Steady-state Packed Bed Reactor Modeling for Methanol

Transcription

rout@nt.ntnu.no
1
Steady-state Packed Bed Reactor
Modeling for Methanol Synthesis
K. R. Rout*, H. A. Jakobsen
NTNU, Norway
2nd Trondheim Gas Technology Conference 2011
rout@nt.ntnu.no
2
Contents
•
Motivation of methanol synthesis
Pseudo-homogeneous model
•
Types of reactor model
Conventional heterogeneous model
Simplified heterogeneous model
•
Comparasion of Psedu-homogeneous-, conventional- and simplified
heterogeneous model
•
Numerics: least squares spectral element method.
•
Results and Discussion
•
Conclussion
rout@nt.ntnu.no
3
Process evaluation for methanol synthesis
Net power
Total input
Total output
Reactor Simulation for methanol synthesis
Reactor type design/size
Best operating conditions
Transport phenomena
Kinetic study for methanol synthesis
Catalyst preparation
Kinetic rate
Catalytic activity
rout@nt.ntnu.no
4
Motivation for the process
•
•
•
•
•
METHANOL is multipurpose based chemical.
It can be used as chemical intermidiates.
Easy for transportation.
It has high octane number.
Good antiknock performance.
Future renewable
energy sources
Fuel
(Methanol + Gasoline)
combustion engine
Methanol: Fuel cell
rout@nt.ntnu.no
5
Motivation for steady-state reactor modeling
METHANOL synthesis:
• Feed treatment purification
Intrinsic nonlinerity
• Reforming
+
= Computer Simulation
• Methanol synthesis
complex phenomena
• Product purification and storage
Dynamic simulation:
Start-up and shut down investigation
System identification
Reactor modeling ?
Safety, control, optimization
Catalytic activity
Pseudo-homogeneous
Conventional- and simplified Heterogeneous
Steady-state model
rout@nt.ntnu.no
6
Reactions involved in Methanol synthesis
•
•
CO + 2H2 = CH3OH……………………(1)
CO2 + H2 = CO + H2O………………...(2)
CO2 + 3 H2 = CH3OH + H2O….......(3)
The over all reaction is exothermic.
Side reactions (even through they are inhibited by high catalyst selectivity
under normal operating conditions):
CO + 3H2 = CH4 + H2O……………………………….(4)
nCO + 2nH2 = CnH2n+1OH + (n-1) H2O………(5)
2CH3OH = CH3OCH3 + H2O……………………….(6)
This reaction occurs at 570 K; might lead to reactor instabilites
due to of its high exothermicity.
•
Catalyst: Cu/Zn/Al2O3 (life cycle: 2 year)
rout@nt.ntnu.no
7
Packed bed reactor models
Derivation of Reactor model
Mass based model :

Mole based model : C
Assumption of total numer of moles
along the reactor axis is constant: this is not true
Modeling of packed bed reactor
Pseudo-homogeneous model
Heterogeneous model
do not account explicitly the
seperate equation for the solid and
Conventional
Simplified interstitial
presence of pellet
fluid.
Mass transport:
Mass transport:
1. Wilke Model
Multicomponent
Efficiency
2. factors
Maxwell-Stefan
Model diffusion models
3. Dusty gas model
rout@nt.ntnu.no
8
Pseudo-homogeneous reactor model
•
•
 g uz
Species Mass balance:
Total continuity equation:
•
Energy balance:
•
Ergun Equation:
 j
z
 (1   ) k Rk jk
k

(u z  g )  0
z
T
4U
 (1   ) (H rk ) k Rk 
Tw  T 
dt
z
2

P
 1     uz  g  1   
   1.75  150 

 3 
z
 Re   d p   

 g Cpg u z
9
rout@nt.ntnu.no
Heterogeneous Model
Conventional
a) Fluid phase equations:
i
 k gi av  g (is  i )

u
Mass balance: g z
z
Energy: g Cpg uz
Ergun Eq.:
T
4U
 h f av (Ts  T ) 
(Tw  T )
z
dt
2
P
1   uz  g 1  
 (1.75  150(
))
( 3 )
z
Re
dp

Simplified
a) Fluid phase equations:

Mass balance:  g u z i  k gi av  g (is  i )
z
Energy: g Cpg uz
T
4U
(Tw  T )
 hf av (Ts  T ) 
dt
z
P
1   uz  g 1  
 (1.75  150(
))
( 3 )
z
Re
dp

