Report

Distillation columns
Sieve tray column

Lene Tapio Pia Haarseth

lenekatr@stud.ntnu.no piakrih@stud.ntnu.no

Room K4-408
Gr B14

Performed 18. september 2012

Contents
1 Theory 3
1.1 Floodingpoint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Weepingpoint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 McCabe-Thiele method . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Gas chromatography . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Experimental 5
2.1 Column startup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Exercise 1: Time required for reaching steady state condition . . . . 6
2.3 Exercise 2: Efficiency vs. vapor velocity . . . . . . . . . . . . . . . . 6
2.4 Column shutdown . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3 Results 6
3.1 Time required for reaching steady state condition . . . . . . . . . . 6
3.2 McCabe-Thiele Diagram . . . . . . . . . . . . . . . . . . . . . . . . 7
3.3 Total column efficiency vs vapor velocity . . . . . . . . . . . . . . . 13

4 Discussion 14

5 Conclusion 15

A Laboratory safety 16

B McCabe-Thiele calculations 19

C Calculations 21
C.1 Water-ethanol mixture . . . . . . . . . . . . . . . . . . . . . . . . . 21
C.2 Calculating the mole fraction of ethanol . . . . . . . . . . . . . . . 22
C.3 Calculating vapor velocity . . . . . . . . . . . . . . . . . . . . . . . 23
C.4 Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

D McCabe-Thiele 27

1
E Assignment calculations 28
E.1 Mole fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
E.2 Vapor velocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
E.3 McCabe-Thiele . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

2
1 Theory
1.1 Floodingpoint
In a distillation column vapour flows up and liquid flows down. The flooding
point is referred to as the vapour velocity when liquid begins to accumulate in
the column. This happens when the vapour velocity gets to high, resulting in the
pressure exerted by the vapour flowing upwards balances the gravitational forces
that pull the liquid down. The flooding point can be observed by a sudden increase
in the pressure drop or when liquid starts flowing over the top of the column.[1].

1.2 Weepingpoint
At steady state conditions the vapour velocity through the perforations prevents
the liquid from dropping down to the next stage through the perforations. When
the vapour velocity gets too low the liquid is able to weep down through these per-
forations. This gives us less vapour/liquid contact, which result in lower efficiency
off the column[1].

1.3 McCabe-Thiele method

The McCabe-Thiele method was presented by MaCabe and Thiele in 1925. This is
perhaps the simplest and most instructive method for analysis of binary distillation[1].

Assumptions[3]:

• the molar heats of vaporization of the feed components are equal

• for every mole of liquid vaporized, a mole of vapour is condensed

• heat effects such as heats of solution and heat transfer to and from the dis-
tillation column are negligible.

3
The MaCabe-Thiele diagram can be used to determine the theoretical number of
trays in a distillation column. To construct the diagram the mass balance and VLE
(vapor-liquid-equilibrium) data for the lower boiling point component, component
A, is used. The x-axis in the diagram represents the mole fraction of A in gas
phase, and the y-axis represents the mole fraction of A in liquid phase. Plotting
the VLE data gives the equilibrium line, and the material balance gives the two
operating lines [1]:

In the enriching section the mole fraction of A in gas phase, yn+1 is given by

R xD
yn+1 = xn + (1.1)
R+1 R+1

where R is the reflux ratio, xn is the mole fraction of A in liquid phase at step n
and xD is is the mole fraction of A in liquid phase in the distillate.

In the stripping section the mole fraction of A in gas phase, ym+1 is given by

Lm W xW
ym+1 = xm − (1.2)
Vm+1 Vm+1

where Lm is liquid flow from tray m, Vm+1 is vapor flow from tray m+1, W is bot-
toms flow, xm and xW is mole fraction of liquid phase in tray m and the bottoms
flow.

The operating line will be equal to the x=y line when it is total reflux.

