Professional Documents
Culture Documents
02 Flash Distillation PDF
02 Flash Distillation PDF
Flash Distillation
1
After being passed through the valve/nozzle,
the feed enters the tank/drum, whose pressure
is low; thus, there is a substantial pressure drop
in the feed stream, causing the feed to partially
vaporise
2
Vapour
Feed yi,
zi, Tdrum,
T F, Pdrum,
PF, hV
Feed hF
Heater
zi, Valve
T 1, Pump
P1 QH
h1
Liquid
xi,
Tdrum,
Pdrum,
hL
3
The overall material (or mole) balance for
the system within the dashed boundary can be
written as follows
F = L +V (2.1)
4
Thus, Eq. 2.3 can be reduced to
FhF = LhL +VhV (2.4)
5
F =3
[note that F = degree of freedom, C =
number (#) of species, and P = number
(#) of phases]
After being fed into the tank, the feed is
divided into 2 phases (i.e. liquid and va-
pour phases), which results in the degree
of freedom of
F =C -P +2
= 2-2 +2
F =2
6
This means that
before the feed is being fed into the tank,
it requires 3 variables (e.g., z, T, and P)
to identify other properties of the system
(e.g., h, s)
after the feed is passed into the tank, it
requires only 2 variables (e.g., x i and Pdrum
or yi and Tdrum ) to obtain exact values of
other properties
7
2.1 Equilibrium data
8
yA yB
Tvapour Pvapour
Tliquid Pliquid
xA xB
9
Table 2.1 illustrates the equilibrium data for
various temperatures of the system containing
ethanol (EtOH: E ) and water (W ) at the pres-
sure of 101.3 kPa (or 1 atm)
10
Table 2.1: VLE data for an EtOH-water binary
mixture at 1 atm
T (oC) xE xW yE yW
11
The VLE data in Table 2.1 can also be pre-
sented graphically as
a y-x diagram (McCabe-Thiele diagram)
(Figure 2.3)
a Txy diagram (temperature-composition
diagram) (Figure 2.4)
an enthalpy-composition diagram
12
1.0
0.8
0.6
yEtOH
0.4
0.2
0.0
0.0 0.2 0.4 0.6 0.8 1.0
xEtOH
13
100
Azeotrope
95
90 Isotherm Vapour
T ( C)
Tx
o
Ty
85
80
Liquid
75
0.0 0.2 0.4 0.6 0.8 1.0
xEtOH or yEtOH
14
2.2 Binary Flash Distillation
Overall balance
F = L +V (2.1)
Species balance
z i F = x i L + yV
i
(2.2)
15
Let’s define
V
ºf: the fraction of the feed that
F
vaporises
L
ºq: the fraction of the feed that
F
remains liquid
L
Thus, the term in Eq. 2.6 can be re-writ-
V
ten, by combining with Eq. 2.1a, as
V
1-
L F -V F
= =
V V V
F
16
L 1- f
= (2.7)
V f
L
The term in Eq. 2.6 can be also re-written
V
as (this time, by combining with Eq. 2.1b)
L
L L F
= =
V F -L L
1-
F
17
L q
= (2.9)
V 1-q
F
Additionally, the term in Eq. 2.6 can be
V
re-written as (once again, by combining with
Eq. 2.1b)
F F 1
= =
V F -L L
1-
F
F 1
= (2.10)
V 1-q
18
Note that Eqs. 2.6, 2.8, and 2.11 are, in fact,
equivalent to one another, and they are the
“operating equations” for the flash tank, in
which
the slopes of Eqs. 2.6, 2.8, and 2.11 are
L 1- f q
- ,- , and - , respectively
V f 1-q
the Y-intercepts of Eqs. 2.6, 2.8, and
F zi æ 1 ö
2.11 are z i , , and çç ÷÷ z , respect-
V f çè1 - q ÷÷ø i
ively
19
The intersection of the equilibrium curve [see
the y-x diagram (Figure 2.3) on Page 13] and
the operating line is the solution (answer) of the
material balances (this plot is called “McCabe-
