Simulation of An Industrial Linear Low D

Chemical Product and Process
Modeling
Volume 6, Issue 1 2011 Article 34
Simulation of an Industrial Linear Low

Density Polyethylene Plant
Ali Farhangiyan Kashani, Islamic Azad University South

Tehran Branch
Hossein Abedini, Iran Polymer and Petrochemical Institute
Mohammad Reza Kalaee, Islamic Azad University South
Tehran Branch
Recommended Citation:
Farhangiyan Kashani, Ali; Abedini, Hossein; and Kalaee, Mohammad Reza (2011) "Simulation
of an Industrial Linear Low Density Polyethylene Plant," Chemical Product and Process
Modeling: Vol. 6: Iss. 1, Article 34.
DOI: 10.2202/1934-2659.1611
©2011 De Gruyter. All rights reserved.

Brought to you by | University of Arizona
Authenticated
Download Date | 5/26/15 7:36 PM
Simulation of an Industrial Linear Low
Density Polyethylene Plant
Ali Farhangiyan Kashani, Hossein Abedini, and Mohammad Reza Kalaee
Abstract
In this paper, an industrial linear low density polyethylene (LLDPE) production process
including two serried fluidized bed reactors (FBR) and other process equipment was completely
simulated in steady state mode. Both of FBRs were considered like two serried continuous stirred
tank reactors (CSTR). In this simulation, a kinetic model that is based on a multiple active site
heterogeneous Ziegler-Natta catalyst was used for simulation of reactions in two FBRs. Simulator
by using this model is able to predict the important attributes of LLDPE like melt flow index (MFI),
density (ρ), polydispersity (PDI), numerical and weight average molecular weight (Mn, Mw) and
co-polymer molar fraction (SFRAC). On the other hand, this simulator can be applied in wide range
of changing in inlet operating conditions. The results of the simulation are compared with industrial
data of LLDPE plant. A good agreement is observed between the simulator predictions and actual
plant data. Finally, by using of the simulator, the steady state operating conditions for producing
different grades of polyethylene are obtained.
KEYWORDS: fluidized bed reactor, Ziegler-Natta catalyst, polymerization, polyethylene,

simulation
Author Notes: The corresponding author is Hossein Abedini, email: [email protected],

telephone: +98-021-44580028, fax: +98-021-44196583.

Authenticated
Farhangiyan Kashani et al.: Simulation of an Industrial Linear Low Density Polyethylene Plant
1. Introduction
Polyethylene is one of the important thermoplastic that used in a wide range of

applications such as films, bags, insulators, pipes and many plastic things.
Scientists and engineers continuously improved the polymerization techniques to
improve polymer properties, increase production capacity and reduce the cost of
material. The first commercial polyethylene was produced under very high
pressure. With the development of catalyst to allow coordination polymerization,
fluidized bed, slurry and solution phase reactors were invented. Gas phase
fluidized bed polymerization reactors were invented in the 1950s. Distinguishing
characteristic of gas phase polymerization is the system does not involve any
liquid phase in the polymerization zone. Polymerization occurs at the interface
between the solid catalyst and the polymer matrix. The gas phase plays a role in
the supply of monomers, mixing of polymer particles, making welled mixing
region and removal of reaction heat (Kiashemshaki et al., 2006). Production of
linear low density polyethylene (LLDPE) through heterogeneous Ziegler–Natta
catalysts is an example of such industrial applications of fluidized bed reactors
(FBR). A simplified flow diagram of gas-phase LLDPE production process is
shown in Figure 1.
Figure 1: Industrial fluidized bed polyethylene reactor (Alizadeh et al., 2004).
Published by De Gruyter, 2011 1

Authenticated
Chemical Product and Process Modeling, Vol. 6 [2011], Iss. 1, Art. 34
The first commercial gas phase polymerization plant using a fluidized bed
reactor was constructed by Union Carbide in 1968 at Seadrift, Texas (Xie et al.,
1994). Another commercialized gas fluidized bed technology was developed by
Naphtachimie in the 1970s (Mun, 2005). The Spherilene technology from
Lyondell-Basell is also a gas-phase process in two serried FBRs (Figure 2) for the
production of the whole density range of polyethylene products, from LLDPE to
MDPE (medium density) and HDPE (high density) (Spherilene Technology,
2010). In this paper, Spherilene technology is simulated.
Figure 2: Spherilene gas-phase technology of Lyondell Basell Co (Petrochemical

Processes, 2003).
Simulating and modeling of LLDPE production in FBRs has recently

received considerable attention. Some research efforts have been conducted on
simulating and modeling of the FBRs for producing of LLDPE. FBRs have
modeled as single, two or three-phase reactors. In case of LLDPE production in
FBRs, models such as single-phase continuous stirred tank reactor (CSTR)
(McAuley et al., 1994), two-phase (Choi and Ray, 1985), and heterogeneous
three-phase plug flow reactor (PFR) for all phases (Fernandes and Lona, 1998),
have been used. Wu and Baeyens (1998) introduced a mixing index, ranging from
0 to 1, for quantifying the extent of mixing in FBRs. For a typical ethylene
polymerization reactor, the mixing index was estimated to be about 0.4–0.5 which
could be explained as a poor mixing (Fernandes and Lona, 1998). This shows that
the FBR of LLDPE production process does not behave either as a CSTR or a
PFR but its hydrodynamics is situated between these two ideal cases (Levenspiel,
1999).
DOI: 10.2202/1934-2659.1611 2

