Docstoc

Catalytic Distillation Modelling

Document Sample
Catalytic Distillation Modelling Powered By Docstoc
					Catalytic Distillation Modelling and Simulation using HYSYS.Process™
Environment

Gheorghe BUMBAC1, Grigore BOZGA 1 , Valentin PLESU 1, Vasile BOLOGA 1 , Ilie. MUJA 2 and
Corneliu Dan POPESCU 2
1
  University “POLITEHNICA” of Bucharest, Department of Chemical Engineering, 1 Polizu
Street, RO-78126, Bucharest, Romania, Tel/Fax:+40 (0)21 21.25.125, email: cttip@chim.upb.ro
2
  S.N.P. PETROM, INCERP Ploiesti Subsidiary, B-dul Republicii Nr. 291 A, RO-2000, Jud.
Prahova, Telefon +40 (0)244 135111, Fax +40 (0)244 198732, email: popescu@serv.incerp.ro

     The catalytic-distillation process for the production of t-amyl-methyl-ether (TAME) from
     methanol and isoamylenes was simulated by developing the process model as a
     combination of unit operations from HYSYS™ operations palette. Geometrical
     characteristics of catalytic-distillation column are those of an industrial pilot plant and the
     results of simulation were compared with experimental data. The experimentally
     determined reactions kinetics was applied in the model. UNIQUAC-UNIFAC model
     equations were selected for the vapour-liquid equilibrium.
     The results show that fair agreements between the calculated and experimental data were
     obtained.

1. INTRODUCTION
        Methyl ethers replace lead compounds in gasoline. One of these ethers obtained by the
etherification of an isoamylenes mixture (2-methyl-1-butene, 2M1B, and 2-methyl-2-butene)
with methanol is t-amyl-methyl-ether (TAME).
        Catalytic distillation is a suitable technique for TAME synthesis due to the
reversibility of the etherification reactions and the difference between the reactants and
products volatilities. These particularities favours the enhancement of the reactant conversion
and the increase of the interphase mass transfer potential.
        Despite of the important number of publications in the field of catalytic distillation,
relatively few of them concern the industrial application of TAME synthesis.
        In this paper we focused on the relevance of the commercial software HYSYS™ for
the simulation of catalytic distillation problems. To predict the behaviour of TAME synthesis
reactor and reactive distillation column HYSYS.Process environment was used. The
simulation results were compared with pilot plant experimental data. The pilot plant system
consists mainly from a tubular, fixed bed pre-reactor (TFBR) and a reactive distillation (RD)
packed column.
        The purpose of this study is: a) to develop a suitable simulation module for
heterogeneous RD with HYSYS and b) to apply the model to an industrial applications.
        The TFBR is used to bring the reaction mixture near its equilibrium composition. The
advantage of using a pre-reactor for TAME synthesis is based on the fact that the greatest part
of reaction components can react before RD column and the throughput of reaction system
increases.
        HYSYS™ provides many built-in modules for simulating various processes.
Unfortunately the COLUMN subflowsheet environment allows simulation of RD with
reactions taking places only in homogeneous phase. In the heterogeneous catalytic distillation
process the solid catalyst particles are placed into many special packing envelopes, serving
also as vapour-liquid contacting. The reactions occur inside the catalytic package where the
liquid contacts the catalyst particles. Then the product flows out of the catalytic zone.
Additional separation takes place on the packing placed below and above the reaction section
of the column.

                                                   1
        In a RD column the reaction and separation actually take place in the different
locations of the column i.e. reaction on the catalyst pellets of the packaging and separation in
the inert packing. Therefore, the parameters of the liquid residence time or the liquid hold-up
on the trays or packings in the heterogeneous process can only be used in the separation
calculation whereas the reaction calculation needs the parameters of contacting time of liquid
with catalyst in the catalytic package. We underline that in the current version of HYSYS™
the built-in RD module is not directly suitable for the simulation of the heterogeneous
catalytic distillation process.
        To overcome the above problem our study concentrated to develop a model for
heterogeneous RD and implement the model in the HYSYS simulation environment. In this
model, the catalyst space velocity appearing in the reaction system equations represents the
contacting time of liquid with catalyst.

