Distillation.

In multistage distillation, one heat source is usually used in the lowest stage, while in the higher part of the tower, a
condenser, partial ot total is used in order to obtain the distillate in liquid form, also, some distillate is recirculated to
enrich the distillate. The rest of the stages are adiabatic. Several withdraw stages are possible.

V
Distillate (D)
L
1

m-1
F m
m+1

n V

L Bottoms (B)

Mass and enthalpy balances can be written for each section.

Rectification
Total: Vi +1 = Li + D
B: Vi +1 yi +1 = Li xi + D xD
H: Vi +1 Hi +1 = Li hi + D xD + qC
Stripping
Total: Li -1 = V i + B
B: Li -1 xi -1 = V i yi + B xB
H: Li -1 hi -1 + qB = V i Hi + B hB
Feed
Total: F + V m+1 + Lm-1 = Vm + Lm
B: F zF + V m+1 ym-1 + Lm-1 xm-1 = Vm ym + Lm xm
H: F hF + V m+1 Hm-1 + Lm-1 hm-1 = Vm Hm + Lm hm
Total
Total: F = B + D
B: F z = B xB + D xD
H: F hF + q R = B hB + D h D + q C
Condenser
Total: V1 = L0 + D
B: V1 y1 = L0 x0 + D xD
H: V1 H1 = L0 h0 + D hD + qC

McCabe -Thiele

If the flows between stages were constant, the balances would give straight lines in a y vs x plane,
Li = L Li = L
Vi = V Vi = V
this would be the case if,
2 4.3.1. Distilation - McCabe.nb

1. The heat of vaporization for both substances is constant and the same
2. The changes in sensible heat and the heat of mixing are both negligible
3. The process is adiabatic everywhere but at the boiler and condenser
4. The pressure is constant throughout the column

Indeed, in this case the vaporization and condensation in each stage would be compensated each one by the other, all the
required heat would be compensated and no further heat would be loss nor produced.

So mass and heat balances can be solved independently. For the mass balances, in the rectification
V yi +1 = L xi + D xD
V =L+D
and
RD 1
yi +1 = 1+RD
xi + 1+RD
xD
where RD = L / D is the "internal reflux ratio". It is clear that this equation represents a straight line.

For the stripping section, by a similar argument,

L B
yi = xi -1 - xB
L-B L-B

For the feed we use the enthalpy balance considering that, because of the assumptions the changes in enthalpy are small,
Hm+1 ≈ Hm hm-1 ≈ hm
so
F hF + V Hm + L hm = Vm Hm + L hm
F hF + V - V  Hm = L - L hm
and from the total mass balance
V -V =L-L-F
F hF + L - L - F  Hm = F (hF - Hm ) + L - L Hm = L - L hm
L - L (Hm - hm ) = F (Hm - hF )
so
L-L H m -h F
F
= H m -h m
≡q
but also
V -V H m -h F
F
-1= H m -h m
≡q

this is defined as the "thermal condition of the feed", it can be interpreted as,

Heat used to vaporize one mol of the feed

q≡ Latent heat (change of phase)

If the heat is below the boiling temperature of the stage (sub-cooled), it is necessary to heat the feed stream up to the
boiling temperature and then give the required heat for the vaporization, then q>1. If the feed is at the boiling temperature
of the stage (bubble point), only the vaporization heat is required then q=1. If the feed is vapor at its dew point, no heat is
required, q=0. It the feed is overheated steam, q<0. At intermediate values, when a mixture of vapor and liquid is feed, if f
is the vapor fraction in the feed,
L + (1 - f ) F = L
and
L-L
q= F
=1-f
 4.3.1. Distilation - McCabe.nb 3

Notice that Hm and hm are saturated, so they are at the same temperature or isotherm in enthalpy diagram, while hF not
necessarily is saturated, but, according to the definition of q its value satisfies:
hF = Hm - q(Hm - hm )
the precise value of q can be calculated as follows: In the case of overheated steam,
H m -h F Hm -(Hm +Cp,vap (TF -Tdew )) CV (TF -Tdew )
q= H m -h m
= λ
=- λ

and
H m -h F Hm -hm -Cp,líq (Tbubble -TF ) CL (Tbubble -TF )
q= H m -h m
= H m -h m
=1+ λ

Having defined and detailed the Thermal Condition of the Feed, we can now see how it is incorporate into the
balances. Adding the mass balance of the rectification to that of the stripping section one gets,
V yi +1 = L xi + D xD with i=m leads to V ym+1 = L xm + D xD
L xi -1 = V yi + B xB with i=m+1 leads to L xm = V ym+1 + B xB
so, we can write (sub-indices dropped)
V - V  y = L - L x - (D xD + B xB )
and from a total mass balance we see that the last sum must equal the feed,
F zF = D xD + B xB
so
V -V L-L
F
y= F
x - zF
where we find the definitions for q, so
(q - 1) y = q x - zF
q zF
y= q-1
x- q-1

which is a straight line.