Authenticated
On the other hand, optimization of LLDPE production process has

recently become more important problem. This trend has led the polyethylene
industry to move away from large continuous production of a single polymer
grade to a more flexible production plan including a number of polymer grades of
high quality but low volume (Chatzidoukas et al., 2003). The calculation of
optimal grade transition strategies in catalytic polymerization processes has been
the subject of various publications. Cozewith (1988) studied the effect of step
changes in chain transfer agent and monomer feed rates on numerical and weight
average molecular weight (Mn,Mw), polydispersity (PDI) and co-polymer molar
fraction (SFRAC) for a CSTR. He clearly showed that the direction and magnitude
of the transition greatly affected the transient responses of the polymer quality
attributes such as melt flow index (MFI) and density (ρ). McAuley and
MacGregor (1992) investigated the optimal grade transition problem for a FBR.
Debling et al. (1994) applied a heuristic approach based on industrial practice to
solve the optimal grade transition problem for solution, slurry and gas-phase
polymerization processes.
In the present study, for the first time an industrial linear low density
polyethylene (LLDPE) production process including two serried fluidized bed
reactors (FBR) and other process equipments was completely simulated. The
results of the steady state simulation are compared with industrial data of LLDPE
plant. Finally, by using of the simulator, the steady state operating conditions for
producing different grades of polyethylene are obtained.
2. Model Description
2.1 Kinetic modeling
In this paper, a kinetic model that is based on a double active site heterogeneous
Ziegler-Natta catalyst was used for producing LLDPE in two FBRs. This kinetic
model was developed by McAuley et al. (1990) and simulator by using this model
is able to predict the important attributes of LLDPE like MFI, ρ, PDI, Mn, Mw.
Catalysts for ethylene polymerization are mostly heterogeneous, but some
processes also use soluble catalysts. There are now four types of catalysts for
ethylene polymerization: Ziegler-Natta, Phillips, Metallocenes and late transition
metal catalysts (Soares, 2001; Winter et al., 2002; Mun, 2005). Ziegler-Natta
catalysts have many variations but are generally TiCl4 supported on MgCl2,
combined with a variety of electron donors and co-catalysts (Khare, 2003;
Ghafelebashi Zarand and Mortazavi, 2005). In operations, triethyl aluminum as
co-catalyst is added separately to the reactor feed. During chain growth, the
double bonds in monomer molecules coordinate to the active catalyst sites. As a
result, the monomer units added at early times are farther down the chain

Authenticated
extending from the catalyst site, relative to monomer added more recently. Figure
3 shows the generally accepted addition mechanism, according to Cossee (1964).
Figure 3: Mechanism for ethylene addition to a growing chain for a Ziegler-Natta

catalyst (Khare, 2003).
There are two existing theories that explain the ability for Ziegler-Natta
catalysts to produce polydisperse polymers. The first is diffusive and the second is
kinetic. The diffusive theory involves the assumption that mass-transfer
limitations affect the ability of the monomer molecules to reach some of the
active sites on the catalyst particle, due to the presence of pores or other structural
effects that hinder transport to the sites. As a result, different sites polymerize
monomer at different rates, leading to a polydisperse polymer product. The
kinetic theory involves the assumption that there exist different catalyst site types,
each with its own relative reactivity, due to variations in the local chemical
composition of each site type. Experimentation has provided compelling evidence
that this second theory is a more accurate description of the physical reality
(Khare, 2003). Therefore the kinetic equations are used in this work is based on
the kinetic theory, rather than the mass-transfer (diffusive) theory.
Ethylene polymerization can be envisioned as occurring at the interface
between the solid catalyst and the polymer matrix, where the active centers are
located. From gas-state monomer to solid-state polymer, ethylene experiences a
dramatic physicochemical transition within a very short time. The polymerization
environment changes with the composition of catalyst, polymerization process,
reactant composition, reactor operating conditions, and extent of polymerization.
Nevertheless, the key elementary reactions have been established, which include
formation of active centers, insertion of monomer into the growing polymer
chains, propagation, chain-transfer reactions, and catalyst deactivation. Most of
the proposed mechanisms are based on information about polymerization rate,
molecular weight and its distribution, polymer chain microstructure, and active
DOI: 10.2202/1934-2659.1611 4

Authenticated
center concentrations (Mun, 2005). It possible to assume that all of the active
centers perform the same reaction mechanisms, but with different reaction rates
for each elementary reaction (Kiashemshaki et al., 2006), then elementary
reactions which are commonly adopted in modeling studies can be summarized as
in Table 1.
In the present study, a comprehensive mechanism was considered to
describe the copolymerization kinetics of ethylene with butene over a Ziegler–
Natta catalyst with double catalyst sites based on the kinetic model proposed by
McAuley et al. (1990). For multicomponent polymerizations, the use of pseudo
kinetic rate constants can considerably simplify the kinetic rate expressions
(Decarvalho et al., 1989; McAuley et al., 1990). Based on the proposed kinetic
mechanism (Table 1) and the definition of the moments of the “live” (λv) and
“dead” (µv) total number chain length distributions,
, , ∑ , ∑ , (1)
μ ∑ (2)
The net production/consumption rates of the various molecular species in

the FBR can be derived. The symbol P , denotes the concentration of “live”
copolymer chains of total length ‘n’ ending in an ‘i’ monomer unit, formed at the
‘k’ catalyst active site. P and D denote the concentrations of the activated
vacant catalyst sites of type ‘k’ and “dead” copolymer chains of length ‘n’
produced at the ‘k’ catalyst active site, respectively.
Table 1: Summary of elementary reactions for ethylene copolymerization

