Sultan Qaboos University

College of Engineering
PCE Department
Chemical Reaction Engineering

PBR Design for Phthalic

Anhydride Production

Ahmed Hamed Al-Qasmi 88805

Abdullah Ghusen Al-Abri 89672
Abdullah Mohammed Al-Kharosi 90033
Abstract

These papers illustrate a reactor design to produce Phthalic anhydride (PAN) in

continuous plant through the oxidation of o-xylene. Maleic anhydride is also obtained
as by-product in this process. Pressure drop, side reactions and maintaining a
relatively constant driving force for heat transfer are considered in this design. The
production of phthalic anhydride is approximately 20,000 tons/year as required. The
process was simulated using the process simulator Aspen plus.

I
Table of Contents

Abstract ................................................................................................................................. I
Table of Contents .................................................................................................................. II
List of Tables ....................................................................................................................... III
List of figures ....................................................................................................................... IV
Introduction .......................................................................................................................... 1
Uses .................................................................................................................................. 1
Manufacturing company .................................................................................................... 2
The feed stock ................................................................................................................... 2
Process description ............................................................................................................ 3
Reactions path way............................................................................................................ 3
Type of reactors ................................................................................................................ 4
Catalyst ............................................................................................................................. 5
Kinetics ............................................................................................................................. 6
First Design: Reactor Volume Using Simple Reaction Rate Expression .................................... 7
Mass balance: .................................................................................................................... 8
Calculating the volume ...................................................................................................... 8
The simulation .................................................................................................................. 9
Second Design: Pressure Drop and Reactor Configuration .................................................... 11
Splitting the Reactor into Tubes ....................................................................................... 13
Tube Length effect .......................................................................................................... 13
Particle Diameter effect ................................................................................................... 14
The simulation ................................................................................................................ 16
Third Design: Multiple Reactions ........................................................................................ 18
Plot Flow ........................................................................................................................ 19
Effect of Temperature in Selectivity ................................................................................ 20
Aspen simulation............................................................................................................. 21
Final Design: Energy Balance............................................................................................. 23
Temperature profile ......................................................................................................... 23
Dynamic stability ............................................................................................................ 24
Aspen Simulation ................................................................................................................ 25
Economic analysis............................................................................................................... 30
Design summary.................................................................................................................. 31
Conclusion .......................................................................................................................... 31

II
References........................................................................................................................... 32
Appendix ............................................................................................................................. 33

List of Tables

Table 1: Component Price ..................................................................................................... 2

Table 2: Fixed-Bed. vs. Fluidized-Bed Reactors for Selective Hydrocarbon Oxidation
Reactions .............................................................................................................................. 4
Table 3:Catalyst Properties and Behavior in o-Xylene Oxidation .......................................... 5
Table 4: Mass Balance Results .............................................................................................. 8
Table 5:Polymath design results ............................................................................................ 8
Table 6: Heat and material balance ...................................................................................... 9
Table 7: Design Results without Sipliting to Tube ................................................................ 11
Table 8: Heat and material balance .................................................................................... 16
Table 9: Heat and material balance .................................................................................... 17
Table 10: Third Design Results ........................................................................................... 19
Table 11: Heat and Material Results in Aspen Plus ............................................................. 21
Table 12: Components Properties ....................................................................................... 23
Table 13: Gain Calculation ................................................................................................. 24
Table 14: Gain Calculation after Modification .................................................................... 25
Table15 :Aspen Results ....................................................................................................... 29
Table 16: Total Cost Results................................................................................................ 30
Table 17: Reactor Configurations ....................................................................................... 31
Table 18: Feed Rates........................................................................................................... 31
Table 19: Products Rates .................................................................................................... 31
Table 20 : Coolant Specification ......................................................................................... 31

III
List of figures

Figure 1: Major uses of phthalic anhydride[1] ...................................................................... 1

Figure 2: Major product chain of phthalic anhydride ............................................................ 2
Figure 3: Reactions Pathway ................................................................................................ 4
Figure 4:Effect of temperature in the conversion ................................................................... 9
Figure5 : Aspen Reactor Model............................................................................................. 9
Figure6 : conversion versus heat duty ................................................................................. 10
Figure 7:Effect of temperature in the conversion ................................................................. 12
Figure 8:Effect of Pressure Drop in the conversion at Different Temprature ....................... 12
Figure 9:The effect of number of tubes in the pressure drop ................................................ 13
Figure 10:Effect of Tube Length in Pressure Drop. ............................................................. 14
Figure 11: Comparing two paticle diameters effect in pressure drop. .................................. 14
Figure 12: Final design results............................................................................................ 15
Figure 13: Shows the Catalyst Weight Effects in Pressure Drop and Conversion ................. 15
Figure 14: Molar flow rate of each component versus the catalyst weight............................ 19
Figure 15: Selectivity of Phthalic Anhydride at Different Temperature ................................ 20
Figure 16: Selectivity of Phthalic Anhydride at Different Temperature ................................ 20
Figure 17: Molar Flow of each Component with Reactor Length obtained from Aspen ........ 22
Figure 18: Temperature Profile........................................................................................... 23
Figure 19:Coolant Temperature Profile .............................................................................. 24
Figure 20: Reactor and Coolant Temperature Profiles ........................................................ 24
Figure 21:Aspen Flowsheet ................................................................................................. 25
Figure 22 : Model Type ....................................................................................................... 26
Figure 23 : Design Dimensions ........................................................................................... 26
Figure 24: Pressure Drop Calculation ................................................................................ 27
Figure 25: Catalyst specification......................................................................................... 27
Figure26 : Reactions in Aspen............................................................................................. 28
Figure 27: Kinetics Specification ........................................................................................ 28
Figure28 : Temperature Profile........................................................................................... 29

IV
Introduction

Phthalic anhydride (PAN) is an organic compound with the formula C6H4(CO)2O. It is

the anhydride of phthalic acid. This colorless solid is an important industrial chemical.
It is one of the most important products of modem large-scale organic synthesis, and
it has a wide application in various branches of chemical industry.

