School of Chemical and Biomedical Engineering

Division of Chemical and Biomolecular Engineering

NANYANG TECHNOLOGICAL UNIVERSITY

AY 2015/2016
YEAR 3/SEMESTER 1
CH3802

Experiment CE2
Distillation

ALDI STEFANUS FELIM U1320089H

LAB GROUP: LA435
EXPERIMENT DATE: 8 OCTOBER 2015

1
Distillation
1 AIM
This experiment aims to investigate the distillation process, a very essential
separation process in chemical industry. It also studies the applications of developed
and widely used methods such as Fenskes method and McCabe-Thieles method.

2 ABSTRACT
This experiment utilizes a tray-type distillation column connected with a
controller and a data logger to investigate the overall efficiency of the column under
total reflux conditions, and to determine the effect of reflux ratios on the result of
separation of a binary product. The overall efficiency of a column under total reflux
conditions can be determined by using Fenskes method, a formula relating the
mole fractions of the components and relative volatility value to obtain the
theoretical number of trays required. Meanwhile, the McCabe-Thieles method can
be used to investigate the theoretical number of trays required and, simultaneously,
the optimum feed location for that particular column. Deviations from theoretical
values are expected, and the experiment itself serves as a tool in investigating the
reasons and the magnitude of the deviations themselves.

2
3 PRINCIPLES
Distillation, as mentioned above, is a method of separating compounds from
a mixture. In the tray-type distillation column that is used in this experiment, the
feed enters from a certain level (i.e. the feed
tray). The plates located above this level function
as the rectifying section of the column, while the
ones located below the feed level function as the
stripping section of the column.
Typically, two streams of product are obtained
for this type of distillation. The overhead product is
symbolized by D and its mole fraction xD.
Meanwhile, the bottom product is symbolized by B,
with its mole fraction symbolized as xB. As the
names suggest, the rectifying section will enrich the
mixture in the more volatile component of the initial
mixture, and stripping section will do the opposite.
The feeds flow rate itself is denoted by F and has
an initial mole fraction xF. All streams are in the unit
of mol/hour and all the mole fraction is based on the
more volatile component of the mixture itself.
Fig 1. Schematic Diagram of Distillation Column

By combining total and any component material balances in the control

surfaces, relations between the feed rate F, the overhead product D, and the
bottom product B can be obtained as shown below,

D x Fx B B x Dx B
= & =
F x D x B F x Dx B

And by combining the total and component material balances both on control
surface I and II, the equations for the operating lines of rectifying and stripping
sections can be generated as follows,

Ln D xD
Rectifying Section - y n+1= xn + and Stripping Section -
Ln + D Ln + D

Lm B xB
y m+1 = x n
Lm B L mB

In order to simplify the above equations, the Constant Molal Overflow (CMO)
approximation is usually used. CMO assumed that all liquid streams are flowing with

3
the same flow rate, and all vapor streams are flowing with the same flow rate. In
other words,
L1=L2=L3==Ln=L & V 1=V 2=V 3 ==V n=V

Lf +1=Lf +2=Lf +3==Lm = L & V f +1=V f +2=V f +3==V m=V

This results in the simplification of the equations for operating lines, and the
equations become:
L V
+1 V +1
D 1 R 1 B 1 1
y n+1= xn + xD= xn + x y m+1 = x m x B = B x m xB
L L R+1 R+1 D V
V V B V B
+1 +1
D D B B

Where R is the reflux ratio used during the distillation, and V B is the boil-up
ratio, the ratio of vapor flowing in to the column from the reboiler to the liquid
flowing out from the column.

In a system using a total condenser, the concentrations of vapor leaving the

top plate, the reflux stream and the overhead product are the same, with the value
xD. This denotes the end point of the rectification section operating line, which is the
point (xD,xD), and the line itself can be sketched as the gradient and one point on
the line are known. The same method can be applied to the stripping section
operating line, with the end point being (xB,xB). Together with the feed line, they will
serve as the basis for calculating the theoretical number of plates using the
McCabe-Thieles method.

The feed line starts from the intersection point between the line x=y and the
point xF at the equilibrium x-y diagram. The equation for feed line is stated below,

q x
y= x+ F
1q 1q

Where q is the variable that represents the thermal condition of the feed
itself. Method for calculating q will be further elaborated in the discussion section.

