See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/272158529

Polypropylene production simulation with

cape-open interfacing of pro/ii and gPROMS

Article in Journal of Engineering Science and Technology January 2015

CITATIONS READS

0 269

5 authors, including:

Oware Sarfo Kofi Seong Hun Kim

Oregon State University Hanyang University
1 PUBLICATION 0 CITATIONS 737 PUBLICATIONS 8,963 CITATIONS

SEE PROFILE SEE PROFILE

Min Oh
Hanbat National University
32 PUBLICATIONS 278 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Coloration of nano fibrous webs View project

All content following this page was uploaded by Oware Sarfo Kofi on 14 June 2015.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
Journal of Engineering Science and Technology
Special Issue on SOMCHE 2014 & RSCE 2014 Conference, January (2015) 48 - 61
School of Engineering, Taylors University

POLYPROPYLENE PRODUCTION SIMULATION WITH

CAPE-OPEN INTERFACING OF PRO/II AND gPROMS
1 2 2 2 2,
J. C. LEE , O. S. KOFI , S. H. KIM , S. U. HONG , M. OH. *
1
Schneider-electric Co., Ltd., 13F, Jei Platz, 186, Digital 1st, Geumcheon-gu, Seoul, Korea
2
Dept. of Chemical & Biological Engineering, Hanbat National University,
Daejeon 305-719, Korea
*Corresponding Author: minoh@hanbat.ac.kr

Abstract
A dynamic pseudo-homogeneous model was made expressing liquid-phase
homo-propylene polymerisation in a stirred reactor with reaction kinetics based
on the Ziegler-Natta catalyst. An energy balance was made to analyse heat
management in this exothermic reaction. The model was simulated using
gPROMS process software and validated with a reference paper to ensure the
models reliability. Results of the simulation and the reference agreed with a
reasonable degree of error. The method of moments was employed to predict
the average molecular weights and the polydispersity and melt flow indices.
Meanwhile, models for the separation and treatment of the polymer formed
were formulated using PRO/II software. Consequently, the steady state form of
model from gPROMS was connected by CAPE-OPEN interface to PRO/II
modelling environment. The flowsheet from this multi-scale modelling
approach was simulated with industrial data. It is concluded, this approach
enables the multi-scale plant simulation for various equipment and rigorous
reactor models simultaneously, without much modelling efforts while we
maintain accuracies at reduced computational cost.
Keywords: Homo-polymerisation, Spheripol process, Polymer properties,
Multi-scale modelling, CAPE-OPEN.

1. Introduction
Polypropylene is one of the most widespread polymers that has found its daily
usage in many applications for domestic and commercial areas. The use of
mathematical models to study and analyse chemical industrial processes is
common as it affords us the privilege to predict the behaviour of the process.
Fundamentally, a rigorous polymerisation process model, which expresses the

48
Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 49

Nomenclatures
A Heat transfer surface area, m2
AlR3 Triethylaluminium concentration, mol/L
C* Catalyst active site concentration, mol/L
Cd Deactivated site concentration, mol/L
Cp Specific heat capacity, J/kg-K
E Activation energy, J/mol
Ef Flexural modulus, GPa
f Mass flow rate, kg/h
i Number of chains
kim Monomer initiation rate constant, L/mol-s
kp Chain propagation rate constant, L/mol-s
ktrA Chain transfer to AIR3 rate constant, L/mol-s
ktrH Chain transfer to hydrogen rate constant, L/mol-s
ktrm Chain transfer to monomer rate constant, L/mol-s
kd Chain deactivation constant, s-1
M Monomer concentration, mol/L
m0 Monomer molecular weight, g/mol
MaxCL Maximum chain length number
MN Number average molecular weight, g/mol
MW Weight average molecular weight, g/mol
NOSI Number of sub-intervals
Pi* Live polymer of i number of chains, mol/L
Pi Dead polymer of i number of chains, mol/L
qin Inlet volumetric flow rate, L/s
q Outlet volumetric flow rate, L/s
rp Rate of polymerisation, mol/L-s
Se Energy generation source term, J/s
Sm Mass source generation term, mol/L-s
T Temperature, oC
Tj Temperature of Jacket, oC
U Heat transfer coefficient, W/m2-K
VR Volume of reactor, L