Finally, the condenser charges can be obtained from its total and enthalpy balances,
Total: V = L + D
H: V H1 = L h0 + D hD + qC = (L + D) H1
combining and rearranging leads to,
qC
D
=  DL + 1 H1 - L
D
h0 - hD = (RD + 1) H1 - RD h0 - hD
and the boiler charge can be obtained from the total enthalpy balance,
F hF + q R = B hB + D hD + q C
from which
q R = B h B + D hD + q C - F h F

Recapitulating, the McAbe-Thiele method uses these three equations as straight lines:

Feed:
q zF
y= q-1
x- q-1

When x = zF then y = zF so it passes through (zF ,zF ) and has slope q/(q-1). If q=1 (saturated liquid), the slope is ∞; if
q=0 (saturated vapor), the slope is 0. If the feed is overheated steam, q<0 and the slope is positive; It the feed is sub-
cooled liquid, q>1 and the slope is positive. When the feed is a saturated vapor-liquid mixture, 0<q<1 and the slope is
negative.

Rectification:
RD 1
yi +1 = 1+RD
xi + 1+RD
xD
When xi = xD then (x , y ) = (xD , xD )
4 4.3.1. Distilation - McCabe.nb

Stripping:
L B
yi = xi -1 - xB
L-B L-B
When xi -1 = xB then (x , y ) = (xB , xB )

These equations are solved simultaneously with the equilibrium distribution diagram and the stage efficiencies.

A simple algorithm can be devised:

1. Locate (zF , zF ), (xD , xD ), and (xB , xB ).
2. Plot the rectification operating line using (xD , xD ) and the reflux ratio.
3. Plot the q line using (zF , zF ) and the thermal condition of the feed.
4. Plot the stripping operating line using (xB , xB ) and the intersection between the q-line and the rectification line.
5. If the efficiency is not 100% use the operating lines and the equilibrium to plot an auxiliary, efficiency line.
7. Plot the stages.
8. Calculate the thermal charges of the boiler and the condenser

It should be noted that the reflux ratio has limits.

a. Total reflux.
If all the distillate is recycled, V=L and D=0 so RD = L / D = ∞ and
xD
1+RD
=0
rectification and stripping lines have the same slope as the reference line (45°). This gives the minimum number of stages
but no product (distillate) is obtained.

b. Minimum reflux ratio.

It corresponds to an infinite number of stages. Usually it can be obtained from the slope of the rectification line that
intersects the q line where the latter intersects the equilibrium line.

However, if the rectification line can not get until that intersection without touching the equilibrium line, the minimum
reflux is obtained from the slope of the line that touches the equilibrium line.
Example: Let’s take the ethanol-water equilibrium at 1 atm,
 4.3.1. Distilation - McCabe.nb 5

0 0 7.24 47.9 99.1

0.05 0.32 7.1 47.85 97.45
0.1 0.42 6.9 47.8 95.49
0.15 0.48 6.8 47.75 93.55
0.2 0.525 6.6 47.7 91.62
0.25 0.56 6.8 47.65 89.84
0.3 0.59 7 47.6 88.26
In[1]:= {xB, yB, hs, Hs, Ts} = 0.4 0.63 7.3 47.5 85.83 ;(*ΔH(kJ/mol), T(°C)*)
0.5 0.67 7.75 47.4 84.02
0.6 0.71 7.9 47.2 82.57
0.7 0.765 8.05 47.15 81.12
0.8 0.825 8.16 47.1 79.78
0.9 0.895 8.28 47. 78.33
0.95 0.935 8.34 46.95 77.87
1 1 8.4 46.9 77.60

ye = Interpolation[Thread[{xB, yB}], Method  "Spline", InterpolationOrder  3];

xe = Interpolation[Thread[{yB, xB}], Method  "Spline", InterpolationOrder  3];
He = Interpolation[Thread[{yB, Hs}], Method  "Spline", InterpolationOrder  1];
he = Interpolation[Thread[{xB, hs}], Method  "Spline", InterpolationOrder  1];
TVs = Interpolation[Thread[{yB, Ts}], Method  "Spline", InterpolationOrder  1];
TLs = Interpolation[Thread[{xB, Ts}], Method  "Spline", InterpolationOrder  1];
{Tz = Plot[{TLs[z], TVs[z]}, {z, 0, 1}],
xy = Plot[{ye[x], x}, {x, 0, 1}, AspectRatio  1,
ImageSize  Small, Epilog  {Point[Thread[{xB, yB}]]}],
Hxy = Plot[{he[x], He[x]}, {x, 0, 1},
Epilog  {Gray, Line[Thread[{Thread[{yB, Hs}], Thread[{xB, hs}]}]]},
AspectRatio  1, ImageSize  Small]}