The parameter q represents the condition of the feed and is defines as follows

heat needed to vaporize 1 mole of feed at entering conditions

q= (1.3)
molar latent heat of vaporization of feed

And the q-line in the diagram is the line with slope q/(q-1) through the cross

4
point between the operating lines as well as the line x = y. To determine the
theoretical number of stages in a distillation steps are made in the McCabe-Thiele
diagram, starting at the top tray where x = xD . The steps are made by making
straight lines, alternately horizontal and vertical between the equilibrium line and
the operating lines, all the way down to x = xB , where xB is the mole fraction
of A in liquid phase at the bottom. These steps represent the theoretical trays
needed in the distillation[1].

1.4 Gas chromatography

Gas chromatography (GC) is used in analytical chemistry for separating and
analysing compounds that can be vaporized without decomposition. GC can be
used to test the purity of a particular substance, in this experiment, the purity of
ethanol after distillation[1].

2 Experimental
2.1 Column startup
We started the experiment with turning on the computer and the cooling water,
and made sure that the bottom and distillate tap lines were closed. Then we
charged the column with mixture(5800 ml) of water(3954 ml) and ethanol (1846
ml). Then we sat the column to total reflux by using the control box and sat
T=105◦ C. We also had to set the boilerńs power duty to 40% when the top tem-
perature substantially changed or the pressure difference over the column rised
quickly. Then we waited until the vapor stream started condensing(t=0).

In case of flooding conditions, we switched the heater off immediately, and in

any emergency situation we did turn off the heater but leaved the cooling water
on.

5
2.2 Exercise 1: Time required for reaching steady state con-
dition
We sat the reboiler to a desired value(30 %) and waited until the column started
to give condensate. Then we took samples from the top every 5 minutes within
1 hour, and 1 sample after 1 hour from the bottom. In the end we analyzed the
samples with GC.

2.3 Exercise 2: Efficiency vs. vapor velocity

We run the column with total reflux for five different boiler powers. Each time
we sat the power to the desired in the reboiler. Then we waiter until the column
reached the steady-state, we assumed this was after the time found in Exercise 1.
After that we took samples from the top and from the bottom, 20 ml each sample,
and wrote down the reflux rate at the time. Then we analyzed the samples with
GC and calculated the efficiency of the column and the theoretical number of
stages using the McCabe-Thiele method. We also calculated the vapor velocity.
In the end we identified the flooding and weeping points and wrote down the power
of the boiler at these two points.

2.4 Column shutdown

When we were finished with the exercises we turned the heater off and let the
column cool down. Then we emptied the column and turned off the cooling water
and the rest of the equipment.

3 Results
3.1 Time required for reaching steady state condition
The time required to reach study state condition was approximately 30 minutes,
see Fig. 1.

6
 Figure 1: The volum percent, %vol ethanol, plottet as a function of Time.

3.2 McCabe-Thiele Diagram

Making McCabe-Thiele diagrams, see Figure 2-6, for the different power used in the
experiment gave a total number of stages, including the reboiler, shown in Tabel
1. In the same table is also the theoretical stages in the column and referring to
the McCabe-Thiele diagrams for the different reboiler powers. The equilibrium
line in the McCabe-Thiele diagram is yet plotted as a function of xet , with values
from table 2 in appendix B [5].

It was impossible to make an McCabe-Thiele to reboiler power equal to 25% be-

cause this result was over the azeotropic point between ethanol and water.

Weeping point was observed to be when the reboiler power was 18 %, and the
flooding point at 45 %.

7
Table 1: Total number of stages and theoretical stages in the column at the dif-
ferent powers.

Reboiler Total number Theoretical McCabe-Thiele Di-

power of stages stages in the agram
column
20% 10 9 Figure 2
25% - - -
28% 6 5 Figure 3
30% 6 5 Figure 4
32% 4 3 Figure 5
35% 5 4 Figure 6

8
Figure 2: Red line is equilibrium line, green line is x ethanol = y ethanol, blue
line is number of stages. x ethanol is the mole fraction of ethanol in liquid phase,
y ethanol is mole fraction of ethanol in gas phase.

9
Figure 3: Red line is equilibrium line, green line is x ethanol = y ethanol, blue
line is number of stages. x ethanol is the mole fraction of ethanol in liquid phase,
y ethanol is mole fraction of ethanol in gas phase.