Uses

PAN is an intermediate petrochemical used for the production of plasticizers

especially for polyvinylchloride (PVC), unsaturated Polyesters, dyes and alkyd resins
which are used in the manufacture of reinforced plastics and in specialized
applications such as surface coatings. It is trade in either molten form or as a white
powder. About 50% of all phthalic anhydride produced is used in the manufacture of
plasticizers for PVC processing. Unsaturated polyester resins are uses (15-20) % of all
PAN produced and (10-15) % used in alkyd resins.[1]

Figure 1: Major uses of phthalic anhydride[1]

The figure below shows the major product chain of phthalic anhydride in chemical
industry.

1
Figure 2: Major product chain of phthalic anhydride

Manufacturing company

PAN is currently produced in many region of the world. In Australia one company
has two plants were built in 1960s and use naphthalene as the feed. Several large
producers of PAN operate in Japan, Korea, and Taiwan. The major company that
produce PAN are NCP coatings inc , Arkema inc, Esco company and Durcon inc .

In 2006, the worldwide production of PAN has slowly increased with production
approaching the estimated capacity of 3,600,000 tones/year.[2]

The selling price has varied due fluctuating demands and production rates but is
currently around $(850-950)/ton with purity of 99.5%.[3]

The feed stock

PAN is produced from o-xylene feedstock or naphthalene. In 1896 the feed stock for
the liquid phase oxidation was naphthalene. In the United State and Germany they try
to come up with process to oxide naphthalene using vapor phase and they success in
1930.The shortage of naphthalene with availability of o-xylene the company forced to
use it. Since the 1960s o-xylene has replaced naphthalene as feedstock more and
more. The selling price of pure o-xylene around $5.05/kg while the price for
naphthalene around $5.96/kg[1].Currently 90% of phthalic anhydride is produced
from o-xylene because of the yield since it is the main factor affecting the choice of
the feed stock and the price.

Table 1: Component Price

Component Price($)
O-xylene 5.05/kg
Naphthalene 5.96/kg

2
Process description

The manufacturing of Phthalic Anhydride (PAN) is done in several steps. As

simplicity, mixture of O-xylene and air is preheated and converting o-xylene to vapor
knowing that the boiling point of the o-xylene is 144.4 C. Then, the mixture is fed to a
catalytic reactor. The reaction is highly exothermic, so to maintain the temperature,
the reactor is designed as packed bed reactor with heat exchanger. Water is used to
cool the reactor. The products then purified using other units like distillation.

There are five reactions happen beside the production of phthalic anhydride by
oxidation of o-xylene. These reactions are listed below.[6]

+ → +

+ . → +

+ . → +

+ . → + +

+ → +

The first reaction is the main reaction and it is assumed 70% selectivity. The second
Reaction refers to the formation of the by-product maleic anhydride (MA) and a 10%
selectivity is considered. The third and the fourth reactions represent the complete and
incomplete combustions of o-xylene with 15% and 5% selectivity, respectively.

Reactions path way

In the reactor, the o-xylene is subject to a variety of oxidation reaction to the desired
product of Phthalic anhydride; by product Maleic anhydride and products of
combustion. The oxidation of o-xylene feed with the air produce Phthalic anhydride.
The reactor effluent containing pure Phthalic anhydride reacts with excess oxygen to
produce water and carbon dioxide. Also, o-xylene reacts with the oxygen to produce
the Maleic anhydride as by product that also react with excess oxygen to produce
water and carbon dioxide.

3
 Figure 3: Reactions Pathway

Type of reactors

In the past, phthalic anhydride (PAN) was produced in a slurry. After o-xylene has
replaced naphthalene as feedstock, the reactors were developed and fluidized bed and
fixed bed reactors were used. Naphthalene-based feedstock is made up of vaporized
naphthalene and compressed air. It is transferred to the fluidized bed reactor and
oxidized in the presence of a catalyst.

In PAN production using o-xylene as the basic feedstock, filtered air is preheated,
compressed, and mixed with vaporized o-xylene and fed into the fixed-bed tubular
reactors. Nowadays, fixed bed catalytic reactor is commonly used. The table below
shows comparison between fluidized bed and fixed bed for many parameters.[5]
Table 2: Fixed-Bed. vs. Fluidized-Bed Reactors for Selective Hydrocarbon Oxidation Reactions

Fixed-bed Fluidized-bed reactor

reactor
Hydrocarbon Below flammability limit Flammable region possible
concentration
Oxygen concentration Large excess Near stoichiometric
Temperature control Hot spot Nearly isothermal
Catalyst effectiveness Poor to average Good
Catalyst cost Least expensive More expensive
Capital investment Expensive Less expensive
Raw material Cost Lower Higher
Raw material availability More available Less available
CO2 emission Less More
PAN Yield Higher Lower

4
Catalyst

There are different types of catalysts which are used for the oxidation of o-xylane to
PAN. Types that contain V2O5 supported on TiO2 are common and affective
catalysts for this process. The amount of V2O5 is used affects the selectivity of PAN
and carbon dioxide and the temperature for total conversion. Table (3) shows some
comparison. It seems that the catalyst contain 5.6% is more active. It has selectivity of
50% PAN when 100% conversion of o-xylene and the best temperature is 553K.[8]

Table 3:Catalyst Properties and Behavior in o-Xylene Oxidation

Catalyst Vanadium Surface area Yield of PA Temp.