Greek Symbols
Reaction species
k
*
kth moment of live polymer chain
k kth moment of dead polymer chain
s Density of slurry mixture
H r Heat of polymerisation, J/mol
Abbreviations
MFI Melt Flow Index, g/10min
MW Molecular Weight, g/mol
PDI Polydispersity Index

Journal of Engineering Science and Technology Special Issue 2 1/2015

50 J.C. Lee et al.

physical and thermodynamic properties, material and energy balances and

polymerisation kinetics, is crucial for exploring and evaluating changes in feed
composition and operation conditions.
The modelling of the polymerisation process in available publications
generally, focuses on aspects like, reactor modelling, polymerisation mechanism
and molecular weight distribution, as well as, polymer property calculations. The
advantages of liquid phase loop reactors are first recognised as being their
capabilities to promote high mixing of reactants in the reaction vessel and to
allow for high heat transfer rates with the cooling jacket due to their high heat
transfer area [1]. Similarly, Zacca and Ray [2] modelled loop reactors as two
tubular sections interconnected by perfectly mixed inlet and outlet zones. Again
they demonstrated that at high recycle ratios, the loop reactor can be modelled as
CSTR as there is perfect mixing of material.
The implications of this assumption meant that all state variables (e.g.,
temperature, pressure, slurry density) in the reactors are uniform and change only
with polymerisation time [3]. Further corroborations were also done to confirm
the earlier suggestions that the residence time distribution of the tubular loop
reactors under these fast cycling conditions resemble that of the CSTR [4].
Whilst most publications focus on either reactor modelling or kinetic
mechanisms of polymerisation, not much has been done on the modelling of the
entire polypropylene production process on the multi-scale level. In this work, a
multi-scale simulation of polypropylene production was executed in two software
packages (gPROMS [5]/ PRO/II [6]) connected in the framework of CAPE-
OPEN. The polypropylene production process is described in section 2. This is
followed by the elucidation of the modelling topology and the reactor model in
section 3. Section 4 states and discusses the results of the simulation whilst
conclusions for the work are stated in Section 5.

2. Process Description
In the present study, the Spheripol loop technology [7], which involves the
continuous loop reactors at the centre is studied. The Spheripol process is a
modular technology consisting of three main process steps catalyst and raw
material feeding, polymerisation and finishing. One of the heralding advantages
of the Spheripol process, in combination with Ziegler-Natta catalysts, is its
unique ability to produce polymer spheres directly, in the reactor. Propylene
polymerisation leading to the formation of polypropylene homopolymer is
schematically, described in the process flow diagram in Fig. 1.
It mainly, consists of the propylene storage drum, the loop reactors, the
scrubbers to strip propylene from product mixture, the flash drum, the gas
filter, the steamer, the dryer, the hoppers and the extruder. At the heart of
the polypropylene production process is the loop reactor in which
polymerisation of propylene takes place [8]. The loop reactor is used due to
the low cost, high heat transfer and maintenance of uniform temperature,
pressure and catalyst distribution.

Journal of Engineering Science and Technology Special Issue 2 1/2015

Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 51

Fig. 1. Simplified Process Flow Diagram

of Polypropylene Homopolymer Process [9].

3. Modelling and Simulation Approach

3.1. Model topology
One major point of this study is to work on the different levels of modelling
abstraction of the polypropylene production process including:
micro-scale modelling for chemical kinetics and reactor of propylene
polymerisation in gPROMS
macro-scale modelling for fractionation of slurry polymer in PRO/II
multi-scale modelling for overall polypropylene production plant in the
CAPE-OPEN framework.

