Solution 1.

See section 11.3.2 and the dew and bubble point calculations in example 11.9.

This type of problem is best solved using a spread-sheet, see the solution to problem 11.2.

Solution 11.2

This problem has been solved using a spread-sheet (MS WORKS). The procedure set out in
example 11.1 was followed.

The L/V ratio is made a variable in the spread-sheet and progressively changed until
convergence between the assumed and calculated value is achieved.

i zi Ki Ki.zi zi/Ki
C3 0.05 3.3 0.165 0.02
iC4 0.15 1.8 0.27 0.08
nC4 0.25 1.3 0.325 0.19
iC5 0.2 0.7 0.14 0.29
nC5 0.35 0.5 0.175 0.70
sum 1 1.075 1.28

So feed is 2-phase.

Try L/V = 4.4

Ki Ai = L/VKi Vi = Fzi/(1+Ai)
C3 3.3 1.33 2.14
iC4 1.8 2.44 4.35
nC4 1.3 3.38 5.70
iC5 0.7 6.29 2.75
nC5 0.5 8.80 3.57
sum 18.52

L= 81.48 L/V = 4.40

convergence test % -0.02

1
Solution 11.3

As the relative volatility is low, this problem can be solved using the Smoker equation.

The required recovery of propylene overhead is not specified. So, the Smoker equation
program was used to determine the relationship between recovery and the number of stages
needed.

Mol masses propane 44. Propylene 42

Feed composition, mol fraction propylene = ___90/42____ = 0.904

10/44 + 90/42

The bottoms composition can be determine by a material balance on propylene:

D = overheads, B = bottoms , F = feed = 100kmol/h

0.904 x 100 = 0.995D + xb B

D + B = 100

Recovery, Q = D/F = 0.995D/(100x0.904)

So, D = 90.854Q, B = 100 - 90.854Q

and xb = (90.4 - 90.854Q)/(100 - 90.854Q)

Q = 0.95 0.99 0.995

xb = 0.299 0.045 0.000028

Number of stages calculated using the Smoker equation program

Rectifying section 33 33 33

Stripping section 52 73 152

Total 85 106 185

The number of stages required increases markedly when the recovery is increased to above
0.95. The higher the recovery the lower the loss of propylene in the bottoms. The loss of
revenue must be balanced against the extra cost of the column. A recovery of 0.99 would
seem to be a good compromise.

Loss of propylene per 100 kmol/h feed = 0.045(100 – 90.854 x 0.99) = 0.45 kmol/h

2
Solution 11.4

Outline solution only

1. Make rough split between the tops and bottoms.

Overheads, 98% recovery of nC4 = 270 x 0.98 = 265 kg/h

Bottoms, 95% recovery of iC5 = 70 x 0.95 = 67 kg/h

feed tops btms kg/h

C3 910 910
i C4 180 180
nC4 270 LK 265 5
iC5 70 HK 3 67
nC5 90 90
nC6 20 20

2. Estimate the bubble and dew points of the feed, tops and bottoms using the methods given
in section 11.3.3, equations 11.5a and 11.5b. See also example 11.9.

3. Relative volatility of each component = Ki / KHK. K values from the De Priester charts,
section 8.16.6.

3. Determine Nm from equation 11.58.

4. Determine Rm using equations 11.60, 11.61.

5. Find N for a range of reflux ratios. Erbar and Madox method, Fig 11.1; see example 11.7.

6. Select the optimum reflux ratio.

7. Find the number of theoretical plates need at the optimum reflux.

8. Determine the feed point using the Kirkbride equation, 11.62.

9. Estimate the column efficiency using O’Connell’s correlation, Fig. 11.13.

The liquid viscosity can be estimated using the method given in Appendix D.

The problem asks for the stage efficiency, but as a rigorous method has not been used to
determine the number of theoretical plates, an estimate of the overall efficiency will be
good enough. The stage (plate) efficiency could be estimated using the AIChemE method
given in section 11.10.4.