After all three operating lines are present in the equilibrium x-y diagram,
McCabe-Thieles method can be used to find the theoretical amount of plates
needed for the desired degree of separation. Starting from the point (xD,xD), draw a
stair-shaped line within the area bounded by the equilibrium curve and both

4
operating lines for rectifying and stripping section until the final point of the
stripping section is reached. Feed plate is located by the triangle that has one of its
corner in the stripping section, and the other one in the rectifying section.

For a special case of total reflux, Fenskes method can be used in determining
the theoretical amount of plates needed. Fenskes method will be more elaborated
in the discussion section.

The ultimate aim is to calculate the columns overall efficiency. Columns

overall efficiency, denoted by E, is defined by the following expression:

Number of Theoretical Plates

E= 100
Number of Actual Plates

5
4 MATERIALS AND APPARATUS

4.1 MATERIALS
1. Ethanol and Water mixture

4.2 APPARATUS
1. UOP3CC Distillation Column Unit (8 Plates)
2. Gas Chromatography Apparatus

5 PROCEDURE

5.1 TOTAL REFLUX CONDITION

1. Measure the actual concentration of ethanol-water mixture using the gas
chromatography apparatus.
2. Pour 7L of the mixture to the reboiler, and 5L to each feed tank.
3. Ensure that all control selector switches are on manual.
4. Ensure that all valves other than valve number 10 are closed.
5. Switch on the control panels power.
6. Set the feed pump controller value to 70 and change from manual to I/O
port. Switch on the pump motor.
7. After the specimen has flown through the tubes and has reached the point of
the feed plate, switch off the pump motor.
8. Set the temperature selector switch to T9, switch on the main water pump,
and open valve number 5 to start the flow of cooling water to the condenser
at a flow rate of F11, approximately 3L/minute.
9. Turn the reboiler power controller anticlockwise until its limit, and turn on the
reboiler heater control.
10.To control the reboiler, change the value of T7 to 80 (for total reflux) or 83
(for partial refluxes). Change the control property from manual to
automatic.
11.The system will take 30-40 minutes in reaching steady state, at which all
measured temperatures at the plates will reach a steady value.
12.Once the system is at steady state, open valve number 3 to collect 10ml of
the sample. This 10 ml should be collected after the drainage of the first 5ml
that flows upon the opening of the valve. Store the collected sample in the
vials provided. Close the valve afterwards.
13.Examine the sample collected using GC and convert the data obtained to
mole fraction or concentration value using the calibration curve provided.
14.Take a little amount of sample from valve number 2 in the same manner as
step 12, and again, examine the sample using the GC in the same manner as
the previous step.
15.Record the temperature of T1 until T4 once.

6
 16.Repeat steps 12-16 once more with the interval of 5 mins to wait the system
to equilibrate again

5.2 CONTINUOUS DISTILLATION UNDER NORMAL OPERATING

CONDITIONS
17.Set the reflux ratio to 5:1 on the controller.
18.Turn on the reflux valve switch on the control panel. A clicking sound should
be heard from the reflux motor, and condensates will start flowing to the
receiver of top product.
19.Turn on the feed pump motor, and set the feed pump to 40 at the computer.
20.After approximately 1.2 L of the feed has been fed to the system, take the
overhead samples through valve number 3, and bottom samples through
valve number 2. Store in vials provided and take the samples 5 minutes later
after the system has equilibrated again. Record the temperatures of T1 to T4
once. Examine the samples collected using GC, and find the
concentrations/mole fractions using the calibration curve provided.
21.Change the reflux ratio to 3:1 and repeat step 21.

5.3 FINISHING THE EXPERIMENT

22.Turn off the feed pump motor, reboiler heater and the reflux value controller
in this order. Turn of the control panels power switch, and close the valve F11
for the water supply and turn off the water main.
23.Wait approximately for 20 minutes to allow the distillation column and the
reboiler to cool down. Discard the specimen from all containers inside the
distillation column through valve number 2, 3 and 4.