2. Reaction kinetics and thermodynamics
        Industrial processes for TAME synthesis are based on the reversible reactions of
isoamylenes (2M1B and 2M2B) with methanol. The equilibrium conversion of isoamylenes
to TAME, at 60°C, is 56% if stoechiometric amounts of isoamylenes and methanol are used 6
and increases slightly with the increase of methanol/isoamylenes ratio. In Table 1 the
equilibrium conversion as a function of temperature, for stoechiometric methanol/ isoamylenes
molar ratio, is presented 6 .
        A typical industrial process for TAME synthesis involves at least 8 components:
isoamylenes, n-pentane, i-pentane, methanol, 1-pentene and trans-2-pentene. Since methanol
associates almost all hydrocarbon components into simple and complex azeotropic pairs, the
system shows strong non-ideal properties.
        The property package used to calculate the liquid activities of the considered
components is based on UNIQUAC-UNIFAC model.

Table 1. Equilibrium isoamylenes conversion              Table 2. The feed composition.
         as a function of temperature
   Temperature             Conversion
          o
         ( C)
          50                  0.618
          60                  0.560
          70                  0.491
          80                  0.446


        The two reactive olefins (2M1B and 2M2B) are contained in the hydrocarbon mixture,
resulted as a C5 fraction from Fluid Catalytic Cracking (FCC) unit. In Table 2 the
composition of the feeding mixture used in the simulated process scheme is presented.
Residual TAME is present in this stream from recycled stream. All components from the C5
fraction are producing azeotropes with methanol and the composition of these azeotropes is
presented in Table 3. The methanol concentration in these azeotropes increases with pressure.




                                               2
Table 3. Azeotrope compositions in TAME synthesis.
                                  p=2.5 bar            p=4 bar                                 p=5.5 bar
                                           o
 Component 1 Component 2         x1      t, C       x1       t, o C                          x1       t, o C
 methanol       2M1B           0.21      53.76     0.243      69.24                         0.268      80.07
 methanol       2M2B           0.28      58.69     0.31       73.78                         0.331      84.69
 methanol       n-pentane      0.295     58.64     0.328      73.96                         0.347      85.20
 methanol       i-pentane      0.21      51.22     0.252      66.61                         0.280      77.85
 methanol       1-pentene      0.22      53.64     0.267      69.06                         0.283      80.75
 methanol       2-pentene      0.265     56.73     0.301      72.29                         0.322      82.72
 methanol       TAME           0.763     87.56     0.793     102.45                         0.802     113.20

The synthesis of TAME from methanol and isoamylenes, catalysed by acid ion-exchange
resin catalyst is a reversible process as shown in following reaction mechanism containing the
main and secondary reactions (1÷7):
                                                                         O − CH 3
                                                          r1              |
CH 2 = C − CH 2 − CH 3 (l ) + CH 3 − OH ( l)                       CH 3 − C       − CH 2 − CH 3 (l )   (1)
       |                                                                   |
      CH 3                                                r2                 CH 3
 2 Methyl 1 Butene ( 2M1B )                                               TAME


                                                                       O − CH 3
                                                     r3
                                                                        |
                                                                 CH 3 − C       − CH 2 − CH 3 (l )     ( 2)
CH 3 − C = CH − CH 3 (l ) + CH 3 − OH (l )
       |                                                                  |
                                                     r4
      CH 3                                                                 CH 3
 2 Methyl2 Butene (2 M 2 B )                                            TAME

                                    r5         CH 3 −        C = CH − CH 3 ( l )                       ( 3)
CH 2 = C − CH 2 − CH 3 ( l )                               |
       |
      CH 3                          r6                    CH 3

  2 M 1B                                         2 M 2B


C 5 H 10 ( l ) + C 5 H 10 ( l )       →         C 10 H 20 ( l )                                        (4)
    2 M 1B          2M 2B                   di − iso − amylene


