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

net/publication/260087717

Emulsion polymerization process control

Conference Paper · November 2011

CITATIONS READS
2 589

4 authors, including:

Francisco Lopez-Serrano L. Martínez-Suástegui

Universidad Nacional Autónoma de México Instituto Politécnico Nacional
46 PUBLICATIONS   335 CITATIONS    35 PUBLICATIONS   145 CITATIONS   

SEE PROFILE SEE PROFILE

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

Understanding/modeling the Pickering emulsion polymerization SiO2-supported View project

Mixed convection heat transfer View project

All content following this page was uploaded by Francisco Lopez-Serrano on 09 February 2014.

The user has requested enhancement of the downloaded file.

PAPER NO. VI INTERNATIONAL CONFERENCE ON ELECTROMECHANICS
AND SYSTEMS ENGINEERING (VI CIIES)

Emulsion polymerization process control

Ruben Mil, Francisco López-Serrano, René O. Vargas, Lorenzo A. Martínez-Suástegui


Abstract – A dynamical model for the emulsion polymerization
of styrene within a batch reactor has been studied. This model
includes the main steps for styrene polymerization, along with the II. INTRODUCTION
relevant balance equations. The resulting set of differential
equations was solved for temperature profile and monomer Emulsion polymerization has grown to become one of the
consume as a function of time. Emphasis is put on the reaction
temperature, which is controlled by the jacket temperature. The
major means for the production of synthetic polymers. There
controller loop, which determines the flow rate of the cooling are substantial incentives for improved design and control of
liquid, is also included. Finally, the advantage of having a emulsion polymerization reactors [1, 2]. To achieve these aims
rigorous control system in order to have the desires properties in the problem of the scale up laboratory data is crucial [3].
the final product is analyzed. Polymerization reactions are very complex reaction systems
characterized by a strong change in the viscosity and high
Index Terms—Batch reactor; Control; Emulsion; exothermicity, which is often accompanied by autoacelerating
Polymerization, Scale-up. kinetics that can lead the system into uncontrollable situations
and nonlinearity characterize polymerization processes [4].
Many industrial polymerization reactions are carried out in
batch reactors. In emulsion polymerization, monitoring and
I. NOMENCLATURE control of monomer conversion is crucial both for proper
process operation and for obtaining products with desired
properties (due to the influence of conversion of polymer
A Heat transfer are (m2) molecular weight and particle size distributions)[5,6].
CPr Heat capacity of reactor contents (J/g°C) Polymer reaction engineering is a discipline that deals with
CPj Heat capacity of water in the jacket (J/g°C) various problems concerning the fundamental nature of
FC Cold flow water (L/s) chemical and physical phenomena in polymerization
FH Hot flow water (L/s) processes. Mathematical modeling is a powerful tool for the
kp Propagation kinetic constant (L/mol s) development of process understanding and advanced reactor
M Monomer (mol) technology in the polymer industry [7-9]. The application of
n Average number of radicals per particle (-) process models to the design of model-based reactor
NT Particle concentration (Paricle/L) optimizations and controls is considered in this paper .
P Power input by stirred (W)
QJ,Loss Heat losses from jacket (W)
QR,loss Heat losses from reactor (W)
RP Reaction rate (mol/L s)
TJ Jacket temperature (°C)
III. GLOBAL BALANCE ENERGY OF THE BATCH
Tr Reactor temperature (°C)
RECTOR.
U Heat transfer coefficient (W/m2 K)
V Reactor volume (m3) The global balance energy, that was made in the batch
x Conversion (-) reactor, see fig 1., includes the dynamic behavior of energy
∆H Heat of polymerization (J/mol) system, which it considers the heat generated from the
V Viscous dissipation exotermic reaction; as well as the loss energy to the
ρ Density (kg/L) surroundings, the interchange of heat between the jacket and
the reaction system, including the heat induced by the stirrer,
and the global difference temperature that exists in the reactor,
according to the material properties of this.
Rubén Mil Martinez ESIME-UA IPN (thebeatles28@hotmail.es)
Francisco López Serrano FQ-UNAM (lopezserrano@unam.mx)
René O. Vargas, ESIME-UA IPN(e-mail: rvargasa@ipn.mx)
Lorenzo A. Martínez, ESIME-UA IPN( martinezlorenzo@gmail.com)

MEXICO CITY, MEXICO, NOVEMBER 2011 1

PAPER NO. VI INTERNATIONAL CONFERENCE ON ELECTROMECHANICS
AND SYSTEMS ENGINEERING (VI CIIES)

IV. REACTOR DYNAMICAL MODEL

The reaction takes place in a batch reactor and the conversion

equation is as follows:

dx k p M p nN T V
 x (0) =0
dt m 0 N av , (3)

The reactor temperature control is achieved by means of a

cooling fluid through a jacket. Energy balances lead to the
following set of ordinary differential equations, describing the
batch reactor and jacket temperature dynamics.