The first level constituted polymer reactions in the loop reactor related to the
kinetic model in which the best approach involved a detailed species balance. It
also included the heat and population balance, as well as the polymer end-use
properties. The second level also comprised of the modelling of process units for
the feed preparation, as well as the fractionation of the slurry polymer from the
reactor. The highest level of abstraction, the multi-scale model, then employed a
consideration of the entire production of polypropylene from the feed and catalyst
treatments through, the polymerisation reactors and the separations of monomer
from polymer to the final part of the polymer treatment involving drying of the
polymer product. The aforementioned levels were therefore, categorised into four
major blocks as shown in Fig. 2.
The feed preparation was done in the feeding block and information was sent
by streams to the reaction block which contains the reactor unit exported by
CAPE-OPEN. Polymerisation calculations were executed here and the results sent
onward to the treatment block for the separation and treatment of the polymer in
the final stages.

Journal of Engineering Science and Technology Special Issue 2 1/2015

52 J.C. Lee et al.

Fig. 2. The Modelling Topology for the Process Simulation.

3.2. Mathematical model of a loop reactor

Propylene polymerisation occurs in the loop reactor which is a closed tube with
slurry flowing under high recycle rates driven by a recycling pump [10]. Hence,
under the prevailing circumstance, it is possible to refer to this loop reactor as
operating in the scheme of the CSTR with constant volume but variable reactor
density. It is also assumed that due to high mixing, all state variables are uniform
anywhere in the reactor and the catalyst particles are assumed to have equal
activity. For a living or active chain of length i, the following mass balance
equation can be formulated for these species [10].
d[ ]
VR qin [ ]in q[ ] VR S m , Pi* , Pi , M , C * , H 2 , AlR3 , Cd , 0 (1)
dt
where
kiM [C * ][ M ] k p [ Pi * ][ M ] ktrM [ M ] ktrH [ H 2 ]0.5 ktrA [ A1R3 ] kd [ Pi * ], i 1; for Pi *

k P [ M ][ Pi *1 ] ktrH [ M ] ktrH [ H 2 ]0.5 ktrA [ A1R3 ] k p [ M ] kd [ Pi * ], i 2..n 1; for Pi *

ktrH [ M ] ktrH [ H 2 ]0.5 ktrA [ A1R3 ] kd [ Pi * ], for Pi

kiM [ M ][C * ] (k p ktrM )[ M ]0 , for M

S m (ktrM [ M ] ktrH [ H 2 ]0.5 ktrA [ AIR3 ])0 kiM [ M ][C * ] kd [C * ] kd 0 , for C *

ktrH [ H 2 ] 0 , for H 2
0.5

ktrA [ AIR3 ]0 , for AIR3

kd 0 kd [C ], for Cd
*

kiM [C ][ M ] ktrM [ M ] ktrH [ H 2 ] ktrA [ AIR3 ] kd 0 , for 0 (2)
* 0.5

Energy balance of the reactor constituents is given as [3]:
dT
VR s C p
dt

f in C pinT in f out C pT UA T T j VR S e (3)

S e (H r )rp (4)

Journal of Engineering Science and Technology Special Issue 2 1/2015

Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 53

Chen and Ray [11] consolidated much of the existing mechanisms in literature
to arrive at a reaction composed of several reaction steps. The Ziegler-Natta
catalyst is assumed to go through the common activation - polymerisation - decay
pathway and finally the chain transfer reactions occur with hydrogen, monomer
and AIR3 [12]. The comprehensive kinetic scheme comprising of the reactions as
stated are shown in Table 1.

Table 1. Kinetic Mechanism [3, 8, 10 and 11].

Reaction step Kinetic constant, k=ko exp(-E/RT)
Activation
C * M
k
P1* iM
kiM 4.97 107 exp(5e4 / RT )
Propagation
k
P1* M
p
P2*
k p 4.97 107 exp(5e4 / RT )
k
P1* M
p
Pi *1
Transfer
Transfer to monomer
P1* M
ktrM
Pi C * ktrM 6.16 103 exp(5e4 / RT )
Transfer to hydrogen
P1* H 2
ktrH
Pi C * ktrH 4.4 106 exp(5e4 / RT )
Transfer to cocatalyst (AlR3)
P1* AIR3 Pi C * ktrA 7.04 102 exp(5e4 / RT )
ktrA

