methods of multicomponent distillation...
Post on 30-May-2020
35 Views
Preview:
TRANSCRIPT
Azeotropic Distillation
Dr. Jungho Cho, Professor
Department of Chemical Engineering
Dong Yang University
Azeotropic Distillation Slide 2
Limit of Distillation by Azeotrope
Liquid Mole Fraction of Ethanol
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Va
po
r M
ole
Fra
ctio
n o
f E
tha
no
l
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
� Distillation range is restricted by
the azeotropic point.
� Binary azeotropic mixtures, such
as ethanol/water and IPA/water,
can be separated into their pure
components by distillation by the
addition of a third component, so
called the entrainer, which forms
a ternary azeotrope with a lower
boiling point than any binary
azeotrope
Ethanol / Water System
Azeotropic Distillation Slide 3
Separation of Azeotropic Mixture
� Shift the Azeotropic Point by Changing Pressure.
� Add the Third Component. � Azeotropic Distillation: Volatile Addition
� Extractive Distillation: Non-volatile Addition
Azeotropic Distillation Slide 4
Azeotropic vs. Extractive Distillation (1 of 2)
� Azeotropic distillation � By forming a ternary heterogeneous azeotrope lower than any
other binary azeotropic temperatures, nearly pure ethanol can be obtained as a bottom product in an azeotropic distillation column.
� Ethanol is obtained as a bottom product from an azeotropic distillation column using an entrainer such as benzene or normal pentane.
� Extractive distillation
� By adding a solvent which is exclusively familiar with a wanted component in a feed mixture, a desired component can be obtained in an extractive distillation column overhead.
� Ethanol is obtained as a top product from an extractive distillation with ethylene glycol solvent.
Azeotropic Distillation Slide 5
Azeotropic vs. Extractive Distillation (2 of 2)
FEEDFEEDFEEDFEED
RECYCLERECYCLERECYCLERECYCLE UPPERUPPERUPPERUPPER PHASEPHASEPHASEPHASE
LOWERLOWERLOWERLOWER PHASEPHASEPHASEPHASE
PURE ETHANOLPURE ETHANOLPURE ETHANOLPURE ETHANOL
FEEDFEEDFEEDFEED
SOLVENTSOLVENTSOLVENTSOLVENT
PURE ETHANOLPURE ETHANOLPURE ETHANOLPURE ETHANOL
Azeotropic Distillation Extractive Distillation
Azeotropic Distillation Slide 6
Principle of Azeotropic Distillation
0.0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.00.0
.1
.2
.3
.4
.5
.6
.7
.8
.9
1.0
WaterBenzene
EthanolP
A
F
W
R D
B
C
V
G
IIII
II
A : 78.07 oC B : 67.99 oC
C : 69.31 oC D : 63.88 oC
Only mixtures in region II will give the desired products of pure ethanol as a bottom product.
Aqueous ethanol can be separated into their pure components by
distillation by the addition of a third component, so called the
entrainer, which forms a ternary heterogeneous azeotrope
with a lower than any other binary azeotropes.
Azeotropic Distillation Slide 7
How to Select Entrainer
In this case, pure ethanol cannot be obtained since
G is in region III.
0.0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0
0.0
.1
.2
.3
.4
.5
.6
.7
.8
.9
1.0
WaterWater
Ethanol
Ether
A
F
W
B
C
RD
V
P
G
A : 78.07 oC B : 75.08 oC
C : 73.44 oC D : 69.94 oCII
III
I
Azeotropic Distillation Slide 8
Homogeneous Azeotrope (I): EtOH-H2O
Mole Fraction of Ethanol
0.0 0.2 0.4 0.6 0.8 1.0
Tem
per
ature
( oC )
75
80
85
90
95
100
105
Azeotropic Distillation Slide 9
Homogeneous Azeotrope (II): EtOH-Benzene
Mole Fraction of Ethanol
0.0 0.2 0.4 0.6 0.8 1.0
Tem
pera
ture
( oC
)
66
68
70
72
74
76
78
80
82
Azeotropic Distillation Slide 11
Theory of Azeotropic Distillation (I)
Ethanol / Benzene Benzene / Water Ethanol / Water
Azeotropic Distillation Slide 12
Theory of Azeotropic Distillation (II)
Ethanol
H2O Benzene
Ethanol
Benzene
H2O
Highly
interaction
Water is
concentrated on
vapor side.
Dew point
temperature
valley
Azeotropic Distillation Slide 13
Theory of Extractive Distillation
Ethanol
EG
H2O
Highly
interaction
Ethanol is
concentrated on
vapor side.