7
6 RESULTS

6.1 TABULATED DATA OF COLLECTED SAMPLES

6.1.1 Total Reflux
Trial Number Mole Fractions of Overhead Mole Fractions of Bottom
Product Product
Ethanol Water Ethanol Water
1 0.6549 0.3451 0.2135 0.7865
2 0.6460 0.3540 0.2029 0.7971
Average 0.6505 0.3495 0.2082 0.7918
Table 1. Mole Fraction values of products in total reflux operation

Column
Position 1 2 3 4 5 6 7
T (C) 78.4 78.8 79.1 79.7 79.6 79.5 80
Column
Position 8 9 10 11 12 13 14
T (C) 80.8 84.3 79.3 28.5 34.8 64.6 22.1
Table 2. Temperature values at various plate positions

6.1.2 Partial Reflux (3:1)

Trial Number Mole Fractions of Overhead Mole Fractions of Bottom
Product Product
Ethanol Water Ethanol Water
1 0.4370 0.5630 0.1509 0.8491
2 0.5044 0.4956 0.1484 0.8516
Average 0.4707 0.5293 0.1496 0.8504
Table 3. Mole Fraction values of products in partial reflux (3:1) operation

Column
Position 1 2 3 4 5 6 7
T (C) 80.4 80.8 81.2 82 81.7 81.7 82.3
Column
Position 8 9 10 11 12 13 14
T (C) 83.1 85.5 81.3 29.2 35.9 57.9 26.3
Table 4. Temperature values at various plate positions

8
6.1.3 Partial Reflux (5:1)
Trial Number Mole Fractions of Overhead Mole Fractions of Bottom
Product Product
Ethanol Water Ethanol Water
1 0.4242 0.5758 0.1128 0.8872
2 0.4360 0.5640 0.0931 0.9069
Average 0.4301 0.5699 0.1029 0.8971
Table 5. Mole Fraction values of products in partial reflux (5:1) operation

Column
Position 1 2 3 4 5 6 7
T (C) 80 80.5 80.8 81.5 81.4 81.4 82.1
Column
Position 8 9 10 11 12 13 14
T (C) 83.7 86.7 81.2 28.6 35.3 61.6 23.2
Table 6. Temperature values at various plate positions

9
7 DISCUSSION

7.1 OVERALL EFFICIENCY UNDER TOTAL REFLUX CONDITIONS

7.1.1 Determining Column Efficiency Using Fenskes Method

Fenskes method uses the following formulas:

xA xB

log
[ ( )( )
xB D xA B
]

log [ ( AB ) avg ]
n+1=

( AB )avg= ( AB ) D ( AB )B =
( P A
PB )( )
d
P A
P B b

Where n equals to the number of theoretical plates, xA is the mole fraction of

the more volatile component, xB is the mole fraction of the less volatile component,
and ( AB )avg is the average relative volatility. The subscripts D and B denote the
source of the sample from which the mole fraction is obtained. D means that the
sample was obtained from overhead product, and B means that the sample was
obtained from bottom product. ( AB )D & ( AB )B are calculated from the ratio of
vapor pressure of the more volatile component to the vapor pressure of the less
volatile component at overhead and bottom temperatures.
The calculation of Vapor Pressures will be done using the Antoine equation.
The parameters used are obtained from literatures.

B
log P (mmHg)= A
C+T ( C )

Species A B C Temperature
Range
Ethanol 7.68117 1332.04 199.2 77-243oC
Water 8.07131 1730.63 233.426 1-100oC
Table 7. Antoine equation parameters for water and ethanol

10
 Temperature of overhead product is recorded as T1, and for the bottom
product, it is recorded as T9 in the results. According to the tabulated temperature,
in the total reflux experiment,
T1 = 78.4oC and T9 = 84.3oC
Sample calculation will be shown for calculating the vapor pressures of
ethanol and water at overhead product temperatures. The other vapor pressure
values are obtained using the same method.

At overhead temperature,

1332.04
log P A (mmHg)=7.68117 =2.883
199.2+78.4

P A =763.405 mmHg

1730.63
log PB (mmHg)=8.07131 =2.521
233.426+78.4

PB =332.142 mmHg

Temperature P A (mmHg) PB (mmHg) P A / P B = AB

78.4oC 763.405 332.142
2.298 ( AB )D
(Overhead Product)
84.3oC 960.763 421.099
2.282 ( AB )B
(Bottom Product)
Table 8. Values of vapor pressure of ethanol and water at overhead and bottom
products

Therefore,

( AB )avg= ( AB ) D ( AB )B =2.2982.282=2.29

Recalling Table 1,

Trial Number Mole Fractions of Overhead Mole Fractions of Bottom

Product Product
Ethanol Water Ethanol Water
1 0.6549 0.3451 0.2135 0.7865
2 0.6460 0.3540 0.2029 0.7971
Average 0.6505 0.3495 0.2082 0.7918
Table 1. Mole Fraction values of products in total reflux operation