Deactivation
P1* k
Pi Cd
d
kd 7.92 103 exp(5e4 / RT )
C * k
Cd d

3.3. Polymer properties

The method of moments is used to determine the average molecular weights.
Using the basic polymer properties like the average molecular weights, the end-
use properties of the polypropylene homopolymer can be predicted by means of
empirical correlations. The following expressions elaborate the determination of
molecular weight averages and their relationship with end-use properties.
Molecular weight moments for live and dead polymers [13]:
max_chain max_chain
k* i 1
i k Pi* ; k
i 2
i k Pi (5)

Number and weight average molecular weight [13]:

1* 1 2* 2
M n m0 ; M w m0 (6)
0* 0 1* 1
Melt flow index [14] and polydispersity index [15]:
Mw
MFI a (M w )b ; PDI (7)
Mn

Journal of Engineering Science and Technology Special Issue 2 1/2015

54 J.C. Lee et al.

where a = 3.391022 and b= -3.92

Flexural modulus (stiffness) [16]
a1
Ef a2 XS a3 (8)
Mw
a1 1.45 105 ; a2 3.37 102 ; a3 2.2530 XS= 4.13 for homopolymer

4. Results and Discussion

The dynamic model was simulated with conditions from the reference [2] to
validate the model. The simulation conditions for this study are illustrated in
Table 2.
Table 2. Simulation Conditions for Multi-scale Model [9].
Variable Value
Temperature 318 K
Pressure 38.2 bar
Mass flow rate 53,000 (kg/hr)
Mass fractions
Propylene 0.904
Propane 0.095
Catalyst 6.1510-5
Hydrogen 3.7210-5
AlR3 1.6110-3

4.1. Transformation of the chain length domain to sub-interval domain

For the purpose of simulation, the number of chains is to be in the finite domain.
In this study, the maximum number of chains was assumed to be 106. Considering
each chain number for the simulation in terms of array would, however increase
the system size enormously and lead to an expensive computational cost. In
avoiding this situation, a sub-interval concept, the transformation of the axis from
chain domain into sub-interval domain was employed such that;

zCL 1...MaxCL zSI 1...NOSI (9)

In this study, we used 100 sub-intervals in transformation of chain length axis
of which maximum value is 106. It is also worth noting that using the uniform
interval size for the whole domain will mislead the simulation especially
considering the unequal distribution of chain length. As a remedy, the size of sub-
interval was determined by log transformation such as;

log zc 1
zI (10)
log 1

The results of the log transformation based on Eqs. (9) and (10) are shown in
Fig. 3. Each sub-interval has a different interval size in chain length. As the sub-

Journal of Engineering Science and Technology Special Issue 2 1/2015

Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 55

interval domain (x-axis) increases, the corresponding chain length interval (y-
axis) also increases, which implies the lesser resolution for higher chain length
region. With this approach, accurate simulation can be carried out for the wider
range of chain length without paying extensive computational cost as well as
losing accuracy.
14000

12000

10000
Chain length interval

8000

6000

4000

2000

0
0 20 40 60 80 100

Sub-interval

Fig. 3. The Result of the Log Transformation.

4.2. Validation of reactor model

The time transient profile of the monomer concentration of the simulation is
compared with the results from the reference [2] as shown in Fig. 4. At steady
state the monomer concentration for the simulation is 2.78 mol/L, which is
0.22mol/L below that of the reference data (3.0mol/L). In likewise manner, the
profile for the conversion of monomers into polymer is also compared with the
reference is illustrated with maximum conversion being attained at approximately,
0.7 after about 8,000s.
12 0.8
Monomer Concentration, mol/L

10 0.6

Simulation (Concentration)
Reference (Concentration)
8
Conversion

0.4
Simulation ( Conversion)
Reference (Conversion)

6 0.2

4 0.0

2
0 2000 4000 6000 8000 10000 12000 14000 16000 18000

Time, s

Journal of Engineering Science and Technology Special Issue 2 1/2015

56 J.C. Lee et al.

Fig. 4. Time Transient Profile of Monomer Concentration and Conversion.