(Chatzidoukas et al., 2003).
Activation
,
Chain initiation
,
,
Propagation
,
Spontaneous deactivation
,
Spontaneous chain transfer
,
,
Chain transfer by hydrogen
,
,
Chain transfer by monomer (Mi)
,

Authenticated
2.2 Choosing rate constants
Experimental determination of rate constants for multiple site systems is an

important task. Decarvalho et al. (1989) discusses one methodology which
involves cross-fractionation of polymer samples by TREF and SEC followed by
NMR analysis of the cross-fractions. Important rate parameters can then be
determined from the NMR results and reactor operating conditions. Since rate
constants have not been determined for the silica-supported magnesium and
titanium-based industrial catalyst system used, initial order of magnitude
estimates for rate constants were obtained from Kissin (1987) for similar catalyst
systems. Propagation and transfer rate constants at individual sites were then
manipulated to produce results corresponding to those determined experimentally
by Usami et al. (1986).
In this paper, double active site catalyst was employed for simulation and
all reaction rate constants have been shown in Table 2. Transfer rate constants
were also manipulated in order to achieve MW corresponding to the MFI
laboratory data and to achieve a PDI between 3.5 and 4.5.
Table 2: Rate parameters for model predictions (McAuley et al., 1990)

Reaction Rate constant Unit Site type 1 Site type 2
Activation Lit/mol.s 0.1 0.1
, 1 1
Initiation Lit/mol.s
,
0.14 0.14
85 85
,
2 15
,
Propagation Lit/mol.s
,
64 64
, 1.5 6.2
0.0021 0.0021
, 0.006 0.11
, 0.0021 0.001
Lit/mol.s
,
Chain transfer 0.006 0.11
,
0.88 0.37
,
, s-1
0.0001 0.0001
-1
Deactivation s 0.0001 0.0001
DOI: 10.2202/1934-2659.1611 6

Authenticated
3. Simulation
3.1 Modeling
In case of low or moderate activity of the catalyst, heat transfer and diffusion
resistances do not play an important role at the particle level in the gas-phase
polyethylene reactors. In the limiting case, where either bubbles are small or
interphase mass and energy transfer rates are high and catalyst is at low to
moderate activity, intraparticle temperature and concentration gradients are
negligible. When modeling a FBR or a recycle reactor with high conversion per
pass, one must account for gas-phase concentration gradients between the bottom
and top of the bed and for mass transfer between bubble and emulsion phases
(McAuley et al., 1994). It is important to determine whether heat and mass
transfer resistances between the bubbles and the emulsion phase are significant in
typical industrial reactors, or whether the entire contents of the FBRs can
reasonably be treated as well-mixed.
The modeling of the input/output characteristics of recycle-dominated
FBRs for polymers can be simplified greatly by making suitable assumptions.
Since the industrial FBRs system under consideration has a sizable recycle stream
and a low conversion per pass through the bed, the vertical concentration gradient
through the bed is very small and can be neglected. Back mixing of both the gas
and solid phases in the fluidized bed does occur. However, even if the gas
experienced pure plug flow through the fluidized bed, the large recycle to fresh
feed ratio would make the well-mixed assumption valid for the gas phase since
recycle PFR dynamics approach those of CSTRs consist of emulsion phase as the
recycle to fresh feed ratio becomes very large (McAuley et al., 1990;
Chatzidoukas et al., 2003). In normal industrial reactor operation, the recycle to
fresh feed ratio is approximately 40:1(McAuley et al., 1994). Thus modeling the
FBR plus recycle system as a CSTR containing a well-mixed solid phase
interacting with a well-mixed gas phase is justified (Levenspiel, 1999). The above
assumption holds true for the majority of industrial FBRs.
Since no separate bubble phase is included in the model, the bed voidage,
(ε bed), accounts for the overall gas volume fraction in the bed. The assumption of
perfect mixing in the bed implies that the temperature and concentrations of the
various molecular species will be in dependent of their position in the bed.
Furthermore, it was assumed that mass and heat transfer resistances between the
polymer particles and the gas phase were negligible and the catalyst contained
two types of active sites. Based on the above assumptions, the following dynamic
molar balances for the monomer and hydrogen is derived.

Authenticated
Monomer i:
,
∑ (3)
Hydrogen:
,
∑ (4)
Where R M , R H are the monomer and hydrogen consumption

rates at the catalyst active site of type‘k’. Xi and Xi, in are the mass
fractions of species ‘i’ in the recycle and input streams, respectively.
Similarly, the mass balances for the potential catalyst sites, S , and all
other molecular species Y : P , λ , λ , λ , μ , μ , μ can be derived:
/ ,
(5)
(6)
Where R denotes the net formation rate of the molecular species Y . The
symbols h, Vbed, and A denote the bed height, the volume and the cross-sectional
area of the bed, respectively. Accordingly, one can derive the unsteady-state mass
balance for the polymer in the bed,
∑ ∑ (7)
Where ρ and ρcat are the corresponding densities of polymer and catalyst.
The dynamic energy balance for the reaction mixture in the bed is written as:
, , ,
(8)
Where the terms Hgas,in, Hgas,out, Hprod,out, and Hgenr denote the enthalpies of
the input, output and product removal streams and the heat of polymerization,
respectively. Assuming that the dynamic behavior of the external heat exchanger
can be approximated by a series of Nz well-stirred zones for the recycle stream
and a single well-stirred zone for the coolant, the following energy balances can
be written:
DOI: 10.2202/1934-2659.1611 8

Authenticated
, , , 1,2, . . . , (9)
∑ ∑ (10)
, , , (11)
In Table 3, the numerical values of the reactor and heat exchanger design
parameters are reported.
Table 3: Reactor and heat exchanger design parameters

