)

2.5

60

1.5

40
30
20

10
0.5
0
0

10

12

14

16

18

20

22

24

Coil length(m)

Fig. 7 Effect of coil outlet temperature on coking rate.

26

2
3
Number of Operating Cycle

Fig. 10 Effect of reduction in furnace outlet temperature


on furnace run length.

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

ARTICLE IN PRESS

CHERD-1674; No. of Pages 10

chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

4
x 10
5.2

5800

Fuel gas flow rate (kg/h)

before COT correction

100% of furnace capacity


after COT correction

5600

5400

4.8

5200

5000

Feed flow rate (kg/h)

90% of furnace capacity

4.6

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

4.4
35

Time (day)

Fig. 11 Effect of reduction in furnace outlet temperature


on furnace capacity.
corrected temperature, furnace run length increased to more
than 60 days while before the correction, run length was less
than 50 days.
The inuence of coil outlet temperature correction on furnace capacity and fuel gas consumption is shown in Fig. 11.
As seen, signicant increase in furnace capacity and decrease
in fuel gas are the results of COT correction.

8.

Conclusion

Furnace outlet temperature operating at 90% capacity were


reduced according to the modeling result so that coil outlet
temperature at furnace roof does not exceed design temperature. According to the temperature correction, the followings
results can be concluded:
Propylene production increased while gasoline and
methane production (with less economic values) reduced.
Furnace capacity increased to 100% which led to enhancement in the unit capacity and its production as a
consequence of coil outlet temperature correction.
Reduction in coil outlet temperature and its skin temperature as a result, led to reduction in coke formation rate and
increase in furnace run length.
Because of the increase in furnace run length, the number
of decoking procedure during a year reduced and using of
utilities decreased.

Nomenclature

C
C1 , C2
C p
D
Fi
f
G
g
Hi
ki
Mj
Mc
Pt
p
R

concentration of coke (mol/m3 )


constants containing emissivity factors, view factors
and StefanBoltzmann constant (kJ/s m2 K4 )
process gas specic heat (kJ/kg K)
inside tube diameter (m)
molar ow rate of component i (mol/s)
friction factor
mass ux (kg/m2 s)
dimensional constant (kg m/kgf s2 )
Heat of reaction for reaction i (kJ/kmol)
rate coefcient of reaction i (1/s or m3 /mol s)
molecular weight of component j (kg/kmol)
coke molecular weight (kg/mol)
total pressure (Pa)
pressure (kgf /m2 )
universal gas constant (J/mol K)

Rb
Re
rn
rc
rri
S(i,j)
T
Te
Tg
T
t
tc
U
U0
z

radius of the tube bend (m)


Reynolds number
rate of reaction n (mol/m2 s)
rate of coking (mol/m2 s)
rate of reaction i (kmol/m3 s)
stoichiometric coefcient of component j in reaction
i
temperature (K)
outside tube wall temperature (K)
ue gas temperature (K)
refractory wall temperature (K)
time (s)
coke thickness (m)
ue gas convective heat transfer coefcient
(kJ/s m2 K)
overall inside heat transfer coefcient (kJ/s m2 K)
length along coil (m)

Greek letters
heat of reaction (kJ/mol)
H

parameter of tube bend
process gas specic gravity (kg/m3 )