(wt%) (m2/g) (%) (K)
VPTGA(1) 0.68 111 26 558
VPTGA(5) 2.85 96 17 513
VPTGA(10) 5.6 104 50 553
VPTGB 0.58 152 25 568
VP25 0.13 56 18 593

5
Kinetics
The kinetic of the five reactions are:[7]

6
First Design: Reactor Volume Using Simple Reaction Rate Expression
In this design we develop an isothermal reactor model to estimate the reactor volume
for the production of phthalic anhydride by oxidation of o-xylene based on the
reaction below

C H + 3O → C H O + 3H O

A catalyst weight of a packed bed reactor model is obtained and then converts the
weight to a reactor volume using catalyst bulk density. In this design we will explore
a reactor with a single reaction. However, reactions to byproducts will be added in the
third design.

The designing of fixed bed reactor depend on the following rate of the reaction that
obtained from literature:

−r = k P P ln = − + 19.837

k =4.1219e8 kmole /(kg catalyst).atm2 hr. The activation energy is 27000 cal/mol[1].

The design differential mole balance equation is

d(x) −r
=
d(w) Fa

With the following specifications:

T=673K P=3atm W (f) =21071kg

Polymath solved the differential equation and gave the relation between the weight of
catalyst and the conversion. The typical conversion is 85%. The calculations were
repeated at different temperatures to visualize the effect of temperature in the
conversion and the weight of the catalyst required to achieve the specified conversion.
The calculations can be obtained in matlab which is easier and the code is in the
appendix. Table (4) shows the results at 400 C and 3 atm. Figure (4) shows the effect
of temperature in the conversion. From figure (4), the weight of the catalyst required
to reach 85% conversion decrease with increasing in temperature. At 633 K, 75 000
kg of catalysts is needed however; 21000 kg is only required to achieve the specified
conversion.

7
Mass balance:

Table 4: Mass Balance Results

species Initial Change Final

(Kmol/s) (kmol/s) (Kmol/s)
0.0523 -0.04453 7.858× 10
0.7259 -0.1333 0.5925
0 0.0444 0.0444
0 0.1333 0.1333
2.7308 0 2.7308
Total 3.509 0 3.509

Table 5:Polymath design results

Variable Final value

1 w (kg) 2.107E+04
2 X 0.8500067
3 T (K) 673
4 P (atm) 3

Calculating the volume

To calculate the volume of the packed bed reactor, we use the weight catalyst, the
density of the particle and the porous media coefficient by the following equation:

w 21071
V= = = 12.55m
ρ (1 − ∅) 3357(1 − 0.5)

Where the (w) is the weight of catalyst in kg

The ρ is the density of the particle in kg/m3 [2]

∅ is the porose media of the catalyst (V2O5/TiO2)[2]

8
Figure 4:Effect of temperature in the conversion

The simulation

We simulate the packed bed reactor in Aspen Plus using the RPLUG reactor. The
specifications of the feed are 673K and 3 atm. The table below shows the result
obtained of heat and material balance.

Table 6: Heat
Heaand
t anmaterial balance
d Material Ba la nce Table
Strea m ID FEED P RODUCT
Tem perature C 39 9. 9 399. 9
P ressure ba r 3.040 3.00 0
Vap or Frac 1.000 1.00 0
Mole Flo w km ol/hr 12 686.36 0 12686.360
Ma ss Flo w kg/hr 38 0541.6 02 380541 .602
Volum e Flow cu m /hr 23 3529.3 66 236623 .630
Enthalpy Gca l/hr 37 .683 -4.82 6
Mole Flo w km ol/hr
C8H10 -01 188. 597 30 .1 43
O2 2612.819 21 37.456
C8H4O-01 15 8. 454
H2O 47 5. 363 Figure5 : Aspen Reactor Model
N2 9884.945 98 84.945

9
The conversion can be calculating using the flow rate of the o-xylene obtained from
aspen plus by the equation below, and we get very close answer to the calculated
result.

F F 188.597 − 30.143
x= = = 0.84
F 188.597

To construct the plot of conversion versus heat duty we use the data composition
obtained and multiplied by the total flow of the outlet and use it to find the
conversion.

0.9

0.8

0.7

0.6

0.5
X

0.4

0.3

0.2

0.1

0
0 10000 20000 30000 40000 50000 60000
Q (KW)

Figure6 : conversion versus heat duty

The figure above shows the relationship between the conversions versus heat duty. As
can be seen from the figure, as we increase the conversion the heat duty increase.
Actually there is a proportional linear relationship between them.

