AND Optimization OF Three Existing Ethylbenzene Dehydrogenation Reactors in Series
AND Optimization OF Three Existing Ethylbenzene Dehydrogenation Reactors in Series
AND Optimization OF Three Existing Ethylbenzene Dehydrogenation Reactors in Series
00
Printed in Great Britain Pergamon Press plc
SIMULATION AND OPTIMIZATION OF THREE EXISTING ETHYLBENZENE
DEHYDROGENATION REACTORS IN SERIES
I. Onal
Middle East Technical University, Ankara, TURKEY
N. Yalcm, E. Uygun, H. Oztllrk
PETKIM Petrochemicals Research Center Yarimca, TURKEY
263
Proc. 7th Int. Conf: on Mathematical and Computer Modelling
F0(l-x,-~2-~3-~4-~5-~6)
dX2
_=- k2
pT
dw F.
FT
* Fo(eE+~I-~2-~4+3x5+6x6)
(16)
pT
FT
FO(I-x,-~2-~3-~4-~5-~6)
dx3
-=- k3
(17)
pT
dw F.
FT
dx4 k4 F0(l-xl-~2-~3-~4-~5-x6)
_=-
pT
dw F.
FT
* FO(eE+x,-x2-x4+3x5+6x6)
(18)
Fig.1. Reactor System pT
FT
experiments that the main reaction (I) was
limited by pore diffusion resistance. The rate F0(1-xl-~2-~3-~4-~5-x6)
dx5
_= -k5
of styrene reaction was therefore found by 09)
pT
multiplying the intrinsic rate with an effective- dw F.
FT
ness factor. The rate equations for the main
and five side reactions were:
FO(I-xI-~2-~3-~4-~5-x6)
dx6
-=- k6
pT
(20)
dw F.
r, = k 1 [I+-] (8) FT
K
eq where;
Fitting the Kinetic Parameters to the Plant Data at fixed activity of catalyst and it can give
insight into the interplay of variables. Catalyst
The activation energies of reactions 1 to 4 were deactivation can be incorporated into the model,
found from literature (Shell 1969, Ralph 1948, but it requires more plant data.
Sergio 1965), and activation energies of the last
two reactions were estimated. During optimization, REFERENCES
the estimated activation energies were taken as
constant and the reaction rate constants were Carl Y.Yaves (1976). Correlation constants
found by adjusting frequency factors F. in the for chemical compounds. Chemical
Arrhenius equation. The change in activation Engineering, 9, 79-87.
energies did not change the values of reaction Charles N. Satterfield (1970). Diffusion. Mass
rate constants of reactions. Transfer in Heterogeneous Catalysis, 1-77,
David M. Himmelblau (1972). Unconstrained
The six material balances were solved Simultaneously nonlinear programming methods. Applied
with the semi-implicit Runge-Kutta Method of Nonlinear Programming, 63-184.
integration (William 1986, Himmelblau 1972). Emerson H.Lee (1973). Iron oxide catalysts
Conjugate Direction Method (William 1986) was for dehydrogenation of ethylbenzene in
used as a multivariable search technique to the presence of steam. Catalysis Reviews,
minimize the sum of the squares of differences 8(2), 285-305.
between the predicted and actual conversions. Ralph R. Wenner and Ernest C. Dybdal (1948).
As a result of this optimization routine, the Catalytic dehydrogenation of ethlybenzene.
kinetic rate parameters F. were obtained. Data Chemical Engineering Progress, 4, 275-286.
for 20 days were available. For each day, the Robert W. Wansbrough (1985). Modelling chemical
optimization was performed and frequency factors r aactors. Chemical Engineering, August
were obtained for each reactor. The arithmetic 5. 95-102.
average of these frequency factors were calculated Sergio Carra and Lucia Farni (1965). Kinetics
and the reaction rate constant equations were of catalytic dehydrogenation of ethyl-
obtained for each reactor. benzene in the presence of steam. Catalysis
Reviews, 8(2), 281-285.
Because Shell-105 (1/8”)/CCI-97 (l/8”) mixture Shell I.P.G. and Crowe C.M. (19691. Simulation
was loaded in the third reactor, a different and optimization of an existing ethylbenzene
method was applied for the examination of the dehydrogenation reactor. The Canadion
third reactor. Third reactor was assumed to Journal of Chemical Engineering, 47,
consist of two reactors. In reactor 3-A, the 183-187.
kinetic parameters of first reactor were used William H. Press, et.al. (1986). Minimization
so that the inlet data for reactor 3-B were or maximization of functions. Numerical
obtained. The kinetic parameters were optimized Recipes, 274-326.
in reactor 3-B for CCI-97(1/8”) such that the
error coming from the use of parameters of
Reactor 1 in Reactor 3-A, was lumped into the
kinetic parameters obtained for Reactor 3-B.