dT r UA Q loss  HR p V
  ( Tr  T j )   
dt  r C pr V  r C pr V  r C pr V
Fig 2. Reactor heating/cooling system description V
 ,Tr ( 0 )  Tr 0
 r C pr V (4)
Therefore the energy balance is presented as follows:

dT j
 j C pjV j  UA(Tr  T j )  Qlsj  Fh SC pj Th  Fc (1  S ) C pc Tc 
dt
∙ 1
FC pj T j ; T j (0)  Tj0 (5)

Where, is the mass reaction, heat capacity as a Two types of controllers are used in this dynamical system,
function of the reactor temperature, global rate of reactor the simplest one on-off, which the manipulated variable is
temperature, heat reaction, heat of mixing, heat either at maximum flow or at zero flow. The other one is a
induced by the stirred, heat interchange with the proportional-integral-derivative (PID) controller. In this case
surrounding and loss heat. is used where tight dynamic response is important [5-8].

For having a closed-loop system the PID controller is:

1 de ( t )
III. BALANCE ENERGY OF THE JACKET. T j , in
 Z p
(e (t ) 
Ti  e ( t ) dt  T d
dt
); ( 6 )

The jacket temperature has its own behavior, which depends

on the coolant inlet and outlet, as well as the type of material With
that was made.
e ( t )  T r  T set (7)

∙ ∙ ∙ ∙ , , ∙ ∙∆
2 V. ANALYSIS AND RESULTS

Where, coolant mass, coolant heat capacity, overall The operation conditions for the emulsion polymerization
batch reactor are given in table 1. Most of the data were taken
rate of coolant temperature, , inlet coolant temperature,
from [7]. The first step of this research is to evaluate two
, outlet coolant temperature, overall heat transfer
different kinds of controllers On-Off and PID, for
coefficient, area of heat transfer, ∆ temperature rise if the understanding how they work, the advantage and
mass and heat loss. disadvantages. On the other hand, some of the reaction
variables are considered constant for simply the model, which
will be included in the next steps of the project for scaling-up
or -down the process for having a complete pilot and industrial
analysis [8].

MEXICO CITY, MEXICO, NOVEMBER 2011 2

PAPER NO. VI INTERNATIONAL CONFERENCE ON ELECTROMECHANICS
AND SYSTEMS ENGINEERING (VI CIIES)

TABLE I
OPERATION CONDITIONS

Condition Value
Initial temperatures Tr and Tj 25°C
Tset 60 and 70°C
Initial styrene mol 1
Reactor volume 3000 (L)
Jacket volume 600(L)
Flow rate 8 L/s
Hot water 100°C
Cold water 20°C
Heat transfer coefficient U 293.076 (W/m2 K)
∆H -69,900 (J/mol)
Kp 0.45E7exp(-26000/(RT))
n 0.5
MP 5.5 (mol/L)
Nav 6.023E23
Styrene weight 104.15 g/mol

Fig 3. Reactor temperature evolution using a PID control

In fig. 3 the reactor temperature is shown using a PID

controller, less than 1000s takes to get at Tset, but a noise
behavior is experimented at the beginning of the process. It is
very important to take care this transitory part of the process
because the material can get to degradation temperature and
damage the final product. This means that parameters of the
controller have to adjust properly.

Fig 2. Reactor and jacket temperatures evolution using a simple On-Off

control.

The reactor and jacket temperatures evolution are shown in

Fig. 2, the desires temperature for the process (Tset) was
chosen 60°C. The on-off controller was used this is the
simples one considering its kind of working. The reactor
temperature takes almost 10000 seconds to get at Tset, with a
monotonically behavior. In the case of the jacket temperature
it has an oscillating behavior around Tset, the cold and hot
valves are open and close to maximum capacity and zero, that
is the reason of this behavior.
Fig 4. Reactor temperature evolution using a PID control
The fig. 4 shows the reactor temperature. A PID controller
was used, for this run the controller parameters were adjusted.
The reactor temperature takes less than 1000s to get Tset,
(70°C) ten times less than on-off controller. The higher
temperature reached before steady-state is 75°C which it is far
away to degradation temperature of the styrene. On the other
hand, it requires less water to control the process and it spends
less money in services.

MEXICO CITY, MEXICO, NOVEMBER 2011 3

PAPER NO. VI INTERNATIONAL CONFERENCE ON ELECTROMECHANICS
AND SYSTEMS ENGINEERING (VI CIIES)

There are problems in scale-up from laboratory data to pilot or

industrial scales, to get a material with the same characteristics
at different scales often is a very difficult task. The viscosity
changes during the polymerization process could be the end of
the process due to limitations of the equipment, in this case
power of the stirred. The reactor material is important for an
efficient heat transfer between reactor and jacket, it is
common to use glass reactors at laboratory scale and metal
reactor at pilot or industrial scale. The overall heat transfer
coefficient is different for every reactor material, so the
control of the process can be altered as well. In fig. 5 exhibits
the effect on the process with different heat transfer
coefficients (U), as the coefficient is less efficient the process
takes more time to get at Tset .

Fig 6. Monomer consume during the reaction.

