¡Descarga Optimización y simulación de reacciones enlazadas: Metanol y deshidrogenación IPA. y más Guías, Proyectos, Investigaciones en PDF de Procesos Químicos solo en Docsity!
Jurnal Teknik Kimia Indonesia Vol. 10, No. 3, 2011 , 127 - 133
SIMULATION AND OPTIMIZATION OF COUPLING
REACTION OF METHANOL SYNTHESIS AND ISOPROPYL
ALCOHOL DEHYDROGENATION
Jenny Rizkiana
1
, Yogi W. Budhi
2
, Azis Trianto
**1 ***
1
Energy and Chemical Engineering Processing System Research Group
2
Chemical Engineering Process Development Research Group
Chemical Engineering Program, Faculty of Industrial Technology
Institut Teknologi Bandung
Jalan Ganesha No 10, Bandung 40132, Indonesia
Email: trianto@che.itb.ac.id
Abstract
A study on simulation and optimization of coupling reaction between methanol synthesis
and isopropyl alcohol (IPA) dehydrogenation was performed. The analysis is carried out
theoretically to obtain the optimum operation conditions which give the best performance.
The reactions are just interacting thermally. In this study, both reactions are held
catalytically in a heat-exchanger type reactor. As a high pressure reaction, methanol
synthesis is placed in the inner side of reactor tube while dehydrogenation of IPA is in the
opposite. Tube wall acts as a heat transfer media. The reactor is modeled by a steady state
heterogeneous equation for a fixed bed reactor. Optimization is done in order to find the
optimum value of operation conditions, those are the inlet temperature of both side of
reactor and the molar feed flow ratio between the exothermic side and the endothermic side.
Sum of weighted reaction conversion is considered to be the objective function that is
maximized. The simulation result shows that coupled reactor makes the reaction conversion
higher than a conventional adiabatic reactor and the optimum operation conditions give the
maximum value of the conversion. This study presents a theoretical proof that coupling
reaction is feasible.
Keywords: coupling reaction, IPA dehydrogenation, methanol synthesis, optimization,
simulated annealing
Abstrak
Telaah mengenai simulasi dan optimisasi reaksi perangkaian ( coupling reaction ) antara
sintesis metanol dengan dehidrogenasi isopropil alkohol (IPA) telah dilakukan. Analisis
dilaksanakan secara teoretik guna mendapatkan kondisi optimum yang akan memberikan
hasil terbaik. Pada penelitian ini, kedua reaksi dilaksanakan secara katalitik dalam reaktor
bertipe buluh-cangkang. Karena bertekanan tinggi, sintesis metanol ditempatkan pada sisi
buluh, sedangkan dehidrogenasi IPA ditempatkan pada sisi cangkang. Dinding buluh
berperan sebagai media perpindahan panas. Reaktor dimodelkan dengan reaktor heterogen
tunak unggun tetap. Optimisasi dilakukan dalam rangka mendapatkan nilai optimum dari
kondisi operasi yang mencakup temperatur inlet sisi eksotermik dan endotermik serta rasio
umpan molarnya. Jumlah total konversi reaksi terbobotkan dipilih sebagai nilai objectif yang
akan dioptimumkan. Hasil simulasi menunjukkan bahwa reaktor perangkaian termal mampu
meningkatkan konversi reaksi jika dibandingkan dengan reaktor adiabatik dan pada kondisi
operasi yang optimum diperoleh konversi maksimal. Penelitian ini menunjukkan bahwa
reaksi perangkaian layak untuk dilaksanakan.
Kata kunci : reaksi perangkaian, dehidrogenasi IPA, sintesis methanol, optimisasi, simulated
annealing
*corresponding author
Jurnal Teknik Kimia Indonesia Vol. 10 , No. 3, 2011
1. Introduction
Thermally coupled reactor can be
considered as an effort to do energy
conservation. Exothermic reaction acts as a
heat source and endothermic reaction will be
a heat sink. Hunter and McGuire (1 978 ) were
the first to do the study about coupling of
exothermic and endothermic reaction. On
their patent, they suggested to use catalytic
combustion or other highly exothermic
reaction as a heat source for any endothermic
reaction.
Another study about reaction coupling
was done by Itoh and Wu (1997). They used
membrane reactor made from palladium to
carry out dehydrogenation of cyclohexane.
