Limit Cycles in Homogeneous

Azeotropic Distillation
Zürich, June 18, 2001.

Urs Zimmerli

Advisors:
Eleonora Bonanomi
Prof. Marco Mazzotti
Prof. Manfred Morari
ETH Zürich

Limit Cycles in Homogeneous Azeotropic Distillation – p.1

Scope

Preparation of the experimental validation of sustained

oscillations

Detailed, dynamic analysis

Selection of a system for experimental investigation
Definition of design guidelines for the experimental
set-up

Limit Cycles in Homogeneous Azeotropic Distillation – p.2

Outline

1. Oscillations in Distillation Columns - Background

2. Detailed Analysis of Dynamic Behaviour
3. Preparations of Experimental Set-Up
4. Conclusions and Outlook

Limit Cycles in Homogeneous Azeotropic Distillation – p.3

Distillation Column
Given:
Design
Qc

Distillate
Feed conditions:
D, xD
Reflux L
, , and quality



Feed F, xF
2 degrees of freedom:
Boilup V
E.g. and

 


Qh Outputs:
Bottoms B, xB
Compositions: ,



Column profiles

Limit Cycles in Homogeneous Azeotropic Distillation – p.4

Nonlinear Behavior of Distillation Columns

xD Output MSS xD Limit Cycles

D Time
Theoretically predicted Theoretically predicted

Validated with simulations Validated with simulations only

Validated experimentally Experimental validation still missing

Limit Cycles in Homogeneous Azeotropic Distillation – p.5

Residue Curve Maps

Evolution of liquid phase

(residue) composition for a
simple batch distillation

V,y

M,x

Limit Cycles in Homogeneous Azeotropic Distillation – p.6

Residue Curve Maps

Evolution of liquid phase

(residue) composition for a
Aceton − Benzene − n−Heptan

simple batch distillation Aceton

1
Residue Curve @ p=101325 Pa

Residue Curve
Starting point
0.9 Ending point

0.8

0.7

0.6

mol fraction
0.5

0.4

V,y 0.3

0.2

M,x 0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Benzene mol fraction n−Heptan

Limit Cycles in Homogeneous Azeotropic Distillation – p.6

Prediction of Limit Cycles (Dorn 2000)

2 neighbouring
L
saddles in residue
Region Ω
curve diagram
Non-monotonous
in all three
components
Distillate composi- Az LH

tion must lie within

-region

Limit Cycles in Homogeneous Azeotropic Distillation – p.7

Status Quo

Mixtures:
Acetone - Benzene - Heptane
Methanol - Methylbutyrate - Toluene
( Expermental Validation of Multiple Steady States )
Models used:
Detailed investigations with CMO
Validation with ASPEN Dynamics

Limit Cycles in Homogeneous Azeotropic Distillation – p.8

Outline

1. Oscillations in Distillation Columns - Background

2. Detailed Analysis of Dynamic Behaviour
3. Preparations of Experimental Set-Up
4. Conclusions and Outlook

Limit Cycles in Homogeneous Azeotropic Distillation – p.9

Oscillations

We want to measure oscillations:

0.12

0.11

0.1

0.09
x(Toluene) (Distillate)

0.08

0.07

0.06

0.05

0.04

0.03

0.02
0 100 200 300 400 500 600
Time [h]

Limit Cycles in Homogeneous Azeotropic Distillation – p.10

Measurement of Oscillations

Measurement:
Amplitude
Period of oscillations

Limit Cycles in Homogeneous Azeotropic Distillation – p.11

Measurement of Oscillations

Measurement:
Amplitude
Top
Period of oscillations

Amplitude is not a problem:

Bottom
Temperature

Limit Cycles in Homogeneous Azeotropic Distillation – p.11

Characteristic Time of Plate Columns


Design, Operating Parameter, Thermodynamics






Limit Cycles in Homogeneous Azeotropic Distillation – p.12

Characteristic Time of Plate Columns


Design, Operating Parameter, Thermodynamics






 



L G

Limit Cycles in Homogeneous Azeotropic Distillation – p.12

Characteristic Time of Plate Columns

Validation through simulation upon two systems

ABH
120 MMBT

100
Period of oscillations [h]

80

60

40

20

0
0 1 2 3 4 5 6 7 8 9
H / L [h] x 10
−3

Limit Cycles in Homogeneous Azeotropic Distillation – p.13

Holdup in Packed Columns

Holdup in a packed column:














Therefore:











Limit Cycles in Homogeneous Azeotropic Distillation – p.14

Holdup in Packed Columns

Holdup in a packed column:














Therefore:











Conclusion:
Operate at highest possible reflux.

