CHAPTER 2

ANALYSIS OF DISTILLATION COLUMNS

The major emphasis of this book is the use of rigorous steady-state and dynamic simulation
for the design and control of distillation columns. However, there are several simple
approximate methods that provide significant insight into how the various design and
operating parameters impact separation. Some of these methods employ graphical techniques that give visual pictures of the effects of parameters. Although some of the methods
are limited to binary systems, the relationships can be extended to multicomponent systems.
2.1 DESIGN DEGREES OF FREEDOM
The design of a distillation column involves many parameters: product compositions, product
flow rates, operating pressure, total number of trays, feed-tray location, reflux ratio, reboiler
heat input, condenser heat removal, column diameter, and column height. Not all of these
variables are independent, so a degrees of freedom analysis is useful in pinning down exactly
how many independent variables can (and must) be specified to completely define the system.
A rigorous mathematical degrees of freedom analysis involves counting the number of
variables in the system and subtracting the number of equations that describe the system.
For a multicomponent, multistage column, this can involve hundreds, if not thousands, of
variables and equations. Any error in counting is grossly amplified because we are taking
the difference between two very large numbers. A simple intuitive approach is used below.
The normal situation in distillation design is that the feed conditions are given: flow rate
F (mol/h), composition zj (mole fraction component j, temperature TF, and pressure PF. The
desired compositions of the product streams are also typically known. We consider a twoproduct column, so that the normal specifications are to set the heavy-key impurity in the
distillate xD,HK, and the light-key impurity in the bottoms xB,LK. These specifications apply
in binary and multicomponent systems.
Distillation Design and Control Using AspenTM Simulation, Second Edition. William L. Luyben.
2013 John Wiley & Sons, Inc. Published 2013 by John Wiley & Sons, Inc.

29

30

ANALYSIS OF DISTILLATION COLUMNS

The design problem is to establish the operating pressure P, the total number of trays NT,
and the feed-tray location NF that produces the desired product purities. All other
parameters are then fixed. Therefore, the number of design degrees of freedom is five:
xD,HK, xB,LK, P, NT, and NF. Therefore, if the desired product purities and the pressure are
given, there are two degrees of freedom.
Just to emphasis this point, the five variables that could be specified might be the distillate
flow rate D, reflux ratio RR R/D, P, NT, and NF. In this case, the product compositions
cannot be specified but depend on the distillate flow rate and reflux ratio selected.
The steps in the design procedure will be illustrated in subsequent chapters. Our purpose
in this chapter is to discuss some of the ways to establish reasonable values of some of the
parameters, such as the number of stages or the reflux ratio.
2.2 BINARY McCABETHIELE METHOD
The McCabeThiele method is a graphical approach that shows very nicely in pictorial
form the effects of vaporliquid equilibrium (VLE), reflux ratio, and number of trays. It is
limited to binary systems, but the effects of parameters can be extended to multicomponent
systems. The basic effects can be summarized
1. The easier the separation, the fewer trays required and the lower the required reflux
ratio (lower energy consumption).
2. The higher the desired product purities, the more trays required. But the required
reflux ratio does not increase significantly as product purities increase.
3. There is an engineering trade-off between the number of trays and the reflux ratio. An
infinite number of columns can be designed that produce exactly the same products,
but have different heights, different diameters, and different energy consumptions.
Selecting the optimum column involves issues of both steady-state economics and
dynamic controllability.
4. There are minimum values of the number trays (Nmin) and the reflux ratio (RRmin) that
are required for a given separation.
All of these items can be visually demonstrated using the McCabeThiele method.
The distillation column considered is shown in Figure 2.1 with the various flows and
composition indicated. We assume that the feed molar flow rate F and composition z are
given. If the product compositions are specified, the molar flow rates of the two products D
and B can be immediately calculated from the overall total molar balance and the overall
component balance on the light component.
F DB
zF DxD BxB
) D F z  xB =xD  xB 
For the moment, let us assume that the pressure has been specified, so the VLE is fixed. Let
us also assume that the reflux ratio has been specified, so the reflux flow rate can be calculated
R (RR) (D). The equimolal overflow assumption is usually made in the McCabeThiele
method. The liquid and vapor flow rates are assumed to be constant in a given section of the
column. For example, the liquid flow rate in the rectifying section LR is equal to the reflux

BINARY McCABETHIELE METHOD

Rectifying
section

Feed
F, z, q

Condenser

{
{

Stripping
section

31

Liquid reflux,R
NT

LR

VR

Reflux drum

NF

Distillate
D, xD
VS

LS
2

Reboiler

Energy
Bottoms
B, xB

Figure 2.1 McCabeThiele method: distillation column.

flow rate R. From an overall balance around the top of the column, the vapor flow rate in the
rectifying section VR is equal to the reflux plus the distillate (VR R D).
The method uses an xy diagram whose coordinates are the mole fraction of the light
component in the liquid x and the mole fraction of the light component in the vapor phase y.
The VLE curve is plotted for the selected pressure. The 45 line is plotted. The specified
product compositions xD and xB are located on the 45 line, as shown in Figure 2.2.