c
coke specic gravity (kg/m3 )
coking factor

angle of bend


References
Basu, B., Kunzru, D., 1992. Catalytic pyrolysis of naphtha. Ind.
Eng. Chem. Res. 31, 146155.
Bajus, M., Vessely, V., Leclercq, P.A., Rijks, J.A., 1980. Steam
cracking of hydrocarbons. 3. Straight-run naphtha. Ind. Eng.
Chem. Prod. Res. Dev. 19, 556563.
Choudhary, V.R., Mulla, S.A.R., 2001. Energy efcient
simultaneous oxidative conversion and thermal cracking of
ethane to ethylene using supported BaO/La2 O3 catalyst in the
presence of limited O2 . Appl. Energy 68, 377386.
Edwin, E.H., Balchen, J.G., 2001. Dynamic optimization and
production planning of thermal cracking operation. Chem.
Eng. Sci. 56, 989997.
Gao, G.Y., Wang, M., Pantelides, C.C., Li, X.G., Yeung, H., 2009.
Mathematical modeling and optimal operation of industrial
tubular reactor for naphtha cracking. Comput. Aid. Chem.
Eng. 27, 501506.
Geng, Z., Cui, Y., Xia, L., Zhu, Q., Gu, X., 2012. Compromising
adjustment solution of primary reaction coefcients in
ethylene cracking furnace modeling. Chem. Eng. Sci. 80,
1629.
Ghashghaee, M., Karimzadeh, R., 2011. Multivariable
optimization of thermal cracking severity. Chem. Eng. Res.
Des. 89, 10671077.
Heynderickx, G.J., Schools, E.M., Marin, G.B., 2006. Simulation of
the decoking of an ethane cracker with a steam/air mixture.
Chem. Eng. Sci. 61, 17791789.
Jianxin, Zh., Hong, X., Xiaojian, L., Xiang, L., 2012. Inuence of the
SiO2 /S coating and sulfur/phosphorus-containing coking
inhibitor on coke formation during thermal cracking of light
naphtha. Fuel Process. Technol. 104, 198203.
Maciel Filho, R., Sugaya, M.F., 2001. A computer aided tool for
heavy oil thermal cracking process simulation. Comput.
Chem. Eng. 25, 683692.
Mei, C.S., Wen, P.Y., Liu, Z.C., Liu, H.X., Wang, Y.D., Yang, W.M.,
Xie, Z.K., Hua, W.M., Gao, Z., 2008. Selective production of
propylene from methanol: mesoporosity development in high
silica HZSM-5. J. Catal. 258, 243249.
Pinter, A., Tungler, A., Nagy, L., Vida, L., Kovacs, I., Kerezsi, J., 2004.
A laboratory steam-cracking reactor to characterize raw
materials. Int. J. Chem. Reactor Eng. 2, 15426580.

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010

CHERD-1674; No. of Pages 10

10

ARTICLE IN PRESS
chemical engineering research and design x x x ( 2 0 1 4 ) xxxxxx

Shokrollahi Yancheshmeh, M.S., Seifzadeh Haghighi, S.,


Gholipour, M.R., Dehghani, O., Rahimpour, M.R., Raeissi, S.,
2013. Modeling of ethane pyrolysis process: a study on effects
of steam and carbon dioxide on ethylene and hydrogen
productions. Chem. Eng. J. 215216, 550560.
Shu, W.R., Ross, L.L., 1982. Cracking severity index in pyrolysis of
petroleum fractions. Ind. Eng. Process. Des. Dev. 21, 371377.
Van Geem, K.M., Hudebine, D., Reyniers, M.F., Wahl, F., Verstraete,
J.J., Marin, G.B., 2007a. Molecular reconstruction of naphtha
steam cracking feedstocks based on commercial indices.
Comput. Chem. Eng. 31, 10201034.

Van Geem, K.M., Zajdlk,


R., Reyniers, M.F., Marin, G.B., 2007b.
Dimensional analysis for scaling up and down steam cracking
coils. Chem. Eng. J. 134, 310.

Wang, G., Xu, Ch., Gao, J., 2008. Study of cracking FCC naphtha in
a secondary riser of the FCC unit for maximum propylene
production. Fuel Process. Technol. 89, 864873.
Xu, T., Zhang, Q., Song, H., Wang, Y., 2012. Fluoride-treated
H-ZSM-5 as a highly selective and stable catalyst for the
production of propylene from methyl halides. J. Catal. 295,
232241.
Xu, X., Li, Ch., Shan, H., 2011. Effect of phosphorus on novel
bifunctional additives for enhancing the production of
propylene and removal of SO2 in FCC process. J. Mol. Catal. A:
Chem. 340, 99107.
Zou, R., Lou, Q., Mo, S., Feng, S., 1993. Study on a kinetic model of
atmospheric gas oil pyrolysis and coke deposition. J. Ind. Eng.
Chem. 32, 843847.

Please cite this article in press as: Barazandeh, K., et al., Investigation of coil outlet temperature effect on the performance of naphtha cracking
furnace. Chem. Eng. Res. Des. (2014), http://dx.doi.org/10.1016/j.cherd.2014.08.010