C 5 H 10 ( l ) + C 5 H 10 ( l )   →         C 10 H 20 ( l )                                            (5)
   2M 2B            2M 2B                di − iso − amylene
2CH 3 − OH
             ( l)                             CH 3 − O − CH        + H 2 O ( l)                        ( 6)
                                                            3 ( l)
                                                  di − methyl − ether
C5 H10 (l)   + H 2O                        C 5H12 − O (l)                                              (7)
 iso − amylenes                                tert − amyl − alcohol

        There are three main reactions (reactions 1÷3, one for etherification of 2M1B, one for
etherification of 2M2B and an isomerisation reaction between 2M1B and 2M2B) and four
secondary reactions (4÷7). Both etherification reactions, (1) and (2), are exothermic, i.e. the
equilibrium conversion of TAME decreases with temperature. The isomerisation reaction at
operation temperature (between 60°C and 120°C) favours the 2M2B formation and this
component will have the greatest concentration in the reaction mixture. From kinetic point of

                                                          3
view this situation is not advantageous because a faster reaction (1) is replaced by a slower
one (2).
         High temperatures and low methanol concentrations are favourable conditions for
isoamylenes’ oligomers formation. On the other hand, excess methanol produces higher
dimethyl-ether concentration, whereas t-amyl-alcohol formation is very limited, in
equilibrium conditions, due to a very small water concentration. Frequently used catalysts are
sulphonic acid ion-exchange resins (Amberlyst 15 or 35, Levatit SPC 118). The kinetic
mechanism is based on the consideration that methanol and TAME are stronger adsorbed on
the catalyst’s surface, compared to isoamylenes.
         According to our knowledge kinetic studies for TAME synthesis were published by
Muja et. al 7 , Randriamahefa et al 8 , Piccoli and Lovisi , Oost and Hoffmann 3 and Rihko et
al. 2 etc.
         In this work the TAME synthesis kinetic model of Rihko et al. 3 on Amberlyst 16 was
used. The 2M1B and 2M2B consumption rates have the expressions:

                                      1      aT       
                                  K ⋅ a ⋅a
         − k B1 ⋅ aM ⋅ a 2M 1B ⋅  1 −                 
                                                       
                                               2 M 1B                          1 a2 M 2 B 
                                                         − k B5 ⋅ a2 M 1B ⋅  1 −             (8)
                                        1   M
r1B    =                                                                           ⋅
                         KT                                                    K3 a2 M 1B 
                                                                                             
                        
                         K ⋅ aT + aM    
                         M               
                                        1       aT         
         − k B3 ⋅ aM ⋅ a 2M 2 B ⋅  1 −
                                          ⋅                
                                                            
                                       K 2 a M ⋅ a 2M 2B   +k                          1 a2 M 2 B 
r2 B   =                                                               ⋅ a 2M 1B ⋅  1 −
                                                                                           ⋅           (9)
                                                                                         K 3 a 2M 1B 
                                                                  B5
                         KT                                                                       
                        
                         K ⋅ aT + a M     
                         M                 

The activation energies for the reactions (1), (2) and (3) are:

       EkB1 = 76 kJ / mol ; EkB3 = 94.1 kJ / mol ; E kB5 = 81.6 kJ / mol
and                        mol                          mol                         mol
       k B1 ( 343 K ) = 0.286   ; k B3 ( 343K ) = 0.125     ; k B5 ( 343K ) = 0.107
                           g ⋅h                         g⋅h                         g ⋅h
are reaction rates constants at 343 K.

           The equilibrium constants in the above reaction rates as temperature functions are:
                K1 = exp (− 8.3881 + 4041.2 / T )                                 (10)
                K 2 = exp (− 8.2473 + 3225.3 / T )                                (11)
                K 3 = exp (− 0.1880 + 833.3 / T)                                  (12)

                     = 0.1405 − 0.00061⋅ (T − 334)
                KT
                                                                                  (13)
                KM

        The ion exchange capacity Amberlyst 15 is 5 mequiv/g 8,9. The reaction kinetic data
have been verified using pilot plant synthesis of TAME in which the fixed bed reactor
packing of the same catalyst was used. Molar ratio between methanol and isoamylenes in the
feed stream was 1.256.