Design parameters Reactor 1 Reactor 2
Bed diameter Dbed = 3.4 m Dbed = 4.8 m
Bed voidage εbed = 0.5 εbed = 0.5
Catalyst density Ρcat = 2092 kg/m3 Ρcat = 2092 kg/m3
Overall heat exchanger area A = 700 m2 A = 1600 m2
Overall heat transfer coefficient Uj = 1000 J/K/m2/s Uj = 1000 J/K/m2/s
Coolant inlet temperature Tw,in = 290 K Tw,in = 290 K
The average polymer properties of interest can be calculated in term of the

bulk moments of the total number chain length distributions and the monomer
consumption rates at the various catalyst sites. Thus, the instantaneous copolymer
composition, φ , for a catalyst having Ns active sites, will be given by the
consumption rate of the ‘i’ monomer over the total consumption rate of the Nm
monomers:
∑ ∑ ∑ (12)
To calculate the cumulative copolymer composition, Φ , in the reactor,

during a transient operation, the following dynamic mass balance equation needs
to be solved:
, , , 1 (13)
Where Mp, Fp,in and Fp,out denote the total polymer mass in the bed and the
input and output polymer mass flow rates, respectively. Accordingly, the number
and weight average molecular weights of the copolymer will be given by the
following equations:

Authenticated
∑ μ ∑ μ (14)
∑ μ ∑ μ (15)
Where MW is the average molecular weight of the repeating unit in the

copolymer chains.
∑ (16)
Finally, the PDI will be given by the ratio of the weight average over the
number average molecular weight.
M
PDI (17)
M
This simulator is able to calculate the important polymer attributes such as

MFI، ρ، PDI، Mw، Mn and SFRAC. In this simulation for calculating MFI and ρ, we
use Sinclair’s correlations (Sinclair, 1983) that are given by the following
equations:
.
ρ 0.966 – 0.02386 SFRAC (18)
M – .
MFI (19)
In addition, for all polymeric streams were used POLY-SRK equation of

state and for all non-polymeric streams were also used NRTL-RK equation of
state. POLY-SRK equation of state (Holderbaum and Gmehling, 1991; Fischer
and Gmehling, 1996; Aspen Polymer Plus User’s Manual, 2001):
∑ 1.546 ∑ (20)
,
0.08664
,
,
0.42748
,
. . .
1 1 , 1 , 1 ,
DOI: 10.2202/1934-2659.1611 10

Authenticated
,
,
To use the polymer SRK equation of state several pure component

parameters are required, including the critical constants Tc,i, Pc,i and the Mathias-
Copeman constants (c1, c2, c3) (Mathias and Copeman, 1983).
3.2 Process description
In this paper, the whole LLDPE production process (base on Basell’s Spherilene
technology) in two serried CSTR and other process equipment such as coolers,
mixers, splitters, separation columns and compressors was completely simulated.
Figure 4 shows the process of LLDPE production. Industrial data of LLDPE plant
was used for all operation conditions and input quantities (Table 4). In this
process, reactants (monomers and hydrogen) and inert gas (ethane, propane) are
fed into the bottom of two reactors through a distributor. The distributor maintains
the fluidization and supplies the reactants for growing polymer particles. The
catalyst is loaded continuously into the bed and the polymer product is taken out
from two reactors at a rate such that the bed height is held constant. Unreacted
gases together with fine particles exit from the top of the bed through a curved
disengaging zone at the upper part of the reactors (Alizadeh et al., 2004). Settling
of solid particles is facilitated in this section.
Figure 4: Process flow diagram of LLDPE industrial plant based on Spherilene

gas-phase technology.

Authenticated
Table 4: Industrial data of input quantities and operating conditions of LLDPE

plant
Industrial Industrial
Quantity Unit
Reactor 1 Reactor 2
M hydrogen Kg/h 16.43 49.88
M ethylene Kg/h 11385.40 29957.05
M ethane Kg/h 169.04 551.97
M propane Kg/h 10675.68 8950.69
M butene Kg/h 1564.64 5437.34
M catalyst Kg/h 4.36 4.36
M co-catalyst Kg/h 26.16 26.16
M polymer Kg/h 13100 43600
T reactor K 348 353
P reactor barg 25 25
3
V reactor m 267 697
Residence time hr 2 1.5
In this way, as the catalyst reaction of ethylene polymerization, the

polymer will grow on the catalyst sites, increasing its weight and size as the
catalyst-polymer particles fall down (Fernandes and Lona Batista, 1999). There
will be a degree of segregation of these particles in the reactor according to size
and weight. Low mass catalyst-polymer particles will tend to stay at the top of the
reactors, whereas heavy, full-grown particles will be at the bottom of two reactors
near the outflow stream. Finally, the gas then enters a cyclone in order to separate
the remainder particles and is combined with fresh feed stream after heat removal
in the heat exchanger and then recycled to the base of the reactors.
4. Results and Discussion
In this work, unsteady state equations (1-13) for LLDPE production process
including two serried FBRs are generally given, but simulation is done in steady
state mode. This simulator can be applied in wide range of changing in inlet
operating conditions. In order to demonstrate the predictive capabilities of the
proposed model, simulations were carried out at the operating conditions that
were shown in Table 4. Such operating conditions are typical of industrial
polyethylene reactors for producing the desired grade indicated in this table. A
comparison between the results of the simulation presented in this work and the
actual plant data is shown in Table 5 in term of the important properties of the
produced polymer and some operating conditions. It has been acceptably shown
DOI: 10.2202/1934-2659.1611 12