10. Calculated the actual number of plates required and the feed point.

11. Estimate the column diameter using equation 11.79.

3
Solution 11.5

As this is a binary system, the McCabe-Thiele method described in section 11.5.2 and
illustrated in example 11.2 can be used.

Compositions

Feed 60% mol acetone, overheads 99.5%mol acetone.

Material balance on 100 kmol/h feed.

Acetone overhead, 95 % recovery, = 60 x 0.95 = 57 kmol/h

Acetone in bottoms = 60 - 57 = 3 kmol/h

Total overheads = (57/99.5) x 100 = 57.3 kmol/h

Total bottoms = 100 - 57.3 = 42.7 kmol/h

Mol fraction acetone in bottoms = 3/42.7 = 0.070 (7% mol)

q line

The feed is essentially at its boiling point, 70.2 °C, so the q-line will be vertical.

McCabe-Thiele method

1. Draw the diagram using the equilibrium data given in the problem, use a large scale.

2. Determine the minimum reflux ratio.

3. Draw in the operating lines for a reflux ration 1.5 times the minimum

4. Step off the number of theoretical plates.

5. Step off the number of real plates using the plate efficiency given; see Fig.11.6.

The accuracy of the determination of the number of plates required in the rectifying section
can be improved by plotting that section of the equilibrium diagram on a log scale; see
example 11.2.

4
Solution 11.6

As this is to be treated as a binary system, the McCabe-Thiele method can be used to

determine the number of theoretical stages; see section 11.5.2 and example 11.2.

The stage efficiency can be estimated using Van Winkle’s correlation or the AIChemE
method; see section 11.10.

The design of sieve plates is covered in section 11.13 and illustrated in example 11.11.

In practice, a side stream containing the fusel oil would be taken off a few plates from the
bottom of the column. The acetaldehyde in the feed would go overhead and be recovered in a
separate column.

Solution 11.7

Summary

Feed 0.9 MEK, Bottoms 0.99 Butanol, 0.01 MEK,

Feed rate 20 kmol/h, feed temperature 30 oC, boiling point 80 oC

Reflux ratio 1.5 x Rmin.

Properties

Latent heats: MEK 31284 kJ/kmol, 2-butanol 40821 kJ/kmol

Specific heats: MEK 164 kJ/kmol, 2-butanol 228 kJ/kmol

Mol mass: MEK 72.11, 2-butanol 74.12

Solution

(a), (b) minimum reflux ratio and number of theoretical stages

Binary system, so use McCabe-Thiele method to find the minimum reflux ratio and number
of stages; see example 11.2.

Latent heat of feed = 0.9 x 31234 + 0.1 x 40821 = 32,193 kJ/kmol

Sensible heat to bring feed to boiling point = (0.9 x 164 + 0.1 x 228)(80 – 35)

= 7668 kJ/kmol

5
 q = (32193 + 7668) / 32193 = 1.24

Slope of q line = 1.24 / (1.24 – 1) = 5.2

From McCabe-Thiele plot φmin = 0.66

Rmin = (0.99/0.66) - 1 = 0.5 (11.24)

R = 1.5 x 0.5 = 0.75, φ = 0.99 / (1 + 0.75) = 0.57

For this reflux ratio, stepping off the stages on the McCabe-Thiele diagram gives

8 stages below the feed and 8 above, total 16 theoretical stages.

The diagram was enlarged by a factor of 8 above the feed to accurately determine the number
of stages.

(c) Plate efficiency

The question asked for the stage efficiency. I will estimate the overall column efficiency
using O’Connell’s correlation. The individual stage efficiency could be estimated, after the
designing the plates, using Van Winkle’s correlation, (11.69) or the AIChemE method,
section 11.10.4.

Liquid viscosity’s at the average column temperature:

MEK 0.038 Nm-2 s, Butanol 0.075 Nm-2 s

µa at feed composition = 0.9 x 0.038 + 0.1 x 0.075 = 0.042 Nm-2 s

αa = 2.6 ( α can be estimated from the equilibrium data using (11.23)).