Produced hydrogen permeated trough reactor
wall and reacted with oxygen giving a huge
number energy to support the reactor itself,
so the operation can be done adiabatically.
Based on that study, Elnashaie et al. (2000)
utilized heterogeneous kinetic model to
simulate reaction coupling between
dehydrogenation of ethyl benzene to become
styrene and hydrogenation of benzene to
become cyclohexane. They focused the study
on the effect of co-current and counter-
current flow.
Khademi et al. (2009) chose to couple
dehydrogenation of cyclohexane and
methanol synthesis. They optimized the inlet
temperature for both side of reaction and
molar flow by using differential evolution
method optimization.
This study will couple methanol
synthesis and dehydrogenation of isopropyl
alcohol (IPA). Methanol is a multipurpose
base chemical. It is the simplest alcohol in the
world and can be produced by several ways.
Recently, most of methanol is made from
syngas (Elkamel et al., 2009). Conventional
reactor of methanol synthesis has very low
conversion because of the nature of its
equilibrium state. To accommodate the kinetic
and equilibrium constraint, methanol
synthesis should be conducted at relatively
low temperature and high pressure (Kirk-
Othmer, 1967).
IPA dehydrogenation process can be
viewed from two sides. First, IPA
dehydrogenation is an alternative route to
make acetone. Acetone made from IPA is free
from aromatic compound so that it is
preferred to use by pharmaceutical industries
which has very tight regulation about the
solvent they used (Turton et al., 2009).
Second, IPA dehydrogenation is a way to
produce hydrogen. Hydrogen is predicted to
be the future fuel because of its cleanness and
easy to use. Currently, most of hydrogen is
produced from natural gas via steam
reforming process. Due to the decrease of gas
reserve in the world, alternative route to
produce hydrogen must be made. Hydrogen
production from IPA dehydrogenation is very
potential since IPA itself can be made from
renewable resources by fermentation.
Conventional optimization methods are
mostly based on gradient method. From this
point, at least there are two weakness of this
method. First, they can easily trapped on local
optimum depends on the degree of
nonlinearity and initial guest (Khademi et al.,
2009). Secondly, they can handle non-
differentiable objective function, such as a
step function.
A number of optimization methods are
developed in order to solve the weakness of
conventional optimization method. One of
them is called simulated annealing (SA)
method. It proposed independently by
Kirkpatrick et al. (1983) and Cerny (1985).
This method started from implementing
Metropolis algorithm combined with
combinatorial optimization method and
statistical mechanic to analogize and optimize
annealing process. SA can be easily applied to
many optimization problems (Bertsimas and
Tsitsiklis, 1993).
1.1. Kinetics Model
Methanol synthesis actually consists of
three main reactions. They are carbon oxide
reduction and water-gas shift reaction as
follows:
CO + 2 H
2
⇄ CH
3
OH (1)
CO
2
+ 3 H
2
⇄ CH
3
OH + H
2
O (2)
CO
2
+ H
2
⇄ CO + H
2
O (3)
From those three reactions, only two
are independent; the one is a combination of
other ones. Van den Bussche and Froment
(1996) have proposed a kinetics model based
only from reaction (2) and (3) utilizing
Cu/ZnO/Al 2
O
3
as the catalyst as follows:
ଵ
ସ
ை
మ
ு
మ
,ଵ
ு మ
ை
ு య
ைு
ு మ
ଷ
ை మ
ଷ
ு
మ
ை
ு
మ
ு
మ
ଶ
ு
మ
ை
ଷ
Jurnal Teknik Kimia Indonesia Vol. 10 , No. 3, 2011
Table 3. Auxiliary Correlation for Estimating Fluid Properties and Transport Coefficient
Parameter Correlation Reference
Heat capacity for pure
component
ܥ
= ܣ + ܤ ൬
ܥ
ܶ
sinh ൬
ܥ
ܶ
൘ ൰൰
ଶ
ܧ
ܶ
cosh ൬
ܧ
ܶ
൘ ൰൰
ଶ
Green and Perry, 2007
Mix heat cap. Based on local composition
Viscosity Chung correlation Reid et al., 1987
Mix viscosity Wilke correlation Reid et al., 1987
Thermal cond. Chung correlation Reid et al., 1987
Mix thermal cond. Korelasi Brokaw Reid and Sherwood, 1958
Mass transfer
coefficient
݇
= 1. 17 ܴ݁
ି .ସଶ
ܿܵ
ି .
