Info

Wall

24"

40"

72"

Figure 9.3 Liquid spread profile in a packed column (20-in-ID column, packed with 6 ft of 2-in Pall® rings, water study. Data from P. J. Hoek, Ph.D. thesis, University of Delft, The Netherlands, 1983.)

Wall «ow (66,87,140,141-149). The tendency of liquid to flow toward the walls of packed columns is a fundamental phenomenon associated with packed-column hydraulics. The development of wall flow is illustrated in Fig. 9.4 using typical measurements by Hoek (140) in a pilot-scale column. The column diameter was 20 in, and the outer distributor nozzle was located about 1.5 in from the wall. In these experiments, "wall flow" was defined as the flow in the outer ring of the column (with an area of 16 percent of the column cross section).

Figure 9.4 shows little wall flow near the top of the bed. This is because the liquid distributor drip points stop short of the wall (the unirrigated ring at the top of the bed was 1 in wide in Hoek's measurements). With increasing depth below the top of the bed, liquid

Distillation Alcohol

Distance from top of bed,feet

Distance from top of bed,feet

Figur« 9.4 Fraction of flow in the wall region of a packed column, (a) Effect of liquid rate and packing size. (20-in-lD column, water study. Data from P. J. Hoek, Ph. D. thesis, University of Delft, The Netherlands, 1983).

Distance from top of bed,ft (b)

Figure 9-4 (Continued) Fraction of flow in the wall region of a packed column. (6) Effect of packing type (20-in-ED column, water study. Data from P. J. Hoek, Ph.D. thesis, University of Delfi, The Netherlands, 1983.)

Distance from top of bed,ft (b)

Figure 9-4 (Continued) Fraction of flow in the wall region of a packed column. (6) Effect of packing type (20-in-ED column, water study. Data from P. J. Hoek, Ph.D. thesis, University of Delfi, The Netherlands, 1983.)

spreads through the packing and on to the wall. The liquid reaching the wall is partially deflected back into the bed and partially flows down the wall. The wall flow therefore increases until an equilibrium is reached.

Figure 9.4a shows that packings that promote uniform liquid spread (Pall® rings; small-diameter packings) also deflect the liquid toward the wall at a slower rate. For these packings, it takes a greater distance into the bed before the wall flow becomes equal to the bed average flow. It follows that for such packings, it is important that the initial distributor irrigates the packing near the column wall (140). Figure 9.4 also shows that wall flow is more of a concern at low than at high liquid flow rates. Other conclusions reached, based on experi mental work in laboratory and pilot-scale columns as well as theoretical predictions are

1. In large columns, the amount of liquid flowing down the wall at equilibrium is very small (141,144).

2. In small columns, the proportion of liquid flowing down the wall at equilibrium can be high. It is therefore recommended to design for a ratio of column to packing diameter {DjlDp) of at least 10 (1,14,56,141,148).

3. Redistributors and wall wipers counteract the spread of liquid toward the wall. There is uncertainty regarding their effectiveness in keeping the liquid off the walls of randomly packed towers (144). It appears that their main value is in promoting lateral mixing (Sec. 9.2.3). In structured packings, well-designed wall wipers were shown (67) to effectively keep liquid off the wall.

4. The factors affecting the development of wall flow are the same as those that affect the liquid spread. These include height into the bed, packing type, packing size, and liquid flow rate (66,140).

9.2.5 The zone-stage model

Zuiderweg, Kunesh et al. (131,136,150,150a) combined the effects of local L/V ratio, lateral mixing, and liquid spread into a single model that describes the effect of liquid maldistribution on packing efficiency. The work leading up to this model was performed at Fractionation Research Inc. (FRI) and at Delft University in The Netherlands. The model is still undergoing development.

The model postulates th t in the absence of maldistribution, there is a "basic" (or "true" or "inherent") HETP which is a function of the packing and the system only. This HETP can be inferred from data for small towers, in which lateral mixing is strong enough to compensate for any pinching (Sec. 9.2.3).

The model divides the column axially into a finite number of sections, each section being one "basic" theoretical stage (Fig. 9.5). The length of each section is therefore equal to the "basic" HETP of the packing. Radially, the column is divided into a number of zones, each zone having the width 2 to 3 times the size of a packing particle (Fig. 9.5). Each zone-stage (i.e., each rectangle on Fig. 9.5) is considered to be a region in which liquid and vapor are mixed ideally, and the vapor stream leaving the zone-stage is in equilibrium with the liquid stream leaving the zone-stage.

A fraction K^ of the liquid stream in every stage flows downward into the next lower stage. The balance (i.e., 1 - KJ flows sideways into the adjacent zones and splits according to the ratio of the interfa-

iiout j to i Particle muxneis

iiout j to i Particle muxneis

4 t I I

Was this article helpful?

0 0

Post a comment