European Symposum on Computer Arded Aded Process Engneerng 15 L. Pugjaner and A. Espuña (Edtors) 2005 Elsever Scence B.V. All rghts reserved. Modellng the Two-Stage Pyrolyss Gasolne Hydrogenaton Navd Mostouf*, Mohammad Ahmadpour and Rahmat Sotudeh-Gharebagh Process Desgn and Smulaton Research Centre, Department of Chemcal Engneerng, Faculty of Engneerng, Unversty of Tehran, P.O. Box 11365/4563, Tehran, Iran Abstract A model s developed based on a two-stage hydrogenaton of pyrolyss gasolne to obtan a C 6 -C 8 cut sutable for extracton of aromatcs. The frst stage hydrogenaton takes place n a trckle bed reactor whle the second stage hydrogenaton s carred out n a two compartment fxed bed reactor n whch hydrogenaton of olefns takes place n the frst compartment whle the sulphur s elmnated n the second compartment. The key component n ths stage s consdered to be cyclohexene whose hydrogenaton was found to be the most dffcult among other olefns n the feed. The Langmur- Hnshelwood knetc expresson was adopted for hydrogenaton of cyclohexene and ts knetc parameters were determned expermentally n a mcro-reactor. The model was solved for the whole process of hydrogenaton, ncludng hydro-desulphurzaton. The predctons of the model were compared wth the actual plant data of an ndustral scale pyrolyss gasolne hydrogenaton unt and satsfactory agreement was found between the model and the plant data. Keywords: pyrolyss gasolne, hydrogenaton, reactor modellng 1. Introducton The pyrolyss gasolne, whch s a by-product n the olefn plants, has a hgh content of unsaturated components such as olefns and d-olefns. The exstence of these unstable components prevents the occurrence of desulphurzaton at hgh temperatures. Furthermore, under such condtons, foulng occurs due to polymer beng deposted on both the catalyst and the equpments. Therefore, the pyrolyss gasolne s hydrogenated n a two-stage process to resolve these dffcultes. In addton, the pyrolyss gasolne should be practcally free of sulphur compounds. In the frst stage, a selectve hydrogenaton of d-olefns to mono-olefns takes place, whle at the second stage, the hydrogenaton of the mono-olefns to saturated hydrocarbons would be completed. An attempt has been made n ths work to smulate the two stage pyrolyss gasolne hydrogenaton process based on the nformaton found n the lterature and the results of the experments conducted durng the study. Some plant data are also used to evaluate the performance of the model proposed n the present study. * Author to whom correspondence should be addressed: mostouf@ut.ac.r
2. Knetcs 2.1 Frst stage reactor In the frst stage reactor, d-olefns are beng hydrogenated. Cheng et al. (1986) studed the knetcs of pyrolyss gasolne hydrogenaton over the supported palladum catalyst. The hydrogenaton reacton s proposed as the followng consecutve way: k k Conjugated dene 1 2 monoolefn alkane Cheng et al. (1986) found that the knetc rate constants k * 1 and k * 2 are equal to 132 and 37 ml.g -1.hr -1, respectvely. The bg dfference between these values ndcates that hydrogenaton of conjugated denes can be selectvely controlled to favour of monoolefn formaton at the reacton condtons. Hanka and Lederer (1999) also studed the knetcs of pyrolyss gasolne hydrogenaton. Typcal pyrolyss gasolne contans C 5 -C 12 hydrocarbons and specally s rch n C 5 -C 6 cut. Accordngly, Hanka and Lederer (1999) consdered 18 ndvdual reactons for selectve hydrogenaton of dens, usng Pd/Al 2 O 3 catalyst. The followng assumptons have been consdered n ths work for the frst stage reactons knetcs: Eghteen reactons reported by Hanka and Lederer (1999) were used to model the frst stage reactor. The rate constants are consdered to be the same as those reported by Hanka and Lederer (1999). However, n the absence of nformaton about the actvaton energes, these values are consdered to be the same as that reported by Cheng et al. (1986) who studed the same reacton system. The actvaton energy and reacton rate constants for styrene and soprene hydrogenaton are obtaned from Cheng et al. (1986). 