Thiele diagram”) for the flash distillation, as
the intersection of the equilibrium line (curve)
and the operating line is the point where the
system (i.e. the flash tank) reaches the equili-
brium
æV + L ö
çç ÷÷ y = F z (2.12)
çè V ÷÷ø i V i
Therefore, x i = yi = z i
21
Consider the McCabe-Thiele diagram of the
V
EtOH-water system for the case where =f
F
2
= and z EtOH = 0.40
3
= 0.60
22
Equilibrium curve
1.0
Solution (answer)
0.8 xE = 0.17; yE = 0.53
0.6
yEtOH
xE = yE = zE = 0.40
0.4
0.2
y = x line Operating line
0.0
0.0 0.2 0.4 0.6 0.8 1.0
xEtOH
23
z i F = x i F - x iV + yV
i
z i F - x i F = -x iV + yV
i
F (z i - x i ) = V (yi - x i )
V (zi - x i )
=f = (2.14)
F (yi - x i )
24
In the case that an analytical solution is
desired (in stead of the graphical solution—as
illustrated in Figure 2.5), an equilibrium curve
must be translated into the form of equation
yA (1 - x A )
or aAB = (2.17b)
x A (1 - yA )
results in
yi Pi *
= (2.19)
xi P
26
Substituting Eq. 2.19 into Eq. 2.16 for species
A and B yields
PA*
*
P
aAB = P* = A* (2.20)
PB PB
P
27
Re-arranging Eq. 2.18c for yi yields
xia
yi = (2.21)
é1 + (a - 1) x ù
êë i úû
28
xia 1- f zi
=- xi +
é1 + (a - 1) x ù f f
êë i úû
(2.22)
êë úû
(2.23)
29
2.3 Multi-component VLE
30
For light hydrocarbons, the value of K i of
each species can be obtained from the graph
(called the “K chart” – Figures 2.6 and 2.7)
prepared by DePriester when the temperature
and pressure of the system are specified
T2 T 6 1
P2 P
(2.25)
in which the values of aT , aT , aT , aP , aP , and
1 2 6 1 2
32
Figure 2.7: The K chart (high temperature range)
[from Introduction to Chemical Engineering Thermodynamics (7th ed)
by J.M. Smith, H.C. Van Ness, and M.M. Abbott]
33
The meaning of K i on the K chart can be
described mathematically by the original Raoult’s
law and the modified Raoult’s law as follows
Re-arranging 2.24:
yi = K i x i (2.24)
results in
yi
= Ki (2.26)
xi
34
In the case of non-ideal systems, Raoult’s
law is modified to
yi P = x i gi Pi * (2.28)
where gi is an activity coefficient of species i,
which can be calculated using, e.g., Margules or
Van Laar equations
35
2.4 Multi-component Flash Distillation
36
Re-arranging Eq. 2.32 for x i gives
ziF
xi = (2.33)
F + (K i - 1)V
37
Substituting Eq. 2.35 into Eq. 2.34 and re-
arranging the resulting equation yields
yi zi
=
Ki V
1 + (K i - 1)
F
ziKi
yi = (2.36)
V
1 + (K i - 1)
F
åx i
= 1 and åy i
= 1, we obtain the following
relationships (by tanking summation to Eqs.
2.34 and 2.36, respectively):
38
æ ö
çç ÷÷
çç zi ÷÷
å çç ÷÷ = 1 (2.37)
çç1 + (K - 1)V ÷÷÷
çè i
Fø
æ ö
çç ÷÷
çç ziKi ÷÷
and å çç ÷÷ = 1 (2.38)
çç1 + (K - 1)V ÷÷÷
çè i
Fø
39
V
Hence, if the first guess of is not good
F
enough (i.e. too far from the real answer), the
correct solution may not obtained
Since both åx i
and åy i
are equal to
unity (1), subtracting Eq. 2.37 from Eq. 2.38
gives
æ ö æ ö
çç ÷
÷ çç ÷÷
çç ziKi ÷÷ çç zi ÷÷
å çç ÷÷ - å ç
çç
÷÷ = 0
çç1 + (K - 1)V ÷÷÷ ç
V ÷÷
1 + (K i - 1) ÷
çè i
Fø ç
è Fø
(2.39)
40
Let’s define the left hand side (LHS) of Eq.