The kinetic parameter estimation results are
listed in Table 1.
CONCLUSIONS
Reaction No. u
1 2 3 4 5 6 (cal/hr-cc-K)
------
REACTOR I
CATALYST: SHELL-105
AVERAGE F 8.555 7.391 18.826 5.888 2.504 7.616 2.05
STANDARD DEVIATION FOR F 0.123 0.282 0.445 0.659 0.542 0.207
ACTIVATION ENERGY 21372 21857 49575 22500 19500 27500
(cal/gmol)
REACTOR 2
CATALYST: Ccl-97
AVERAGE F 6.549 7.309 19.113 5.805 2.680 7.842 11.1
STANDARD DEVIATION FOR F 0.206 0.447 0.372 0.830 0.697 0.461
ACTIVATION ENERGY 16462 21857 49575 22500 19500 27500
(cal/gmol)
REACTOR 3-A
CATALYST: SHELL-105
AVERAGE F 8.555 7.391 18.826 5.888 2.504 7.616 28.0
STANDARD DEVIATION FOR F 0.123 0.282 0.445 0.659 0.542 0.207
ACTIVATION ENERGY 21372 21857 49575 22500 19500 27500
(caI/gmol)
REACTOR 3-B
CATALYST: Ccl-97
AVERAGE F 10.877 9.802 20.686 7.199 5.428 9.578 28.0
STANDARD DEVIATION FOR F 0.252 0.576 0.457 1.216 0.314 0.605
ACTIVATION ENERGY 16462 21857 49575 22500 19500 27500
(caI/gmol)
DAY ACTUAL
~__~ MODEL ACTUAL MODEL
- ACTUAL
~--- MODEL ACTUAL MODEL ACTUAL MODEL
15.09.88 49.81 45.40 93.20 92.77 815.00 818.00 851.00 856.00 851.00 853.00
20.10.88 47.17 49.15 92.60 91.43 817.00 819.00 849.00 854.00 854.00 853.00
25.10.88 47.89 48.82 86.65 91.38 818.00 821.00 853.00 853.00 856.00 853.00
27.10.88 46.19 48.03 93.10 91.67 814.00 817.00 852.00 851.00 851.00 853.00
01.11.88 47.90 48.90 92.28 91.33 818.00 819.00 853.00 854.00 856.00 855.00
08.11.88 49.85 48.29 92.26 91.76 814.00 817.00 852.00 852.00 853.00 854.00
10.11.88 49.27 48.39 91.49 91.68 816.00 819.00 854.00 852.00 854.00 855.00
15.11.88 49.61 48.36 90.64 91.63 815.00 817.00 854.00 855.00 855.00 853.00
17.11.88 47.54 47.28 93.04 92.07 813.00 815.00 858.00 852.00 851.00 850.00
22.11.88 46.74 47.76 92.23 91.76 813.00 812.00 859.00 850.00 852.00 853.00
24.11.88 46.81 48.08 91.81 91.77 813.00 814.00 858.00 851.00 851.00 853.00
29.11.88 47.53 47.97 92.20 91.90 813.00 813.00 860.00 852.00 853.00 854.00
01.12.88 48.09 48.40 92.31 91.93 814.00 816.00 853.00 854.00 854.00 853.00
06.12.88 45.92 47.25 92.82 92.11 815.00 813.00 859.00 853.00 852.00 844.00
08.12.88 45.65 47.76 92.50 91.75 815.00 815.00 861.00 853.00 855.00 853.00
20.12.88 47.63 47.83 91.93 92.00 814.00 818.00 855.00 853.00 853.00 854.00
29.12.88 46.94 47.93 92.69 91.72 816.00 815.00 863.00 855.00 853.00 853.00
10.01.89 46.76 48.37 92.07 91.35 817.00 820.00 863.00 859.00 855.00 855.00
12.01.89 47.75 48.34 92.36 91.67 819.00 818.00 853.00 858.00 857.00 858.00
17.01.89 48.79 49.48 91.91 91.10 820.00 819.00 855.00 857.00 859.00 859.00
Proc. 7th Int. Conf. on Mathematical and Computer Modelling 261
““I 14-J