VI. CONCLUSIONS

The emulsion polymerization control has been studied. For

Fig 5. Influence of the different heat transfer coefficients on the process
control.
Finally, the monomer consume along the reaction is analyzed,
using the same parameter of the table 1. A typical batch
emulsion polymerization reaction contains three distinct
intervals. In the first one, is that where particle formation takes
place and monomer droplets together with the rest of the
ingredients are presented. The second interval occurs after the
conclusion of the particle formation period whereby only
mature latex particles now exist; the particle number density
remains constant and the particles grow by propagation in the
presence of monomer droplets. Finally, the third interval
commences, where the remaining monomer contained within
the particles is polymerized. In this model the particle
formation is done, and from the second interval is captured in
the model and presented in the fig. 6, where the monomer
remains constant and the third interval start when the
monomer start to decrease as is also shown in fig.6.
this first model two types or controllers were considered on-
off and PID. In the first one the reactor temperature increases
monotonically up to the desires temperature Tset, the problem
is that requires more water (coal and hot) and it takes more
time to control the process. In the case of the PID controller, it
takes less time to control the process once adjusted the
controller parameters. The reactors are made of different
materials, it implies that the heat transfer process can be alter
and also the control. The most commons are glass reactors for
laboratory scale and metal for pilot or industrial scale. The
main intervals of the emulsion polymerization are identified
by the consume of the monomer during the reaction.
Variables like heat loss, mixing, styrene and water heat
capacities were takes constant in this first step, they will be
analyzed individually in a future research.
VI. ACKNOWLEDGEMENT
This project was supported by the research grants, no.
20110354 SIP-IPN. FLS gratefully acknowledges support by
PAPIIT-UNAM (IN103109).

MEXICO CITY, MEXICO, NOVEMBER 2011 4

PAPER NO. VI INTERNATIONAL CONFERENCE ON ELECTROMECHANICS
AND SYSTEMS ENGINEERING (VI CIIES)

VII. REFERECENS
IX. BIOGRAPHIES
[1] H.S. Fogler, Elements of Chemical Reaction Engineering, Prentice-
Hall, Englewood Cliffs, NJ, 1992.

[2] F.J. Schork, P.B. Deshpande, K.W. Leffew, Control of Polymerisation

Rubén Mil Martínez was born on February 28,
Reactors, Dekker, New York, NY, 1993.
1991 in Comoapan, Veracruz. In 2006 admission to
the Center for Industrial Technology and service
[3] Barton, J.A., & Nolan, P.F. (1987). J. hazard. mater. 14, 233.
number 110 and in the 2009 he finished his high
school. In the 2009 he has been studied mechanical
[4] Soh, S.K., & Sundberg, S.C. (1982). J. Pol. Sci. Chem. Ed., 20, 1299
engineering in the National Polytechnic Institute.
Ruben who is currently studying the 5th semester at
[5] Won Jung Yoon, Yang Soo Kim, In Sun Kim and Kyu Yong Choi. Recent the School of Mechanical and Electrical Engineering
Advances in Polymer Reaction Engineering: Modeling and Control of (ESIME) National Polytechnic Institute, and recently
Polymer Properties .29 September 2003. participated in a research project in the
postgraduate section.
[6] Sturt C. Thickett and Robert G. Gilbert. Emulsion polimerization: State of
the art in kinetics and mechanisms. Polymer 48 (2007) 6965-6991

[7] Carlos Fortuny Sendrós. Modelización y control de temperatura de un Francisco López Serrano Ramos received his B.Sc.
reactor “Batch” para un proceso exotérmico. November 2004. in Chemical Engineering from the Universidad
Nacional Autónoma de México (UNAM). Afterwards,
he received his M. Sc. from the University of
[8] E. Bruce Nauman. Chemical reactor design, optimization, and scale up.
Minnesota and Ph.D. degree Chemical Engineering
2001 at UNAM. He is currently working at UNAM, where
he is a full time professor in the areas of polymer,
[9] Mariano Asteasuain, Claudia Sarmoria, Adriana Brandolin. Integration of heat and mass transfer.
control aspects and uncertainty in the process design of polymerization
reactors. Chemical Engineering Science 61 (2006) 3362-3378.

René O. Vargas Aguilar received his B.Sc, M.Sc.,

and Ph.D. degrees in Chemical Engineering from the
Universidad Nacional Autónoma de México
(UNAM). He started to work at Instituto Politécnico
Nacional (IPN) in 2009, where he currently works as
an Associate Professor performing research in the
areas of polymer processing, transport phenomena
and rheology.

Lorenzo Alberto Martínez Suástegui was born in Mexico on November 10,

1975. He received his B.Sc, M.Sc., and Ph.D.
degrees in Mechanical Engineering from the
Universidad Nacional Autónoma de México
(UNAM). Afterwards, he worked as a Postdoctoral
Researcher at the University of California Riverside
and later joined the Instituto Politécnico Nacional
(IPN) in 2009, where he currently works as an
Associate Professor performing research in the
areas of fluid mechanics, heat transfer and
biomedical applications.

MEXICO CITY, MEXICO, NOVEMBER 2011 5

View publication stats