2.39 as
æ ö æ ö
çç ÷÷ çç ÷÷
çç ziKi ÷÷ çç zi ÷÷
g = åç ÷÷ - å ç ÷÷
çç V ÷÷ çç V ÷÷
ççè1 + (K i - 1) ø÷ èçç1 + (K i - 1) ø÷
F F
which can be re-arranged to
æ ö
çç ÷÷
çç (K i - 1) z i ÷÷
g = åç ÷÷ (2.40)
çç V ÷÷
ççè1 + (K i - 1) ÷ø
F
V
Note that g is a function of or
F
æV ö
g = g çç ÷÷÷
çè F ÷ø
g=0
41
We can employ a numerical method, such as
a “Newtonian convergence procedure” to solve
for the solution as follows
dgk
Note that is the value of the derivative
æV ö
d çç ÷÷÷
çè F ÷ø
æV ö -gk
D ç ÷÷÷ =
ç (2.43)
çè F ÷ø dgk
d (V / F )
43
Since
æV ö æV ö æV ö
D çç ÷÷÷ = çç ÷÷÷ - çç ÷÷÷
çè F ø÷ èç F ø÷ èç F ÷ø
k +1 k
dgk
By employing Eq. 2.40, the term
d (V / F )
in Eq. 2.44 can be derived as
(Ki - 1) zi
2
dgk
= -å
æV ö é ù
2
d çç ÷÷÷ ê1 + (K i - 1)
Vú
çè F ø÷ ê F úû
ë
(2.45)
44
V
Hence, the next guess of , which may ena-
F
ble gk +1 = 0 , and, thus, makes å x i = 1 and
åy i
= 1 can be obtained from the following
equation:
æV ö æV ö gk
çç ÷÷ ç
= ç ÷÷ -÷ (2.46)
çè F ø÷÷ èç F ø÷k éê dg ù
k +1 k ú
ê ú
êë (
d V / F )úû
or
æ ö
çç ÷÷
çç (K i - 1) z i ÷÷
å çç ÷÷
çç1 + (K - 1)V ÷÷÷
æ ö æ ö è Fø
ççV ÷÷ = ççV ÷÷ +
i
çè F ø÷÷ ç F ÷÷ø
(K - 1) z i
2
k +1
è k
åé i
ù
2
V
ê1 + (K i - 1) ú
ê F úû
ë
(2.47)
45
Example A flash chamber operating at 50 oC
(122 oF) and 200 kPa (29 psia) is separating
1,000 kmol/h of feed containing 30 mol% pro-
pane, 10% n-butane, and 15% n-pentane, and
45% n-hexane. Find the product compositions
and flow rates
V, yi
46
At T = 50 oC (122 oF) and P = 200 kPa (29
psia), the value of K i read from the K chart of
each species is as follows
K1 = 7.0
K 2 = 2.4
K 3 = 0.80
K 4 = 0.30
V
Let’s start with = 0.1, and the value of
F
é ù
ê K -1 z ú
æV ö
ç ÷ ê ( i )i ú
g ç ÷÷ = å ê ú can be calculated as
çè F ÷ø ê Vú
ê
ë
1 + (K i
- 1) F
ú
û
follows
47
(7.0 - 1)(0.30) (2.4 - 1)(0.10)
g (0.1) = +
1 + (7.0 - 1)(0.1) 1 + (2.4 - 1)(0.1)
(0.80 - 1)(0.15) (0.30 - 1)(0.45)
+ +
1 + (0.80 - 1)(0.1) 1 + (0.30 - 1)(0.1)
g (0.1) = 0.88
(Ki - 1) zi
2
dgk
= -å yields
æV ö é ù
2
ç ÷ V
d ç ÷÷
çè F ø÷ ê1 + (K i - 1) ú
ê F úû
ë
ìï ü
ï
ïï (7.0 - 1) (0.30) (2.4 - 1) (0.10)
2 2
ï
ï
ïï + ï
2 2 ï
dgk ïï êë1 + (7.0 - 1)(0.1)úû
ï é ù é
êë1 + (2.4 - 1)( 0.1 ) ù
úû
ï
ï
ï
= -í ý
æV ö ï ï
( ) ( ) ( ) ( )
2 2
ç
d ç ÷÷ ÷ ï 0.80 - 1 0.15 0.30 - 1 0.45 ï
çè F ÷ø ï
ïï+ + ï
ï
2 2ï
ïï é1 + (0.80 - 1)(0.1)ù é1 + (0.30 - 1)(0.1)ù ï
îï ëê ú
û êë úû ïþï
dgk
= -4.63
æV ö
d çç ÷÷÷
çè F ÷ø
48
Then, using Eq. 2.46 (or Eq. 2.47) to com-
V
pute the value of the next guess of gives
F
æV ö æV ö gk
çç ÷ = çç ÷÷ -
÷
çè F ø÷÷ ç F ø÷÷
è é dg ù
k +1 k ê k ú
ê ú
êë (
d V / F )úû
æV ö
çç ÷÷ = (0.1) - 0.88
çè F ÷÷ø (-4.63)
k +1
æV ö
çç ÷÷ = 0.29
çè F ÷÷ø
k +1
V
The corresponding value of that makes
F