11
Substituting the values into Fenskes method formula,

log
[ ( 0.6505
0.3495 )D
( 0.7918
0.2082 )
]
B

log [ 2.29 ]
n+1=

n+1=2.362

n=1.362 1

Therefore, the column efficiency can be calculated as:

Number of Theoretical Plates 1

E= 100 = 100 =12.5
Number of Actual Plates 8

7.1.2 Temperature Distribution in the Distillation Column (Total Reflux)

Temperature Distribution Across The Column

(Total Reflux)
85
84
83
82
81

Temperature (oC) 80
79
78
77
76
75
0 1 2 3 4 5 6 7 8 9 10

n-th Plate

Fig 2. Temperature distribution across the column (Total Reflux Condition)

T1 until T8 are the temperature values for each respective plate number, while
T9 is the temperature for the reboiler.

12
7.2 EFFECT OF REFLUX RATIOS ON THE SEPARATION
7.2.1 Temperature Distribution in the Distillation Column (Partial Reflux)

Temperature Distribution Across The Column

(Partial Reflux 3:1)
86
85
84
83
82
Temperature (oC) 81
80
79
78
77
0 1 2 3 4 5 6 7 8 9 10

n-th Plate

Temperature Distribution Across The Column

(Partial Reflux 5:1)
88

86

84

Temperature (OC) 82

80

78

76
0 1 2 3 4 5 6 7 8 9 10

n-th Plate

Fig 3. Temperature distribution across the column (Partial Reflux 3:1 Condition)

Fig 4. Temperature distribution across the column (Partial Reflux 5:1 Condition)

T1 until T8 are the temperature values for each respective plate number, while
T9 is the temperature for the reboiler.

13
7.2.2 Calculation of Number of Theoretical Plates Using McCabe-Thieles
Method

Using the constant molal overflow (CMO) approximation, the equations for
operating lines are able to be simplified in terms of reflux ratio and boil-up ratio.
However, since the feed line can also be drawn on the x-y equilibrium plot, only one
equation is needed. Since reflux ratio is known, the rectifying section operating line
will be used altogether with the feed operating line.
R 1
y n+1= xn+ x
R+1 R+1 D

q 1
y= x+ x
q1 1q F

The classifications of feed based on its q value is as follows:

Feed q
Characteristic
Liquid (Cold) q>1
Saturated q=1
Liquid
Liquid/Vapor 0<q<
1
Saturated q=0
Vapor
Superheated q<0
Vapor
Table 9. Classification of feed using q values based on its thermal condition
If the feed falls into the category of subcooled liquid, it can be calculated
using the following formula:
T b T f
q=1+C pL

And if it falls into the category of superheated vapor, the calculation follows:
T f T d
q=C pV

Where C pL and C pV are the specific heat of liquid and vapor respectively.

Subsequently, T f is the feed temperature, Tb is the bubble point temperature

and Td is the dew point temperature. Finally, denotes the heat of

vaporization of the feed. These parameters will have their value based on the

14
composition of the feed itself. The equation for q also can be rewritten in terms of
the enthalpy value, namely:
H h f
q=
Hh

Where H is enthalpy of the saturated vapor state, h f is the enthalpy of the

feed at its state, and h is the enthalpy value of the saturated liquid state.

15
7.2.2.1Partial Reflux 3:1 (R=3)

With R = 3 and xD = 0.4707, the operating line equation above can be re-
written as:
3 1
y n+1= xn + 0.4707=0.75 x n+ 0.118
3+1 3+1

In determining the q value for the feed line, according to the method
elaborated above, the various enthalpy values need to be determined first using an
enthalpy diagram for ethanol-water system.

Fig 5.
Enthalpy
diagram of ethanol-
water system,
based on
ethanol weight
fraction

Since the

available
enthalpy graph
is
based
on the
ethanol
weight

fraction, the
feeds ethanol mole fraction (xf = 0.2703) value needs to be converted first into the
weight fraction, wf.

1x
x f M w EtOH +( f )M w H 2 O
x M w EtOH
wf= f

Molecular weights of water and ethanol are 18.015 g/mol and 46.07 g/mol
respectively.