10
Second Design: Pressure Drop and Reactor Configuration
Based on the previous design (isothermal reactor), the pressure drop consideration is
taken into account in this design. The minimum pressure ratio (P/P0) is 0.9 that give
us the outlet pressure of 2.7atm.The consideration of pressure drop is important to get
the suitable outlet pressure that needed for downstream operation. Moreover, if we
have very large pressure drop, we have to increase the pressure before it fed to the
next operation hence increase the capital and operating cost. The optimization will be
done in this design to find the typical conversion which is 0.85 and reasonable reactor
configuration using multiple tubes, changing weight of catalyst and the temperature.
The differential mole balance equation is

d(x) −r
=
d(w) Fa

And the momentum balance Ergun equation is

( )
= (1+εx)
( )

With the following specifications:

T=673K P=3atm W (f) =21071kg = y (0) =1

Polymath solved the differential equations and gave the relation between the weight
of catalyst and the conversion. The calculations were repeated at different
temperatures to visualize the effect of temperature in the conversion with
consideration of pressure drop and compare it with the result of previous design.
Table 7: Design Results without Sipliting to Tube

Variable Final value

1 w (kg) 2.107E+04
2 x 0.7044096
3 y 0.4936886
4 T (K) 673
5 N 1
6 P0 (atm) 3
7 D (m) 3.5
8 V (m3) 12.55347
9 L (m) 1.304257

11
 0.8 Weight VS conversion
0.7
0.6 T=673K
0.5 T=650K
T=630K
X

0.4 T=600K
0.3 T=573K
0.2
0.1
0
0 5000 10000 15000 20000 25000
W(kg)
Figure 7:Effect of temperature in the conversion

Figure (7) above shows the effect of catalyst weight on the conversion at different
range of temperature with consideration of pressure drop. As can be seen, the
conversion increases as the catalyst weight increase. When the pressure drop
increased the concentration decreased resulting in a decreased rate of reaction. As a
result of this smaller reaction rate, the conversion will be less with pressure drop than
without pressure drop as we observed that on figure (8).

0.9
Weight VS conversion
0.8
0.7
0.6 T=673K without (y)
0.5 T=673K with (y)
X

0.4 T=630K without(y)

T=630K with(y)
0.3
0.2
0.1
0
0 5000 10000 15000 20000 25000
W(kg)
Figure 8:Effect of Pressure Drop in the conversion at Different Temprature

12
Splitting the Reactor into Tubes

After the calculation of conversion and pressure drop, it can be seen that the pressure
drop is less than 10%. The proper solution is to split the flow in tubes to decrease the
pressure drop since it function of mass flow rate. From figure (9), it is clear that
dividing the flow rate into tubes affect the pressure drop. Pressure drop is decreasing
with increasing the number of tubes. So, using banks of tubes will solve the pressure
drop problem and the design will meet the specifications.

Figure 9:The effect of number of tubes in the pressure drop

Tube Length effect

After optimizing the design to achieve the required conversion 85% and the pressure
drop constrain 10% maximum, the number of tubes was chosen to be 25140 tubes.
The weight of catalyst is 5.8 kg in each tube. The tube diameter was selected as 1.5
inches. The length of the tube is changing to discuss the effect of tube length in the
pressure drop. Figure (10) shows the results of different tube length in pressure drop.
As can be seen from the figure, increasing the tube length will increase the pressure
drop and that due to the friction. Also, since the volume of the reactor is fixed,
increasing the length will decrease the diameter and as a result the friction will be
greater and pressure drop will increase. From the calculation, a length of three meters
gives a pressure drop of 0.2885 atm which is a proximately 10%.

13
 Figure 10:Effect of Tube Length in Pressure Drop.

Particle Diameter effect

As well known theoretically that particles with smaller size will minimize the spacing
between them and as a result the fluid needs more force to pass through them. A large
pressure drop will gain if smaller particle is used. For studying the particle size effect,
the particles diameter that was used is doubled and the calculation is repeated. Figure
(11) gives the result of both diameters used to calculate the pressure drop. As
expected, the higher diameter has a less pressure drop of 0.1365 atm which is 5%
approximately.

Figure 11: Comparing two paticle diameters effect in pressure drop.

14
After optimizing all parameters, the result shows below

Variable Final value

1 w (kg) 2.152E+04
2 x 0.8507969
3 y 0.9009362
4 T (k) 617
5 N 2.514E+04
6 P0 (atm) 3
7 D (in) 1.5
8 L (m) 0.653
9 Vt (m3) 86.87042
Figure 12: Final design results

The number of tubes is considered to be 25140 tubes and contains 5.8 kg of catalyst in
each one. The conversion is calculated and gives 85% and the pressure ratio is 0.9.
Figure (13) shows the catalyst weight effects in pressure drop and conversion. The
tube diameter of each tube is 1.5 inches and a length of 0.653 meter.

Figure 13: Shows the Catalyst Weight Effects in Pressure Drop and Conversion

15
The simulation

Based on the simulation done in memo 2 of packed bed reactor in Aspen Plus using
the RPLUG reactor, the pressure drop was including in this memo. We specify Ergun
friction correlation to calculate process steam pressure drop. The specifications of the
feed in the first run are 673K and 3 atm with the following configurations and catalyst
weight.

N (number of tube) =1 D=3.5m L=1.304257m

W=21071kg

The table below shows the result obtained of heat and material balance.