Ethanol
H2O EG
Azeotropic Distillation Slide 14
Configuration of Azeotropic Distillation
Feed
Waste water
Concentrated ethanol
Recycle stream
Upper
phase
Lower phase
Pure ethanol Waste water
Concentrator
Azeo Column Stripper
Decanter
Azeotropic Distillation Slide 15
Example 1: Ethanol Dehydration
� Feed � Feed Composition
1) Ethanol : 60mole%
2) Water : 40mole%
� Feed Condition
1) Inlet Temperature : 25oC
2) Inlet Pressure : 3.5 bar
� Flow Rate
1) 100 Kg-mole/hr
� Entrainer: Benzene
� Cooling Medium: Mono-ethylene Glycol
1) In/Out Temperature : 10oC /18oC
2) Decanter Operating Temperature : 45oC
Azeotropic Distillation Slide 16
Simulation Procedure
� Step 1: Selection of Proper Thermo Model & Construction
of Binodal Curve
� Step 2: Finding Binary & Ternary Azeotropic Points
� Step 3: Construction of Binary & Ternary Residual Curves
� Step 4: Concentrator Simulation
� Step 5: Azeotropic Column Simulation
� Step 6: Azeo + Dryer(Stripper) Column Simulation
� Step 7: Simulation of Overall Azeotropic Distillation Unit
� Step 8: Optimization of Ethanol Dehydration Process
Azeotropic Distillation Slide 17
Step 1a: Selection of Proper Thermo Model
� NRTL. This model has up to 8 adjustable binary
parameters that can be fitted to data.
( )[ ]ijijijij
TG τβα +−= exp
−+=
∑∑
∑∑∑
∑
k
kkj
l
ljljl
ij
j
k
kkj
ijj
k
kki
j
jjiji
ixG
Gx
xG
Gx
xG
xG ττ
τγln
2T
c
T
ba
ijij
ijij ++=τ
Azeotropic Distillation Slide 18
Step 1b: Construction of Binodal Curve
� Binodal Curve Construction
using NRTL
1) Water
2) Ethanol
3) Benzene
� Temperature = 45oC
� Binary Interaction Parameters
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
WaterBenzene
Ethanol
I J A(I,J) A(J,I) αααα (i,j)
1 2 483.324 195.495 0.291
1 3 2990.025 758.494 0.200
2 3 -60.000 660.00 0.300
Azeotropic Distillation Slide 19
Step 2a: Finding Azeotrope (Ethanol + H2O)
M1 F1 M2
GUESS
VAP
L1
L2
LIQ
Stream Name Stream Description
Phase
Temperature Pressure
Flowrate
Composition H2O ETHANOL
C KG/CM2
KG-MOL/HR
VAP
Vapor
78.07 1.033
1.000
0.12 0.88
L1
Liquid
78.07 1.033
1.000
0.12 0.88
L2
Unknown
n/a n/a
n/a
n/a n/a
LIQ
Liquid
78.07 1.033
1.000
0.12 0.88
Azeotropic Distillation Slide 20
Result for Step 2
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
WaterBenzene
Ethanol
A
B
C
D
A: 78.07oC, B: 69.77oC
C: 69.31oC, D: 63.88oC
Entrainer Selection
Criteria 1
Azeotropic Distillation Slide 21
Step 3: Construction of Residual Curves
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
WaterBenzene
Ethanol
A
B
C
D
I
II
III
Only mixtures in region II will give the desired products of pure ethanol as a bottom product.
Entrainer Selection Criteria 2
Azeotropic Distillation Slide 22
Step 3: Feed having composition in region I
Feedstock
Composition Mole %
Water 10.00
Ethanol 10.00
Benzene 80.00
Total Flow 100.00
Top Product
Composition Mole %
Water 50.63
Ethanol 46.57
Benzene 2.80
Total Flow 19.76
Bottom Product
Composition Mole %
Water 0.00
Ethanol 1.00
Benzene 99.00
Total Flow 19.76
Azeotropic Distillation Slide 23
Step 3: Feed having composition in region II
Feedstock
Composition Mole %
Water 50.00
Ethanol 20.00
Benzene 30.00
Total Flow 100.00
Top Product
Composition Mole %
Water 20.10
Ethanol 31.59
Benzene 48.31
Total Flow 62.11
Bottom Product
Composition Mole %
Water 99.00
Ethanol 1.00
Benzene 0.00
Total Flow 37.89
Azeotropic Distillation Slide 24
Step 3: Feed having composition in region III
Feedstock
Composition Mole %
Water 10.00
Ethanol 70.00
Benzene 20.00
Total Flow 100.00
Top Product
Composition Mole %
Water 24.74
Ethanol 27.23
Benzene 48.03
Total Flow 40.41
Bottom Product
Composition Mole %
Water 0.00
Ethanol 99.00
Benzene 1.00
Total Flow 59.59
Azeotropic Distillation Slide 25
Step 4: Concentrator Simulation
Feed
Waste water
Concentrated ethanol
Concentrator
Basis: Feed = 100 Kg-mole/hr
xF = 0.60
Ethanol Mole Balance
F xF = D xD (Nearly Pure Water at
Column Bottom)
( ) ( )( )
18.6888.0
6.0100=
⋅=
⋅=
⋅=
azeo
F
D
F
x
xF
x
xFD
Azeotropic Distillation Slide 26
PRO/II Keyword Input for Step 4
UNIT OPERATIONS
COLUMN UID=CONC, NAME=Concentrator
PARA TRAY=25,CHEMDIST=150,DAMP=0.4
FEED 1,8
PRODUCT OVHD(M)=2,68.18, BTMS(M)=3
COND TYPE=TFIX, PRES=1.033, TEMP=45
DUTY 1,1/2,25
PSPEC PTOP=1.379, DPCOLUMN=0.21
ESTIMATE MODEL=CHEM
CESTIMATE(L) 1,0.1152,0.8848/25,1,0
SPEC RRATIO, PHASE=L, VALUE=3.0
SPEC STREAM=3,FRACTION, COMP=1,WET, VALUE=1
VARY DUTY=1,2
Azeotropic Distillation Slide 27
Distillation Algorithm Selection (I)
� Inside Out (I/O)
� Relatively Ideal Thermodynamics including Hydrocarbon with Water Decant
� Incorporates sidestrippers into column: No recycle!