16
 0.270346.07
wf= =0.4865
0.270346.07+(10.2703)18.015

Feed temperature Tf can be obtained from the value of T14, therefore,

Tf = T14 = 26.3oC
Based on the graph, the values of H, h f, and h are approximately:

H (wf=0.4865, saturated 450 kcal/kg

vapour)
hf (wf=0.4865, Tf = 26.3oC) 11.8 kcal/kg
H (wf=0.4865, saturated 67.7 kcal/kg
liquid)
Table 10. Enthalpy values of various conditions of the binary mixture at the feed
composition (3:1)
Therefore, q can be calculated as:

H h f 45011.8
q= = =1.146
Hh 45067.7

As q is known, the feed line equation can be obtained:

1.146 1
y= x+ 0.2703=7.849 x1.851
1.1461 11.146

Values of mole fractions at overhead and bottom product can be obtained by

referring to Table 3. Based on Table 3,

17
 xD = 0.4707 and xB= 0.1496
Proceeding with the McCabe-Thieles procedure,
the result obtained is as follows:

The graph shows that the result obtained from

McCabe-Thieles method is that the theoretical
number of plate is one. This number already
includes the partial reboiler. The optimum feed
stage can also be concluded to be at the first
plate.

Fig 6. McCabe-Thieles
method applied to
distillation column under
partial reflux (3:1)

18
7.2.2.2Partial Reflux 5:1 (R=5)

With R = 5 and xD = 0.4301, the operating line equation can be re-written as:

5 1
y n+1= xn + 0.4301=0.833 x n +0.072
5+1 5+1

As the feed composition was not changed, therefore, xf and wf are constant.
xf = 0.2703 and wf = 0.4865
Based on Figure 5 and the feed temperature (T 14 = 23.2oC), the enthalpy values are
approximately:

H (wf=0.4865, saturated 450 kcal/kg

vapour)
hf (wf=0.4865, Tf = 23.2oC) 10.2 kcal/kg
H (wf=0.4865, saturated 67.7 kcal/kg
liquid)
Table 11. Enthalpy values of various conditions of the binary mixture at the feed
composition (5:1)

Therefore, q can be calculated as:

H h f 45010.2
q= = =1.15
Hh 45067.7

And feed operating line can be written as:

1.15 1
y= x+ 0.2703=7.667 x1.802
1.151 11.15

Values of mole fractions at overhead and

bottom product can be obtained by referring
to Table 5. Based on Table 3,
xD = 0.4301 and xB= 0.1029
Proceeding with the McCabe-Thieles
procedure, the result obtained is as follows:
The graph above also shows that the result
obtained from McCabe-Thieles method is that
the theoretical number of plate is one. Similar

19
with the previous case, this number already includes the partial reboiler, and the
optimum feed stage is at the first plate.

Fig 7. McCabe-Thieles method applied to

distillation column under partial
reflux (5:1)

7.3 DIFFERENCES BETWEEN THEORETICAL AND EXPERIMENTAL

NUMBER OF PLATES

It can be observed that all theoretical calculations yield the same result; they
all give a very low number of plates required. Both Fenskes method and McCabe-
Thieles method yield one plate as the final answer. McCabe-Thieles method always
include one stage as the reboiler. Therefore, rightfully, the number of plates
obtained through Fenskes method should not be the same as the one obtained
through McCabe-Thieles method. Moreover, the equipment itself is equipped with
eight plates, and with the reboiler, it will be considered as nine plates when plotted
in a McCabe-Thieles method. Together with the theoretical contradiction, this huge
deviation suggests that this equipment has many flaws in it.
In the calculation process, many assumptions were invoked. The deviations in
the number of plates may have risen from the assumptions that are rendered
invalid in this experiment.
The calculation process assumed constant molal overflow. In CMO, all liquid
streams have the same flow rate, and all the vapor streams also have the same flow
rate. However, its validity is dependent upon many factors. One of them is that the
molar enthalpy of vaporization of the binary mixture must not be a function of its
composition. Referring back to Figure 5, it is clear that that is not the case in this
experiment. Molar enthalpy of vaporization of the ethanol-water mixture will change
with composition accordingly. Usage of such invalid assumption can cause many
discrepancies in the result.
Ideality of all liquids and gas was also assumed. It can be seen from the
calculation process of Fenskes method. The relative volatility was calculated
straightly as the ratio of vapor pressures obtained from Antoine equation. While
these values can provide a reasonable approximation, this may not always be the
case. Under the experiment conditions, the gases and the liquids involved were very
likely to exhibit a great degree of non-ideality. Therefore, this assumption is also
likely to be rendered invalid, and its usage will introduce more errors to the final
result.
In doing the McCabe-Thieles procedure, many steps were also prone to
uncertainties and mistakes. The number of stages obtained will depend on the
location of all operating lines. In this experiment, the operating line of rectification

