Chapter 14.

Topics in Phase Equilibrium

The simplest models for vapor/liquid equilibrium - Raoults law - Henrys law

y i P = x i Pisat

yi P = x i Hi

=> for a species whose Tc <= T of applicants

- Modified Raoults law

y i P = x i i Pisat => Applied only to low pressure

14.1. The / formulation of VLE.

Extension of modified Raoults Law Overcome the limitation of modified Raoults Law Introduce vapor-phase fugacity to explain nonideality

From eqn (11.52) From eqn (11.90) at equilibrium

v = y f i i iP l = x f f
i i i i

=> Vapor phase => Liquid phase

v = f l f i i

i P = x i ifi yi

From eqn (11.44)

Vil (P - Pisat ) f i = P exp[ ] RT

sat sat i i

l sat V ( P P ) i i sat sat y i i P = x i i i Pi exp[ ] RT

i Vil ( P - Pisat ) y i sat exp[]P = x i i Pisat i RT

y i P = x i i Pisat
Usually poynting factor 1

(14.1)

=> gamma/phi formulation

i i sat i

(14.2)

If If

i = i = 1 i = 1

=> (14.1) becomes raoults Law => (14.1) becomes modified Raoults Law

 To use eqn(14.1) for the analysis of VLE => need to know the value of 1)

Pisat , i , i

sat i

=> obtained by the Antoine eqn

ln P
2)

sat i

Bi = Ai T + Ci

(14.3)

i and sat i => need to obtain i

= exp

P 1 [Bii + y i y k (2 ji - jk )] RT 2 j k

(14.4)

Values of the pure species vivial coeff. of the species in the mixture (11.69) ~ (11.74)

sat i : fugacity coeff for pure i as a saturated vapor in eqn(14.4) ji , jk = 0

sat B P ii i sat i = exp RT 1 sat Bii (P - Pi ) + P y y (2 ji - jk ) 2 j k j k

i i = sat = exp i

(14.6)

RT
(14.5)

For a binary system comprised of species 1 and 2, this becomes

B11 (P - P1sat ) + Py 2 2 12 1 = exp RT

2 B22 (P - P2sat ) + Py1 12 2 = exp RT

(14.7a)

(14.7b)

3) i : can be obtained from the various models ( Van Laar, Wilson, UNIFAC, etc ..)

 Dew point and Bubble point Calculation using Gamma/Phi formulations formulations => need iteration

i = (T, P, y1 , y 2 , i = (T, x 1 , x 2 ,
Pisat = f (T)
x i i Pisat yi = i P
since

, y N -1 )

, x N-1 )

(14.8)

yi i P xi = i Pisat

(14.9)

= 1, y i = 1
(14.10)

x i i Pisat P= i i
1) BUBL P calculation

1 P= yii P sat i i i

(14.11)

=> Calculation (yi) and P , given xi and T

Bubble P calculations

Fig. 14.1 Block diagram for the calculation BUBL P

2) Dew P calculation => Calculation xi and P , given yi and T - BUBL P and Dew P calculation : temperature is given => can calculate

Pisat

- BUBL T and Dew T : T is unknown => need initial estimate for T to do iteration i)

T= x i Tisat
i

(for BUBL T)

T= y i Tisat
i

(for Dew T)

where

Tisat

B = -C A i - lnP i

(14.12)

ii) Multiply eqn(14.10) (14.11) by (inside summation)

Pisat

(outside summation) and didice by

sat y i i Pj = P ( sat ) i Pi i

Pisat
(14.14)

sat j

(x
i

i i

/ i )(P / P )

sat i

sat j

(14.13)

Pjsat

iii) Calculate corresponding T from eqn(14.3)

T=

Bj A j - lnPjsat

- Cj

Dew P calculations

Fig. 14.2 Block diagram for the calculation DEW P.

BUBL T calculation
=> Calculate (yi, T) given xi and P

Fig. 14.3 Block diagram for the calculation BULB T.

DEW T calculation
=> Calculate (xi , T) given yi and P

Fig. 14.4 Block diagram for the calculation DEW T.