� Thermosiphon Reboilers
� Flash Zone Model
� Very forgiving of bad initial estimates
� Fast!
� No VLLE
Azeotropic Distillation Slide 28
I/I Column Features
2 phase condenser +
water decant
N-1
2 1
Side Columns
Side Streams
Heater/Cooler
Heat Source/Sink
Multiple Feeds
Kettle and Thermosiphon
Reboilers
Pumparounds
N
Azeotropic Distillation Slide 29
Distillation Algorithm Selection (II)
� Chemdist
� Mechanically simple columns, complex thermo
� True VLLE
� Azeotropic and Reactive distillation
� Sidestrippers solved by recycle
� No Pumparounds or Thermosiphons
� More sensitive to bad initial estimates
Azeotropic Distillation Slide 30
Distillation Algorithm Selection (III)
� Sure
� Very general: Complex column and thermo
� Use when I/O and Chemdist do not apply
� Newton Method
� Very sensitive to bad initial estimates
Azeotropic Distillation Slide 31
Distillation Algorithm Selection (IV)
• Generality
• Total Pumparounds
• VLWE on any tray
• Water draw any tray
• Slow
• Sensitive to initial guesses
• Free water or water draw on trays other than condenser
• Total pumparounds or vapor bypass
Inside/Out (I/O) CHEMDIST
• Very fast
• Insensitive to initial estimates
• Thermo non-ideality
• NO VLLE capability (VLWE at condenser)
• Hydrocarbons
• EOS & Slightly non-ideal LACT Thermo
• Interlinked columns
• Side & main columns solved simultaneously
• Reactive Distillation
• VLLE on any tray
• Highly Non-Ideal Systems
• No Pumparounds
• Side columns solved as recycles
• Non-Ideal Systems
• Mechanically simple columns
• VLLE within column
Unique Features
Strengths
Limitations
Applicability
SURE
Azeotropic Distillation Slide 32
Specifications and Variables
� Specifications are constraints to be met by the column.
� Variables are calculated to meet specifications.
� Column always balances equations and unknowns.
� To impose a specification, you must add a variable, otherwise equations and unknowns don’t balance.
� Example: Impose two product specifications by declaring reboiler & condenser duties as variables.
Azeotropic Distillation Slide 33
Column Status at Initialization
� Fixed quantities remain at their current values unless you declare them as variables.
� If no specs/variables provided, default status used:
QUANTITY STATUS
Overhead and Bottoms Rates
Side Draw Rates
Duties
Feed Rates
Tray Temperatures
Tray Pressures
Vapor and Liquid Rates
Product Properties (e.g. Viscosity)
Tray Vapor or Liquid Properties
Calculated
Fixed
Fixed
Fixed
Calculated
Fixed
Calculated
Calculated
Calculated
Azeotropic Distillation Slide 34
Improper Specification
� 0% methane in crude column bottoms � Infinitely many solutions
� 300 lb-mole/hr propylene in overhead
� No solutions if column feed only 250 mol/hr propylene
� 98% ethanol product
� No solutions if Water-Ethanol Azeotrope present
Azeotropic Distillation Slide 35
You Can Help I/O or Chemdist using DAMP
� DAMP reduces iteration step and suppresses oscillation.
� Conventional Columns: DAMP = 1.0 (default)
� Columns with Steam: DAMP = 0.6 - 0.8 � Crude, Vacuum, FCC Main Fractionator
� Less-ideal: DAMP = 0.2 - 0.6 � N Increase number of allowed iterations
� If oscillations persist, use Chemdist
Azeotropic Distillation Slide 36
Condenser Options (I)
Overhead � Available in all algorithms
� Partial
� Vapor overhead product
� Liquid product is totally
refluxed to the column.
� Overhead vapor product is
a dew point state at the
overhead reflux operating
temperature.