10
Figure 4: Red line is equilibrium line, green line is x ethanol = y ethanol, blue
line is number of stages. x ethanol is the mole fraction of ethanol in liquid phase,
y ethanol is mole fraction of ethanol in gas phase.

11
Figure 5: Red line is equilibrium line, green line is x ethanol = y ethanol, blue
line is number of stages. x ethanol is the mole fraction of ethanol in liquid phase,
y ethanol is mole fraction of ethanol in gas phase.

12
Figure 6: Red line is equilibrium line, green line is x ethanol = y ethanol, blue
line is number of stages. x ethanol is the mole fraction of ethanol in liquid phase,
y ethanol is mole fraction of ethanol in gas phase.

3.3 Total column efficiency vs vapor velocity

Total column efficiency is plotted as a function of vapor velocity in Figure 7.

13
 Figure 7: Total column efficiency plotted as a function of vapor velocity.

4 Discussion
The results presented is the results another group got on the Sieve-tray column.
This is because this column leaked when we run the experiment, which gave us
wrong results. After looking on the GC results from the first exercise we saw
that they were not as expected, and the system never reached steady-state when
we started exercise two. The temperature in the column was over 100 ◦ C, which
means that the ethanol had vapoured out of the system.

As we can see from Figure 1 it took approximately 30 minutes for the concen-
tration of ethanol to stabilize and the column to reach steady state conditions
in the first part of the experiment. When the system reached steady state, and
the system runs with total reflux, the concentration of the distillate will remain
constant.

The volume percent we got from the GC in the first part of the experiment seems
reasonable. After the column reached steady state, the volume percent of the dis-

14
tillate was stable around 93-94%, which was expected from the column. In the
second part the volume percent of the distillate varied a bit more, from 90-97%.
To get 97 vol% is impossible for this distillation, because this result is over the
azeotrope of a mixture between ethanol and water. There must have been an error
in the analysis of this sample. This error may come from bad measurements of the
solutions that were mixed before analyzing the sample with GC.

When the column efficiency was calculated for the different power duties, the
McCabe-Thiele diagram was used to find theoretical number of trays needed for
the distillation. For the sample that showed 97 vol% the McCabe-Thiele diagram
could not be drawn, because it was over the azeotropic point in the diagram. In
the plot of column efficiency vs. vapour velocity it seems that the column efficiency
decreases when the vapour velocity gets higher. It seems that the column is most
efficient right above the weeping point. This makes sense because this is the point
when there is most contact between vapour and liquid.

One possible error source in this experiment can be some flooding. Flooding
will give the wrong volume fraction of ethanol in the distillate when undistilled
solution mixes with the distillate. Another error source could be weeping, which
will give lower contact surface between vapor and liquid phase. This will result in
a lower mole fraction of ethanol in the distillate. Un mixed samples tested with
GC because of too little shaking, will give wrong results of the volume percent
ethanol in the sample. During the experiment 12 samples were taken. Because the
concentration in these samples were high, this could have altered the concentration
of ethanol in the system. When the samples were taken the stabilization of the
system were interrupted, and could also lead to wrong results. When the samples
from the bottom flow were taken the system were not totally homogeneous because
of lack of stirring. This may also affected the results.

5 Conclusion
The time needed to reach steady state condition was approximately 30 minutes.
The total column efficiency decrease with increasing vapor velocity, so at the lowest

15
vapor velocity the the column is most efficient . Weeping point was at 18 % of
total reboiler power, and flooding point at 45 %.

References
[1] Felles lab: Distillation Columns, 2012. Viewed 13. September 2012.
http://www.nt.ntnu.no/users/preisig/Repository/TKP4110
FellesLab/experiment%20descriptions/DistInstruct.pdf

[2] Gordon Aylward, Tristan Findlay, SI chemical data 6th edition, 2008

[3] Geankoplis, C. J., 2003. T ransport P rocesses and Separation P rocess P rinciples. 4th
ed. Prentice Hall. pp. 706-717.

