Slovak Society of Chemical Engineering Institute of Chemical and Environmental Engineering Slovak University of Technology in Bratislava PROCEEDINGS rd 33 International Conference of Slovak Society of Chemical Engineering Hotel Hutník Tatranské Matliare, Slovakia May 22 – 26, 2006 Editors: J Markoš and V Štefuca ISBN 80-227-2409-2 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia Le-Tu-3, 061p.pdf MODELING OF ETHYLBENZENE PRODUCTION PROCESS WITH INTERNAL RECYCLE Peter Burian **, Marcell Horváth *, Peter Mizsey * and Alois Mészáros ** (**) P Burian, A Mészáros Slovak University of Technology, Faculty of Chemical and Food Technology, Radlinského 9, 81237 Bratislava, Slovak Republic E-Mail: peter.burian@stuba.sk (*) M Horváth, P Mizsey, Budapest University of Technology and Economics, Faculty of Chemical Engineering, Műegyetem rkp 3-9, H-111 Budapest, Hungary E-mail: mhorvath@mail.bme.hu Keywords: recycle system, dynamic analysis, separation process Introduction: Ethylbenzene is a very important industrial substance: it is mainly used as an intermediate in the production of styrene Ethylbenzene is also a major component of mixed xylenes used as solvents in agricultural and home insecticide sprays, rubber and chemical manufacturing, and household degreasers, paints, adhesives, and rust preventives [1] The overall dynamics of chemical processing plants with material recycle can be very different from the dynamics of the individual processing units Material recycle may dramatically alter the overall gain and time constants of the plant, and may give rise to oscillatory or instable behaviour, even when the individual processing units are stable by themselves In most cases, recycle leads to positive feedback effects For example, increasing the concentration of a chemical species in a process stream will normally increase the amount of this species in the recycle stream, and, thus, lead to a reinforcement of the original increase It refers to a self-reinforcing mechanism associated with the recycle [3,5] The aim of this paper is to give a comprehensive picture on dynamic behaviour of processes with internal recycle The analysis is carried out applying theoretical and simulation experimental tools on the model of the ethylbenzene production process 061–1 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia Le-Tu-3, 061p.pdf Ethylbenzene Production Ethylbenzene is a clear, colourless liquid with a characteristic sweet, gasoline-like, aromatic odour It is a flammable and combustible liquid Its combustion may produce irritants and toxic gases It is exclusively used as an intermediate for the manufacture of styrene Currently, the primary source of ethylbenzene is the alkylation of benzene with ethylene [4] Process Description The ethylbenzene process consists of two subsystems (Fig 1): • • a reactor, where ethylene is reacted with benzene a distillation section, where the rest of benzene and polyethylbenzenes are separated from the reactor effluent to produce ethylbenzene of high purity Fig Ehtylbenzene production system Benzene is alkylated with ethylene to form ethylbenzene in the reactor in present of AlCl3 catalyst Diethylbenzenes, triethylbenzenes and other heavier polyethylbenzenes are also formed The benzene feed is heated to reaction temperature To maximize ethylbenzene formation, the diethylbenzene is usually transalkylated with benzene to form more ethylbenzene The reactor effluent, a mixture of benzene, ethylbenzene and polyethylbenzenes, flows to a distillation block Untreated benzene is recovered in the first distillation column, which is called benzene column The benzene from the top of the column is recycled to the reactor The bottoms containing ethylbenzene, diethylbenzene and triethylbenzene is fed to the second column called ethylbenzene column It separates ethylbenzene from the polyethylbenzenes as a final product The ethylbenzene in a high purity (99,9%) is led away from the top of the second column The mixture of diethylbenzene and triethylbenzene takes out from the