Azeotropic Distillation Slide 37
Condenser Options (II)
Overhead
� Bubble
� Liquid product at bubble point
� Fixed temperature
� Sub-cooled at a specified
temperature
� Delta T below bubble point
� Specify degree of sub-cooling
Azeotropic Distillation Slide 38
Condenser Options (III)
Overhead
Side Draw
� Mixed
� Vapor product at dew point
� Liquid product at bubble point
Azeotropic Distillation Slide 39
Condenser Options (IV)
� Mixed with Decanter
� Vapor product at dew point
� Overhead vapor sub-cooled
and liquid phase split
� Upper phase refluxed or
produced as a column top
product
� Lower phase decanted
Overhead
Liquid 1
Liquid 2
Azeotropic Distillation Slide 40
Reboiler Models
� Most reboilers can be simulated as: � Kettle
� Thermosiphon with Baffles
� Thermosiphon without Baffles
Azeotropic Distillation Slide 41
Kettle Reboilers
N-1
Bottom Tray
N
Reboiler BTMS Q
VN-1 LN-2
VN LN-1
BTMS
LN-1
VN
Vapor in Equilibrium with Bottoms
Azeotropic Distillation Slide 42
Single Pass (Once Through) Thermosiphon
BTMS
LN-1
VN
Bottom Sump Reboiler
Sump
Baffle
LN-1
VN
BTMS
Bottom
Sump
� Equivalent to a Kettle Reboiler because Bottoms
is in Equilibrium with VN
Azeotropic Distillation Slide 43
Circulating Thermosiphon Adds 2 Stages
� Simulate as TS without Baffles.
N-2
Bottom Tray
N-1
Combined Sump
N
Reboiler
BTMS
Q
VN-1 LN-2
RL RF
R V
LN-2
VN-1 R V
RL
RF BTMS
Combined Sump
Azeotropic Distillation Slide 44
Circulating Thermosiphon
BTMS
VN-1
Bottom Sump
RF
RV
RL
LN-2
Reboiler Sump
N-2
Bottom Tray
N-1
Reboiler Sump
N
Reboiler
BTMS
Q
VN-1 LN-2
RL RF
R V
� Simulate as TS without Baffle.
Azeotropic Distillation Slide 45
Preferential Thermosiphon
BTMS R L
R F
L o
R V
N-2 Bottom Tray
N-1 Reboiler Sump
N Reboiler
Bottom Sump
Q
V N-1 L N-2
BTMS
L N-2
V N-1
Bottom Sump Reboiler
Sump
R F
R V
R L
L O
� Simulate as TS with Baffle.
Azeotropic Distillation Slide 46
Step 5a: Azeotropic Column Simulation
Azeo Column
F
F2 R
W
P
V
Composition of Stream F
Component Mole %
Ethanol 81.15
Water 18.85
Flow Rate 73.9 K-mol/hr
Azeotropic Distillation Slide 47
Step 5b: Decanter Simulation
V
R
W
Assume OVHD Vapor Composition, V around ternary azeotrope
Mole %
Benzene 53.00
Ethanol 31.00
Water 16.00
Azeotropic Distillation Slide 48
Step 5b: Decanter Simulation (Continued)
TITLE PROJ=AZEOTROPE, PROB=FLASH,USER=J.H.CHO
PRINT INPUT=ALL, RATE=M, FRACTION=M, PERCENT=M
DIMENSION METRIC
COMPONENT DATA
LIBID 1,BENZENE/2,ETHANOL/3,WATER
THERMODYNAMIC DATA
METHOD SYSTEM(VLLE)=NRTL, SET=NRTL01
STREAM DATA
PROP STRM=V, TEMP=70, PRES=1.033, RATE(M)=100, & COMPOSITION(M)=1,53/2,31/3,16
UNIT OPERATIONS
FLASH UID=COND, NAME=Condenser, KPRINT
FEED V
PRODUCT L=R, W=W
ISO TEMPERATURE=45, PRESSURE=1.033
END
V (Mole %) R (Mole %) W (Mole %)
Benzene 53.00 73.3072 3.1511
Ethanol 31.00 24.0964 47.9467
Water 16.00 2.5965 48.9022
Flow Rate 100 % 71.05 % 28.95 %
Azeotropic Distillation Slide 49
Step 5c: Component Mass Balance around
Azeo Column • Water balance around Azeotropic Column
( ) ( )8115.018115.019.73_ 2 −×+−×= FfeedAzeo
21885.093.13 F+=
( ) ( )2895.0489022.0 ××=VW
Azeo Column
F
F2 R
W
P
V
V⋅= 141572.0
21885.093.13141572.0 FV ⋅+=⋅ - (1)
• Ethanol balance around Azeotropic Column
( ) ( ) ( ) 28115.0479467.02895.0 FV ×=××VF ⋅= 171048.02 - (2)
• Combining Eq. (1) & (2):
4.127=V 79.212 =FK-mol/hr , K-mol/hr
• Benzene flow from the decanter
( ) ( ) ( ) 162.1031511.02895.04.127 =⋅⋅=K-mol/hr
Azeotropic Distillation Slide 50
