Distillation Theory

Jesse Sharifi, Fall 2010

Theory
For this experiment, fractional distillation was used to separate a binary mixture containg components
with a relatively similar boiling point. Distillation is a significant process in the fuels, pharmaceutical, and
chemical manufacturing. Distillations applications for both purification and chemical separation are lucrative
in industry Separation typically makes up to 70% of the capital and operating costs for manufacturing (Ruiz,
Kim, and Linninger). Designing a column is by no means a simple task. In non-ideal conditions, rigorous
calculations are required which account for plate temperature, flow rate of vapor and liquid inside the column,
feed composition, and the intrinsic properties of the distillate. For this reason, column design is often
simplified by assuming a binary mixture under ideal conditions. When design is complete, a correlation factor
is found to accurately depict the columns activity (Bhethanabotla). However, there are several other methods
that can be used when designing a column. The concurrent paragraphs outline how to calculate the theoretical
number of trays, column efficiency, and ways to improve column efficiency.
One of the first methods to facilitate column design is the utilization of The Fenske Equation:

avg
b
b
d
d
x
x
x
x
N
o log
1
1
log
(

|
|
.
|

\
|
|
|
.
|

\
|

=
(1)
Where, for a binary mixture, x
d
is the mole fraction of the more volatile component, or light key, x
b
is the mole
fraction of the less volatile component, or heavy key,
avg
is the relative volatility, with regards to the more
volatile and less volatile components, and N is the minimum number of theoretical trays needed. Relative
volatility is used in the Fenske Equation as volatility throughout the column will vary due to temperature
effects.
There are a few things to note about the Fenske Equation. The first is that the calculation does not
include the reboiler. The Fenske Equation is valid for calculating the number of theoretical trays during total
reflux. During total reflux, product is not withdrawn until a separation that is near the desired value is
achieved. Total reflux is commonly when starting up the column (Price).
Another method that can be used to simplify tray calculations is the McCabe-Thiele Method. This
method assumes that the distillation column operates under constant pressure and that the flow rates of
liquid and vapor do not change throughout the column. With these assumptions, the number of plates and
feed stage can be estimated graphically. This McCabe-Thiele Method is often used over the Fenske Equation as
it allows for the number of theoretical stages to be used at any reflux value.
The first step to utilizing the McCabe-Thiele Method is to plot the equilibrium curve for the feed
composition. After this, q, the moles of liquid flowing in the stripping section:

(2)
where is the heat of vaporization, C
p
is the specific heat of the liquid, Tb is the bubble point of the feed and
Tf is the feed temperature.
The slope of the q line can then be found:

(3)
Now that the slope is known, the feed line can be plotted on the equilibrium curve using x
f
, the feed
composition, and the slope of the line. The y-intercept of the rectifying line is then calculated using the
following equation:

(4)
where x
d
is the distillate composition and Rd is the reflux rate defined by equation five:

(5)
where V is the vapor flow rate in the column, D is the distillate flow rate, and L is the liquid flow rate.
The rectifying line, or the section in the column above the feed tray is vapor rich, is then plotted on the
equilibrium curve using (x
d
, x
d
) as a point. The intersection of the rectifying line and the feed line is then
connected to the point (x
b
, x
b
). This line is known as the stripping line, which represents the region below the
feed tray (Price).
The number of equilibrium stages can now be found by starting at either the top or bottom, using the
composition of the operating line, respectfully. Draw horizontal lines to the equilibrium curve, vertical lines to
the operating curve, and continue until the desired composition is reached.