bottom of the second column and it is fed into the third column Diethylbenzene from the third, the so-called polyethylbenzene column, is recycled back to the reactor [1,2] 061–2 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia Le-Tu-3, 061p.pdf Dynamic model The system of the reactor and the first distillation column is studied in the first step a Reactor Benzene, with flow rate of 327,612 kmol/h and ethylene with flow rate of 121,802 kmol/h is fed to the reactor (at the temperature of 180 ºC) The temperature in the reactor is 180 ºC and the pressure is 10 bars Three exothermic reactions occur in liquid phase, in presence of AlCl3 catalyst + + + CH2=CH2 CH2=CH2 CH2=CH2 k1 ∆rH1 k-1 k2 ∆rH2 k-2 k3 ∆rH3 k-3 where ki stands for reaction rate k i = k 0i e − E RT The conversion of the benzene in the reactor is approximately 28% The reactor effluent contains benzene, ethylbenzene, diethylbenzene and triethylbenzene This mixture is led out from the bottom of the reactor The model of the process can be derived from the material and energy balance laws Material balance: Benzene: F1 − ν B ξ V 1V + ν EB ξ V −1V = F3 x B + d [Vx B ] dt (1) Ethylene: F2 − ν E ξ V 1V + ν EB ξ V −1V − ν E ξ V 2V + ν DEB ξ V − 2V − ν E ξ V 3V + ν TEB ξ V −3V = F3 x E + Ethylbenzene: ν E ξ V 1V − ν EB ξV −1V − ν E ξV 2V + ν DEB ξ V − 2V = F3 x EB + d [Vx EB ] dt Diethylbenzene: ν E ξ V 2V − ν DEB ξV −2V − ν E ξV 3V + ν TEB ξ V −3V = F3 x DEB + Triethylbenzene: ν E ξ V 3V − ν TEB ξV −3V = F3 xTEB + d [VxTEB ] dt d [Vx DEB ] dt d [Vx E ] (2) dt (3) (4) (5) 061–3 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia Le-Tu-3, 061p.pdf Energy balance: Reactor effluent: F1c p1T1 + F2 c p 2T2 = F3 c p 3T + Ak [T − Tc ] + ξ V 1V∆ r H + ξ V 2V∆ r H + ξ V 3V∆ r H + dVc p 3T dt (6) Coolant: Fc c pc Tc + Ak [T − Tc ] = Fc c pc Tc + dVc c pc Tc (7) dt b First column The first column (see Fig 2) is used to separate benzene from the other product components The column consists of 12 trays, a reboiler and a condenser The top-product of the column is benzene with purity higher than 99% The other three components leave the first column at the bottom, and are fed into the second column The feed is saturated liquid Equimolal overflow and constant relative volatility are assumed The model of the column is derived from the material balance law Condenser: nv y1 = FD x + n L x + d [H x ] dt 1st tray: n L x + nV y = nV y1 + n L x1 + y1 = f ( x1 ) d [H x1 ] dt ith tray: n L xi =1 + nV y i +1 = nV y i + n L xi + y i = f ( xi ) d [H i xi ] dt 7th (feed) tray: n L x + nV y8 + FF x F = nV y + n L x + Fig Distillation column y = f (x7 ) kth tray: (FF + n L )x k =1 + nV y k +1 = nV y k + n L x k + y k = f (x k ) Reboiler: (FF + n L )x n = nV y n +1 + FW xW + y n +1 = f (x n +1 ) d [H k x k ] dt d [H x ] dt (15) (16) d [H n +1 x n +1 ] dt (17) (18) 061–4 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia Le-Tu-3, 061p.pdf List of symbols: nL – flow rate of liquid phase in column nV – flow rate of vapour phase in column FF – fresh feed to column (flow rate of reactor mixture) FW – flow rate of bottoms FD – flow rate of distillate xi – composition in liquid phase yi – composition in vapour phase – flow rate of benzene – flow rate of ethylene – flow rate of reactor mixture – stoichiometric coefficient – volume of the reactor – reaction enthalpy – heat capacity – temperature of the reactor effluent – input temperature of the coolant – output temperature of the coolant – heat transfer surface F1 F2 F3 νi V ∆rHj cpi T TC0 TC A A simulation schema, created in MATLAB SIMULINK environment can be seen in Fig Fig Simulation model in Matlab - Simulink environment Simulation results Simulation experiments, applying input variables step changes to the process, were carried out to demonstrate the dynamic behaviour of the system First, process transient behaviour with recycle, as responses to step-wise changes in inlet benzene flow rate (change of + 7%), ethylene flow rate (–10%), bottoms flow rate (+10%) and coolant temperature (+10%), are shown in Fig