g = 0 for this Example is found to be 0.51
49
V
After obtaining the correct value of , the
F
values of x i and yi of each species can be calcu-
lated as follows
50
Then, the value of yi can be computed using
Eq. 2.24:
yi = K i x i (2.24)
V
Since = 0.51 and F = 1,000 kmol/h,
F
V = 510 kmol/h , and L = 490 kmol/h
51
Thus, the flow rates of each species in the li-
quid and gas phases can be computed as follows
Liquid phase (li = x i L ):
Propane: (0.074)(490) = 36.26 kmol
n-butane: (0.058)(490) = 28.42 kmol
n-pentane: (0.167 )(490) = 81.83 kmol
n-hexane: (0.700)(490) = 343.0 kmol
TOTAL = 489.5 kmol
52
Another approach to solve this Example is as
already studied from ChE Thermodynamics II;
the details are as follows
ziF
xiL = (2.48)
æ ö
çç1 + K V ÷÷
çè i
L ø÷÷
53
For the feed to become liquid, åx i
=1
Thus,
ziF
å x L = Lå x = L (1) = L = å
i i æ V ö
÷
ççç1 + K i ÷÷÷
è Lø
(2.49)
54
z i K i F = yi L + yi K iV
= yi (L + K iV )
æL ö
z i K i F = yV çç + K ÷÷
çèV i÷
i
÷ø
ziKiF
yV = (2.50)
i æL ö
çç + K ÷÷
çèV i÷÷ø
Hence,
ziKiF
å yV = V å y = V (1) = V = å
i i æL ö
çç + K ÷÷
çèV i÷
ø÷
(2.51)
55
To solve for the value of L , the trial & error
technique must be employed such that the value
of L is guessed, and the corresponding value of
V is obtained from the fact that V = F - L
56
The value of the guessed V and the corres-
ponding value of L are then used to calculate
the value of V using Eq. 2.51
57
The value of L can then be computed using
Eq. 2.49 as illustrated in the following Table:
zi F
li =
Species zi Ki æ ö
çç1 + K V ÷÷
çè i
L ÷÷ø
(0.30)(1, 000)
= 4.7
Propane 0.30 7.0 é æ 900 ö÷ù
ê1 + (7.0)çç ÷÷úú
ê çè 100 ÷ø
ë û
n-Butane 0.10 2.4 4.4
n-Pentane 0.15 0.80 18.3
n-Hexane 0.45 0.30 121.6
∑= 1.00 L= 149.0
58
The iteration is performed until the guessed
value of L is equal to the value of L computed
using Eq. 2.49, and the resulting value of L is
found to be 488.6 kmol (and the corresponding
value of V is 1,000 – 488.6 = 511.4 kmol), which
yields
zi F
li =
Species zi Ki æ ö
çç1 + K V ÷÷
çè i
L ÷÷ø
(0.30)(1, 000)
= 36.0
Propane 0.30 7.0 é æ öù
ê1 + (7.0)çç 511.4 ÷÷ú
ê çè 488.6 ÷ø÷ú
ë û
n-Butane 0.10 2.4 28.5
n-Pentane 0.15 0.80 81.6
n-Hexane 0.45 0.30 342.5
∑= 1.00 L= 488.6
59
The composition of each species in each phase
can, thus, be calculated as follows
li
Species zi fi = ziF li xi =
L
36.0
Propane 0.30 300 36.0 = 0.074
488.6
n-Butane 0.10 100 28.5 0.058
n-Pentane 0.15 150 81.6 0.167
n-Hexane 0.45 450 342.5 0.701
∑= 1.00 1,000 488.6 1.001
vi
Species fi = ziF li vi = fi - li yi =
V
264.0
Propane 300 36.0 300 – 36.0 = 264.0 = 0.516
511.4
n-Butane 100 28.5 71.5 0.140
n-Pentane 150 81.6 68.4 0.134
n-Hexane 450 342.5 107.5 0.210
∑= 1,000 488.6 511.4 1.000
60