Step 7: Simulation of Overall Azeotropic Unit
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
1
25
CONC
DCNT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
AZEO
CTRL
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
1
20
STRP
1
2
3
V
R
W
F2
P 7
Azeotropic Distillation Slide 51
Step 8: Optimization of Overall Process
Mole fraction Ethanol in Conc Top
0.70 0.75 0.80 0.85 0.90
MM
KJ / K
g-m
ole
140
160
180
200
220
Mole fraction Ethanol in Conc Top
0 .7 0 0.75 0 .8 0 0.85 0 .9 0
MM
KJ /
Kg-m
ole
35
40
45
50
Concentrator, azeotropic tower and stripper reboiler duties vs. conc. EtOH in feed
Concentrator
Azeotropic tower
Stripper
Mo le frac tion E tha no l in C o nc T op
0.70 0 .7 5 0.80 0.85 0.9 0
MM
KJ / K
g-m
ole
200
210
220
230
240
250
Azeotropic tower
Azeotropic Distillation Slide 52
Step 8: Objective Function
M o le fr a c t io n E th a n o l in C o n c T o p
0 .7 0 0 . 7 5 0 .8 0 0 .8 5 0 .9 0
To
tal re
bo
iler
du
ties M
M K
J /
Kg-m
ole
EtO
H p
rodu
ct
4 1 0
4 1 5
4 2 0
4 2 5
4 3 0
4 3 5
4 4 0
4 4 5
4 5 0
4 5 5
Azeotropic Distillation Slide 53
Conclusion
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
WaterBenzene
Ethanol
A
B
C
D
I
II
III
F
P
R V
W
G
Optimal concentration of ethanol in the feed
to the azeotropic tower is approximately
80.6 mole percent.
Azeotropic Distillation Slide 54
Future Works
� Comparison of Two-columns and Three-columns Configuration in Ethanol Dehydration Using Benzene as an Entrainer
� Application of an Environmentally Friendly Entrainer, NC5 to Obtain an Absolute Ethanol in Azeotropic Distillation
� Experimental Works for Ternary LLE Rather Than Each
Binary VLE’s & LLE
� Replacing a New Equation of State Approach with a LACT
Model
Azeotropic Distillation Slide 56
Performance Comparison between Three-
column & Two-columns Configuration
Three-columns Scheme Two-columns Scheme
Column Diameter (Valve Tray)
Concentrator: 1,370mm
Azeo Tower: 1,680mm
Stripper: 800mm
Azeo Tower: 1,500mm
Stripper: 1,670mm
Reboiler Duties 8.2865 MM Kcal/Hr 9.2203MM Kcal/Hr
Condenser Duties 8.4462 MM Kcal/Hr 9.0236MM Kcal/Hr
Steam Consumption 16.4 Ton/Hr 18.0 Ton/Hr
Cold Utility Consumption 1,027 Ton/Hr 1,127 Ton/Hr
Ethanol Purity 99.9493 wt% 99.9665 wt%
Other Impurities Benzene: 500 ppm (wt)
Water: 0.04 ppm (wt)
Benzene: 335 ppm (wt)
Water: 0.04 ppm (wt)
Azeotropic Distillation Slide 57
Performance Comparison between Benzene
& NC5 as an Entrainer
Benzene NC5
Top Press. 1.38 Kg/cm2 3.5 Kg/cm2
Entrainer Benzene NC5
RRATIO 651 Kgmol/hr 439 Kgmol/hr
Theo. Trays 25 25
Feed Tray 5 5
Reboiler Heat Duty
5.2268x106 Kcal/hr
2.6964x106 Kcal/hr
Condenser Heat Duty
3.3290x106 Kcal/hr
2.8552x106 Kcal/hr
Column
Diameter
1600 mm
(valve tray)
120 mm
(valve tray)
Impurities Water : 16ppm
Benz. : 12ppm
Water < 3ppm
NC5 < 1ppm
Ethanol
Recovery
99.99 mole % > 99.99 mole%
Stripping Column
Theo. Trays 25 25
Reboiler Heat Duty 2.2512x106 Kcal/hr
0.1543x106 Kcal/hr
Condenser
Heat Duties
2.0437x106 Kcal/hr
0.1369x106 Kcal/hr
Total Reboiler Duties
7.7580x106 Kcal/hr
0.1369x106 Kcal/hr
Total Condenser
Duties
5.3727x106 Kcal/hr
2.9912x106 Kcal/hr
Azeotropic Distillation Slide 58
Azeotropic Distillation
� Although the azeotropic distillation scheme, using NC5, operates at a higher pressure, comparative calculation results through this study show that azeotropic distillation using NC5 as an entrainer is better than that using benzene as an entrainer.
