Page 279 - Applied Process Design For Chemical And Petrochemical Plants Volume II

268 Applied Process Design for Chemical and Petrochemical Plants

less sensitive to the uniformity of liquid distr;,bution than bed has a reasonable tolerance for both a uniform or
the smaller sizes. smooth variation in liquid distribution and for one that is
Norton [83] classifies liquid distributors as: totally random.” Once a system has become maldistrib-
uted, the recovery to natural and constant HETP is a very
1. High performance Intalox@: 90-100% quality complex problem.
2. Intermediate performance: 75-90% quality Perry et al. [85] point out that packed columns are more
3. Historic standard: generally 30-65% dependent on liquid distribution than trayed columns, as
can be appreciated by the differences in the way the liquid
The quality grading system relates to uniformity of dis- must flow down the two types of columns. Liquid distribu-
tribution across the tower cross-section, where 100% qual- tion quality is measured or described as [85] :
ity indicates ideal uniform distribution. Distributors are
designed to suit the system, particularly the packing type 1. Number of distribution points (distribution density).
and size. 2. Geometric uniformity of distribution points across
The type of distribution to select depends on the sensi- the cross-section of the tower.
tivity of the tower performance to the liquid distribution 3. Uniformity of liquid flow from the distributor points.
as discussed earlier. Norton’s [83] data indicate that the
sensitivity of tower performance to liquid distribution Currently, most designs use 4 to 10 distribution points
quality depends only on the number of theoretical stages per square foot of tower cross-section, with 9 points being
in each bed of packing achievable at its “System Base generally considered useful for a wide variety of random
HETP” [83]. Tower beds of high efficiency packing are and structured packings [85]. The distribution demands
more sensitive to liquid distribution quality than shorter of small random packings are greater than for the large
beds of medium efficiency packing [83]. It is important to sizes due to the lower radial spreading coefficients, i.e.,
extend the uniformity of the distributor all the way to with- the larger the radial spreading coefficient the more quick-
in one packing particle diameter of the tower wall [85]. ly the initial liquid distribution will reach an equilibrium
Good liquid distribution starting right out of the dis- with the “normal” or “natural frequency” of distribution
tributor and onto the top layer of packing is essential to (see Figure 91 1).
develop the full usefulness of the packing bed [85]. In Hoek [86] proposed a radial spreading coefficient to
principle this applies to all types and sizes of packing. characterize the liquid distribution. This coefficient is a
Kunesh, et al. [84, 851 present FFU (Fractionation measure of how quickly a packing can spread a vertical liq-
Research, Inc.) studies on distribution that reflect the uid stream radially as the liquid progresses down the col-
importance of maintaining level distributor trays umn [86]. Radial mixing tends to reduce the effects of
(devices), and eliminating discontinuities or zonal flow
(Figure 9-10). Their results further show that a “packed

FRI C6 - C7 24 PSlA
I I
NOTCHED TROUGH (WEIR)

1.5

%... ... ... ... ... ._. ....-....
..
*...
.. .....+..
.*
(9.6 Pwft2)
Kunesh (84) 0
0.5 I I I 0702030405060
0 25 50 75 100 Packing SI-, mm
PCT OF USABLE CAPACITY Figure 9-11. Radial spreading coefficients for several types and sizes
of packings. Small packings require better initial distribution. The
Figure 9-10. Effect of liquid maldistribution on efficiency; FRI data for larger the radial spreading coefficient, the more rapid the initial dis-
25-mm Pall rings in cyclohexane/n-heptane distillation with two dif- tribution will reach its natural equilibrium of flow distribution. After
ferent quality distributors. Used by permission of the American Insti- Hoek, P. J., Wesselingh, J. A., and Zuidennreg, [86], and Nutter, D.
tute of Chemical Engineers, Chemical Engineering Progress, Peny, [88]; reproduced with permission of the American Institute of Chem-
D. and Nutter, D., Jan. (1990) p. 30, and by special permission of ical Engineers; Perry, D., Nutter, D. E., and Hale, A., Chemical Engi-
Fractionation Research, Inc., all rights reserved. neering Progress, Jan. (1 990), p. 30; all rights reserved.