[4] Green, D. W. and Perry, R. H., 2008.

P errys chemical Engineers handbook. 8th ed. McGraw-Hill. pp. 13-18-13-19,
14-36.

[5] Gmelin Handbook of Inorganic and Organometallic Chemistry p. 171

A Laboratory safety

16
 NTNU
Norges teknisk-
naturvitenskapelige
universitet

Kjemikaliekort

STOFFNAVN FORMEL
Etanol C2H5OH

GENERELL HELSEFARE
Ingen utpreget helsefare.
Molar
FYSIKALSKE Sm.p/f r.p Kokepkt. fl.p. masse Ekspl.gr. BRANNFARE
Meget brannfarlig
DATA
-117° C 79°C 425°C 46,07 g/mol 3-19 % REAKSJONSFARE
-
Alt arbeide med farlige kjemikalier skal utføres i avtrekkskap eller i
godt ventilert og godkjent rom. Nøddusj og mulighet for øyeskylling
VERNETILTAK skal finnes på arbeidsplassen. ADMINISTRATIV NORM
-
Skift forurensede klær. Vask hender og ansikt etter arbeid med
produktet. Åndedrettsvern (type A) skal anvendes ved forekomst av
damper/aerosoler. Ved risiko for direkte kontakt eller sprut skal
øyebeskyttelse benyttes. Bruk ikke kontaktlinser ved håndtering.
Vernehansker skal benyttes ved fare for direkte kontakt og sprut.

HELSEFARE
Vil irritere slimjimmene I nese og øvre luftveier.
BRANNSLUKKING
Innånding Slukkes med pulver, skum, karbondioksid
Vil kunne føre til brekninger, kvalme, absorpsjon av toksiske
eller vann.
kvantiteter kan gi symptomer som ved alkoholpåvirkning. Eufori.
Svimmelhet. Bevisstløshet. Åndedrettstopp.
Svelging
Lett irritasjon. Ved langvarig hudkontakt med etanol kan føre til
avfetninga av huden og hudbetennelse. Hudsprekker.
Hud
MERKNADER
Irriterer øyne.
Øyne -

FØRSTEHJELP ØYNE SVELGING

Skyll lenge med rennende vann. Dersom irritasjon
vedvarer gå til spesialist.
HUD
Ta av forusenset tøy og skyll godt med vann.
Drikk mye vann. Kontakt lege. Gi aktivt kull om mulig.
Fremkall ikke brekninger. Ikke melk/matolje. Hold
luftveier åpne.
INNÅNDING
Frisk luft og hvile. Gi kunstig åndedrett ved behov.

SPESIELLE MERKNADER

Destrueres etter lokale forskrifter. Håndter forurenset forpakning på samme måte som substansen. Klassifisert
som farlig avfall. Avfallsgruppe: Organiske løsemidler uten halogen. Forhindre utslipp til kloakk:
SPILL/
Eksplosjonsrisiko.
RESTER
Forpakningen oppbevares lukket og tørt ved romtemperatur (+15 °C til +25 °C). Holdes vekk fra
antennelseskilder.
LAGRING
B McCabe-Thiele calculations
In table 2 are the equilibrium values for mole fraction of ethanol in gas phase, x1 ,
and mole fraction of ethanol in liquid phase, y1 [5].

Table 2: Values for mole fraction of ethanol in gas phase, x1 , and mole fraction of
ethanol in liquid phase, y1 at equilibrium.

19
%empirical data for ethanol
xe_e=[0 0.0051 0.0069 0.0162 0.0316 0.0823 0.1065 0.1368 0.1450 0.1770 0.4034
0.5733 0.7152 0.7715 0.8160 0.8180 0.8386 0.8780 0.9167 0.9910 1];
ye_e=[0 0.0486 0.0752 0.1442 0.2429 0.3985 0.4513 0.4812 0.4805 0.5095 0.6120
0.6849 0.7607 0.7961 0.8246 0.8322 0.8450 0.8789 0.9117 0.9892 1];