Figure 1. A McCabe-Thiele Diagram constructed in Mathematica

There are two limiting conditions to keep in mind when using the McCabe-Thiele Method: Maximum
Reflux and Minimum Reflux. Maximum reflux is depicted on the equilibrium curve by positioning the upper
and lower operating lines on the diagonal, which is as far away from the equilibrium curve as possible. This
means that the maximum possible separation is occurring at each stage (according to the diagram). This also
means that using this method will calculate the minimum required stages for a selected purification. This also
implies that there is no product, as all of the overhead is being returned as reflux.
The second limiting condition is minimum reflux. During minimum reflux, the q-line and the upper
operating line intersect on the equilibrium curve. This is strictly a conceptual limit, as it implies that both the
streams entering and exiting the column are at equilibrium. This is impossible, as it would take an infinite
number of stages to accomplish this theoretically. In practice, when a McCabe-Thiele diagram is graphed for
minimum reflux, the upper operating line is disregarded. Instead, the operating line is shifted to (x
d
, y
1
) and
where the q-line comes in contact with the equilibrium curve. Calculating the stages at these conditions would
yield an infinite number of stages. Since this condition cannot be met, columns are designed so that minimum
and maximum reflux are proportional to the operating conditions (Bhethanabotla).
Naturally, regardless of the method chosen to model a fractional distillation column, the overall
efficiency of the column will be determined by its throughput. However, every column has its intrinsic
properties which affect its efficiency. These properties are determined by the column height, tray spacing, and
size of downcomers, or openings in the tray which allow the flow of liquid down the column from the feed.
Column efficiency is often affected by vapor flow rate. A high vapor flow rate can cause the column to
foam in the stripping region. This will drop the overall efficiency of the column as vapor which is meant to flow
up the column and two the decanter gets stuck in the liquid phase. Column flooding is also possible at high
flow rates as excess liquid cannot overcome the capacity of the downcomer. Flooding is also possible if the
liquid flow rate is set too high for the column At low vapor flow rates, the column may begin to weep. This is
when liquid flows through the tray openings from plate to plate. Any of these factors can cause the calculated
number of plates to deviate from the actual number required. Where the overall efficiency is defined as:

(6)
where N
ideal
is the number of plates calculated using either the Fenske equation or McCabe-Thiele Method,
N
actual
is the actual number of plates required, and E
o
is the overall efficiency of the column.
As noted earlier, the efficiency of the column is typically proportional to the vapor flow rate through
the downcomers. The following image from University of Pittsburgh roughly shows how vapor flow rate effects
efficiency in the weeping, entrainment, and flooding regions:
Figure 2. A general plot of how vapor flow rate effects column efficiency
One of the easiest ways to increase or decrease column efficiency is by changing the diameter of the
column, as the flow rate through the column is a function of cross sectional area and vapor velocity. Columns
are typically designed to a diameter that minimizes weeping; however, increasing the diameter of column can
greatly affect its cost.
Rather than modifying the column, the velocity of the vapor required to cause flooding, v
flood
can be
calculated using the following equation:

(7)
where

is the density of the vapor,

is the density of the liquid, and K is a factor. The K factor can be found
using the following equation:

(8)
where is the surface tension and C
sb
is a function of F
lv
, the flow parameter, the thickness between plates,
the liquid mass flow rate L, and the vapor mass flow rate V:

(9)
Now that all the necessary parameters are known, v
flood
can be found using the Fair correlation, a plot found
below:

Figure 3. A plot of the Fair correlation
Knowing the vapor velocity required allows for the operating vapor velocity to be calculated:

(10)
where f is the fraction of flooding and v
op
is the operating vapor velocity.
If it is so desired, the theoretical diameter for the column can now be calculated for a desired vapor
velocity:

)
(11)
where D is the diameter of the column, V
mass
is the molar flow rate of the vapor per hour, f
design
is the fraction
of flooding desired, A
d
is the area of the down comers and A is the total area of the tray.
Utilizing either the McCabe-Thiele Method or Fenske Equation in combination and calculating the
column diameter or required vapor flow rate will allow for the quick design of a column with minimal
calculation.




Works Cited
Bennett, Douglas L., and Kenneth W. Kovak. "Optimize Distillation Columns." American Institute of
Chemical Engineers. Web. 6 Oct. 2010.
<http://userpages.umbc.edu/~dfrey1/ench445/0005benn.pdf>.
"Distillation." CU Boulder Organic Chemistry Undergraduate Courses. Web. 06 Oct. 2010.
<http://orgchem.colorado.edu/hndbksupport/dist/dist.html>.
Johnson, J. K. "ChE1008." University of Pittsburgh. Web. 06 Oct. 2010.
<http://puccini.che.pitt.edu/~karlj/Classes/CHE1008/>.
"McCabe-Thiele Background." CofE | College Of Engineering | University of South Florida. Web. 06
Oct. 2010. <http://www.eng.usf.edu/~bhethana/McTh/start2.html>.
Price, R. M. "RMP Lecture Notes." Christian Brothers University. 14 Feb. 2003. Web. 06 Oct. 2010.
<http://www.cbu.edu/~rprice/lectures/distill4.html>.
"Vapor Liquid Separation." University of Connecticut. Web. 06 Oct. 2010.
<http://www.engr.uconn.edu/~suzy/cheg%20224/Vapor%20Liq%20separation%20notes.pdf>
.