4
Flash drum
V, yi, hv
PV, TV
Heater
Liquid
feed
(a)
V, yi, hV
PV, TV
Partial
condenser
(1) PV PL
(2) TV TL
(3) yi Kixi
(4) Fzi Vyi Lxi
(5) F V L
(6) hFF Q = hVV hLL
P
P
(7) i yi " i xi 0
F, zi
hF
TF, PF
Number of
Equations
Equation
Valve
PL, TL
L, xi, hL
Vapor
feed
F, zi
hF
TF, PF
Flash drum
147
PL, TL
L, xi, hL
Ki Ki{TV, PV, y, x}
hV hV{TV, PV, y}
(mechanical
equilibrium)
(thermal equilibrium)
(phase equilibrium)
(component material
balance)
(total material balance)
(energy balance)
(summations)
1
1
C
C
1
1
1
E 2C 5
hF hF {TF, PF, z}
hL hL {TL, PL, x}
(b)
Isothermal flash
Bubble-point temperature
Dew-point temperature
Bubble-point pressure
Dew-point pressure
Adiabatic flash
Nonadiabatic flash
Percent vaporization flash
CK i " 1
i1
for C V/F, where Ki = Ki{TV, PV}.
(4) V FC
zi
(5) xi
1 CK i " 1
zi K i
(6) yi
xi K i
1 CK i " 1
(7) L F " V
(8) Q hVV hLL " hFF
148
Chapter 4
0.8
f {}
0.6
0.4
0.2
0.0
0.2
0.4
EXAMPLE 4.1
0
0.2
0.4
0.6
V
=
F
0.8
A 100-kmol/h feed consisting of 10, 20, 30, and 40 mol% of propane (3), n-butane (4), n-pentane (5), and n-hexane (6), respectively,
enters a distillation column at 100 psia (689.5 kPa) and 200+ F
(366.5+ K). Assuming equilibrium, what fraction of the feed enters
as liquid, and what are the liquid and vapor compositions?
1.0
by substituting Eq. (5) into Eq. (4) to eliminate L and combining the result with Eq. (3) to obtain Eqs. (5) and (6) in
Table 4.4. Here (5) is in xi but not yi, and (6) is in yi but not
x
Pi. Summing
P these two equations and combining them with
yi
xi 0 to eliminate yi and xi gives Eq. (3) in Table
4.4, a nonlinear equation in V (or C V/F) only. Upon solving this equation numerically in an iterative manner for C
and then V, from Eq. (4), the remaining unknowns are
obtained directly from Eqs. (5) through (8) in Table 4.4.
When TF and/or PF are not specified, Eq. (6) of Table 4.3 is
not solved for Q.
Equation (3) of Table 4.4 can be solved iteratively by
guessing values of C between 0 and 1 until the function
f{C} 0. A typical function, encountered in Example 4.1, is
shown in Figure 4.10. The most widely employed procedure
for solving Eq. (3) of Table 4.4 is Newtons method [6]. A
value of the C root for iteration k 1 is computed by the
recursive relation
Ck1 Ck
f fC g
0
f fCk g
4-10
Solution
At flash conditions, from Figure 2.4, K3 4.2, K4 1.75, K5
0.74, K6 0.34, independent of compositions. Because some
K-values are greater than 1 and some less than 1, it is necessary first
to compute values of f{0} and f{1} for Eq. (3) in Table 4.4 to see if
the mixture is between the bubble and dew points.
C
X
i1
zi 1
1
0:31
1(2
1:75
1
0:74 0:41 0:34
1
1
0:21
0:128
0:11 4:2
0:21 1:75
1 4:2 1 1 1:75 1
f f1g
0:31 0:74
0:41 0:34
0:720
1 0:74 1 1 0:34 1
Since f{1} is not less than zero, the mixture is below the dew point.
Therefore, the mixture is part vapor. Using the RachfordRice procedure and substituting zi and Ki values into Eq. (3) of Table 4.4
gives
0
4-11
The iteration can be initiated by assuming C(1) 0.5. Sufficient accuracy is achieved by terminating the iterations when
jCk1 Ck j=Ck < 0:0001.
The existence of a valid root (0 * C * 1) must be checked
before employing the procedure of Table 4.4, by checking if
the equilibrium condition corresponds to subcooled liquid or
superheated vapor rather than partial vaporization or condensation. A first estimate of whether a multicomponent feed
gives a two-phase mixture is made by inspecting the Kvalues. If all K-values are > 1, the phase is superheated
vapor. If all K-values are < 1, the single phase is a subcooled
liquid. If one or more K-values are greater than 1 and one or
more K-values are less than 1, the check is made by first
4:2
Since f{0} is not more than zero, the mixture is above the bubble
point. Now compute f{1}:
0:11 4:2
0:21 1:75
1 C4:2 1 1 C1:75 1
K i 2
1 Ck K i
0:11
f f0g
0:31 0:74
0:41 0:34
1 C0:74 1 1 C0:34 1
0.5000
0.0982
0.1211
0.1219
0.2515
0.0209
0.0007
0.0000
0.6259
0.9111
0.8539
0.8521
0.0982
0.1211
0.1219
0.1219
0.8037
0.2335
0.0065
0.0000
4.4
x
0.0719
0.1833
0.3098
0.4350
1.0000
Propane
n-Butane
n-Pentane
n-Hexane
Sum
149
y
0.3021
0.3207
0.2293
0.1479
1.0000
EXAMPLE 4.2
P
However, zi 1. Therefore, the bubble-point equation is
P
zi K i 1
4-12
i
$
zi 0
f f 1g
Ki
Ki
i
Therefore, the dew-point equation is
P zi
1
i Ki
4-13
C
X
zi Psi
4-14
i1
C
X
i1
gi zi Psi
4-16
Bubble-Point Temperature.