2.2 Second stage reactor In the second stage reactor hydro-treatng of mono-olefn components and desulphurzaton are performed. The feed to ths stage s the C 6 cut from the frst stage reactor. The reactor conssts of two fxed catalyst beds n seres. In the frst bed, monoolefns are hydrogenated whle desulphurzaton s occurred n the second bed. As mentoned above, cyclohexene s the key component n the C 6 cut. Hydrogenaton of cyclohexene produces cyclohexane as the man product. The catalyst n ths bed s N- Mo/Al 2 O 3. Qu and Prns (2002) showed that the order of hydrogenaton reacton of cyclohexene s between zero and one. Therefore, a Langmur-Henshlwood expresson was selected for the knetcs of cyclohexene hydrogenaton as follows: kk H PH r 1 K P H H In the second bed of the second stage, sulphur s removed from the pyrolyss gasolne by reacton wth hydrogen. Snce the C 6 cut has a narrow bolng pont range, t could be assumed that thophene, whch has almost the bolng pont of the C 6 cut, s the man sulphur compound n ths cut. Thophene dsappearance due to hydro-desulphurzaton s gven as follows (Ganetto and Slveston, 1986): (1
r HDS k P P T H 2 K P K P 2 1 T T H 2S H 2S (2) 3. Expermental 3.1 Materals and catalyst The chemcals employed n the experments were hydrogen (purty>99.99%), ntrogen (purty>99.99%); so-octane (purty>99%) as solvent; cyclohexene (purty>99%). The N-Mo/Al 2 O 3 catalyst sample was suppled by Procatalyse Co. (LD 145). The sample was crushed and seved to obtan catalyst partcles ranged from 300-600 μm. 3.2 Apparatus and procedure A schematc dagram of the expermental set-up s shown n Fgure 1. Ntrogen was passed through cyclohexene to have ntrogen gas saturated wth cyclohexene vapour to form the feed stream. Temperature of cyclohexene contaner was controlled by a water bath equpped wth temperature controller and heater. Hydrogen was fed to the reactor smultaneously. The reactor conssted of a 6.3 mm I.D. U-shaped stanless steel tube contanng 0.9 g of commercal N-Mo/Al 2 o 3 catalyst. The hydrogenaton was carred out at atmospherc pressure. The surroundng temperature of the reactor was kept constant by a melt salt bath equpped wth temperature controller and heater. The experments were carred out at 190, 240 and 290 C accordng to the ndustral condtons. Dfferent partal pressures of cyclohexene hydrogenaton were used at dfferent temperatures wth varyng feed flow n order to acheve a wde range of cyclohexene hydrogenaton conversons. Composton of the feed and product was analyzed usng a gas chromatograph. Fgure 1: Schematc of the expermental set-up for cyclohexene hydrogenaton A=Regulator, B=Saturator, C=Reactor, PI=Pressure ndcator, FIC=Rotameter, TC=Temperature controller 4. Model Development 4.1 Frst stage reactor The frst stage reactor s a trckle bed. The model equatons for the frst stage reactor are:
Mass balance: 1 dfl, f wac ls l, Lqud-catalyst mass transfer: a c k C C 0 n, r 1 kls Cl Cs, b s, (4) Energy balance: F dt C P H r (5) The correlatons used n ths work for evaluatng the hydrodynamc parameters of the trckled bed were adapted from Holub et al. (1992) for pressure drop, Larach et al. (1991) for lqud hold-up, Al-Dahhan, and Dudukovc (1995) for catalyst wettng effcency and Lakota and Levec (1990) for lqud-sold mass transfer coeffcent. 4.2 Second stage reactor The second stage reactor s a gas-phase fxed bed reactor whch could be modelled by the followng equatons: Mass balance: n 1 df r (6) b 1 Energy balance: F dt CP H r 5. Results and Dscusson 5.1 Expermental The Arrhenus and Van t Hoff laws were consdered for the reacton rate constants k and adsorpton constant K H, respectvely. From the expermental results obtaned at dfferent temperatures n ths work, the temperature dependence of the reacton rate and adsorpton equlbrum constants were calculated. Table 1 shows the ftted parameters for hydrogenaton of cyclohexene to cyclohexane over the NMo/Al2O3 catalyst based on the expermental data obtaned n ths work. Table 1 :Constants of knetc rate and adsorpton constants for cyclohexene hydrogenaton Parameter Unt Value Standard devaton k 0 mol/gr cat./mn 0.178 0.012 E a kj/mol 40.47 2.62 K H0 kpa -1 29.89 1.00 H kj/mol -24.48 1.48 (3) (7)