%uses the function polyfit to find the coeffesients for the polynom fitting
the data= eq. line
%uses polyval to find the polynom
x_poly = 0:0.001:1;
n=10;
p = polyfit(xe_e,ye_e,n);
f = polyval(p,x_poly);

%Our polynom, with the coeffesients p

g
=@(x,y)(p(1)*x^10+p(2)*x^9+p(3)*x^8+p(4)*x^7+p(5)*x^6+p(6)*x^5+p(7)*x^4+p(8)*x
^3+p(9)*x^2+p(10)*x^1 - y);

%Plotting the curves

%The eq. line
plot(x_poly,f,'r');
hold on;

% operating line
set(line([0 1],[0 1]),'Color',[0 1 0]);

% distillate, xd, and bottom, xb, mole fractions are defined from
% experimental data
xd=0.744;
xb=0.033;

%The figure
figure(1);
hold on;
axis([0 1 0 1]);

%Labeling the axis

xlabel('x ethanol');
ylabel('y ethanol');

%Drawing the steps

i=1;
xp(1)=xd; %start x-value
yp(1)=xd; % start y-value
y=xd;
while (xp(i)>xb), %gÂr xp(bare x verdien som varierer)helt til vi kommer til
xb(nederst p eq.line)
xp(i+1)=fzero(g,0.5,[],y);
set(line([xp(i) xp(i+1)],[yp(i) yp(i)]),'Color',[0 0 1]);
yp(i+1)=xp(i+1);
y=yp(i+1);
if (xp(i+1)>xb) set(line([xp(i+1) xp(i+1)],[yp(i) yp(i+1)]),'Color',[0 0
1]);
end
i=i+1;
end
% Displaying the (integer) number of necessary steps.
disp('Number of steps:');
disp(i-1);
hold off;
C Calculations
Physical data used in the calculations is found in SI[2] , unless other is stated.

C.1 Water-ethanol mixture

The mole fraction of ethanol, xet , is given by

net
xet = (C.1)
net + nw

where net is the number of moles of ethanol and nw is the number of moles of water.

Moles of ethanol, net , is given by

Vet · ρet
net = (C.2)
Met

where Vet is the volume of ethanol and Met is the molar mass of ethanol.

Moles of water, nw , is given by

(Vtot − Vet ) · ρw
nw = (C.3)
Mw

where Vtot is the total volume, ρw is the density of water and Mw is molar mass of
water.

Substituting Eq. C.2 and Eq. C.3 into Eq. C.1 gives the amount of pure ethanol
Vet ·ρet
Met
xet = (C.4)
Vet ·ρet
Met
+ (Vtot −V et )·ρw
Mw

where ρet is the density of ethanol. With the values xet = 0, 12, ρet = 791 mg3 ,

21
 g
Met = 46, 07 mol , ρw = 998 mg3 , Mw = 18, 02 mol
g
, Vtot = 5800ml and by using Wol-
framalpha we get

Vet =1772 ml

The amount of ethanol needed when we use ethanol with 96% purity, Vet96% , is
given by

Vet
Vet96% = = 1846ml (C.5)
0, 96

Water, Vw , needed in the mixture is given by

Vw = Vtot − Vet96% = 3954ml (C.6)

C.2 Calculating the mole fraction of ethanol

All calculations for the distillate is done with values from Table 4 and calculations
from the bottom of the column is done with values from Table 9. Both tables are
in Appendix C.4.

Calculating number of moles from volume percent, %volet , ethanol by

%volet · V · ρet
net = (C.7)
Met

Number of moles of water is given by

(1 − %volet ) · V · ρw
nw = (C.8)
Mw

Then the mole fraction of ethanol, xet is given by

net
xet = (C.9)
net + nw

22
C.3 Calculating vapor velocity
The vapor velocity, v, is given by

V̇gas V̇gas
v= = π (C.10)
A 4
· D2

where V̇gas is the volume of the sample in gas phase, A is cross section area of
column and D is the inner diameter of the column.