Authenticated
that a good agreement is observed between the simulator predictions and actual
plant data.
Since the high volume flow rate of the non reacted feed with other
neutralize gases are returned to two reactors by recycle streams, thus the smallest
error in predicting flow rate and percentage of components in the output of reactor
is caused a small error in the calculation of parameters such as average heat
capacity and total mass flow rate of recycle stream and consequently a higher
error in the rate of the condenser heat duty (Q = F . Cp,avg . (Tout - Tin), Error in
calculating of Q is obtained from the error in calculating of F and Cp,avg ).Also,
because of the errors in the calculating of output specification of the first reactor
(input of the second reactor), the error in calculating of the condenser heat duty of
the second reactor is more. All operating conditions and input quantities of the
simulation were selected for producing of LL16503 grade as reference grade.
Table 5: Comparison between simulation results and industrial data

Equipment Quantity Unit Industrial data Simulation result Simulation error
REACTOR 1 QLLDPE kg/hr 13100 13023 0.59%
REACTOR 2 QLLDPE kg/hr 43600 43506 0.22%
REACTOR 1 MFI gr/10 min - 38 -
REACTOR 2 MFI gr/10 min 3.00 3.08 2.67%
REACTOR 1 ρ gr/cm3 0.9160 0.9178 0.19%
REACTOR 2 ρ gr/cm 3
0.9160 0.9156 0.04%
REACTOR 1 PDI - - 3.95 -
REACTOR 2 PDI - - 4.76 -
REACTOR 1 Mw kg/kmol - 39027 -
REACTOR 2 Mw kg/kmol - 80553 -
COOLER 1 HEAT DUTY KW -13237 -12880 2.70%
COOLER 2 HEAT DUTY KW -35669 -32500 8.88%
COMPRESSURE 1 WORK KW 2500 2599 3.96%
COMPRESSURE 2 WORK KW 5000 4969 0.62%
4.1 Monomer effects
Ethylene monomer is used in this process as an unsaturated hydrocarbon. Figures

5 and 6 show the effects of the ethylene mass flow rate inlet to the first and
second reactor (Methylene) on Mpolymer, ρ and MFI of outlet polymer from the
second reactor. Generally, with increasing of Methylene to each reactor, Mpolymer and
ρ are increased but MFI is decreased as shown in these figures. As shown in
Figure 5c, the MFI is almost constant by increasing the Methylene of the first reactor
from 8000 to 11000 and then decreases.

Authenticated
Figure 5: The effects of variation of Methylene (feed to reactor 1) on Mpolymer (a), ρ

(b) and MFI(c).
DOI: 10.2202/1934-2659.1611 14

Authenticated
Figure 6: The effects of variation of Methylene (feed to reactor 2) on Mpolymer (a), ρ

(b) and MFI(c).

Authenticated
4.2 Hydrogen effects
Hydrogen is used as chain transfer agent and chains growing controller in olefin
polymerization especially in ethylene polymerization. Figure 7 shows variation of
MFI with changing of hydrogen mass flow rate inlet to the first and second
reactor (Mhydrogen). As it is observed in Figure 7, whit increasing of Mhydrogen to
each reactor, MFI of product (outlet polymer from the second reactor) is
increased. The hydrogen mass flow rate inlet to the second reactor has more effect
on MFI than the first reactor hydrogen mass flow rate which is due to higher
volume and inlet feed mass flow rates of the second reactor (Table 4).
Figure 7: The effects of variation of Mhydrogen (feed to reactor 1 (a) and 2 (b))
on MFI.
DOI: 10.2202/1934-2659.1611 16

Authenticated
4.3 Co-monomer effects
Butene is used in co-polymerization reactions as co-monomer. Butane segments

are placed on long chains as a branch for controlling density of polymer and with
changing of its concentration; it could be possible to produce different grades of
polyethylene. Figure 8 shows clearly variation of ρ versus flow rate of co-
monomer inlet to the first and second reactor. With increasing of inlet flow rates
of co-monomer, polymer density is decreased in outlet polymer of the second
reactor.
Figure 8: The effects of variation of Mbutene (feed to reactor 1 (a) and 2 (b)) on ρ.

Authenticated
4.4 Catalyst effects
Catalyst has an important role in ethylene polymerization process and reactions

are occurred by using it. Figure 9 shows effect of variation of catalyst flow rate on
ρ and MFI of product. With increasing Mcatalyst, ρ of product is decreased and MFI
is increased. It is clear, with increasing of catalyst flow rate; polymer production
outlet from two reactors is increased.
Figure 9: The effects of variation of Mcatalyst on ρ and MFI.
DOI: 10.2202/1934-2659.1611 18

Authenticated
4.5 Molecular weight distribution (MWD)
Figure 10 shows the distribution of molecular weight of polymer with respect to

weight fraction for produced polyethylene from two reactors. As it is observed in
Figure 10, MWD of outlet produced polymer from first reactor is narrow because
of lower molecular weight, but MWD of exited LLDPE from second reactor is
wider due to higher molecular weight.
Figure 10: Molecular weight distribution of produced polyethylene from reactor 1

(○) and 2 (□).
4.6 Operating conditions for production different grades
In this part, the capabilities of producing different grades of polyethylene in an

industrial plant based on present model is investigated and at last a main pattern
for finding the operating conditions for any grade is suggested. Various grades
have different polymer density and MFI. Polymer density is determined by ratio
of Co-monomer (butane) flow rate respect to ethylene in feeds of the reactors
(bi=C4/(C4+C2)). Also ratio of Hydrogen flow rate respect to ethylene in feeds of
the reactors (ai=H2/C2) have main effect on polymer MFI. Thus the variations of
final polymer MFI and ρ versus of ai=H2/C2 and bi=C4/(C4+C2) (where i=1,2
refers to input fraction to reactor number i) as two main variables are shown in
Figure 11.
In this paper, graphical optimization method is used for the finding the
operating conditions for production any desired grade. So, for producing any
grade, by having ρ and MFI the related range of ai and bi are obtained by using