Table 8: Heat and material balance

Heat and Materi al Balance Table

Stream ID FEED P ROD UCT
Temperature C 399 .9 399 .9
P res sure bar 3.0 40 1.4 95
Vapor Frac 1.0 00 1.0 00
Mole Flow kmol/hr 126 86.3 60 126 86.3 60
Mass Flow kg/hr 380 541.602 380 541.602
Volume Flow cum/hr 233 529.366 474 674.606
Enthalpy Gcal/hr 37.683 2.3 01
Mole Flow kmol/hr
C8H10-0 1 188 .597 56.708
O2 261 2.81 9 221 7.15 3
C8H4O-01 131 .889
H 2O 395 .666
N2 988 4.94 5 988 4.94 5

The conversion can be calculating using the flow rate of the o-xylene obtained from
aspen plus by the equation below, and we get very close answer to those obtained in
POLYMATH.

F F 188.597 − 56.708
x= = = 0.699
F 188.597

The conversion obtained using POLMATH was 0.704. As we observed the difference
between them is 0.7%.

16
 In the second run, the specifications of the feed are 617K and 3 atm with the
following configurations and catalyst weight that give us the desired conversion
which is 0.85.

N (number of tube) =25140 D=0.0381m for each tube

L=3.03m W=21520 kg

The table below shows the result obtained of heat and material balance.

Table 9: Heat and material balance

Heat and Material Balance Tab le

Stream ID FEE D PRODUCT
Temperature C 343. 9 343. 9
Pressure bar 3. 040 2. 738
Vapor Frac 1. 000 1. 000
Mole Flow kmol/hr 12686. 360 12686. 360
Mass Flow kg/hr 380541. 602 380541. 602
Volume Flow cum/hr 214097. 502 237669. 115
Enthalpy Gcal/hr 31. 909 -10. 926
Mole Flow kmol/hr
C8H10-01 188. 597 28. 761
O2 2612. 819 2133. 313
C8H4O-01 159. 835
H2O 479. 506
N2 9884. 945 9884. 945

The calculated conversion using flow rate of the o-xylene obtained from Aspen is
0.848 with the difference of 0.2% from POLYMATH result.

In both trails the ratio of the pressure (P/Po) is exactly the same we obtained in
POLYMATH.

17
Third Design: Multiple Reactions
Based on the previous design (isothermal reactor), the side reactions are included in
this design. There are four reactions happen beside the production of phthalic
anhydride by oxidation of o-xylene. The reactions and their kinetics [7] are below.

The differential mole balance equations are

( ) ( )
=r =r
( ) ( )

( ) ( )
=r =r
( ) ( )

( ) ( )
=r =r
( ) ( )

And the momentum balance Ergun equation is

d(y) −α F T
=
d(w) 2y F T

18
With the following specifications:
Table 10: Third Design Results

Variable Final value

1 w (kg) 23055
2 y 0.9005
3 T (k) 673
4 N 2.25E+04
5 P0 (atm) 3
6 D (in) 1.5
7 L (m) 0.653
8 Vt (m3) 86.87042

Plot Flow

Figure 14: Molar flow rate of each component versus the catalyst weight

Figure (14) shows the molar flow of each component with the weight of catalyst. The
molar flow rate of all products is always increasing. Water is produced with high
amount. Pressure drop was fixed to be 10% only.

19
Effect of Temperature in Selectivity

As can be seen from figure (15), the selectivity of phthalic anhydride is increasing
with the decrease of temperature. It is as expected since the activation energy of a
reaction that produce phthalic anhydride has the lowest value than the others. The
selectivity of maleic anhydride is decreasing with increase of temperature. The
relation is shown in figure (16).

2.5
T=673K
2
T=655K
sd_pa

1.5 T=640K

0.5

0
0 5000 10000 15000 20000 25000
W(kg)

Figure 15: Selectivity of Phthalic Anhydride at Different Temperature

0.3
sd_ma Vs W
0.25
T=673K
0.2
T=655K
sd_ma

0.15 T=640K

0.1

0.05

0
0 5000 10000 15000 20000 25000
W(kg)

Figure 16: Selectivity of Phthalic Anhydride at Different Temperature

20
Aspen simulation

Based on the simulation done in memo 3 of packed bed reactor in Aspen Plus using
the RPLUG reactor, the four side reactions are added in the configuration with their
kinetics. The specifications of the feed are 673K and 3 atm with the following
configurations and catalyst weight.

N (number of tube) =1 D=5.35m L=0.558m

W=21071kg

The table below shows the result obtained of heat and material balance.

Table 11: Heat and Material Results in Aspen Plus

Heat and Material Balance Table

Stream ID FEED PRODUCT
Temp erature C 399.9 399.9
Pressure bar 3.040 2.737
Vap or Frac 1.000 1.000
M ole Flow kmol/hr 22547.160 22626.729
M ass Flow kg/hr 676363.053 676363.053
Volume Flow cum/hr 415046.066 462533.781
Enthalp y Gcal/hr 66.980 -47.109
M ole Flow kmol/hr
C8H10-01 335.520 92.947
O2 4646.160 3432.062
C8H4O-01 160.015
H2O 848.566
N2 17565.480 17565.480
CO2 483.391
C4H2O-01 44.268

The conversion can be calculating using the flow rate of the o-xylene obtained from
aspen plus by the equation below, and we get very close answer to those obtained in
POLYMATH.

