Page 55 - Applied Process Design For Chemical And Petrochemical Plants Volume II

44 Applied Process Design for Chemical and Petrochemical Plants

EM = 6.8 (N&Nsc)'.' (NQNS,-)~.~~~ (8-70B) where E, = overall efficiency
= 6.8 [(2.07 x lo4) (?15)]~.~ [(37) (55)I0.ll5 H = Henry's law constant, lb mole/ (atm) (ft3)
= 6.8 (4.04) (2.40) = 66% P = pressure, atmospheres
a = relative volatility
In this example, Equation 8-70B gives a more conserva- p = viscosity, centipoise, cp
tive design basis.
Gerster [176] presents the results of studies on the tray
efficiencies of both tray and packing contacting devices.
where A, B, C, E = constants in equation
D = molecular diffusion coefficient, sq ft/hr Note that Gerster compares his work to the AIChE Manu-
EM = Murphree vapor plate efficiency, % al [21.
FA = fractional free area In terms of the change in gas composition [2]:
h,, = weir height, inches
G = superficial mass vapor velocity based on the E, = EOG = Y -Yn+l (8 - 76)
cross-sectional area of the column, lb/hr-sq ft Y"-Yn+1
M = molecular weight, lb/lb mole
N = dimensionless number where EG = overall column efficiency
P = pressure, consistent units EOG = overall point efficiency in vapor terms (see Ref.
[PI = Sugden parachor 2, page 38)
sg = specific gravity yn + 1 = component mol fraction in the gas to the point
T = temperature, "F considered
U = superficial velocity, ft/hr y = component mol fraction in the gas from the
V = molar volume, ft3/lb mole point considered
u = volume fraction y* = composition the leaving gas would have if it left
the point in equilibrium with the liquid
x = mole fraction in the liquid
y = mole fraction in the vapor In Table 8-2 Proctor [ 1781 compares efficiencies of sieve
p = liquid viscosity, lb/hr-ft and bubble cap trays (plates). He concludes that the sieve
p = density, lb/ft3 design provides a 15% improvement in plate efficiencies.
o = surface tension, dynes/cm
I# = mixture parameter To fully evaluate the actual efficiencies in any particular sys-
tem, the physical properties, mechanical details of the trays,
Subscripts and flow rates must be considered. See Reference 2 also.
i = component
L = liquid Table 8-2
LK = liquid light key Comparative Efficiencies of Sieve and Bubble-Cap
mix = binary mixture Trays/Plates [ 1781
n = plate number
t = total Vapor Throughput, Over4 Plate Efficiency, %
V = vapor Plate Type Lb Mole/= of Dry HzS Cold Tower Hot Tower
~~
Sieve 18,200 69 55 75 *(8)*
Biddulph [go] emphasizes the importance of using Bubblecan 16.200 60 ~5 69 *5
point efficiencies rather than tray efficiencies or overall *See the discussion of accuracy of the plate efficiency results in the text.
~~~~~
~
column efficiencies, due to the wide fluctuations that Used by permission of the American Institute of Chemical Engineers; all
often exist. rights reserved.
Kessler and Wankat [loll have examined several column
performance parameters, and for O'Connell's [49] data Strand [l79] proposes a better agreement between
presented in Figure 8-29 they propose equations that report- experimental and predicted efficiencies when recognizing
edly fit the data generally within about do% limits: a liquid by-passing factor to correct predicted values deter-
mined by the AIChE method. The results suggest that for
A. Distillation Trays the representative systems studied recognition of a liquid
by-passing factor for a tray can lower the AIChE method
E,, = 0.54159 - 0.28531 loglo ap (8-74)
results by as much as 5 to 10% to be in better agreement
with experimental results. A vapor by-passing effect was
B. Plate Absorbers (data fit *5%)
not required to correlate the data. Because the Murphree
vapor efficiencies vary considerably for various systems,
E, = 0.37237 + 0.19339 loglo (HP/p) + 0.024816
(log10 (HP/ PI2 (8-75) the data in Reference 1'79 can only be a guide for other
systems not studied.