Moreover, the behaviour of slurry temperature in the polymerisation process

as a function of time is crucial for the management of heat in this system. This is
because propylene evaporates at about 70oC and hence, will disrupt the liquid-
phase process if heat is not properly managed. The time transient profile for
temperature in the reactor is illustrated in Fig. 5. It must however, be put on
record that the unavailability of certain information required in executing the
simulations in the study led to the marginal variations in the results reported
between the reference and the work done in this paper.
58

56

54
Temperature, C

Simulation
52
Reference

50

48

46

44
0 500 1000 1500 2000 2500 3000

Time, s
Fig. 5. Time Transient Profile of Temperature.

Values and decisions corresponding to such conditions were made guided by other
literature documents and recommendations from experts acquainted with polymer
simulations. These range from the choice of the maximum chain number through the
number of intervals for molecular weight distribution to design conditions like the
heat transfer area of the reactor. This is the reason for the marginal differences in the
results of monomer concentration, conversion and temperature.

4.3. Result of steady state simulation of the overall plant using

CAPE-OPEN
The multi-scale model with the exported reactor as a CAPE-OPEN unit is shown
in Fig. 6. Selection of the appropriate thermodynamic methods is a crucial step in
the execution of the simulation. This is because the process involves both pure
components and polymer components which differ in properties. The involvement
of the polypropylene polymer components led to the choice of the Flory-Huggins
thermodynamic method. This method is useful for modelling polymer/solvent
systems, especially if molecules are non-polar [6]. In addition, the Soave-Redlich-
Kwong (SRK) equation of state was selected due to the involvement of pure
components like H2, N2 and CO2 in the system.

Journal of Engineering Science and Technology Special Issue 2 1/2015

Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 57

Table 3 describes the material balances of the occurring components in the

process and their detail compositions, as well as the operating conditions at feed
stream and some notable equipment on the flowsheet. From this table, it could be
observed that the monomer is separated after the reaction by heating in the heat
exchanger causing it to evaporate at 70oC and the polymer is concentrated as you
go downstream. Moreover, the polypropylene is sent to a steamer to deactivate
the existing catalysts after which, it is further conveyed to the dryer where the
moisture is removed. At this point, the stream composition as could be observed
from Table 3 is about 99% propylene polymer.

gPROMS

Fig. 6. Process Model in PRO/II with CAPE-OPEN.

Table 3. Material Balance of Chemical Components from the Simulation.

Stream Name 104A 104B 105 108 110 112 114
Polymer Polymer
polymer Polymer
Reactant from from
Description Feed1 Feed2 from from
from RX flash bag
steamer dryer
drum filter
Temperature,
44.2 45.0 55.0 70.0 65.0 104.3 89.0
C
Pressure,
39.0 42.0 39.0 2.0 0.5 0.2 0.1
kg/cm2g
Total Mass
33014.3 17325.8 50340.1 27352.6 27247.3 28091.4 27184.3
Rate, kg/h
Weight
Fractions
Butane 0.00349 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
Teal 0.02224 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
Donor 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
Propylene 0.88180 0.90507 0.45970 0.00660 0.00276 0.00004 0.00000
Propane 0.09245 0.09489 0.00054 0.00001 0.00000 0.00000 0.00000
Hydrogen 0.00004 0.00004 0.00000 0.00000 0.00000 0.00000 0.00000
Water 0.00000 0.00000 0.00000 0.00000 0.00000 0.03270 0.00042
Nitrogen 0.00000 0.00000 0.00000 0.00000 0.00002 0.00000 0.00004
Polypropylene 0.00000 0.00000 0.53976 0.99339 0.99723 0.96726 0.99954

Journal of Engineering Science and Technology Special Issue 2 1/2015

58 J.C. Lee et al.

In order to validate the process model, the results of the simulation were
compared with a reference data [9]. The profile of the main feed component,
propylene is first illustrated. This is in order to determine the level of conversion
of the monomer into polymer and also understand the efficiency of the separation
process after the reactor to remove residual propylene for recycling. A bar chart
comparing the propylene mass fraction from simulation with that of the reference
is shown in Fig. 7. Propylene being the main reactant is consumed in the
polymerisation reaction as shown in Fig. 7 but some go unreacted and are
vapourised in the heat exchanger after the reactor and are separated from the
stream in the flash drum for recycling.