F F 335.520 − 92.947
x= = = 0.722
F 335.520

The conversion obtained using POLMATH was 0.726. As we observed the difference
between them is 0.5%.However in this memo we care more about the exit flow rate of

21
 our product (phthalic anhydride) to get the desired amount specified (20,000
tons/year). The outlet flow rate obtained using Aspen is 160.0 kmol/hr which is very
close answer calculated by POLYMATH(160.5 kmol/hr).

Figure (17) shows the flow rate of all species except the flow rate of O2 in the
primary axis with reactor length. As can be seen, the reactant flow rate (O-xylene and
Oxygen) are decreasing with the reactor length while the product species increasing
with the reactor length. These results are expected from the reactions since all it are
producing water and four reactions producing CO2 and only one reaction are
producing phthalic anhydride.

flow rate VS Length

900 5000
800 4500

700 4000
3500
flow rate (kmol/h)

flow rate (kmol/h)

600 f_O-xylene
3000 f_phthalic
500
2500 f_water
400 f_CO2
2000
300 f_o2
1500
200 1000
100 500
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6
L(m)
Figure 17: Molar Flow of each Component with Reactor Length obtained from Aspen

22
Final Design: Energy Balance
Energy balance equation was added to the previous design. All reactions in phthalic
anhydride production are exothermic reactions. The multi tubular reactor acts as a
shell and tube heat exchanger with co-current coolant. The energy balance of the
coolant was considered and steam was used as a coolant. The calculated heat capacity
(Cp) of each component at the inlet temperature ( 673 oC ) are shown in table(12).
Table (13) contains the Heat of formation for each component that was used to
calculate the heat of reactions.

Table 12: Components Properties

Component o-xylene Oxygen Phthalic Maleic Carbon Water Nitrogen

anhydride anhydride dioxide
Cp 458.6 32.73 723.5 269.9 48.96 37.23 30.6
(KJ/kmol.K)
∆H 19.1 0 -371.4 -398.2 -393.5 -241.80 0
(KJ/mol)

The overall heat transfer coefficient was found in literature to be 58 (W/m2.K). the
coolant inlet temperature is 655 K. the pressure drop was maintained not to exceed
10%.

Temperature profile
Figure (18) shows the temperature gradient with the weight of catalyst. A hot spot
appear in the reactor. The hot spot temperature which is the highest temperature is
calculated to be 672.8 oC. Then the temperature decreases and goes to constant. The
outlet temperature of the reactor is 656.1 oC which is almost acceptable. Coolant
temperature increases along the reactor and Figure (19) shows the relation.

Figure 18: Temperature Profile

23
Figure 19:Coolant Temperature Profile

Figure 20: Reactor and Coolant Temperature Profiles

Dynamic stability
The reactor was tested dynamically by considering the gain. The inlet temperature of
the coolant was changed to study the effect of changing in the hot spot temperature.
The following results were obtained:
Table 13: Gain Calculation

Ta ( K ) Th ( K ) Gain
638 715.2 17
636 689.1 9.1

24
From the results, the design is not stable dynamically since the gain is more than 2. As
a solution the amount of inert component should be increased and in this case the ratio
of the air to o-xylene was increased from 20 to 25 ( kg.air/kg.xylene). After the air
amount was changed and other parameters were optimized, the dynamic results were
shown in table (14). The coolant inlet temperature was set to be 643 K and hot spot
temperature was found to be 688.1.
Table 14: Gain Calculation after Modification

Ta ( K ) Th ( K ) Gain
642 681.2 1.9
644 684.8 1.7

Aspen Simulation
After the reactor was designed to obtain the target production of phtalic anhydride, the
reactor was simulated in Aspen Plus. The flow sheet of the model was shown in figure
(21). PBR model was used with co-current coolant and specified heat transfer
coefficient U of 3.5 (kw/m2.K). Figure (22) shows the model specification. The
reactor configuration used in Aspen is the calculated dimensions. A length of 0.6 m,
1.5 in diameter and 36000 tubes was considered.

Figure 21:Aspen Flowsheet

25
Figure 22 : Model Type

Figure 23 : Design Dimensions

The inlet reactor pressure is specified to be 3 atm. The Ergun equation was used to
calculate the pressure drop in the reactor. Catalyst weight, porosity and particles
diameter were added in the model specification. Figures (24) and (25) show the
progress.

26
Figure 24: Pressure Drop Calculation

Figure 25: Catalyst specification

The five reactions were added to the design. POWERLAW type of kinetics was
considered and the all kinetics were added. Rate phase, basis and activation energies
were specified for each reaction. Figure (27) explains the kinetics specification way.

27
Figure26 : Reactions in Aspen

Figure 27: Kinetics Specification

After running the simulation, the results were tabulated and then compared with the
calculated values. The production of phthalic anhydride is consistent. The calculated
phthalic anhydride is 163 kmol/hr and Aspen calculated to be 160 kmol/hr. The
pressure drop was 0.3 atm as same as the calculated value. The temperature profile
through the reactor was plotted and a hot spot can be determined. The hot spot
temperature is 391.5 C, 665 K and calculated is 670 K. The results were shown
below.