µa x αa = 0.042 x 2.6 = 0.109

Eo = 51 – 32.5 x Log (0.109) = 82.3 % (11.67)

Seems rather high, would need to confirm before use in practice.

Table 11.2 gives a value for Toluene – MEK as 85 %. So use 80 % for the remainder of the
question.

(d) number of actual stages

Number of real plates = 16/0.8 = 20

(e) plate design

6
Flow-rates

Feed = 20kmol/h

Mass balance on MEK, 0.9 x 20 = 0.99 D + 0.1B

Overall balance, 20 = D + B,

which gives D = 18.16 and B = 1.40 kmol/h

From the McCabe-Thiele diagram the slope of the bottom operation line, (Lm’ / Vm’)

= 0.95/0.90 = 1.056

Vm’ = Lm’ - B, so Lm’ / ( Lm’ - 1.40) = 1.056 hence,

Lm’ = 1.056 Lm’ - 1.40 x 1.056, = 1.478 / 0.056 = 26.12 kmol/h

Vm’ = 26.12 - 1.40 = 24.72 kmol/h

Densities
-3
2-butanol, at feed temperature, 80 oC, = 748 kgm ,
-3
at bottoms temperature, 99.5 oC, = 725 kgm .

The properties of MEK will be very similar, so ignore the change in composition up the
column.

Design for conditions at the base.

Base pressure

Allow 100 mm WG per plate. Number of plates, allowing for reboiler = 19

∆P = 19 x 100 = 1900 mm WG = 1.9 x 1000 x 9.8 = 18620 N/m2

2 2
Say, allow 19 kN/m , column base pressure = 100 + 19 = 119 kN/m
-3
Vapour density = 74.12 x ___273___ x 119 = 2.9 kgm
22.4 (273 + 99.5) 100

Column diameter

Lm’/ Vm’ ≅ Lw’/ Vw’ = 1.056

FLv = 1.056 √(2.9 / 725) = 0.067 (11.82)

7
The column diameter is likely to be small, as the feed rate is low, so take the plate spacing as
0.45 m.

From Fig. 11.27, K1 = 0.078

Surface tension, estimated using (8.23), = 9.6 mJ/m2 (mN m).

Correction = (0.0096/0.02)0.2 = 0.86

Corrected K1 = 0.078 x 0.86 = 0.067

uf = 0.067 √(725 – 2.90)/2.90) = 1.06 m/s

Take design velocity as 80% of flooding,

Maximum velocity = 1.06 x 0.8 = 0.85 m/s.

Volumetric flow-rate = (24.72 x 74.12)/(2.90 x 3600) = 0.176 m 3/s

So, area required = 0.176 / 0.85 = 0.21 m2

Take downcomer area as 12 %, then minimum column area required

2
= 0.21 / (1 – 0.12) = 0.24 m

Column Diameter = √(4 x 0.24) Π = 0.55 m

Liquid flow pattern

Max. vol. liquid flow-rate = 26.12 x 74.12)/(725 x 3600) = 0.74 x 10-3 m3 /s

Fig. 11.28, column diameter is off the scale but liquid rate is low so try a reverse flow plate.
Adapt design method for across-flow plate

Keep downcomer area as 12% , Ad / Ac = 0.12

From Fig. 11.31, lw / Dc = 0.76

Take this chord for the reverse flow design.

Then down comer width, weir length = (0.76x 0.55 ) / 2 = 0.21 m

Summary, provisional plate design

Column diameter = 0.55 m

2
Column area, Ac = 0.23 m

8
 2
Downcomer area, Ad = 0.06 x 0.23 = 0.014 m
2
Net area, An = Ac - Ad = 0.23 - 0.014 = 0.216 m
2
Active area, Aa = Ac - 2 Ad = 0.202 m

2
Hole area, take as 10% of Aa , Ah = 0.02 m

As column diameter is small and liquid flow-rate low, take weir height as 40 mm, plate
thickness 4 mm. and hole dia. 5 mm.

Check on weeping

Design at rates given in the question, for illustration; turn down ratio not specified.