Azeotropic Distillation Slide 59
Experimental Works for Ternary LLE
Mole Percent of Benzene
0 10 20 30 40 50 60 70 80 90 100
Mo
le P
erc
ent
of E
tha
no
l
0
10
20
30
40
50
60
70
80
90
100
Mole Percent of Benzene
0 10 20 30 40 50 60 70 80 90 100
Mole
Pe
rcen
t of
Eth
an
ol
0
10
20
30
40
50
60
70
80
90
100
Azeotropic Distillation Slide 60
In Renon’s Original Paper
� The following sentence is partially wrong:
“The NRTL equation is applicable to calculate multi-component
VLE and LLE mixtures with only binary parameters.”
“We have to obtain the NRTL constants from the ternary LLE data
for the systems containing ternary components such as ethanol +
water + entrainer system.”
Azeotropic Distillation Slide 61
Comment for Azeotropic Simulation
� “The NRTL parameters obtained (regressed) from the
ternary LLE experimental data give a better result than the
NRTL parameters obtained from the binary experimental
data only. However, the above state holds true when you
simulate a system containing only three components in the ternary system. When you simulate a system containing
only ternary components system like ethanol-water-
benzene azeotropic unit, a better approach should be the
ternary LLE interaction parameters.”
- Recommended by Dr. Twu
Azeotropic Distillation Slide 62
Vapor Pressures using Improved Alpha Form
600 500
Vapor Pressure, Bar
400 300
ETHANOL
Temperature, K
WATER
BENZENE
Azeotropic Distillation Slide 63
Soave’s Original Alpha form
� Good representation of
liquid vapor pressure:
“Proper alpha form”
� Soave’s original alpha
form is wrong since it increases again as
reduced temperature of
hydrogen, Tr approaches
to infinity. Reduced Temperature of Hydrogen
0 20 40 60 80 100A
lpha V
alu
e for
Hyd
rog
en
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
( ) ( )( )[ ]22126992.054226.137464.01
riiTT −−++= ωωα (11)
Azeotropic Distillation Slide 64
Requirements for Alpha form
� Requirements for alpha form:
� The α function must be finite and positive for all
temperature. � The α function must equal unity at the critical point.
� The α function must approach a zero value as the temperature approaches infinity.
� The trend from now is to set the coefficients of
alpha function component dependently by
regressing the experimental vapor pressure data
vs. temperatures.
Azeotropic Distillation Slide 65
Several Alpha functions
Soave (1972)
Peng-Robinson (1980)
Soave (1979)
Boston-Mathias (1980)
Twu (1988)
Twu-Bluck-Cunningham (1990)
(Recommended by SimSci)
( )[ ]α = + −1 110 5
2
C Tt.
( )[ ]α = + −C C TtC
1 2
2
1 3
( )α = + − +
1 1 12T CC
Tr
r
( )[ ]α = −expC TrC
112
( ) ( )[ ]α = −−
T C Tr
C
rC2 1
1
22 21exp
( ) ( )[ ]α = −−
T C Tr
C C
rC C3 2 2 3
1
11exp
Azeotropic Distillation Slide 66
New Alpha Form
� Since 1972, many alpha forms have been
proposed, some better than others.
� PRO/II allows input of parameters for 11 different
forms, including the SIMSCI (TBCC) alpha form.
� This 3 parameter form eliminates the 2 problems
with the Soave alpha for defined components
( ) ( ) ( )[ ]iiii MN
ri
MN
ri TLTT −= − 1exp1α
Azeotropic Distillation Slide 67
Mixing Rules
� The accuracy of correlating vapor-liquid equilibrium data
using a cubic equation of state can be improved by
choosing an appropriate mixing rule for calculating a and b
for mixture.
� Expressions for mixing rules a and b are:
∑∑=i j
ijji axxa ∑=i
iibxb
Azeotropic Distillation Slide 68
New Mixing Rules
� The advanced mixing rules in Aspen Plus or PRO/II are:
� Quadratic (original SRK & PR)
� PR with Panagiotopolous mixing rule
� RKS with Panagiotopoulos mixing rule
� PR with Wong-Sandler mixing rule
� RKS with Won-Sandler mixing rule
� RKS with modified Huron-Vidal-2 mixing rule
� Kabadi and Danner (SRKKD)
� Modified Panagiotopolous & Reid (SRKM & PRM), 1990.
� Twu, Bluck, Cunningham & Coon (SRKS), 1992.
� These mixing rules are very important whenever the
system is highly non ideal
Azeotropic Distillation Slide 69
van der Waals mixing rule (1)
∑∑=i j
ijji axxa ∑=i
iibxb
( )ijjiijkaaa −= 1
Azeotropic Distillation Slide 70
Acetone - Water Using SRK
3000
2500
2000
Pressure, mm Hg
1500
1000
500
0.0 0.2 0.4
Mole Fraction Acetone
0.6 0.8 1.0
Azeotropic Distillation Slide 71
Panagiotopoulos mixing rule (2)
� In 1986 Panagiotopoulos and Reid proposed an
asymmetric mixing rule containing two parameters for SRK
and PR equations of state (denoted as SRKP and PRP).