Authenticated
Figure 11. In Table 6 allowable ranges of two upon fractions (operating

conditions) for some commercial grades are given.
Table 6: Allowable ranges of ai and bi for producing different LLDPE grades

Range of
Grade Application Range of a1 Range of b1 Range of a2 Range of b2 Range of ρ
MFI
LLDPE
FILM 0.255-0.310 0.180-0.195 0.245-0.275 0.185-0.215 0.85-1.05 0.915-0.917
16501
LLDPE
FILM 0.490-0.590 0.180-0.195 0.365-0.425 0.185-0.215 1.8-2.2 0.915-0.917
16502
LLDPE
FILM 0.700-0.845 0.180-0.195 0.475-0.530 0.185-0.215 2.7-3.3 0.915-0.917
16503
LLDPE
FILM 0.195-0.250 0.155-0.165 0.205-0.245 0.155-0.165 0.6-0.8 0.921-0.923
235F7
LLDPE
FILM 0.255-0.310 0.155-0.165 0.245-0.275 0.155-0.165 0.85-1.05 0.921-0.923
22501
LLDPE INJECTION
0.440-0.540 0.155-0.165 0.345-400 0.155-0.165 1.6-2 0.921-0.923
22502 MOULDING
LLDPE INJECTION
2.965-3.290 0.155-0.165 1.200-1.35 0.155-0.165 14-16 0.921-0.923
20516 MOULDING
LLDPE INJECTION
1.000-1.450 0.155-0.165 0.590-0.750 0.155-0.165 4-6 0.921-0.923
20505 MOULDING
LLDPE EXTRUSION
1.450-1.865 0.150-0.160 0.750-0.900 0.145-0.155 6-8 0.922-0.924
23507 COATING
LLDPE ROTO
1.000-1.450 0.120-0.130 0.590-0.750 0.110-0.120 4-6 0.929-0.931
30505 MOLDING
LLDPE ROTO
0.775-1.230 0.110-0.120 0.500-0.675 0.105-0.110 3-5 0.931-0.933
32604 MOLDING
DOI: 10.2202/1934-2659.1611 20

Authenticated
Figure 11: The effects of variation of ai=H2/C2 and bi=C4/(C4+C2) on ρ and MFI
of final product, (×) for feed to the reactor 1, (○) for feed to reactor 2.
5. Conclusions
Polyethylene is the most important thermoplastic and used in a wide range of

applications. Simulating and modeling of LLDPE production in FBRs has
received considerable attention. These attempts have led to a more realistic
understanding of the reactor behavior as well as the properties of the polymer
produced in the reactor.
In this work, LLDPE production process in two serried FBRs was
completely simulated. In this simulation, two serried CSTRs were used instead of
FBRs in series and a dynamic kinetic model that is based on a double active site
heterogeneous Ziegler-Natta catalyst was employed for simulating LLDPE
production in two FBRs. This kinetic model was developed by McAuley’s
dynamic kinetic model (McAuley, 1990), and simulator by using this model is
able to predict the important attributes of LLDPE like MFI, ρ, PDI, Mn, Mw and
SFRAC.

Authenticated
In this work, unsteady state equations for LLDPE production process

including two serried FBRs are generally given, but simulation is done in steady
state mode. Also, the results of the simulation were compared with industrial data
of LLDPE plant in steady state mode that a good agreement is observed between
the simulator predictions and actual plant data. Finally, by using of simulator, the
process was optimized for producing different grades of LLDPE.
Nomenclature
A cross-sectional area (m2)

Cp,mean specific heat capacity of the reaction mixture in the recycle stream
(kcal/kg.K)
Cp, w specific heat capacity of water (kcal/kg.K)
[D ] concentration of “dead” copolymer chains of length n produced at k
catalyst active site (kmol/m3)
Fcat catalyst feed rate (kg/hr)
FH2 hydrogen feed rate into the bed (kg/hr)
Fp polymer flow rate (kg/hr)
Frec recycle flow rate (kg/hr)
[H2] hydrogen concentration in the bed (kmol/m3)
Haccum accumulation enthalpy term (kcal/K.m3)
Hgas,in gas input enthalpy rate (kcal/s.m3)
Hgas,out gas output enthalpy rate (kcal/s.m3)
Hgenr polymerization heat rate (kcal/s.m3)
Hprod,out product output enthalpy rate (kcal/s.m3)
ka kinetic rate constant of activation reaction (m3/kmol.s)
kdsp kinetic rate constant of spontaneous deactivation reaction (s−1)
k0 kinetic rate constant of initiation reaction (m3/kmol.s)
kp kinetic rate constant of propagation reaction (m3/kmol.s)
kt kinetic rate constant of chain transfer reaction (m3/kmol.s)
ktsp kinetic rate constant of spontaneous chain transfer reaction (s−1)
Mbutene mass flow rate of butane (kg/hr)
Mc coolant mass in the heat exchanger (kg)
Mcatalyst mass flow rate of catalyst (kg/hr)
Mco-catalyst mass flow rate of co-catalyst (kg/hr)
Methylene mass flow rate of ethylene (kg/hr)
Methane mass flow rate of ethane (kg/hr)
MFI melt flow index (gr/10 min)
Mgas total mass of gases in the heat exchanger (kg)
Mhydrogen mass flow rate of hydrogen (kg/hr)
[Mi] monomer concentration in the bed (kmol/m3)
DOI: 10.2202/1934-2659.1611 22