V̇gas is given by
ngas · RT
V̇gas = (C.11)
P
where R is the universal gas constant, T =79 ◦ C is the temperature, P =1 bar is
the pressure and ngas is the total number of moles.

ngas is given by
ngas = ntot = net + nw (C.12)

C.4 Values
Number of theoretical stages in the column and the reflux at the different powers
are listed in Table 3, and there is also referred to which McCabe-Thiele Diagram
belonging to them.
Values used in the calculations for the distillate are listed in Table 4, and values
used in the bottom calculations are listed in Table 9.

23
Table 3: Total number of stages and theoretical stages in the column at the dif-
ferent powers.

Reboiler Reflux Theoretical McCabe-Thiele Di-

power stages in the agram
column
20% 10 9 Figure 2
25% - - -
28% 6 5 Figure 3
30% 6 5 Figure 4
32% 4 3 Figure 5
35% 5 4 Figure 6

Table 4: Values used in the calculations for the distillate and vapor velocity. R
is the reflux, topget is gram ethanol, T opvol% is the volume percent of ethanol,
gvann is gram water, xet is the mole fraction of ethanol, ngas is total number of
moles, the first V gas is volume gas, the second V gas is volume gas, and vg as is
the vapor velocity. All values are for the distillate.

24
Figure 8: Values used in the calculations for the bottom of the column.
Bottom vol% is the volume percent ethanol, Bot mlet is mL ethanol, Bot get
is mass ethanol, Bot net is number of moles ethanol, gw is the mass of water, nw
is the number of moles of water, xbot , et is the mole farction of ethanol. All values
are for the bottom of the column.

25
26
D McCabe-Thiele
Figure 9: Hand drawn McCabe-Thiele diagram on mm-paper.

27
E Assignment calculations
E.1 Mole fractions
The volume of the pycnometer, Vi is given by

Wi,H2 O − Wi
Vi = (E.1)
ρH2 O

where Wi,H2 O is the weight of the pycnometer with water, Wi is the weight of the
empty pycnometer and ρH2 O is density of H2 O.
The density of a sample, ρsample is given by

Wi,sample − Wi
ρsample = (E.2)
Vi

where Wi,sample is the weight of the pycnometer with the sample. V1 =4,988 cm3
and V2 =4,999 cm3

ρ1 = 0, 8237 g/cm3 gives a weight percent of 87, 8% ethanol [4]

, and ρ2 = 0, 9847
g/cm3 gives a weight percent of 8, 0% ethanol.
The mole fraction of ethanol is given by

met wet %
net net Met Met
xet = = = met mw = (E.3)
ntot net + nw Met
+ Mw
wet %
Met
+ (100−w
Mw
et %)

where ntot is the total number of moles in the sample, met and mw is the masses
of ethanol and water, and wet % is the weight percent of ethanol in the sample.
This gives

xet,1 = 0, 738
xet,2 = 0, 033

which is approximately the same result we get from GC.

28
E.2 Vapor velocity
The vapor velocity, v, is given by

V̇gas V̇gas
v= = π (E.4)
A 4
· D2

where V̇gas is the volume of the sample in gas phase, A is cross section area of
column and D is the inner diameter of the column.

V̇gas is given by
ngas · RT
V̇gas = (E.5)
P
where R is the universal gas constant, T =105 ◦ C is the temperature, P =1 bar is
the pressure and ngas is the total number of moles.

ngas is given by

mtot ρ1 · Vtot
ngas = ntot = = (E.6)
Mtot xet · Met + (1 − xet ) · Mw

where mtot is the total mass, Mtot is the total molar mass and ρ1 is the den-
sity of the distillate.

By solving Eq. E.4-E.6 with the values D=0,05 m, Vtot =0,883 ml, and values
for ρ1 , xet and the molar masses from section E.1 we find

v=0,232 m/s

E.3 McCabe-Thiele
Values for x ethanol and y ethanol is found in Gmelin [5]

29
Figure 10: Red line is equilibrium line, green line is x ethanol = y ethanol, blue
line is number of stages. x ethanol is the mole fraction of ethanol in liquid phase,
y ethanol is mole fraction of ethanol in gas phase.

Number of stages: 5

30