� The Panagiotopoulos mixing rule
∑∑=i j
ijji axxa ∑=i
iibxb
( ) ( )[ ]ijiijijjiijxkkkaaa −+−= 1
Azeotropic Distillation Slide 72
Fatal Weaknesses of Panagiotopoulos
mixing rule
� The dilution of the mixture with additional components
(reducing all the mole fractions, xi) multiplies the effect of
the second binary parameter kji. In the limit of an infinite
number of components so that all the xi approach zero, the
mixing rules reduces to the original van der Waals mixing rule.
� The mixing rule is not invariant to dividing a component
into a number of identical pseudo-components. For
example, if methane in a mixture is divided arbitrarily into
“alpha” and “beta” methane, the calculated properties of the mixture will be slightly different.
Azeotropic Distillation Slide 73
Modified Panagiotopoulos mixing rule (3)
� In 1990, SimSci has modified Panagiotopoulos mixing rule
in a way that eliminate the first of the two flaws. This
mixing rule has four adjustable binary parameters such as
kij, kji, cij and cji.
� The modified Panagiotopoulos mixing rule
( ) ( )
+−+−=
ijc
ji
ijiijijjiij
xx
xkkkaaa 1
Azeotropic Distillation Slide 74
Huron-Vidal Mixing Rule (4)
� In 1991, Twu et al. proposed another modified mixing rule
that eliminated both of the inconsistencies of the
Panagiotopoulos-Reid mixing rule noted above slide. For a
binary system, the mixing rule can be expressed in the
following form for a12:
( )
++−=
2121
2
212121xGx
xGHkaaa ijjiij
( )122112 kkH −= ( )121212 exp HG β−=
Azeotropic Distillation Slide 75
Requirements for the New Mixing Rule
� Composition dependent mixing rule.
� Does not suffer from the invariant problem associated with
the Panagiotopoulos – Reid mixing rule:
� It is capable of correlating highly nonideal systems (i.e.
hydrocarbon / alcohol)
� It can be used for LLE systems as well.
Azeotropic Distillation Slide 76
Twu-Bluck-Cunningham-Coon Mixing Rule
++
−=
jiji
jijijij
jiijxGx
xGH
T
kaaa
2
1
T
kkH
ijji
ij
−=
( )ijijij HG β−= exp
� Binary form:
where
Azeotropic Distillation Slide 77
3000
2500
2000
Pressure, mm Hg
1500
1000
500
0.0 0.2 0.4
Mole Fraction Acetone
0.6 0.8 1.0
Acetone - Water Using SRKH
Azeotropic Distillation Slide 78
Solubility of Water in Benzene
20 16
Mole Percent of Water
12 8 4 0 0
40
80
120
160
200
120
Temperature, C
Azeotropic Distillation Slide 79
Solubility of Benzene in Water
0.06 0.40
Mole Percent of Benzene
0.20 0.00 0
40
80
160
200
120
Temperature, C
Azeotropic Distillation Slide 80
Ethanol - Benzene VLE at 45 C
1.0 0.8
Mole Percent of Ethanol
0.6 0.4 150
200
250
350
300
Pressure mmHg
0.2 0.0
Azeotropic Distillation Slide 81
Methanol - Heptane at 1 atm.
1.0 0.8
Mole Percent of Methanol
0.6 0.4 50
60
70
90
80
Temperature, C
0.2 0.0
100
110
Azeotropic Distillation Slide 83
Vapor Pressures using Improved Alpha Form
600 500
Vapor Pressure, Bar
400 300
ETHANOL
Temperature, K
WATER
BENZENE
Azeotropic Distillation Slide 84
Ethanol / Water System
Liquid Mole Fraction of Ethanol
Vapor
Mole
Fra
ction o
f E
thanol
Figure:
Ethanol-benzene system: comparison of predicted x-y
curve (line) with experimental
data (symbols) at 760 mmHg from Ellis, S.R.M. and Clark,
M.B. Chem. Age, India, 12, 377 (1961)
Azeotropic Distillation Slide 85
Ethanol / Benzene System
Liquid Mole Fraction of Ethanol
Vapor
Mole
Fra
ction o
f E
thanol
Figure:
Ethanol-benzene system: comparison of predicted x-y
curve (line) with experimental
data (symbols) at 760 mmHg from Ellis, S.R.M. and Clark,
M.B. Chem. Age, India, 12, 377 (1961)
Azeotropic Distillation Slide 86
Solubility of Water in Benzene
20 16
Mole Percent of Water
12 8 4 0 0
40
80
120
160
200
120
Temperature, C
Figure:
Benzene-water system: comparison of predicted water
solubility in benzene (curve) with
experimental data (symbols) from DECHEMA Vol. V. Part 1
(1979)
Azeotropic Distillation Slide 87
Solubility of Benzene in Water
Figure:
Benzene-water system: comparison of predicted
benzene solubility in water
(curve) with experimental data (symbols) from
DECHEMA Vol. V. Part 1 (1979)
0.06 0.40
Mole Percent of Benzene
0.20 0.00 0
40
80
160
200
120
Temperature, C
Azeotropic Distillation Slide 88
Prediction of Ternary System
� The predictions for the ternary system at 1 atm are given
on the next slide.