5.2 Modellng The model equatons for both reactors were solved for an ndustral unt. The reactors at each stage consst of two packed beds n seres (located n a sngle vessel) and a quench space s foreseen between these beds for coolng the product of the frst bed. Fgure 2 llustrates the calculated and measures axal temperature profles of the frst stage reactor. The curcles n ths fgure are the plant data and the sold lne represents the results of smulaton. It s seen n ths fgure that the temperature ncreases along the reactor length due to exothermc hydrogenaton of dens. It s obvous from ths fgure that there s a good agreement between the model predctons and actual plant data for ths reactor. Fgure 2. Axal temperature profle n the 1 st stage reactor Calculated and plant axal profle of temperature n the second stage reactor s shown n Fgure 3. It s seen n ths fgure that a very good agreement exsts between the calculated temperatures and plant data n both beds. The change n the temperature n the frst bed s due to hydrogenaton of olefns to alkanes whle n the second bed, ths change corresponds to hydrodesulphurzaton of the feed. Fgure 3. Axal temperature profle n the 2 nd stage reactor
As dscussed earler, thophene s consdered to be the sulphur contanng component of pyrolyss gasolne. It was found that sulphur s removed effectvely from the feed n ths reactor (Ahmadpour, 2003). 6. Conclusons A model was developed for modellng the process of pyrolyss gasolne hydrogenaton. The proposed model ntegrates hydrodynamc parameters and knetc models necessary to smulate such a process. For modellng the frst stage of hydrogenaton, n whch dens are completely removed from the feed, the knetc expressons from the lteratures were used. In the second stage of hydrogenaton, the olefns are removed from the feed by hydrogenaton to alkanes n the frst bed. The reacton rate of ths process s expermentally determned n ths work by nvestgatng hydrogenaton of cyclohexene to cyclohexane. Hydrodesulphurzaton of the feed occurs n the second bed of the second stage hydrogenaton reactor. For the modellng purpose, the sulphur-contanng component of the pyrolyss gasolne s assumed to be thophene whose hydrogenaton rate s found n the lteratures. The valdty of the model was demonstrated by comparng the model predctons wth plant data n terms of temperature profles n both frst and second stage reactors. Very good agreement between the predctons of the model developed n ths study and the actual ndustral reactor data s obtaned. The results of ths study can be used as a framework to further development of PGH processes. References Ahmadpour, M., 2003, Smulaton of Pyrolyss Gasolne Hydrogenaton Unt of Tabrz Petrochemcal Complex, MSc Thess, Unversty of Tehran, Iran. Al-Dahhan, M. H. and M. P. Dudukovc, 1995, Catalyst wettng effcency n trckle-bed reactors at hgh pressure, Chem. Eng. Sc. 50, 2377. Cheng, Y. M., T. R. Chang and J. C. Wu, 1986, Knetc study of pyrolyss gasolne hydrogenaton over supported palladum catalyst, Appl. Catal. 24, 273. Ganetto A. and L. Slveston, 1986, Multphase chemcal reactors: Theory, desgn, scale up, Hemsphere publshng. Hanka, J. and J. Lederer, 1999, Model hydogenace denu v pyrolyznm benznu, Chem. Lsty 93, 428. Holub, R. A., M. P. Dudukovc and P. A. Ramachandran, 1992, A phenomenologcal model for pressure drop, lqud holdup, and flow regme transton n gas-lqud trckle flow, Chem. Eng. Sc. 47, 2343. Larach, F., A. Laurent, N. Mdoux and G. Wld, 1991, Expermental study of a trckle-bed reactor operatng at hgh pressure: Two-phase pressure drop and lqud saturaton, Chem. Eng. Sc. 46, 1233. Lakota, A. and J. Levec, 1990, Sold-lqud mass transfer n packed beds wth cocurrent downward two-phase flow, AIChE J., 36, 1444. Qu L. and R. Prns, 2002, Hydrogenaton of cyclohexene over n stu fluornated NMoS catalysts supported on alumna and slca alumna, J. Catal. 207, 286. Acknowledgements Ths work was supported by Tabrz Petrochemcal Co. The authors are also grateful to Mr. S. Sharfna for hs help n completng the experments.