20
section and the feed line served as the basis of constructing the operating line of
the stripping section. This means that the operating line of the stripping section is
affected by all variables that constituted the two other lines. One of those variables
is the quality of the feed, q. q is obtained through the calculation using enthalpy
data. The data itself was obtained from an enthalpy chart. This method of obtaining
data is also very prone to mistakes. As the chart provided does not have a detailed
scale, there will be a lot of mistakes and inconsistencies in reading. The error itself
will snowball through the cascade of calculations, causing the final result to differ
greatly from the right value.
On the equipment itself, there may be leftovers of the previous experiments,
and it can cause errors to the measurement value. As it is impossible to drain all
poured materials using pipes and valves alone, some residual materials may end up
to stay in the equipment, and causes discrepancy in the measurement values, as
the purity of the products will be jeopardized. All the graphs obtained from the GC
apparatus also indicated the presence of methanol inside the sample. While it has
been neglected in doing all the calculations above, it should not have been
neglected in obtaining an accurate result. The presence of methanol makes all
calculation steps and data processing to be wrong. All formulas used are intended
for binary mixture, and presence of a third compound will render all these formulas
to be invalid. The enthalpy data was also obtained from the chart dedicated for
binary mixture of ethanol and water. This will further add to the errors introduced to
the obtained values.
Another possible technical issue is the possible problems and defects
associated with the apparatus itself. In an ideal distillation process, each tray
exhibits vapor-liquid equilibrium. However, this may not be the case in real
experiments. Holes in the trays may be clogged by residuals from previous
experiments, and fouling may also have occurred. This causes vapor and liquid
passing the trays to have less contact area, allowing less exchange of materials,
and they may not reach equilibrium because of that. If the liquid throughput by the
reflux is too large, it can cause problems as well. Flooding is possible if the
distillation column is not large enough. If flooding occurs, the liquid may flow in an
opposite direction than intended. This may lead to contamination of the samples,
and therefore discrepancies in the measurement of concentration. A prolonged use
of the apparatus may have also caused damage in the internal parts. Liquid hold-
ups may corrode the parts of the apparatus, and it will further affect the
performance of the distillation column. All these factors may lead to a decrease in
the efficiency of the column itself.

7.4 REFLUX RATIO

It can be seen from the graphs produced in McCabe-Thieles method that for a
given degree of mixture separation, an increase of reflux ratio R will reduce the
stages required to reach a certain separation level. An increase in reflux ratio will
affect the position of the rectification section operating line. As the slope is defined

21
as: R/(R+ 1) , the value of 1 will become insignificant as the value of R grows, and
the slopes value will approach 1. This translates in the operating line being further
and further away from the equilibrium curve.
The driving force of separation is directly proportional with the distance of a
given point from the equilibrium. The further it is, the greater the separation will be.
The number of stages required will then be reduced as it has been compensated by
the increase in the driving force of the separation process.
However, the scenario above is only applicable to the design process of a
distillation column. In this experiment, the number of stages in the column is
already fixed. Therefore, the observable effect of changes in reflux ratio will be in
the form of difference in the degree of mixture separation. As the separation
driving force increases, the mole fraction of the more volatile component in the
overhead product can be expected to increase as well. As the process has a fixed
number of stages, the thing that must be changed in the McCabe-Thieles diagram
is the distance between xD and xB, or in other words, the degree of separation.
While it goes as explained above theoretically, it is not fully possible to prove
the validity of the points mentioned above using the obtained results. Both xD and
xB always change during all three different conditions under which the column was
operated. Therefore, there are no standard or benchmark that can be used to
compare the performance of the column under different reflux ratios. The
fluctuations in all the values suggest that the molar fractions of the overhead and
bottom product are affected not only by the reflux ratio, but by other variables as
well. However, the widening gap between xD and xB observed when the reflux ratio
is increased from 3 to 5 to total reflux, somehow agrees to the theoretical approach,
yet further analysis still need to be done in proving this connection between reflux
ratio and separation results.