� Note that only pure and binary data fitting have been done.
No ternary data have been regressed.
� The same results, as predicted using the NRTL equation, are presented on the following slide.
� Note that the results are between EOS with a new mixing
rule and NRTL extremely similar.
Azeotropic Distillation Slide 89
Liquid Mole Fraction of Ethanol
Mole
Fra
ction o
f E
thanol
Figure:
Binodal curve, azoetropes, tie lines and residue curve
according to NRTL.
temperatures are reported at 101.325 kPa. Dashed
lines are experimental tie lines at 64oC.
Prediction of Ternary LLE (NRTL, I)
Azeotropic Distillation Slide 90
Liquid Mole Fraction of Ethanol
Mole
Fra
ction o
f E
thanol Figure:
Binodal curve, azoetropes, tie lines and residue curve
according to a new mixing
rule (binary parameters regressed from VLE data for
miscible binaries and LLE data for benzene-water).
Temperatures are reported at
101.325 kPa. Dashed lines Are experimental tie lines at
64oC.
Prediction of Ternary LLE (A New EOS, I)
Azeotropic Distillation Slide 91
Liquid Mole Fraction of Ethanol
Mole
Fra
ction o
f E
thanol
Figure:
Binodal curve, azeotropes, tie lines and residue curve
according to a new mixing
rule (ethanol-benzene binary parameters has been
adjusted to fit ternary LLE data). Temperatures are
reported at 101.325 kPa.
Dashed lines are experimental tie lines at
64oC.
Prediction of Ternary LLE (A New EOS, II)
Azeotropic Distillation Slide 92
Simulation Results
� The predictions of our simulation of this system are given on the
next three slides.
� The NRTL results are given, in addition to the results from a new
mixing rule with binary fitting only and also with a new mixing rule
after adjusting the ethanol / benzene binary interaction
parameters.
� The main difference between the results is that the EOS predicts
somewhat higher slopes for the tie-lines than NRTL predicts when
only binary fitted parameters are used. In actually, the tie0lines
are essentially horizontal.
� We have purposely simulated the main column to operate only in
the VLE range. This is faster from a simulation perspective but
should also be the more stable way to operate these columns.
Azeotropic Distillation Slide 94
Simulation Results (I)
(FEED COMPOSITION, MOLE %)
NRTL EOS (binary) EOS (adjusted)
Ethanol 85.00 85.00 85.00
Benzene 0.00 0.00 0.00
Water 15.00 15.00 15.00
(2nd COLUMN RETURN FEED, MOLE %)
NRTL EOS (binary) EOS (adjusted)
Ethanol 77.77 68.54 81.09
Benzene 7.23 16.46 3.91
Water 15.00 15.00 15.00
Azeotropic Distillation Slide 95
Simulation Results (II)
(BOTTOM COMPOSITION, MOLE %)
NRTL EOS (binary) EOS (adjusted)
Ethanol 99.87720 99.897656 99.96782
Benzene 0.00007 0.102340 0.03176
Water 0.12273 0.000004 0.00042
(OVERHEAD VAPOR COMPOSITION, MOLE %)
NRTL EOS (binary) EOS (adjusted)
Ethanol 38.53 37.26 39.91
Benzene 49.45 50.98 46.89
Water 12.02 11.76 13.20
Azeotropic Distillation Slide 96
Simulation Results (III)
(REFLUX COMPOSITION, MOLE %)
NRTL EOS (binary) EOS (adjusted)
Ethanol 36.32 28.22 39.31
Benzene 57.83 67.57 54.13
Water 5.85 4.21 6.56
(DISTILLATE COMPOSITION, MOLE %)
NRTL EOS (binary) EOS (adjusted)
Ethanol 50.32 57.51 43.58
Benzene 4.68 13.81 2.10
Water 45.00 28.68 54.32
Azeotropic Distillation Slide 97
Conclusions
� Advanced alpha functions and advanced mixing rules have
greatly expanded the applicability of the EOS methods to highly
non-ideal systems.
� The results of our simulation of this system using NRTL or the
equation of state approach with a new mixing rule and alpha
function are reasonably close.
� The results of either method could be improved by fitting VLLE
ternary data. We have demonstrated this with a new EOS.
� The EOS approach should be more versatile and should be
superior in multicomponent systems involving light gases and
wide ranges of T and P.
top related