Authenticated
Mn number average molecular weight of polymer (kg/kmol)

Mp total polymer mass in the reactor (kg)
Mpolymer mass flow rate of polymer (kg/hr)
Mpropane mass flow rate of propane (kg/hr)
MW component molecular weight (kg/kmol)
Mw weight average molecular weight of polymer (kg/kmol)
Nm total number of monomers
Ns number of catalyst active sites
Nz number of well-stirred zones in the heat exchanger
P pressure (barg)
PDI polydispersity index
[Pn,i] concentration of “live” copolymer chains of length n ending in an i
monomer unit (kmol/m3)
Q heat transfer rate (kcal/s)
Q0 volumetric product removal rate (m3/s)
SFRAC segment molar fraction
[Sp] concentration of potential catalyst active sites (kmol/m3)
T temperature (K)
Tw, in inlet water temperature to the heat exchanger (K)
Tw water temperature in the heat exchanger (K)
U overall heat transfer coefficient (J/K.m2.s)
V volume (m3)
Vbed bed volume (m3)
XMi mass fraction of monomer i in the bed
XH2 mass fraction of hydrogen in the bed
Greek letters
ε bed void fraction

λ “live” copolymer moment of l order (kmol/m3)
µ “dead” copolymer moment of l order (kmol/m3)
ρ density (kg/m3)
φ instantaneous copolymer composition with respect to the i monomer
Φ cumulative copolymer composition with respect to the i monomer
Subscripts and superscripts
1 ethylene property
2 butene property
k type of catalyst active site
p polymer property

Authenticated
References
Alizadeh, M., Mostoufi, N., Pourmahdian, S., Sotudeh-Gharebagh, R, Modeling

of Fluidized Bed Reactor of Ethylene Polymerization, Chemical
Engineering Journal, 97, 27-35, 2004.
Aspen Polymer Plus 11.1, User’s Manual, Aspen Technology, Inc., 131-136,
2001.
Chatzidoukas, C., Perkins, J.D., Pistikopoulos, E.N., Kiparissides, C., Optimal

Grade Transition and Selection of Closed-loop Controllers in a Gas-phase
Olefin Polymerization Fluidized Bed Reactor, Chemical Engineering
Science, 58, 3643-3658, 2003.
Choi, K.Y., Ray, W.H., The Dynamic Behavior of Fluidized Bed Reactors for
Solid Catalyzed Gas Phase Oleﬁn Polymerization, Chemical Engineering
Science, 40, 2261–2279, 1985.
Cossee, P. Ziegler-Natta Catalysis I. Mechanism of Polymerization of a-Olefins

with Ziegler-Natta Catalysts, Journal of Catalysis, 3, 80, 1964.
Cozewith, C., Transient Response of Continuous flow Stirred tank Polymerization

Reactors, AIChE. Journal, 34, 272, 1988.
Debling, J.A., Han, G.C., Kuijpers, J., VerBurg, J., Zacca, J., Ray, W.H.,
Dynamic Modeling of Product Grade Transitions for Olefin
Polymerization Processes, AIChE. Journal, 40, 506, 1994.
Decarvalho, A.B., Gloor, P.E., Hamielec, A.E., A Kinetic Mathematical Model

for Heterogeneous Ziegler-Natta Copolymerization, Polym., 30, 280,
1989.
Fernandes, F.A.N., Lona Batista, L.M.F., Fluidized-bed Reactor and Physical-

chemical Properties Modeling for Polyethylene Production, Computers
and Chemical Engineering Supplement, 803-806, 1999.
Fernandes, F.A.N., Lona, L.M.F., Heterogeneous Modeling for Fluidized-bed

Polymerization Reactor, Chemical Engineering Science, 56, 963–969,
2001.
DOI: 10.2202/1934-2659.1611 24

Authenticated
Fischer, K., Gmehling, J., Further development, status and results of the PSRK
method for the prediction of vapor-liquid equilibrium and gas solubilities,
Fluid Phase Eq., 121, 185,1996.
Ghafelebashi Zarand, S.M., Mortazavi, S.M.M., Mathematical Modeling of

Ethylene Polymerization with Ziegler-Natta Catalyst, ESCAPE-15,
Barcelona, Spain, 2005.
Holderbaum, T., Gmehling, J., PSRK: A group contribution c based on UNIFAC,

Fluid Phase Eq., 70, 251,1991.
Khare, N.P., Predictive Modeling of Metal-Catalyzed Polyolefin Processes, PhD

Thesis, Virginia Polytechnic Institute and State University, 2003.
Kiashemshaki, A., Mostoufi, N., Sotudeh-Gharebagh, R., Two-Phase Modeling of

a Gas Phase Polyethylene Fluidized Bed Reactor, Chemical Engineering
Science, 61, 3997– 4006, 2006.
Kissin, Y.V., Isospecifc Polymerization of Olefins with Heterogeneous Ziegler-

Natta Catalysts, Springer-Verlag, 1st ed., New York, 1987.
Levenspiel, O., Chemical Reaction Engineering, 2nd ed., Wiley, New York, 1999.
Mathias, P. M., Copeman, T. W., Extension of the Peng-Robinson equation of

state to complex mixtures: evaluation of the various forms of the local
composition concept, Fluid Phase Eq., 13, 91, 1983.
McAuley, K.B., MacGregor, J.F., Optimal Grade Transition in a Gas Phase

Polyethylene Reactor, AIChE. Journal, 38, 1564, 1992.
McAuley, K.B., MacGregor, J.F., Hamielec, A.E., A Kinetic Model for Industrial
Gas-phase Ethylene Copolymerization, AIChE. Journal, 36, 837-850,
1990.
McAuley, K.B., Talbot, J.P., Harris, T.J., A Comparison of Two-phase and Well-
mixed Models for Fluidized-bed Polyethylene Reactors, Chemical
Engineering Science, 49, 2035–2045, 1994.
Mun, T.C., Production of Polyethylene Using Gas Fluidized Bed Reactor,

National University of Singapore, HT022626U, 1-20, 2005.