ݑ
× 10
ଷ
ܴ݁ =
2 ܴ
ݑ
ߤ
ܿܵ
=
ߤ
ߩܦ
× 10
ି ସ
ܦ
=
1 − ݕ
∑
ݕ
ܦ
ୀ
ܦ
,
=
10
ି
ܶ
ଷ/ଶ
ඨ
1
ܯ
1
ܯ
ܲ ቀݒ
ଷ/ଶ
ଶ/ଷ
ቁ
ଶ
Cussler, 2009
Heat transfer
coefficient
1
ܷ
=
1
ℎ
ܣ
ln
ܦ
ܦ
2 ߨܭ
௪
ܮ
ܣ
ܣ
1
ℎ
ቆ
ℎ݀
ߣ
ቇ = 1. 17 ቆ
ݑߩ݀
ߤ
ቇ
.ହ଼ ହ
൬
ܥ
ߤ
ߣ
൰
ଵ/ଷ
McCabe et al., 1993
Sum of weighted reaction conversion is
considered to be the objective function.
Weighted factor for this study is the price of
each component. But, since price study hasn’t
finished yet, we assumed that weighted
factors are equals for all components.
Therefore, the objective function is just sum of
reaction conversion multiplied by (-1)
because most of optimization method require
an objective function to be minimized.
We classify this optimization problem
as a constrained problem. To make it as an
unconstrained one, a penalty function can be
added to the objective function. Penalty
function is function for penalizing the
objective function when the variable violates
its constraint. For example, if variable satisfy
the constraint, penalty function is zero;
otherwise, penalty function will be a finite
value. For this study, the penalty values are:
ଶ
ୀଵ
Where
ଵ
= max{ 0 , (ܶ
ଶ
ଵ
ଶ
= max
ଶ
ଵ
= max{ 0 , (ܶ
ଶ
ଵ
= max
ଵ
ଵ
= max{ 0 , (ܶ
ଵ
3. Result and Discussion
For a base case, all conditions of
methanol synthesis Van den Bussche and
Froment (1996) are used, excluding feed
composition since it is used from Sinadinovic-
Fiser et al. (2001). For IPA dehydrogenation,
reaction condition is taken from Rioux and
Vannice (2005).
Before the thermally coupled reactor is
simulated, first we have to simulate the
conventional adiabatic reactor. Figure 2
shows the effect of inlet temperature to the
yield of methanol. We can see that when
temperature increases, yield of methanol is
also increase. But at some point, methanol
yield will remain constant and then decrease.
This is because at low temperature, methanol
synthesis is controlled by kinetics. However,
as temperature increases, the reaction
changes to be thermodynamically controlled.
The effect of temperature to acetone
yield can be seen on figure 3. Since IPA
dehydrogenation is an endothermic reaction
and it is kinetically controlled, temperature
increase gives positive effect to the yield. The
higher temperature will make acetone yield
higher.
Simulation and Optimization of Coupling Reaction (J. Rizkiana, et al.)
Figure 2. Inlet temperature vs methanol
yield
As described before, optimization is
done in order to get the best conditions which
give the maximum conversion for both side of
reactor. Optimization process was carried out
using Simulated Annealing method. The
results are summarized in Table 4. All value
from optimization results is then used to
simulate the optimized coupled reactor.
Table 4. Optimized Condition for
Thermally Coupled Reactor
Parameter Value
Inlet temperature of
exothermic side (K) 525.
Inlet temperature of
endothermic side (K) 512.
Molar feed flow ratio 0.
Figure 3. Inlet temperatures vs. acetone
yield
Reactor simulation results carried out
by using optimized conditions are shown in
several figures. Figure 4 shows temperature
and yield profile for both side of reactor. From
figure 4 (a), we can see that temperature of
exothermic side is relatively remains constant.
It means that all heat generated from reaction
of methanol synthesis is absorbed and being
used by IPA dehydrogenation to maintain its
process.
As a highly endothermic reaction, IPA
dehydrogenation needs much energy. Energy
can come from additional heat sources or
from the system itself. If heat from additional
sources isn’t enough, reaction will use its
internal energy. It is marked by the decrease
of reactor temperature as shown on figure
4 (a) at earlier section of reactor (between 0
until 0.1). When reaction occurs, it needs a
huge number of energy, but when conversion
almost reaches 100% (reaction rate almost
zero), it doesn’t needs energy anymore (see
figure 4 (b)). So, the energy from methanol
synthesis is used to regain temperature.
