Page 181 - Applied Process Design For Chemical And Petrochemical Plants Volume III

66131_Ludwig_CH10E 5/30/2001 4:33 PM Page 144

144 Applied Process Design for Chemical and Petrochemical Plants

son may be improved. The method uses diffusion coeffi-
cients. An example using vertical tubes is included.
The survey of Marto 181 includes several excellent refer-
ences to this topic. The proposal of Rose 182 is reported to
give good agreement with selected experimental data.

Example 10-14. Chlorine-Air Condenser,
Noncondensables, Vertical Condenser

A chlorine-air mixture is to be cooled and the water vapor
condensed using chilled water. The design conditions are as
follows:

Flow: 92.3 lb gas mixture, per tube. Estimated tube bank: 4.48 lb
water, per tube
Gas in: 110°F (saturated)
Required gas out: 58°F
Water in: 48°F
System pressure: 1 atm
Number of tubes, assumed: 416; 0.75 in. O.D. 20 BWG

The computer print-out of good results is presented in
Table 10-23. A brief interpretation of the result follows:
Water condensed, total 815 lb/hr
Figure 10-86A. Influence of air content on the heat transfer coefficient Partial pressure water vapor in: 0.087
of steam containing air. (Used by permission: Edmister, W. C., and Partial pressure water vapor out: 0.011
Marchello, J. M. Petro/Chem. Engineer, June 1966, p. 48. © Petro-
leum Engineer International.) Cooling water out (counter flow): 58.14°F
2
Inside film coefficient, Btu/hr (ft ) °F: 13.12 (avg.)
Internal tube surface (calculated): 786 ft 2
Internal tube surface (recommended): 867 ft 2
% Range Non- % Standard
System C condensable Deviation 31, 70
The design method of Colburn and Hougen has with-
Steam-air 0.51 0.64—25.1 9.2 stood many examinations and is considered the best for any
Toluene-nitrogen 0.149 0.71—59.1 8.7 problem of this type. However, it is somewhat long and
Benzene-nitrogen 0.076 07.1—20.3 14.3 tedious and several approximation methods have been pro-
posed. 9, 10, 11, 12, 23, 79, 123
Figures 10-85, 10-86, and 10-86A and Equation 10-115A The graphical methods of Bras 9, 175, 176 provide helpful
represent the effective reduction of the pure component short-cuts to avoid the tedious trial-and-error solutions
(condensable) when inert gases are present, resulting in the required of the rigorous methods. Reference 9 is the most
reduced effective heat transfer for condensing the mixture. recent and perhaps the easier to use. The results agree in
Although it is not stated in the study, from a practical general within about 10%.
68
industrial standpoint, the effects of air, nitrogen, and other The graphical method of Hulden is also helpful as it is
common inert gases can be expected to be about the same not as tedious as the arithmetic methods, and based on his
for other organic systems. comparison with the Colburn-Hougen method, the pro-
A computer program developed by Volta 121 handles the posed results are within 1%.
problem of condensing in the presence of a noncondens- All of these have some limitations and have not been thor-
able gas for down-flow of either a saturated or superheated oughly compared against the Colburn method, which is
gas-vapor mixture inside vertical tubes. The program is based considered to be within 10% of any correct solution.
on a modification of Colburn-Hougen and Bras and is cer- Cairns 23, 176 has compared his proposal with 6 different sys-
tainly more accurate and easier to use than the lengthy man- tems and 4 other approximation methods. In general, the
ual calculations. Although the program was written for agreement with the Colburn-Hougen method is excellent.
vertical tubes, it can be used to approximate the results in a The selection of the number of temperature increments
horizontal unit, and if the correction factor between vertical is important as it affects the accuracy of the final heat trans-
and horizontal tube condensation is applied, the compari- fer area. In the majority of cases, the selection of a limited