22
7.5 TEMPERATURE VARIATION ACROSS THE COLUMN

Theoretically, an increase in temperature should be observed down across the

distillation column, from the rectification section to the stripping section. Liquids
evaporate as it goes down in the column, and vapors condense as it goes up. This
means that the higher section of the column is always at a higher temperature than
the bottom ones.
However, the temperature values recorded does not always exhibit this
predicted trend. All three processes exhibit a decrease in temperature between the
fourth and the fifth tray. The total reflux process also exhibit the decrease between
the fifth and the sixth tray. This is inconsistent with the theoretical statement above
that suggested a sustained increase in temperature down across the column.
The feed entered the column at the fourth tray, and this is very likely to be the
reason of the slight temperature drop between the fourth and the fifth tray. The feed
temperatures, Tf, recorded show that the feed is a subcooled liquid. Therefore, it is
possible that when the feed enters the system, the heat carried by the flowing
streams are partially transferred to the feed stream. This loss of heat is therefore
responsible for the temperature drop.
As the liquid needs to be heated up until its bubble point when it entered the
column, again, the defects in the apparatus may pose some issues on the process.
Defects in the apparatus may interfere with the heat transfer, causing the liquid to
be at a temperature below its bubble point when it travels across the column. This
will cause a disruption in the vapor-liquid equilibrium that are supposed to be
achieved in all trays present, affecting the efficiency of the separation, as the
equilibrium point serves as one of the factors governing the separation process.
Furthermore, it may introduce errors to the measurement that ultimately causes
deviation in the final result.

7.6 SUGGESTED IMPROVEMENTS

Following the sources of errors that have already been mentioned in the
section above, there are several improvements that can be done to the experiment.
In order to increase the accuracy and the integrity of the experimental data,
the validity of every assumption used must be checked with the real experimental
conditions. The experiments can be modified in such ways so that it enables the
assumption to be used. For example, the constituents of the binary mixture can be
changed in obtaining a mixture with constant molar enthalpy of vaporization,
enabling constant molal overflow approximation to be used safely.
Great care must also be taken in ensuring the integrity of the mixture used.
The presence of methanol in the binary mixture disrupted the integrity of the
mixture itself, and it might have rendered all the calculation formulas and data

23
measured to be invalid. The presence of such compounds may also affect the vapor-
liquid equilibrium in each tray, thus the separation process that relies on it. This also
applies to the cleanliness of the apparatus. As residuals from previous experiments
also can introduce flaws to the experiment, it will be better if all parts of the
apparatus are to be made clean before the commencement of the experiment itself.
Maintenance of the apparatus itself is also important. Defects such as fouling can
affect the performance of the column itself, reducing the overall column efficiency
as it prohibits efficient mass transfer, hence separation from occurring.
Although the manual stated that students are supposed to wait around 40
minutes for the system to reach steady state, this may not always be the case. If
the measurement data are recorded when the system is yet to reach steady state,
there will be errors in values, and it will bring heavier impact to the integrity of the
result as it is carried forward in the calculations.
Random errors are also very likely to be responsible for the discrepancies in
values. One way to alleviate the effect is to take more sample and do more
measurements. By taking the average of more samples, the adverse effects
generated from random errors can be reduced, improving the overall result.

24
8 CONCLUSION

This experiment enabled the students to investigate the performances of a

distillation column under three different conditions, namely total reflux, and two
partial reflux with reflux ratios of 3 and 5. The product samples were then collected
and measured, and the temperatures across the column were recorded. Then, the
experiment also enabled the students to apply the concepts of unit operations, such
as using Fenskes method and McCabe-Thieles method in determining the
theoretical number of plates required for a certain degree of separation.
Furthermore, this experiment also provided the students with the concept of
determining the efficiency and performance measure of a distillation column. The
experimental results, albeit having heavy deviations from the theoretical values,
have also helped the students in identifying the flaws contained in the experiment
by using the fundamental concepts of the distillation column itself. Therefore, the
experience and insights gained by the students have proven that this experiment
has successfully completed its aim in equipping the students with more knowledge
regarding the distillation column.