(a)
(b)
Figure 4. Yield and temperature profile at
optimized condition; (a) temperature
profile, (b) yield profile
Figure 5 shows comparison between
endothermic side temperatures of optimized
coupled reactor (OCR) with temperature of
conventional adiabatic reactor (CAR) at the
same inlet condition. From that figure, it is
shown that by using OCR, maximum acetone
yield can be achieved quicker than by using
CAR. On the OCR, maximum acetone yield
achieved at about 0.1 lengths of reactor and
on the CAR at about 0.8 (see figure 6 ). It’s
because on OCR, IPA dehydrogenation gets
additional energy from methanol synthesis. It
is cleared that OCR gives better performance
than CAR.
0,
0,
0,
0,
0,
0,
0,
0,
470 480 490 500 510 520 530
methanol yield
temperature (K)
0
0,
0,
0,
0,
1
420 440 460 480 500 520
acetone yield
Temperature (K)
400
420
440
460
480
500
520
540
0 0,2 0,4 0,6 0,8 1
temperature (K)
reactor length
exothermic
endothermic
0
0,
0,
0,
0,
1
1,
0 0,2 0,4 0,6 0,8 1
yield
reactor length
acetone
methanol
Simulation and Optimization of Coupling Reaction (J. Rizkiana, et al.)
Elkamel, A.; Zahedi, G. R.; Marton, C.; Lohi, A.,
Optimal fixed bed reactor network
configuration for the efficient recycling of co
into methanol, Energies , 2009 , 2 (2), 180 - 189.
Elnashaie, S. S. E. H.; Moustafa, T.; Alsoudani,
T.; Elshishini, S. S., Modeling and basic
characteristics of novel integrated
dehydrogenation-hydrogenation membrane
catalytic reactors, Computers & Chemical
Engineering , 2000 , 24 (2-7), 1293 - 1300.
Han, Y.; Shen, J.; Chen, Y., Microkinetic analysis
of isopropanol dehydrogenation over Cu/SiO 2
catalyst, Applied Catalysis A: General , 2001 ,
Hunter, J. B.; McGuire, G., Method and
Apparatus for Catalytic Heat Exchange , US
Patent 925,862, 18 Juli 1978.
Itoh, N.; Wu, T., An adiabatic type of palladium
membrane reactor for coupling endothermic
and exothermic reactions, Journal of
Membrane Science , 1997 , 124(2), 213 - 222.
Khademi, M. H.; Setoodeh, P.; Rahimpour, M.
R.; Jahanmiri, A., Optimization of methanol
synthesis and cyclohexane dehydrogenation
in a thermally coupled reactor using
differential evolution (DE) method,
International Journal of Hydrogen Energy ,
Kirk-Othmer, Encyclopedia of Chemical
Technology ; John Wiley & Sons: New York,
Kirkpatrick, S.; Gelatt, C. D.; Vecchi, M. P.,
Optimization by simulated annealing, Science ,
McCabe, W. L.; Smith, J. C.; Hariott , P., Unit
Operation of Chemical Engineering ; McGraw-
Hill: New York, 1993.
Sinadinovic-Fiser, S. V.; Jankovic, M. R.;
Radicevic, R. Z., Simulation of the fixed bed
reactor for methanol synthesis, Petroleum and
Coal , 2001 , 43 (1), 31 - 34.
Green, D. W.; Perry, R. H., Perry's Chemical
Engineers' Handbook ; McGraw-Hill: New
York, 2007.
Reid, R. C.; Prausnitz, J. M.; Poling, B. E., The
Properties of Gases and Liquids ; McGraw-Hill:
New York, 1987.
Reid, R. C.; Sherwood, T. K., The Properties of
Gases and Liquids: Their Estimation and
Correlation ; McGraw-Hill: New York, 1958.
Rioux, R. M.; Vannice, M. A., Dehydrogenation
of isopropyl alcohol on carbon-supported Pt
and Cu–Pt catalysts, Journal of Catalysis , 2005 ,
Turton, R.; Bailie, R. C.; Whiting, W. B.;
Shaeiwitz, J. A., Analysis, Synthesis, and Design
of Chemical Processes ; Prentice-Hall: New
Jersey, 2009.