28
Table15 :Aspen Results

392
391
390
Temperature (c)

389
388
387
386
385
384
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Tube Length (m)

Figure28 : Temperature Profile

29
Economic analysis
The cost of production phthalic anhydride per year depends on purchase cost of the
packed bed reactor with heat exchanger, annual cost of the raw material (air and o-
xylene); cooling cost and catalyst cost (V2O5).The cost of production of
20000tons/year can be calculated using the following equation

Cost=Purchase cost of the reactor + Annual cost of raw material + Annual cooling
cost + catalyst cost- annual revenue of by product (malic anhydride)- Annual revenue
of phthalic anhydride

Purchase cost of the packed bed reactor with heat exchanger can be calculated using
the following equation [1] ($) =1.25*1600(As) .

As = πDLN

Where N is number of tubes

Cooling cost calculated by obtained first the heat duty then calculate mass flow rate
using the following equation

Q̇
ṁ =
C ΔT

Cooling cost ($) = ∗ 6.70 [1]

Annual cost of the raw material and revenue of by product and the product depend on
the flow rates multiply by it price.
Table 16: Total Cost Results

Cost per year($)

Purchase cost of the reactor 3244210
cost of raw material 106,143
cooling cost 365730
catalyst cost 670220
revenue of by product (malic anhydride) 3531407
revenue of phthalic anhydride 18,000,000
Total 17,145,104

30
Design summary
Table 17: Reactor Configurations

N (tubes) D (in) L (m) W ( kg ) T(K) P ( atm)

36000 1.5 0.6 40000 656 3

Table 18: Feed Rates

Component O-xylene Air

Rate ( kg/s) 10 250

Table 19: Products Rates

Component O-xylene Oxygen PAN MAN CO2 Water

Rate (kmol/s) 0.0256 1.4706 0.0454 0.0125 0.1352 0.2395

Table 20 : Coolant Specification

Type of coolant Inlet temperature (K) Flow rate (kg/s)

Water 640 0.18

Conclusion

The packed bed reactor design that used for phtalic anhydride production has been
done. Pressure drop, side reactions, temperature variation were considered. After
optimizing all parameters to achieve 20 ton/year of phthalic anhydride, reactor
configurations listed in table (15). The reactor looks like a double pipe heat exchanger
with 36000 tubes with 1.5 in. water steam at 640 K was used to maintain the
temperature with a flow rate of 0.18 kg/s per tube. The feed rates to the reactor shown
in table (16) and products rate tabulated in table (17).

31
References

1. Chemical Engineering Design Project

2. Kirk-Othmer Encyclopedia of Chemical Technology Volume 18
3. Encyclopedia of Chemical Processing
4. Handbook of Industrial Catalysts By Lawrie Lloyd
5. CHEMSYSTEMS PERP PROGRAM Report Abstract Phthalic anhydride.
6. A novel route to produce phthalic anhydride by oxidation of o-xylene with air
over mesoporous V-Mo-MCM-41 molecular sieves.
7. Kinetics of the Selective Oxidation of o-Xylene to Phthalic Anhydride
Doctoral Thesis (Dissertation) to be awarded the degree of Doctor of
Engineering (Dr.-Ing.) submitted by Dipl.-Ing. Robert Marx.
8. THE USE OF MATHEMATICAL MODELLING TO INVESTIGATE THE
EFFECT OF CHEMISORPTION ON THE DYNAMIC BEHAVIOR OF
CATALYTIC REACTORS. PARTIAL OXIDATION OF o-XYLENE IN
FLUIDIZED BEDS By S. S. E. H. ELNASHAIE.
9. Simulation of a Reactor for the Partial Oxidation of o-Xylene to Phthalic
Anhydride Packed with Ceramic Foam Monoliths By A. REITZMANN, A.
BAREISS and B. KRAUSHAAR-CZARNETZKI.
10. Reactors and Separations Design Project, Phthalic Anhydride Production.
11. http://www.alibaba.com/showroom/price-of-phthalic-anhydride.html
12. http://www.metalprices.com/metal/vanadium/vanadium-pentoxide-v2o5-fob
13. Catalyst Deactivation 1994b edited by B. Delmon, G.F. Froment

32
Appendix

 Final Design Matlab Code.

function main ()
clear all;close all;clc

%-----------------------------------------------
% OPERATING CONDITION
%-----------------------------------------------
To=656; % k
R1=1.987; % cal/mol.k
Po=3; % atm
N=36000; % Tubes
Ta0=640; % K
m_c=.01; % kmol/s
U=3.5; % Kw/m2.k
Tr=298; % K
D=.0254*1.5; % m
w=40000; % kg
%------------------------------------------------
% O-Xylene FEED
%------------------------------------------------
m_xy=10/N; % kg/s
%------------------------------------------------
%-----------------------------------------------
%Reactions
%-----------------------------------------------
% (1) XY + 3 O2 ---> PA + 3 H2O
% (2) PA + 7.5 O2 ---> 8 CO2 + 2 H2O
% (3) XY + 10.5 O2 ---> 8 CO2 + 5 H2O
% (4) XY + 7.5 O2 ---> MA + 4 CO2 + 4 H2O
% (5) MA + 3 O2 ---> 4 CO2 + H2O
%
% Stoic
% XY O2 PA MA CO2 W
v=[ -1 -3 1 0 0 3
0 -7.5 -1 0 8 2
-1 -10.5 0 0 8 5
-1 -7.5 0 1 4 4
0 -3 0 -1 4 1];
%----------------------------------------------
% Cp
% XY O2 PA MA CO2 W
Cp=[ 458.588 32.7362 723.474 269.858 48.96 37.23 ]; % KJ/kmol.K
Cpn2=30.5798;
%------------------------------------------------
dCp=v*Cp';
%----------------------------------------------
% H
% XY O2 PA MA CO2 W
Ho=[19.1 0 -371.4 -398.2 -393.51 -241.83]*1e3; % KJ/kmol
dHrx_o=v*Ho';
%-----------------------------------------------
% K CALCULATION
%-----------------------------------------------
ko=[19.837 20.86 18.98 19.23 20.47];
E=[27000 31000 28600 27900 30400];