2
Ergun Eq.:
b) Solid phase equations:
Mass balance:
av  g k g ,i (i  i , s )  (1   ) k Rkik  0
b) Solid phase equations (Wilke,
Maxwell-Stefan, Dusty gas model):
k
Mass balance:
Ji,r
i,s Energy:
 (1)k Rkik where Ji,r g,s Dr,i
r
r av h f (T  Ts )  (1   ) (H rk )k Rk  0
k
Energy:
Qs
T
 (1   ) (H rk )k Rk where Qs  s s
r
r
rout@nt.ntnu.no
10
Numerical scheme
•
Least square spectral method (LSM)*.
Because, for a given accuracy the high order spectral method requires
only a few collocation points where as the low order methods such as
the finite volume method (FVM), require a large number of grid
points.
•
The basic idea in the LSM is to minimize the integral of the square of
the residual over the computational domain.
*Rout K. R., Solsvik J., Nayak A. K., Jakobsen H. A. (2011) A numerical study of multicomponent
mass diffusion and convection in porous pellets for the SE-SMR and desorption processes. Chemical
Engineering Science 66:4111-4126
rout@nt.ntnu.no
11
Overall Purpose
•
Comparison of pseudo-homogeneous reactor model, conventional- and
simplified heterogeneous reactor models with experimental data.
•
The proposal of best suitable reactor model under normal operating
conditions of fixed bed Lurgi reactor.
•
Comparasion of different multicomponent diffusion models.
rout@nt.ntnu.no
12
MoleTemperature[K]
fraction of CH OH
3
Comparasion with Rezaie et al.
530
0.06
525
Simplified heterogenous(gas phase)
Simplified heterogenous(gas phase)
Simplified heterogenous(solid phase)
Literature heterogenous(gas phase)
Literature heterogenous(gas phase)
520
0.04
515
510
0.02
505
50000
0
22
44
Axial direction
direction[m]
[m]
Axial
66
88
Rezaie N., Jahanmiri A., Moghtaderi B. Rahimpour M.R (2005) A comparasion of homogeneous
and heterogeneous dynamic models for industrial methanol Reactors in the presence of catalyst
deactivation. Chemical Engineering and Processing 44: 911-921
rout@nt.ntnu.no
13
Comparasion with different packed bed reactor models
Mole fraction
of [bar]
CH OH
Pressure
Temperature[K]3
Psedu-homogeneous
Psedu-homogeneous
Psedu-homogeneous
Simplified
heterogenous(gas
phase)
Simplified
heterogenous(gas
phase)
Simplified
heterogenous
Simplified
heterogenous(solid
phase)
heterogenous(solid
phase)
6 Simplified
Conventional
heterogenous
x 10 Conventional
Conventional
heterogenous(gas
phase)
heterogenous(gas
phase)
7.7
Conventional
heterogenous(solid
phase
Conventional
heterogenous(solid
phase)
0.06
530
7.695
0.04
520
7.69
0.02
510
7.685
0
500
0
7.68
00
1
2
3
4
5
22
4
4
Axial direction [m]
Axial direction
direction [m]
[m]
Axial
66
6
7
88
rout@nt.ntnu.no
14
Comparasion with Experimental data by
Rezaie et al
Time Plant
(day) (tonne/day)
PseudoSimplified
homogeneous heterogeneous
(tonne /day)
(tonne/day)
Conventional
heterogeneous
(tonne/day)
0
308.86
300.00
295.0
308.80
Rezaie N., Jahanmiri A., Moghtaderi B. Rahimpour M.R (2005) A comparasion of homogeneous
and heterogeneous dynamic models for industrial methanol Reactors in the presence of catalyst
deactivation. Chemical Engineering and Processing 44: 911-921
rout@nt.ntnu.no
15
Comparasion of different multicomponent
diffusion models in conventional heterogeneous
model
Maxwell-Stefan Model
Maxwell-Stefan Model
Dusty
Dusty
gasgas
ModelModel
Wilke Model
Wilke Model
0.06
Temperature [K]
3
Mole fraction of CH OH
530
525
0.04
520
0.02
515
0
0
r
2
510
505
0
2
4
Axial direction [m]
6
4
Axial direction [m]
8
6
8
rout@nt.ntnu.no
16
Conclusion
• Both Pseudo-homogeneous- and simplified heterogeneous
reactor model slightly overestimates with experimental data.
• However, the percentage of deviation is quite small.
• Conventional heterogeneous model is computationally
expensive.
• As there is no significant deviation between gas- and solid
phase concentrations of different components along the
reactor axis, the pseudo-homogeneous reactor model seems
to be best choice for methanol synthesis process.
rout@nt.ntnu.no
17
Thank You

Steady-state Packed Bed Reactor Modeling for Methanol

Transcription

Similar documents

Physics of Oil recovery

Büchi Parallelreaktorsystem PPR - C3 Prozess

alkoxylation - Buss ChemTech

Use of models at central banks: Comments by

Document 6577123

user`s Guide - Bulk Reef Supply

AUTHORS AND PAPERS