Limit Cycles in Homogeneous Azeotropic Distillation – p.14

Dependence on Distillate Flow Rate

Effects:
90
Period

80

Nonlinear 70

60

Period of oscillations [h]

Decreases at limits of 50

40

oscillation 30

20

Weak dependence in the 10

middle 0
0.93 0.935 0.94 0.945
D/F
0.95 0.955 0.96 0.965

Limit Cycles in Homogeneous Azeotropic Distillation – p.15

Dependence on Distillate Flow Rate

Effects:
90
Period

80

Nonlinear 70

60

Period of oscillations [h]

Decreases at limits of 50

40

oscillation 30

20

Weak dependence in the 10

middle 0
0.93 0.935 0.94 0.945
D/F
0.95 0.955 0.96 0.965

Conclusion:
Cannot be exploited to reduce time of oscillations.

Limit Cycles in Homogeneous Azeotropic Distillation – p.15

Summary of Influences on




Effects:
Strong dependence on holdup/reflux
Weak dependence on distillate flow rate
No effect of feed tray position
Weak dependence on feed composition
Conclusion:
Period of oscillations is very high and cannot be
reduced as desired for the given system.

Limit Cycles in Homogeneous Azeotropic Distillation – p.16

Screening for a new System

Another system might have a smaller period of oscillation

Requirements:
No reactions between the components
Harmless components (toxicity, acidity, flammability)
Ease of separation
Big -region low sensitivity to operating conditions
Systematic screening of a database with about 50
components was not successful.

Limit Cycles in Homogeneous Azeotropic Distillation – p.17

Outline

1. Oscillations in Distillation Columns - Background

2. Detailed Analysis of Dynamic Behaviour
3. Preparations of Experimental Set-Up
4. Conclusions and Outlook

Limit Cycles in Homogeneous Azeotropic Distillation – p.18

Distillation Column
Design:
Qc

Distillate
Reflux L
D, xD Tray or packed
Feed F, xF column
Number of Stages
Boilup V
( )



Qh
Reboiler and con-
Bottoms B, xB
denser

Limit Cycles in Homogeneous Azeotropic Distillation – p.19

Influence of Holdup

Phenomenom:
Oscillations disappear if holdup in reboiler or
condenser becomes much higher than on stages
Ratio of Interest:


  

and



 



Critical Values:
10 (ASPEN) to 50 (CMO)








 









Limit Cycles in Homogeneous Azeotropic Distillation – p.20

Influence of Holdup
1

0.9

0.8

0.7

0.6

x(Methanol)
0.5

0.4

0.3

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x(Toluene)

Limit Cycles in Homogeneous Azeotropic Distillation – p.21

Influence of Holdup
1

0.9

0.8

0.7

0.6

x(Methanol)
0.5

0.4

0.3

0.2

0.1

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x(Toluene)

Conclusion:
Reflux drum as small as possible
Additional stages favourable
Keep reboiler holdup small
Limit Cycles in Homogeneous Azeotropic Distillation – p.21
Influence of

0.96 0.96
46, period ~55h 69, period ~80h

When increasing : 0.95 0.95

0.94 0.94

0.93 0.93

x(Methanol)

x(Methanol)
longer 0.92 0.92




0.91 0.91

Limit cycle behaviour 0.9 0.9

reached quicker 0.89 0.89

0.88 0.88
0.02 0.04 0.06 0.08 0.1 0.12 0.02 0.04 0.06 0.08 0.1 0.12
x(Toluene) x(Toluene)

Limit Cycles in Homogeneous Azeotropic Distillation – p.22

Influence of

0.96 0.96
46, period ~55h 69, period ~80h

When increasing : 0.95 0.95

0.94 0.94

0.93 0.93

x(Methanol)

x(Methanol)
longer 0.92 0.92




0.91 0.91

Limit cycle behaviour 0.9 0.9

reached quicker 0.89 0.89

0.88 0.88
0.02 0.04 0.06 0.08 0.1 0.12 0.02 0.04 0.06 0.08 0.1 0.12
x(Toluene) x(Toluene)

Conclusion:

Optimum must be found

Limit Cycles in Homogeneous Azeotropic Distillation – p.22

A Rate based Approach

Change to a rate based model, RateFrac:

Heat transfer
Mass transfer
Motivation:
Validation of our column design
Retrieve efficiencies which can be used in dynamic
simulations with a simpler model

Limit Cycles in Homogeneous Azeotropic Distillation – p.23

RateFrac / Results

Results:
Practically sensible Size around 3m
Height of a theoretical stage cm


 

h for MMBT, thus h









Disadvantages:
No validation with dynamic simulation possible

Limit Cycles in Homogeneous Azeotropic Distillation – p.24

Suggested Experimental Setup

Mixture MMBT: ,





















Feed flow rate 1 kg/h
Packing: Sulzer DX, height 3.5m, diameter 0.1m
Condenser as small as possible













Temperature measurements 0.25 -1.75m from top

Expected period of oscillation: 30h

Limit Cycles in Homogeneous Azeotropic Distillation – p.25

Conclusions

Experimental validation rather hard

Very high



More checks are necessary before the experimental
validation

Limit Cycles in Homogeneous Azeotropic Distillation – p.26

Limit Cycles in Homogeneous Azeotropic Distillation – p.27