3. Process flowsheet
       In figure 1 TAME synthesis process flowsheet using catalytic distillation is presented.
The C5 fraction is mixed with methanol and the resulting stream is fed to the preliminary


                                                            4
reactor (IV). In the Table 4 the composition of the mixture at the preliminary reactor exit is
presented.
        The resulting product is mixed with a recycled methanol stream and is fed to a
catalytic distillation column, with three zones, below the reaction zone. The stripping zone of
the catalytic-distillation column is simulated as reboiled absorber, a standard operation in
HYSYS™.
        The second part is the reaction separation zone, represented in our model by a
backflow cell model (BCM) with forward flow of the liquid and backward flow of the vapour
in the reactive part of RD zone. The BCM consist of series of five continuous stirred tank
reactors (CSTR) units with the same geometry and size of the individual unit. The third part is
another pure mass transfer unit, representing the rectifying zone of the reactive distillation.
This zone is simulated as refluxed absorber a HYSYS™ standard operation. Both stripping
and the rectifying zones are represented as non-catalytic packed columns.
        From the computational point of view, each cell of the series was assumed to be at V-
L equilibrium, the increase of conversion being calculated as in a CSTR reactor. Here
reactions take place in liquid-solid interface, following the kinetic law mentioned above.
        System characteristics: The main characteristics of the catalytic distillation column
are: pre-reactor volume: 0.12 m3 , stripping zone: 6 theoretical stages, rectifying zone: 3
                                                                                             3
theoretical stages and in cell model there are five CSTRs, a CSTR vessel has 0.02 m of
catalyst. Catalyst particles average diameters considered in the simulation were of 1 mm as in
the pilot plant case. These characteristics are in agreement with those of the pilot plant. More
explicit characteristics for stripping and rectifying zones of the RD column are presented in
Tables 6 and 7.

Table 4. Feed conditions for the RD           Table 5. Column heat exchangers
column system




Table 6. Stripping zone characteristics      Table 7. Rectifying zone characteristics




                                               5
4. Results and discussion:
        Several results from the simulation of main streams are presented in Table 8. The
isoamylenes conversion in the reactive distillation column and pre-reactor is 80.76 %. Pilot
plant scheme is presented in Figure 2. The characteristics of the experimental pilot plant are:
pre-reactor volume 120 l; rectifying zone packing height 2 m; stripping zone packing height
3.5 m. Feeding condition and thermodynamic regime was the same as in the simulation.
        The simulation results with HYSYS™ for the TAME synthesis reactive distillation
module set-up, presented in this work, allow drawing the following conclusions:
     - From the chemical transformation point of view it is profitable to place the reaction
        zone as close as possible to the top of the column. However, above the reaction zone
        a separation zone is needed to separate TAME from the distillate.
     - It is recommended to place the column feed bellow the reaction zone in order to
        ensure high concentration for the reactants in this zone (as there are more volatile
        compared with the reaction product).
     - The best structure for the RD column, obtained from this simulation study, involve 15
        theoretical plates. As we denoted plates from top to bottom, the best position for the
        reaction zone are the theoretical plates 3 and 4, and the feed plate is the 5-th plate.
        The optimal reflux ratio is 2, as result of the trade-off between separation degree and
        energy saving.

        To describe the flow and fluid phase mixing in the reaction zone, a classical, multi-
cellular, model was used, considering the back-flow of the vapour phase. In each cell the
conversion increase was calculated considering uniform distribution of the catalyst and
vapour-liquid equilibrium.
        The results obtained (over 90% conversion, much bigger than the equilibrium
conversion) emphasises the advantage of catalytic distillation compared with the classical
scheme, because the chemical transformation is not limited by the chemical equilibrium as
result of continuous separation of the reaction product from the mixture.