Flash calculation
In chap 10. Flash calculation is based on Raoults law and K-value correlation => Use the gamma/phi formulation

yi =

zi K i 1 + (K i - 1) zi xi = 1 + (K i - 1)

(i = 1,2,
(i = 1,2,

, N)
, N)

(10.16)

First,

y i = 1, y i - 1 = 0 = Fy
zi K i Fy = -1 = 0 1 + ( K 1) i i
(14.17)

Second,

x i = 1, x i - 1 = 0 = Fx
zi Fx = -1 = 0 1 + ( K 1) i i
(14.18)

14.2 VLE from Cubic Equation of State

For VLE,

Since,

v = f l (11.48) f i i f iv = x i P il y i iv = x i il i = i , y i P xiP

Vapor pressure for a pure species

- Usually -

Pisat , can be obtained by experimental measurement

Pisat at equilibrium can be obtained using EOS

Generic Cubic Equation of state

RT a (T ) P= v - b ( v + b)(v + b)
If P -> , V -> b , if P -> 0, V ->

Fig. 14-7

- For pure species, and

ln iv = ln il ln il - ln iv = 0
P P sat
in that case where

i = i (T, P )

- For a saturated liquid or vapor

i = i (T, Pisat )

Pisat = f (T)

- Two widely used cubic eqn of state Sorve / Redlich / kwung (SPK) eqn Peng/Robinson (PR) eqn -

ZRT Vapor & vapor-like Roots of the Generic Cubic EOS (use v = ) P Zi - i (3.52) -> (14.31) Z i = 1 + i - q i i ( Zi - i ) + ( Zi + i )
Liquid & liquid like Roots of the Generic Cubic EOS

1 + i - Zi Zi = i + ( Zi + i )( Zi + i ) q i i
Parameter : eqn (3.45), (3.46), (3.50), (3.51)

(3.56) -> (14.35)

(a i , b i , i , q i )

Pure number : From (11.37)

, , , , (Tri )

=> Table 3.1

(11.37)

ln i = Zi - 1 - ln(Zi - i ) - q i I i

- Procedure to obtain 1) Calculate

Pisat for pure for a give T, P

Zil - dZiv using eqn (14.35)(14.36) before that, calculate parameters

l v

2) Using eqn(11.37), calculate ln i , ln i before that, calculate qi, Ii 3) Check if ln i - ln i = 0 If yes, P -> Psat, if no, iteration Eqn for calculating vapor pressure -> table 14.3
l v

14.3 Equilibrium and Stability

If system surrounding => thermal and mechemical equilibrium heat exchange (Q) and expansion work (w) are reversible. Under these circumstance

dSsurr =

dQsurr - dQ = Tsurr T

dQ : heat transfer for system T : system = T

surr

By thermodynamic 2nd law ,

dSt + dSsurr 0
dQ TdSt
(14.65)

dSt -

dQ 0 T

By thermodynamic 1st law

dU t = dQ + dW = dQ - PdV t dQ = dU t + PdV t

dU t + PdV t TdS t dU t + PdV t - TdS t 0

(14.66)

Inequality : apply to every incremental change of the system between nonequilibrium states equality : holds for changes between equilibrium states (reversible)

eqn(14.61) : too general to apply to practical problem => need more restricted version

(dU t )St ,V t 0 or (dSt ) U t ,V t 0

At const, T, P

dU t + PdV t - TdS t 0 d( U t + PV t - TS t ) T ,P 0 G t = H t - TS t = U t + PV t - TS t
d (G t ) T , P 0
Determination of equilibrium states 1) Express Gt as a function of the numbers of moles of the species in the several phases 2) Finds the set of values for mole number that minimize Gt subject to the constraints of mass conversion At equilibrium
(14.67)

d (G t ) T , P = 0

(14.68)

At constant T, P, G , G , G must be continuous function of x1, and the

'

''

G '' > 0 => Criterion of stability for single-phase

d 2 G 2 >0 dx1
and (const. T, P)

d 2 (G / RT ) >0 2 dx1

(const. T, P)

(14.65)

From eqn(12.30)

G E = G - RT x i ln x i
d (G / RT ) d (G E / RT ) = ln x1 - ln x 2 + dx1 dx1

G GE = x1 ln x1 + x 2 ln x 2 + RT RT

d 2 (G / RT ) 1 d 2 (G E / RT ) = + >0 2 dx1 x1x 2 dx1

d 2 (G E / RT ) 1 >(const T, P) 2 dx1 x1 x 2

(14.65)

From eqn(12.6)

GE = x1 ln 1 + x 2 ln 2 RT

d(G E / RT) d ln 1 d ln 2 = ln 1 - ln 2 + x1 + x2 = ln 1 - ln 2 dx1 dx1 dx1

zero Gibbs-Duhem eq.

d 2 (G E / RT ) d ln 1 d ln 2 = 2 dx1 dx1 dx1

In combination with eqn(14.70)

1 d ln 1 = x 2 dx1

d ln 1 1 > dx 1 x1

(const T, P)

df d1 1 >0 , > 0 (const T, P) dx1 dx1