Page 101 - Applied Process Design For Chemical And Petrochemical Plants Volume II

90 Applied Process Design for Chemical and Petrochemical Plants

5. Calculate (yi/yh) for next tray, using the yi values of al convergence techniques with some requiring consider-
tray one (m - 1) in the equation to solve for (Yi/Yh)m. ably less computer running time than others.
6. Test to determine if a is varying significantly by & =
Z (yi/ai). Evaluate temperature of heavy component Example 8-29: Tray-to-Tray Design Multicomponent
at the column bottoms pressure (estimated) using K Mixture
charts or the equivalent. If necessary, calculate new ai
values for each component at the new temperature. A column is to be designed to separate the feed given
Recheck every two or three trays if indicated. below into an overhead of 99.9 mol % trichloroethylene.
7.Introduce components lighter than the light key The top of the column will operate at 10 psig. Feed tem-
which are not found in the bottoms in the same gen- perature is 158°F.
eral manner as discussed for the rectifying section.

~
XF~/X, = [ (1 + D/L)Ki]P’ (8-183) Overhead Bottoms
Mol Mol Mol
where p’ is the number of trays below the feed tray Frac- FIIIC- FraO
where the component, i, is introduced in an assumed Feed tion Mols tion Mols tion
amount (usually small) xa. Then XF~ is the mol frac- ~-
tion of the component in the feed. (A) Trichloroethylene 0.456 0.451 0.999 0.00549 0.010
0.101
(B) p Trichloroethane 0.0555
0.00045 0.001 0.05505
8. Continue step-wise calculations until ratio of light to (C) Perchloroethylene 0.3623 . . . . . . 0.36250 0.661
heavy keys in the liquid portion of the feed essential- (D) Tetras (1) 0.0625 . . . . . . 0.0625 0.114
ly matches the same component ratio in the liquid on (E) Tetras (2) 0.0633 . . . . . . 0.0625 0.114
one of the trays. 1.0000 0.45145 1.000 0.54804 1.000
9. The total of theoretical trays in the column is the sum Note: the material balance for overhead and bottoms is based on:
of those obtained from the rectifying calculations, (a) 99.9 mol % uichlor in overhead
plus those of the stripping calculations, plus one for (b) 1.0 mol % trichlor in bottoms
the feed tray. This does not include the reboiler or (c) 1.0 mol feed total
partial condenser as trays in the column. (d) Light key = trichloroethylene
Heavy key = p trichloroethane
Tray-by-Tray: Using a Digital Computer
Detmine Overhead Temperature
Multicomponent distillation is by far the common
requirement for process plants and refineries, rather than Because trichlor is 99.9% overhead, use it only to select
the simpler binary systems. There are many computer pro- boiling point from vapor pressure curves at 10 psig over-
grams which have been developed to aid in accurately head pressure = 223°F (1,280 mm Hg abs).
handling the many iterative calculations required when
the system involves three to possibly ten individual com- Detmine Bottoms Temperature (Bubble Point)
ponents. In order to properly solve a multicomponent
design, there should be both heat and material balance at Allowing 10 psig column pressure drop, bottoms pres-
every theoretical tray throughout the calculation. sure = 20 psig (1,800 mm Hg abs)
To accommodate the stepby-step, recycling and check-
ing for convergences requires input of vapor pressure rela-
Try t = 320°F
tionships (such as Wilson’s, Renon’s, etc.) through the =iB -
Vapor Press.
previously determined constants, latent heat of vaporiza- Hg
tion data (equations) for each component (or enthalpy of Component xi (vp.) WB
liquid and vapor), specific heat data per component, and A 0.01 4,500 45 0.0249
possibly special solubility or Henry’s Law deviations when B 0.101 2,475 250 0.1382
the system indicates. C 0.661 1,825 1,210 0.67
There are several valuable references to developing and D 0.114 1,600 183 0.1012
applying a multicomponent distiUation program, including E 0.114 1,050 120 0.0664
Holland [26,27,169], Prausnitz [52,53], Wang and Henke 1,808 1.0007
[76], Thurston [167], Boston and Sullivan [6], Maddox and mm Hg abs.
Erbar [115], and the pseudo-K method of Maddox and
Fling [116]. Convergence of the iterative trials to reach a This compares quite well with the selected 1,800 mm
criterion requires careful evaluation [ 1141. There are sever- bottoms pressure. Bottoms temperature is 320°F.