Liquid rate = (26.12 x 74.12) / 3600 = 0.54 kg/s

how = 750(0.54 / (725 x 0.21))2/3 17.5 mm (11.85)

. how + . hw = 17.5 + 40 = 57.5

From Fig 1130, K2 = 33

uh min = [33 - 0.90 (25.4 – 5)] / (2.9)1/2 = 8.5 m/s (11.84)

Vapour rate = 0.176 m3/s, so velocity through holes, uh =

0.176 / 0.02 = 8.8 m/s

Just above weep rate. Need to reduce hole area to allow for lower rates in operation.

Try 8% , Ah = 0.0202 x 0.08 = 0.014

uh = 0.176 / 0.014 = 12.6 m/s - satisfactory

Plate pressure drop

Plate thickness / hole diameter = 4/5 = 0.8
From Fig. 11.34, Co = 0.77

hd = 51(12.6 / 0.77)2 x (2.9 / 725) = 54.6 mm (11.88)

hr = 12.5 x 103 / 725 = 17.3 mm (11.89)

ht = 54.6 + 57.5 + 17.3 = 129.4 mm (11.90)

Downcomer liquid back-up

Head loss under downcomer

9
Take hap at 5 mm below the top of the weir (see Fig. 11.35)
2
Then Aap = 0.21 x (40 – 5) = 0.00735 m
2
So, Aap < Ad and Am = Aap = 7.35 x 10-3 m

hdc = 166 [ 0.54 / (725 x 7.35 x 10-3)]2 = 1.7 mm (11.92)

Back-up, hb = 57.5 + 129.4 + 1.7 = 188.6 mm

Which is less than half the plate spacing plus the weir height, so the design is satisfactory.

Check residence time

tr = (0.014 x 188.6 x 10-3 x 725) / 0.54 = 3.5s, acceptable

Check entrainment

uv = Vol. Flow-rate / net column area = 0.176 / 0.216 = 0.815 m/s

uf (flooding vel.) = 1.06 m/s, so percent flooding = 0.815 / 1.06 = 77%

FLv = 0.067 (calculated previously)

From Fig. 11.29, ψ = 3.5 x 10 , satisfactory.

-2

Conclusion

Design using cross flow plates looks feasible.

Solution 11.8

The number of theoretical stages can be determined using the McCabe-Thiele method
illustrated in example 11.2, section 11.5.

For the plate column, the column efficiency can be approximated using the value given in
Table 11.2. The column diameter can be estimated using equation 11.79.

For the packed column, the HETP value given in Table 11.4 can be used to estimate the
column height. The column diameter can be calculated using the procedure given in section
11.14.4.

Having sized the columns, the capital costs can be compared using the procedures and cost
data given in Chapter 6.

10
The column auxiliaries and operation costs are likely to be more or less the same for both
designs

Solution 11.9

See section 11.16.2, example 11.15.

Feed 2000 kg/h, 30 % MEK

Solvent 700 kg/hr, pure TCE

MEK in feed = 600 x 0.3 = 600 kg/h

Water in raffinate = 2000 – 600 = 1400 kg/hr

At 95% recovery, MEK in final raffinate = (1 – 0.95) x 600 = 30 kg/hr

Composition at the point o = (600 ) / (2000 + 700) = 0.22 MEK, 22%

Composition of final raffinate = 30 / ( 1400 + 30) = 0.21 MEK, 2.1%

Following the construction set out in section 11.16.2 gives 3 stages required, see diagram.

Diagram

11
Solution 11.10

See section 11.14 and example 11.14.

For this design, as the solution exerts no back-pressure the number of overall gas phase
transfer units can be calculated directly from equations 11.107 and 11.108.

∆y = y

so, ylm = (y1 - y2)/ ln(y1/y2) (11.108.)

and, NOG = (y1 - y2)/ ylm = ln(y1/y2) (11.107)

When estimating the height of an overall gas phase transfer unit, note that as there is no back
pressure from the liquid the slope of the equilibrium line, m, will be zero; i.e. there is no
resistance to mass transfer in the liquid phase.

Ceramic or plastics packing would be the most suitable this column.

12