%------------------------------------------------

%================================================
% CALCULATION
%------------------------------------------------
m_air=m_xy*25; % kg/s
m_o2=0.232*m_air; % kg/s
m_n2=(1-.232)*m_air; % kg/s

% XY O2 N2 PA
Mr=[106.16 32 28.02 148.1]; % kg/kmol
n_xy=m_xy/Mr(1);
% kmol/s
n_o2=m_o2/Mr(2); % kmol/s
n_n2=m_n2/Mr(3); % kmol/s

33
Ft0=n_xy+n_o2+n_n2; % kmol/s
% N2 XY O2
y0=[n_n2 n_xy n_o2]/Ft0;
%------------------------------------------------------
% Alpha Calculation
%------------------------------------------------
wf=w/N;
P0=Po*101325; % Pa
R=8.314; % kPa.m3/kmol.k
por=0.5;
Dp=.006; % m
Vis=(-1e-08*To^2+5e-5*To+0.0039)/1000; % Pa.s
rho_c=3357; % kg/m3
rho_b=rho_c*(1-por); % kg/m3
V=wf/(rho_c*(1-por)); % m3
L=4*V/(pi*D^2); % m
M_avg=y0(3)*Mr(1)+y0(1)*Mr(3)+y0(3)*Mr(2); % kg/kmol
Ac=pi/4*D^2; % m2
m=Ft0*M_avg; % kg/s
G=m/Ac; % kg/m2.s
rho_o=P0*M_avg/(R*To)/1000; % kg/m3
B0=G/(rho_o*Dp)*(1-por)/por^3*(150*(1-por)*Vis/Dp+1.75*G);% kg/m2.s2
a=2*B0/(rho_c*Ac*(1-por)*P0); % 1/kg
%------------------------------------------------
%------------------------------------------------------
Fn=n_n2;
Fi=[ Ft0*y0(2) Ft0*y0(3) 0 0 0 0 1 To Ta0];

ws=0:10/N:wf; n=length(ws);
[w,z]=ode45(@(w,z)rate(w,z,Po,a,v,Ft0,dHrx_o,dCp,rho_b,D,Cp,Fn,Cpn2,To,ko,E,R1,m_c,U,T
r,Ta0),ws,Fi);

%------------------------------------------------------

%======================================================
% RESULTS
%-------------------------------------------------------

disp(' Fxy Fo2 Fpa Fma Fco2 Fw y ')

table=[z(:,1:6)*N z(:,7:9)]
yf=z(n,7)
Th=max(z(:,8))
D
L
x=(z(1,1)-z(n,1))/(z(1,1))
mp=z(n,3)*Mr(4)*(30326400/1000)*N % ton/year
q=[z(:,1:6)*N];
plotyy(w,q,w,z(:,7))
plot(w,z(:,8:9))
Vt=V*N
end

%==========================================================================
%==========================================================================
%==========================================================================

function
dzdw=rate(w,z,Po,a,v,Ft0,dHrx_o,dCp,rho_b,D,Cp,Fn,Cpn2,To,ko,E,R1,m_c,U,Tr,Ta0)

for i=1:6
F(i)=z(i);
end

y=z(7);
T=z(8);
Ta=z(9);

Ft=sum(F)+Fn;

34
% Heat
dHrx=(dHrx_o+dCp*(T-Tr)); % KJ/kmol

k=exp(ko - E/(R1*T))/3600; % kmol/(kgcat.atm2.s)

Pxy=F(1)/Ft*Po*y*To/T; % atm
Po2=F(2)/Ft*Po*y*To/T; % atm
Ppa=F(3)/Ft*Po*y*To/T; % atm
Pma=F(4)/Ft*Po*y*To/T; % atm

% Reactions
rs(1)=k(1)*Pxy*Po2; % kmol/kgcat.s
rs(2)=k(2)*Ppa*Po2; % kmol/kgcat.s
rs(3)=k(3)*Pxy*Po2; % kmol/kgcat.s
rs(4)=k(4)*Pxy*Po2; % kmol/kgcat.s
rs(5)=k(5)*Pma*Po2; % kmol/kgcat.s

% Rates Evaluation
for i=1:6
r(i)=0;
for j=1:5
r(i)=r(i)+rs(j)*v(j,i);
end
end

% differential
for i=1:6
dzdw(i)=r(i);
end
dzdw(7)=-a/(2*y)*(Ft/Ft0)*(To/T);
dzdw(8)=(((4*U)*(Ta-T)/(D*rho_b))+(sum(-rs.*dHrx')))/(sum(F.*Cp)+Fn*Cpn2);
dzdw(9)=((4*U)*(T-Ta)/(D*rho_b))/(m_c*Cp(6));

dzdw=dzdw';

end

35