       Table 8. Simulation results.




                                              6
                           III




                                                              n=N     VN
                                                  LN+1

                      II                                                         VN-1
                                                         LN           n=N-1
                                                                                        LN-1
                                                               VN-2

                                                                                n=3     L4
                                                                V3

                                                                                      n=2      V2
                                                                      L3



                                                                           V1                        L2
                                                                                               n=1
                                             I
                 IV                                                                     L1
                                                                                                          V0


                           a) process flowsheet               b) cells in series model


                            Figure 1. Simplified flowsheet for RD column.




               Figure 2. TAME Pilot Plant with Catalytic Distillation Column.

Iso-amylenes conversion data (Table 9) are obtained using the pilot plant presented in fig. 2,
under the same experimental conditions.


                                                  7
Table 9. Experimental results for TAME synthesis.




5. Conclusions

        This paper presents a theoretical study for the modelling of reactive distillation
column operation in TAME synthesis. The simulation procedure is based on a mathematical
model considering chemical reaction kinetics for the main reactions and the vapour-liquid
equilibrium. Phase contact in the reaction zone is described with the back-flow cell model.
The problem statement in HYSYS.Process environment was made considering three zones for
the catalytic distillation column (rectifying, reaction and stripping). Constructive and
operational characteristics of the column are specified as a consequence of the parametric
study: reaction zone position, feed position and reflux ratio, in order to obtain maximum yield
for the transformation of C5 reactive olefins in the pilot plant. The simulation results are in
good agreement with experimental data obtained in the experimental pilot plant at SNP
PETROM, INCERP Ploiesti subsidiary.
        The quality of the results obtained in this paper is limited by the uncertainty
introduced by the phase hydrodynamics in the reaction zone the phase equilibrium hypothesis.
The authors foreseen additional studies in order to better describe phase hydrodynamics, to
consider interphase mass transfer inside the catalyst pallets on process performances.


Nomenclature:
ai – activity of component i;                               ri – reaction rate;
Kj – equilibrium constants in TAME synthesis                p – pressure;
     reactions;                                             T – absolute temperature, K;
kBm – rate constants in reactions (8), (9);                 R – ideal gas constant.



REFERENCES
  1. L.K. Rihko, A.I.O. Krause, Ind. Eng. Chem. Res. 1995, 34, 1172.
  2. L.K. Rihko, P.K. Kiviranta-Pääkkönen, A.I.O. Krause, Ind. Eng. Chem. Res., 1997, 36,
      614.
  3. C. Oost, U. Hoffmann, Chem. Eng. Sci. 1996, 51, 329.
  4. W.B.Su, J.R. Chang, Ind. Eng. Chem. Res, 2000, 39, 4140.
  5. A.P. Higler, R. Taylor, R. Krishna, Chem. Eng. Sci. 1999, 54, 1389.
  6. K. Sundmacher, U. Hoffmann, Chem. Eng. Sci, 1994, 49, 4443.
  7. L. Muja, et. al., Revista de Chimie, 1986, 37, 1047.
  8. S. Randriamahefa, R. Gallo, J. Mol. Catal. 1988, 49, 85.
  9. H. Subawalla, J.R. Fair, Ind. Eng. Chem. Res., 1999, 38, 3696.
  10. J.L. DeGarmo, V.N. Parulekar, V. Pinjala, Chem. Eng. Progress, 1992, 88, 43.
  11. J.C. Gonzales, H. Subawalla, J.R. Fair, Ind. Eng. Chem. Res., 1997, 36, 3845.
  12. J.C. Gonzales, J.R. Fair, Ind. Eng. Chem. Res. 1997, 36, 3833.
  13. S. Ung, M.F. Doherty, Chem. Eng. Sci., 1995, 50, 23.



                                               8

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:22
posted:10/1/2011
language:Romanian
pages:8
Description: Catalytic Distillation Modelling Document