VLE

xD
y

xB
x

Figure 2.2 McCabeThiele method: locate products and VLE.

32

ANALYSIS OF DISTILLATION COLUMNS

2.2.1 Operating Lines

Next the rectifying operating line (ROL) is drawn. This is a straight line with a slope equal
to the ratio of the liquid and vapor flow rates in the rectifying section.
Slope ROL LR =V R R=R D RR=1 RR
The line intersects the 45 line at the distillate composition xD, so it is easy to construct (see
Fig. 2.3). The proof of this construction can be derived by looking at the top of the column,
as shown in Figure 2.4.

VLE

xD
ROL

Slope =

LR
R
RR
=
=
VR R + D 1 + RR

xB
x

Figure 2.3 McCabeThiele method: draw operating lines.

D, xD

Cut above tray n

Overall balance: V R = LR + D
Component balance:
VR yn = LR xn+1 + D xD
yn =(LR/VR) xn+1 + (D xD /VR)

R, xD

y=mx+b
Slope = m = LR

LR, xn+1

VR, yn

/VR

Intercept on 45o line (x = y)

Xint = =(LR/VR) xint + D xD /VR
(VR LR) xint = D xD
D xint = D xD

xint = xD
Figure 2.4 ROL construction.

BINARY McCABETHIELE METHOD

33

The liquid and vapor flow rates in the stripping section (LS and VS) can be calculated if
the thermal condition of the feed is known. Since the temperature, pressure, and
composition of the feed are given, the fraction of the feed that is liquid can be calculated
from an isothermal flash calculation. This fraction is defined as the variable q. Knowing
q, the liquid and vapor flow rates in the stripping section can be calculated. If the feed is
saturated liquid, q is 1. If the feed is saturated vapor, q is 0.
q LS  LR =F
) LS qF LR
V S LS  B
The stripping operating line can be drawn. It is a straight line with slope LS/VS that interacts the
45 line at the bottoms composition xB. The proof of this construction can be derived by looking
at the bottom of the column, as shown in Figure 2.5. Figure 2.6 shows the two operating lines.
2.2.2 q-Line
There is a relationship between the intersection point of the two operating lines and feed
conditions. As shown in Figure 2.7, a straight line can be draw from the location of the feed
composition z on the 45 line to this intersection point. As we will proof below, the slope of
this line is only a function of the thermal condition of the feed, which is defined in the
parameter q. The slope is q/(1  q). This makes the construction of the McCabeThiele
diagram very simple.
1. Locate the three compositions on the 45 line (z, xD, and xB)
2. Draw the ROL from the xD point with a slope of RR/(1 RR).
3. Draw the q line from the z point with a slope of q/(1  q).
4. Draw the SOL from the xB point to the intersection of the q line and the ROL.
Cut above tray n
LS, xn+1

VS, yn

Total molar balance: LS = VS + B

Component balance:
LS xn+1 = B xB + VS yn
yn =(LS /VS) xn+1 - (B xB /VS)
y=mx+b

LS, x1

VS, yB

Q
B, xB

Slope = m = LS

/VS

Intercept on 45o line (x = y)

Xint = (LS /VS) xint - B xB /VS
(VS LS) xint = B xB
B xint = B xB

xint = xB
Figure 2.5 SOL construction.

34

ANALYSIS OF DISTILLATION COLUMNS

VLE

xD
y

ROL

SOL

xB
x

Figure 2.6 Operating lines.

q-Line

VLE
xD

ROL

SOL
xB
x

Figure 2.7 q-Line.

The equations of the rectifying and stripping operating lines are given below in terms of
the point of intersection of the two lines at yint and xint.
ROL :

yint LR =V R xint DxD =V R

SOL :

yint LS =V S xint  BxB =V S

BINARY McCABETHIELE METHOD

35

Subtracting the two equations gives

V R  V S yint LR  LS xint DxD BxB
The last term on the right is just Fz. Using the definition of q leads to
V R  V S 1  qF
LR  LS qF
Substituting these relationships into the previous equation gives
1  qFyint qFxint Fz
yint q=1  qxint z=1  q
This is the equation of a straight line with slope q/(1  q). The q-line is vertical for
saturated liquid feed (q 1), and it is horizontal for saturated vapor feed (q 0). On the 45
line, xint is equal to yint. Define this as x45.
1  qx45 qx45 z
x45 z
Thus, the q-line intersects the 45 line at the feed composition z.
2.2.3 Stepping Off Trays
The number of trays is determined by moving vertically from the xB point on the 45
line to the VLE line. This is the composition of the vapor yB, leaving the partial
reboiler. Then, we move horizontally over to the SOL. This step represents the partial
reboiler. The value of x on the SOL is the composition of liquid x1, leaving Tray 1 (if
we are numbering from the bottom of the column up). This stepping is repeated,
moving vertically to y1 and horizontally to x2. Stepping continues until we cross the
intersection of the operating lines. This is the feed tray. Then the horizontal line is
extended to the ROL. Continuing to step until the xD value is crossed gives the total
number of trays.
2.2.4 Effect of Parameters
We know enough now about the McCabeThiele diagram to make several observations,
which can be applied to any distillation system, not just a binary separation.
1. The farther the VLE curve is from the 45 line, the smaller the slope of the rectifying
operation line. This means a smaller reflux ratio and therefore lower energy
consumption. A fat VLE curve corresponds to large relative volatilities and an
easy separation.
2. The easier the separation, the fewer trays it takes to make a given separation.
3. The higher the product purities, the more trays it takes to make a given separation.
4. Increasing product purities does not have a big effect on the required reflux ratio.