In Figure 1.9, the nC4-rich bottoms product from Column C3 has the
composition given in Table 1.5. If the pressure at the bottom of the
distillation column is 100 psia (689 kPa), estimate the mixture
temperature.
Solution
The bottoms product is a liquid at its bubble point with the following composition:
kmol/h
zi xi
8.60
0.0319
215.80
0.7992
28.10
0.1041
17.50
0.0648
270.00
1.0000
The bubble-point temperature can be estimated by finding the temperature that will satisfy (4-12), using K-values from Figure 2.4. Because the bottoms product is rich in nC4, assume the K-value of nC4
is 1. From Figure 2.4, for 100 psia, T 150& F. For this temperature,
using Figure 2.4 to obtain K-values of the other three components
and substituting these values and the z-values into (4-12),
P
zi K i 0:03191:3 0:79921:0 0:10410:47
0:06480:38
0:042 0:799 0:049 0:025 0:915
Component
i-Butane
n-Butane
i-Pentane
n-Pentane
The sum is not 1.0, so another temperature is assumed and the summation repeated. To increase the sum, the K-values must increase
and, thus, the temperature must increase as well. Because the sum is
dominated by nC4, assume its K-value 1.09. This corresponds to a
temperature of 160& F, which results in a summation of 1.01. By linear interpolation, T 159& F.
EXAMPLE 4.3
Bubble-Point Pressure.
Cyclopentane is separated from cyclohexane by liquidliquid extraction with methanol at 25& C. To prevent vaporization, the mixture
must be above the bubble-point pressure. Calculate that pressure
using the following compositions, activity coefficients, and vapor
pressures:
Methanol
Vapor pressure, psia
Methanol-rich layer:
x
g
Cyclohexane-rich layer:
x
g
Cyclohexane
Cyclopentane
2.45
1.89
6.14
0.7615
1.118
0.1499
4.773
0.0886
3.467
0.1737
4.901
0.5402
1.324
0.2861
1.074
150
Chapter 4
Solution
Assume the modified Raoults law in the form of (4-16) applies for
either liquid phase. If the methanol-rich-layer data are used:
Pbubble 1:1180:76152:45 4:7730:14991:89
3:4670:08866:14
5:32 psia 36:7 kPa
A similar calculation based on the cyclohexane-rich layer gives an
1 1
identical result because the data are consistent; thus giL xi
2 2
giL xi . A pressure higher than 5.32 psia will prevent formation of
vapor at this location in the extraction process. Operation at atmospheric pressure is viable.
EXAMPLE 4.4
Vapor Distillate
Liquid Reflux
0.90
0.10
1.00
0.76
0.24
1.00
Propylene
1-Butene
10
Legend
K-value
Propylene
1
1-Butene
f fT V g
0.1
60
80
100
120
140
160
Pressure, psia
180
200
$
Solution
The column pressure is at the dew point for the vapor distillate. The
reflux composition is that of a liquid in equilibrium with the vapor
distillate at its dew point. The method of false position [8] is used to
perform the calculations by rewriting (4-13) in the form
f fPg
C
X
zi
'1
K
i
i1
4-17
4.5
C
X
i1
zi 1 $ K i
0
1 CK i $ 1
4-18
Then, Eqs. (5) and (6) of Table 4.4 are solved for x and y.
Equation (4-17) is then solved directly for C, since
f fCg
ChV 1 $ ChL $ hF
0
1; 000
from which
C
4-19
h F $ hL
h V $ hL
4-20
Solution
This problem, involving a wide-boiling feed, is best solved by using a
process simulator. Using the CHEMCAD program with K-values and
enthalpies from the PR equation of state (Table 2.5), the result is:
151
XB FA XB S XB FA
4-21
and the distribution of solute at equilibrium is given by
E
4-22
X B KDB X B
where K 0DB is the distribution or partition coefficient in terms
of mass or mole ratios (instead of mass or mole fractions).
E
Substituting (4-22) into (4-21) to eliminate X B ,
R
XB
XB FA
F A K 0DB S
4-23
kmol/h
Component
Feed 120' F
485 psia
Vapor 112' F
165 psia
Liquid 112' F
165 psia
Hydrogen
Methane
Benzene
Toluene
Total
Enthalpy, kJ/h
1.0
27.9
345.1
113.4
487.4
$1,089,000
0.7
15.2
0.4
0.04
16.34
362,000
0.3
12.7
344.7
113.36
471.06
$1,451,000
EB K 0DB S=F A
4-24
Large extraction factors result from large distribution coefficients or large ratios of solvent to carrier. Substituting (4-24)
into (4-23) gives the fraction of B not extracted as
R
X B =X B
1
1 EB
4-25
Thus, the larger the extraction factor, the smaller the fraction
of B not extracted or the larger the fraction of B extracted.
Alternatively, the fraction of B extracted is 1 minus (4-25) or
EB/(1 EB).
Mass (mole) ratios, X, are related to mass (mole) fractions,
x, by
X i xi =1 $ xi
4-26