25
9 APPENDIX

9.1 IMAGES OF THE APPARATUS USED

Fig 8. Gas Chromatography Apparatus

Fig 9. Specimen Injection Site on Gas Chromatography Apparatus

26
Fig 10. Feed Tank

27
Fig 11. UOP3CC Distillation Column Apparatus

Fig 12. Control Panel of the Distillation Column

9.2 RAW GAS CHROMATOGRAPHY DATA AND SAMPLE CALCULATIONS

9.2.1 Gas Chromatography

The data obtained from GC is in the form of signal peaks, with the
corresponding area below the graphs reflecting the amount of that particular
substance in the measured sample.

Fig 13. Example of analysis data

obtained from GC

28
 In the graph, three peaks can be seen. According to the manual, the first
peak is water molecule, the second being methanol molecule, and the last one
represents ethanol molecule. In this experiment, the presence of methanol itself
was neglected. The actual amount of each substance is later obtained from the
conversion of the area below their respective peaks.
The conversion is done through a calibration equation that is unique for every GC,
the equations used are:
For Ethanol

Area= (1371648799 Amount of moles ) +27.507349

Area27.507349
Amount of moles=
1371648799

For Water

Area= (513413665 Amount of moles )49.239188

Area +49.239188
Amount of moles=
513413665

9.2.2 Raw Data from Gas Chromatography Apparatus

Total Reflux

Product Area Number of Moles

2nd
Bottom 1st Run Run 1st Run 2nd Run
1598.0 1578.6
EtOH 86 37 1.14503E-06 1.13085E-06
2116.7 2232.2
H2O 17 82 4.21873E-06 4.44383E-06
Overhea
d
2630.4 1955.0
EtOH 68 38 1.89769E-06 1.40527E-06
464.10 551.18
H2O 11 41 9.99857E-07 1.16947E-06

Partial Reflux (3:1)

Product Area Number of Moles

Bottom 1st Run 2nd 1st Run 2nd Run

29
 Run
1281.5 1206.6
EtOH 2 36 9.14238E-07 8.59643E-07
2591.8 2484.3
H2O 74 36 5.14422E-06 4.93476E-06
Overhea
d
2392.0 2745.8
EtOH 44 6 1.72386E-06 1.98181E-06
1090.8 950.47
H2O 46 55 2.2206E-06 1.94719E-06

30
Partial Reflux (5:1)

Product Area Number of Moles

2nd
Bottom 1st Run Run 1st Run 2nd Run
798.51 898.74
EtOH 22 95 5.62101E-07 6.35179E-07
2221.4 3126.7
H2O 86 06 4.4228E-06 6.18594E-06
Overhead
2792.1 2407.4
EtOH 06 58 2.01553E-06 1.7351E-06
1355.5 1103.1
H2O 25 53 2.73613E-06 2.24457E-06

Conversion from area to number of moles is done using the calibration equation of
the GC.
The sample calculation below will demonstrate the conversion from the area
obtained for ethanol species, in the first trial, under total reflux condition.
Area of ethanol peak = 1598.086

Area27.507349 1598.08627.507349 6
Amount of moles= = =1.1450310 moles
1371648799 1371648799

The mole fraction of the sample is then calculated as follow:

Amount of Ethanol (moles) : 1.14503106

6
Amount of Water (moles) : 4.2187310

1.14503106
Mole fraction of Ethanol= =0.2135
1.14503106 + 4.21873106

And thus, the result is consistent with the tabulated result in Table 1.

Initially, the binary mixture was also analyzed in order to determine its exact
composition. Result from GC:
Area under ethanol peak : 1700.89221
Area under water peak : 1641.76782
Amount of Ethanol : 1.22 10-6 moles
Amount of Water : 3.294 10-6 moles

31
 1.22106
Mole fraction of Ethanol= =0.2703
1.22106 +3.295106

Therefore, the xf used in all calculations above is consistent with the data obtained
from GC.

32
10 REFERENCES
1. http://webbook.nist.gov/cgi/cbook.cgi?
ID=C64175&Mask=4&Type=ANTOINE&Plot=on
2. http://webbook.nist.gov/cgi/cbook.cgi?
ID=C7732185&Mask=4&Type=ANTOINE&Plot=on
3. https://ntulearn.ntu.edu.sg/bbcswebdav/pid-677441-dt-content-rid-
2161684_1/courses/15S1-CH3802-LAB/LabManualComplete-CBE%20Year3-
forupload.pdf
4. http://www.engr.uconn.edu/~ewanders/CHEG237W/Bubblecap_fall01.pdf

33