36

ANALYSIS OF DISTILLATION COLUMNS

5. Increasing the liquid-to-vapor ratio in a section of a column increases the separation

that occurs in that section.
6. These effects apply to all types of separations and distillation columns.
2.2.5 Limiting Conditions
We need to discuss some of the limiting conditions in distillation systems. The minimum
number of trays for a specified separation corresponds to total reflux operation. If the
column is run under total reflux conditions, the distillate flow rate is zero. Therefore, the
reflux ratio is infinite, and the slope of the operating lines is unity. This is the 45 line. Thus,
the minimum number of trays can be determined by simply stepping up between the 45
line and the VLE curve (see Fig. 2.8).
The minimum reflux ratio for a specified separation corresponds to having an infinite
number of trays. This usually occurs when the intersection of the operating lines and the qline occurs exactly on the VLE curve. This is a pinch condition. It would take an infinite
number of trays to move past this point. This is illustrated in Figure 2.9. The minimum
reflux ratio is calculated from the slope of this limiting operating line.
2.3 APPROXIMATE MULTICOMPONENT METHODS
Many years before the availability of computers for rigorous analysis, several simple
approximate methods were developed for analyzing multicomponent systems. These
methods are still quite useful for getting quick estimates of the size of a column (number
of trays) and the energy consumption (reflux ratios and the corresponding vapor boilup and
reboiler heat input).
VLE

xD

y
z

Operating lines
at total reflux

xB

x
Figure 2.8 Minimum number of trays.

APPROXIMATE MULTICOMPONENT METHODS

37

Pinch condition, N

xD

z
VLE
ROL at minimum
reflux ratio

xB

x
Figure 2.9 Minimum reflux ratio.

2.3.1 Fenske Equation for Minimum Number of Trays

The minimum number of trays corresponds to total reflux operation (an infinite reflux
ratio). The Fenske equation relates the compositions at the two end of a column to the
number of stages in the column under this limiting condition.





log xD;LK =xD;HK xB;HK =xB;LK
N min 1
logaLK;HK
where Nmin is the minimum number of stages, xD,LK is the mole fraction of the light-key
component at the top of the column, xD,HK is the mole fraction of the heavy-key component
at the top of the column, xB,HK is the mole fraction of the heavy-key component at the
bottom of the column, xB,LK is the mole fraction of the light-key component at the bottom
of the column, and aLK,HK is the relative volatility between the LK and HK components.
This equation is applicable to multicomponent systems, but it assumes a constant
relative volatility between the two components considered.
An example of the use of the Fenske equation is given in Chapter 4. Results of this
approximate method will be compared with the results found from rigorous simulation.
2.3.2 Underwood Equations for Minimum Reflux Ratio
The Underwood equations can be used to calculate the minimum reflux ratio in a
multicomponent system if the relative volatilities of the components are constant. There
are two equations.
NC
X
a j zj
1q
a
u
j1 j

38

ANALYSIS OF DISTILLATION COLUMNS

RRmin

Infinite number of design will

produce the same products
Tall column, small ID, low energy

NT
Short column
Large ID
High energy

Nmin
RR

Figure 2.10 Trade-off between reflux ratio and number of trays.

NC
X
aj xDj
1 RRmin
a
u
j1 j

The feed composition zj (mole fractions j 1, NC), the desired distillate composition xDj
(j 1, NC), and the feed thermal condition q are specified. The relative volatilities aj (j 1,
NC) of the multicomponent mixture are known.
The first equation contains one unknown parameter u. However, expanding the
summation of NC terms and multiplying through all the denominator terms (aj  u)
give a polynomial in u whose order is NC. Therefore, there are NC roots of this polynomial.
One of these roots lies between the two relative volatility values aLK and aHK. This is found
using some iterative solution method. It is substituted into the second equation, which can
then be solved explicitly for the minimum reflux ratio.
An example of the use of the Underwood equations is given in Chapter 4. The results of
this approximate method will be compared with the results found from rigorous simulation.
2.4 CONCLUSIONS
Several methods for analyzing distillation columns have been presented in this chapter.
Graphical methods provide valuable insight into how various parameters affect separations
in distillation.
Some of the general relationships that hold for both binary and multicomponent
distillation columns are as follows
1. There is a trade-off between energy and number of trays required to make a specified
separation, as illustrated in Figure 2.10.
2. Lower relative volatilities make separation more difficult and require more trays or
higher reflux rations.
3. Higher product purities require more trays but not higher reflux ratios.