Page 30 - Gas Purification 5E

20 Gas Purijication

which can be used for estimating the required number of transfer units, are given in several
publications (e.g., Treybal, 1980; Perry and Green, 1984).
Alternative equations and graphical techniques have been developed to calculate NoG for
other design conditions (Colbum, 1941; White, 1940). A summary of useful design equa-
tions for transfer-unit calculations is presented by Sherwood et al. (1975).
The HTU concept can also be employed for analysis of the contributions of the individual
film resistances although, in general, the individual absorption coefficients are preferred for
basic studies. Values of NOG are particularly useful for expressing the performance of equip
ment in which the volume is not of fundamental importance. In spray chambers, for example
(see Chapter 6), the effectiveness of the equipment is more a function of liquid flowmte and
spray nozzle pressure than of tower volume. The use of volume-based absorption coeffi-
cients for such units is quite meaningless.
An approach frequently used by vendors to describe the mass transfer efficiency of pack-
ing is the “height equivalent to a theoretical plate” (HETP) which is defined as follows:

HETP = Height of packed zone/Number of theoretical plates achieved in packed zone

In this approach the number of theoretical plates required is estimated as described in the
next section for tray columns, and this number is simply multiplied by the HETP value given
for the packing employed to obtain the required packing height. The HETP concept is not
theoretically correct for packed columns, in which contact is accomplished by differential
rather than stagewise action; however, it is very easy to use for column design. For the spe
cial case of parallel equilibrium and operating lines (Le., mGdM = l), HETP and HTU are
eqUal.
The calculation of packed column height by these techniques requires a knowledge of the
overall absorption coefficient (e.g., ha), the height of a transfer unit (e.g., HOG), or the
height equivalent to a theoretical plate (HETP) and estimation of these values is usually the
most difficult column design task. Although some success has been achieved in predicting
packed-column, mass-transfer coefficients from a purely theoretical basis (e.g., Vivian and
King, 1963), the use of empirical correlations and experimental data represents the usual
design practice. Test or operating data relating to absorption coefficients are therefore given
whenever feasible for processes described in subsequent chapters. Examples of &a values
for a number of gas absorption operations are presented in Table 1-5. Data for a variety of
packings operating under similar conditions are given in Table 1-6. The values given in this
table are calculated for the absorption of carbon dioxide in dilute sodium hydroxide solution
by assuming zero equilibrium vapor pressure of carbon dioxide over the solution and using a
log-mean partial pressure over the length of the column.
Generalized correlations for estimating the individual mass transfer coefficients have been
proposed by Onda et al. (1968), Bolles and Fair (1982), and Bravo and Fair (1982). These
correlations cover commonly used packings such as Raschig rings, Berl saddles, Pall rings,
and related configurations. Correlations for structured packings have been developed by
Bravo et al. (1985) for Sultzer BX (gauze) packing, and by Spiegel and Meier (1987) for
Mellapak (sheet metal) packing. Fair and Bravo (1990) suggest that the Bravo et al. (1985)
correlation can be used for sheet metal as well as gauze packing by using a ratio of interface
aredpacking area of less than 1.0, and they provide a simple method of estimating the ratio.
A computer model that makes use of correlations, such as those referred to above for the
individual mass transfer coefficients, to predict the actual performance of small sections of