Fig. 7. Comparison of Propylene Mass Fraction.

This explains the significant reduction in propylene concentration from the

flash drum onwards.
The next figure, Fig. 8, describes a comparison of the polypropylene polymer
mass fractions from the CAPE-OPEN simulation with the reference.The marginal
differences between the simulation and reference data that could be observed in
Figs. 7 and 8 are caused by the numerical errors during simulation and the use of
certain design and operating conditions which could not be found in the reference
data for the simulation.
Furthermore, multi-scale simulation within this CAPE-OPEN framework
allows the possibility of observing polymer end-use properties predicted in the
reactor model. Hence, calculations made within the reactor model in using
empirical correlations to predict the end-use properties could be seen as part of
the simulation results in Table 4.
The polypropylene produced after this process has a molecular weight which
is within the notable range of molecular weight of available for polypropylene
grades as illustrated in Table 4. Meanwhile, the polydispersity of approximately,
2.3 is consistent to the range of polypropylene grades as illustrated. MFI is the
commonest method of testing the flow characteristics of the polymer melt. By its
definition, the MFI is an inverse measure of the melt viscosity of the polymer.

Journal of Engineering Science and Technology Special Issue 2 1/2015

Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 59

This property indeed controls the processability and spinability of the polymer
resin formed [21]. The polypropylene formed has a MFI of 11.1g/10min, which is
well within the range of polypropylene grades [18]. Also, the flexural modulus,
which is proportional to the stiffness of the material is a very important property
of the polymer. The value of the flexural modulus corresponds directly to the
deformability, as well as the durability of the polymer. It is also inversely,
proportional to the molecular weight of the polymer material [21]. The value
obtained for the polypropylene formed in this process, 1.6GPa, is also within the
range of the polypropylene grades as shown in Table 4.

Fig. 8. Comparison of Polypropylene Mass Fraction.

Table 4. Polymer Properties.

Variable Value Typical value Ref
Number average MW 137,567 g/mol 50,000-300,000 [17]
Weight average MW 322,483 g/mol 300,000-600,000 [17]
MFI 11.1 g/10min 0.50-70 [18]
Ef 1.66GPa 1.5-2 [19, 20]
PDI 2.34 2-6 [17]

5. Conclusions
This work involved multi-scale modelling of the polypropylene production
process taking the peculiar benefits of two process simulations software by
simultaneous execution on a unified framework called CAPE-OPEN. The
summary of the work done are the following:
A dynamic model for the loop reactor on the basis of the polymerisation of
propylene mechanism was developed considering the mass balance and
polymerisation reactions.

Journal of Engineering Science and Technology Special Issue 2 1/2015

60 J.C. Lee et al.

The model was validated with a reference data and was found to be
consistent with a reasonable degree of error.
The steady state form of the model was exported via the CAPE-OPEN
interface to connect with other units on PRO/II environment to calculate
the material balance and evaluate the process of polypropylene production.
The fractionation units considered are polymer separation and treatment
unit models including the steam jacket pipes, the degasser, the steamer and
the dryer.
Polymer end-use properties such as the MFI and PDI were also predicted by
calculation of the molecular weight averages using the method of moments.
High level modelling and simulation of the reactor and the overall production
process was achieved simultaneously, using the two powerful simulation software
with a reduced modelling burden whilst maintaining accuracy and paying less
computational cost.

Acknowledgement
This research is funded by the LINC Program of Hanbat National University
(2014) and the authors would like to acknowledge for the assistance.