50
1.0

100 0.8 40

95
0.6 30
90
Out[8]=  , , 
85 0.4 20

80
0.2 10
0.2 0.4 0.6 0.8 1.0

0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0
6 4.3.1. Distilation - McCabe.nb

In[9]:= Manipulate
q z
ym[x_] := Ifq != 1, x- , Infinity; (*Feed*)
q-1 q-1
RD 1
yr[x_] := x+ xD; (*Rectification*)
1 + RD 1 + RD
q (xD - xB) + RD (z - xB)
LB = ; (*Required LB ratio*)
xD - z
LB 1
ya[x_] := x- xB; (*Stripping*)
LB - 1 LB - 1
xD (q - 1) + z (1 + RD)
int = ;
q + RD
(*Intersection between Feed and Rectification*)
y1 = x0 = xD; (*Stages start at xD*)
steps = {{xD, xD}}; (*Stages start at xD*)
eff = {}; (*An array to save the auxiliary efficiency line*)
Do[xei = xe[yi ]; (*Calculate xi * *)
xi = xi-1 - ϵL (xi-1 - xei ); (*Calculate xi *)
eff = Append[eff, {xi , yi }];
(*Append {xi ,yi } as a point in the efficiency line*)
yi+1 = If[xi > int, yr[xi ], ya[xi ]];
(*Calculate yi+1 , with the rectification or striping equation as required*)
If[yi+1 < 0, yi+1 = 0]; (*yi+1 cannot be less than zero*)
steps = Join[steps, {{xi , yi }, {xi , yi+1 }}]; (*Append a new step*)
If[xi < xB, Append[steps, {xi , yi }]; n = i; Break[]], (*Are we done?*)
{i, 1, 50}];
Plot[{ye[x], x}, {x, 0, 1}, AspectRatio  1, PlotLabel  Text[(n) "etapas"],
Epilog  {Line[{{z, z}, {int, ym[int]}}], Line[{{xD, xD}, {int, yr[int]}}],
Line[{{xB, xB}, {int, ya[int]}}], Red, Line[steps], Dashed, Blue, Line[eff]}],
{{z, 0.27}, 0, 1},
{{xD, 0.85}, 0, 1},
{{xB, 0.05}, 0, 1},
{{q, 0.3}, - 0.5, 1.5},
{{RD, 2.459}, 0, 20},
{{ϵL, 0.62}, 0, 1}, ControlPlacement  Right
 4.3.1. Distilation - McCabe.nb 7

z
24 etapas
xD
1.0
xB

q
0.8
RD

ϵL

0.6
Out[9]=

0.4

0.2

0.2 0.4 0.6 0.8 1.0

Global balances allow us to calculate the unknown flows,

F =D+B
F zF = D xD + B xB
from there,
F zF = D xD + (F - D) xB
D = F xzF --xxB
D B

The energy balances are:

qC = D {(RD + 1) H1 - RD h0 - hD )}
and
q R = B hB + D hD + q C - F h F
8 4.3.1. Distilation - McCabe.nb

In[10]:= (*Data*)
F = 150; zF = 0.27; xD = 0.85; xB = 0.05; f = 0.7;
(*mass balances*)
RD = 0.259;
zF - xB
De = F ;
xD - xB
B = F - De;
(*Enthalpy balances, let's start by calculating hF*)
xF = x /. FindRoot[x + f (ye[x] - x) == zF, {x, zF}];
yF = ye[xF];
hF = he[xF] + f (He[yF] - he[xF]);
H1 = He[xD];
h0 = hD = he[xD];
qC = De ((RD + 1) H1 - RD h0 - hD);
hB = he[xB];
qB = B hB + De hD + qC - F hF;
(*Summary*)
TableForm[{{F, zF, hF} , {De, xD, hD}, {B, xB, hB}},
TableHeadings  {{"F", "D", "B"}, {"Flow (mol/s)", "zB ", "H (kJ/mol)"}}]
TableForm[{{qC}, {qB}}, TableHeadings  {{"qC ", "qB "}, {"kJ/s"}}]
Out[22]//TableForm=
Flow (mol/s) zB H (kJ/mol)
F 150 0.27 35.5957
D 41.25 0.85 8.22
B 108.75 0.05 7.1
Out[23]//TableForm=
kJ/s
qC 2017.33
qB - 2210.82