a, b, c and d, respectively 0.045 0.09 0.04 0.08 0.035 0.07 0.03 0.06 0.025 w x 0.05 w x 0.02 0.04 0.015 0.03 0.01 0.02 0.005 0 0.01 Fbenzene+7% time (h) 10 0 15 a) Fethylene-9% 10 time (h) b) 061–5 15 20 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia x 10 Le-Tu-3, 061p.pdf 0.12 -4 T Fbottoms+10% +10% coolant 2.8 0.1 2.6 0.08 2.4 w x 2.2 w x 0.06 0.04 1.8 0.02 1.6 1.4 10 15 time (h) 20 25 0 30 10 c) 15 time (h) 20 25 30 d) Fig Dinamic behaviour of the system with recycle In order to demonstrate the effect of recycle on process dynamics, simulation experiments in form of transient responses were carried out and compared, see Fig It is transparent that the recycle modifies the time constant as well as the static gain of the overall system 0.045 0.04 0.035 0.03 0.025 w x 0.02 0.015 0.01 0.005 0 with recycle without recycle 10 15 time (h) 20 25 30 Fig Comparison of the response of the system with and without recycle Process identification in linearized form Process identification, on basis of transient dynamic responses, was carried out in order to obtain the linearized models of the system with and without recycle, in form of linear transfer functions The transfer functions of the process with recycle, GS, and without recycle, GM (G1, G2), take then the following forms: GS = G1 = 0,00723 1,1806s + 3,3511s + 3,0122s + e −5 s 0,0006759 22,3876s + 23,8301s + 8,4552s + e −5 s 061–6 (19) (20) 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia G2 = Le-Tu-3, 061p.pdf 0,912 −0.003s e (21) Fig Transfer function block diagraml of the ethylbenzeneproduction plant The transfer function of the recycle, (23), was derived from the model configuration in Fig GR = GR = G S − G1 G S G (22) 1,0911s + 1,1522s + 0,4002s + 0,0444 (23) s + 1,0648s + 0,3776s + 0,0447 Fig shows the comparison of the transient responses of the nonlinear system and the linearized model of system with recycle The modeling mismatch problem leads to an oscillatory behaviour of the linearized model 0.07 nonlinear system linear system 0.06 0.05 0.04 w x 0.03 0.02 0.01 0 10 20 time (h) 30 40 50 Fig Comparison of the nonlinear and the linearized systems with recycle 061–7 33rd International Conference of SSCHE May 22–26, 2006, Tatransk´e Matliare, Slovakia Le-Tu-3, 061p.pdf Conclusion Modeling problem of a realistic industrial scaled coupled reactor/column process to produce ethylbenzene is studied in this paper The model was created using Matlab – Simulink environment Simulation experiments were carried out for various step-wise changes to demonstrate the dynamic behaviour of the system It is transparent that the recycle modifies the time constant as well as the static gain of the overall system Process identification was carried out in order to obtain the linearized model of the ethylbenzene producing system In further study, the linear transfer function models, obtained above, will be utilized to eliminate the recycle effect by introduction of a recycle compensator To control the overall process with robust recycle elimination, an adaptive controller with gain-scheduling performance can extend the compensator Acknowledgements: The authors highly acknowledge the financial support of the Scientific Grant Agency of the Slovak Republic under grants No 1/1046/04 and 1/3081/06 References [1] Ullmann’s Encyclopedia of Industrial Chemistry, Vol A 10, 1987 VCH Verlagsgellschaft mbH, D-6940 Weinheim, Germany [2] Encyclopedia of Chemical Processing and Design 371.698/20, 1984 Marcel Dekker Inc., 270 Madison Avenue, New York, 10016 [3] Alois Mészáros, Peter Mizsey, Marcell Horváth, Zsolt Fonyo, Dynamic analysis and control of recycle processes 2004, SSCHE [4] Toxicology and Carcinogenesis Studies of Ethylbenzene (CAS No 100-41-4) in F344/N Rats and B6C3F1 Mice (Inhalation Studies), http://ntp-server.niehs.nih.gov/ntpweb/printFriendly.cfm?objectid=070A7F1A-CD44-B3EE05596CD253F268B6 [5] MORUD, J.; SKOGESTAD S.: Effects of recycle on dynamics and control of chemical processing plants.1994, Computers Chem Engng, 18, Suppl., S529-S534 061–8