Authenticated
Petrochemical Processes, Hydrocarbon Processing, 116-120, 2003.
Sinclair, K.B., Characteristics of Linear LDPE and Description of UCC Gas

Phase Process, Process Economics Report, SRI International, Menlo Park,
CA, 1983.
Soares, J.B.P., Mathematical Modeling of the Microstructure of Polyolefins Made

by Coordination Polymerization, Chemical Engineering Science, 56,
4131- 4153, 2001.
Spherilene Technology, Licensed Polyolefin Technologies and Services,

Lyondell-Basell Co., Germany, www.lyondellbasell.com, viewed on 2010.
Usami, T., Gotoh, Y., Takayama, S., Generation Mechanism of Short-Chain

Branching Distribution in Linear Low Density Polyethylene,
Makromolec., 19, 2722-2726, 1986.
Winter, A., Kueber, F., Spaleck, W., Riepl, H., Herrmann, W.A., Dolle, V.,
Rohrmann, J., Process for the Preparation of an Olefin Polymer Using
Metallocenes Containing Specifically Substituted Indenyl Ligands, US
Patent RE37,573 E, 2002.
Wu, S.Y., Baeyens, J., Segregation by Size Difference in Gas Fluidized Beds,
Powder Technol., 98, 139–150, 1998.
Xie, T., McAuley, K.B., Hsu, J.C.C., Bacon, J.W., Gas Phase Ethylene
Polymerization: Production Processes, Polymer Properties and Reactor
Modeling, Ind. Eng. Chem. Res., 33, 449 – 479, 1994.
DOI: 10.2202/1934-2659.1611 26

Authenticated

Simulation of An Industrial Linear Low D

Uploaded by

Copyright:

Available Formats

Simulation of An Industrial Linear Low D

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Simulation of An Industrial Linear Low D

Uploaded by

Copyright:

Available Formats

Chemical Product and Process

Simulation of an Industrial Linear Low

Ali Farhangiyan Kashani, Islamic Azad University South

©2011 De Gruyter. All rights reserved.

KEYWORDS: fluidized bed reactor, Ziegler-Natta catalyst, polymerization, polyethylene,

Author Notes: The corresponding author is Hossein Abedini, email: [email protected],

Brought to you by | University of Arizona

Polyethylene is one of the important thermoplastic that used in a wide range of

Figure 1: Industrial fluidized bed polyethylene reactor (Alizadeh et al., 2004).

Published by De Gruyter, 2011 1

Brought to you by | University of Arizona

Figure 2: Spherilene gas-phase technology of Lyondell Basell Co (Petrochemical

Simulating and modeling of LLDPE production in FBRs has recently

Brought to you by | University of Arizona

On the other hand, optimization of LLDPE production process has

2.1 Kinetic modeling

Published by De Gruyter, 2011 3

Brought to you by | University of Arizona

Figure 3: Mechanism for ethylene addition to a growing chain for a Ziegler-Natta

Brought to you by | University of Arizona

The net production/consumption rates of the various molecular species in

Table 1: Summary of elementary reactions for ethylene copolymerization

Published by De Gruyter, 2011 5

Brought to you by | University of Arizona

2.2 Choosing rate constants

Experimental determination of rate constants for multiple site systems is an

Table 2: Rate parameters for model predictions (McAuley et al., 1990)

Brought to you by | University of Arizona

Published by De Gruyter, 2011 7

Brought to you by | University of Arizona

Where R M , R H are the monomer and hydrogen consumption

Brought to you by | University of Arizona

Table 3: Reactor and heat exchanger design parameters

The average polymer properties of interest can be calculated in term of the

To calculate the cumulative copolymer composition, Φ , in the reactor,

Published by De Gruyter, 2011 9

Brought to you by | University of Arizona

Where MW is the average molecular weight of the repeating unit in the

This simulator is able to calculate the important polymer attributes such as

In addition, for all polymeric streams were used POLY-SRK equation of

Brought to you by | University of Arizona

To use the polymer SRK equation of state several pure component

3.2 Process description

Figure 4: Process flow diagram of LLDPE industrial plant based on Spherilene

Published by De Gruyter, 2011 11

Brought to you by | University of Arizona

Table 4: Industrial data of input quantities and operating conditions of LLDPE

In this way, as the catalyst reaction of ethylene polymerization, the

4. Results and Discussion

Brought to you by | University of Arizona

Table 5: Comparison between simulation results and industrial data

4.1 Monomer effects

Ethylene monomer is used in this process as an unsaturated hydrocarbon. Figures

Published by De Gruyter, 2011 13

Brought to you by | University of Arizona

Figure 5: The effects of variation of Methylene (feed to reactor 1) on Mpolymer (a), ρ

Brought to you by | University of Arizona

Figure 6: The effects of variation of Methylene (feed to reactor 2) on Mpolymer (a), ρ

Published by De Gruyter, 2011 15