References
1. De Lucca, E.A.; Filho, R.M.; Melo, P.A.; and Pinto, J.C. (2008). Modelling
and simulation of liquid phase propylene polymerisations in industrial loop
reactors. Macromolecules Symposium, 271(1), 8-14.
2. Zacca, J.J.; and Ray, W.H. (1993). Modelling of the liquid phase
polymerisation of olefins in loop reactors. Chemical Engineering Science,
48(22), 3743-3765.
3. Luo, Z.H.; Su, P.L.; and Wu, W. (2010). Industrial loop reactor for catalytic
propylene polymerisation: dynamic modelling of emergency accidents.
Journal of Industrial Engineering Chemical Research, 49(22), 1123211243.
4. Neto, A.G.; and Pinto, J.C. (2001). Steady state modelling of slurry and bulk
propylene polymerisations. Chemical Engineering Science, 56(13), 4043-4057.
5. gPROMS ModelBuilder (2014). Process Systems Enterprise, London,
United Kingdom.
6. Simsci PRO/II. (2013). Schneider Electric, California, USA.
7. Spheripol LyondellBasell group. (2014). Licensed Polyolefin Technologies
and Services. Retrieved July 7, 2014, from, http://www.lyondellbasell.
com/Technology/LicensedTechnologies.
8. Alizadeh, M.; Mostoufia, N.; and Pourmahdian, S. (2004). Modelling of
fluidized bed reactor of ethylene polymerisation. Chemical Engineering
Journal, 97(1), 27-35.
9. Daelim Industrial Company Limited. (1998). 35 MTY polypropylene
industrial plant, Seoul. Republic of Korea

Journal of Engineering Science and Technology Special Issue 2 1/2015

 Polypropylene Process Simulation with CAPE-OPEN Interfacing of Pro/II and . . . 61

10. Luo, Z.H.; Zheng, Y.; Cao, Z.K.; Wen, S.H. (2007). Mathematical modelling
of the molecular weight distribution of polypropylene produced in a loop
reactor. Polymer Engineering Science, 47(10), 1643-1649.
11. Chen, C.M.; and Ray W.H. (1993). Polymerisation of olefins through
heterogeneous catalysis; XI: gas phase sequential polymerisation kinetics.
Journal of Applied Polymer Science, 49(9), 1573-1588.
12. Shaffer, W.K.A; and Ray, W.H. (1997). Polymerisation of olefins through
heterogeneous catalysis; XXVII: A kinetic explanation for unusual effects.
Journal of Applied Polymer Science, 65(6), 1053-1080.
13. Crowley, T.J.; and Choi, K.Y. (1997). Calculation of molecular weight
distribution from molecular weight moments in free radical polymerisation.
Journal of Industrial Engineering Chemical Research, 36(5), 1419-1423.
14. Bremner, T.; and Rudin, A. (1990). Melt flow index and molecular weight
distributions of commercial thermoplastics. Journal of Applied Polymer
Science, 41(7-8), 1617-1627.
15. Shamiri, A.; Hussain, M.A.; Mjalli, F.S.; and Mostoufi, N. (2010). Kinetic
modelling of propylene homo-polymerisation in a gas-phase fluidized-bed
reactor. Chemical Engineering Journal, 161(1-2), 240-249.
16. Latado, A.; Embirucu, M.; Mattos Netto, A.G.; and Pinto, J.C. (2001).
Modelling of end-use properties of poly(propylene/ethylene) resins. Polymer
Testing, 20(4), 419-439.
17. Anthony, L.A. (2003). Plastics and the environment. John Wiley & Sons
Inc., New Jersey.
18. DSIR. (1998). Executive summary. Retrieved July 9, 2014, from
www.dsir.gov.in/reports/techreps/tsr082.pdf
19. MatWeb. (2014). Material property data: flexural strength testing of plastics.
Retrieved September 19, 2014, from http://www.matweb.com/reference/
flexuralstrength.aspx
20. Smithers, Rapra. (2012). Mechanical properties of polymers. Retrieved
September 21, 2014, from http://info.smithersrapra.com/downloads/
chapters/Physical%20testing%20of%20plastics_chapter%201.pdf
21. Fatih Dogan. (2012). Polypropylene. InTech Books and Journals, Rijeka.

Journal of Engineering Science and Technology Special Issue 2 1/2015

View publication stats