Liquid maldistribution. The stagnant zones (Fig. 7.76; Fig. 7.8a,6) have a detrimental effect on efficiency in large-diameter trays. The stagnant liquid composition reaches equilibrium with the rising vapor, and from then on changes no further. Vapor rising through stagnant zones undergoes no change in composition, and tray efficiency drops. A counteracting effect is that the stagnant zones receive fresh liquid from the neighboring active region (Fig. 7.8c), and this reinstates mass transfer. Lockett (12) states that when the maximum width of the stagnant region is less than about 20 in, transverse mixing is sufficient to overcome the detrimental effects of stagnant regions on tray efficiency.
A far more detrimental effect of stagnant zones on tray efficiency develops in a group of trays stacked above each other (171-173). Vapor from the very bottom of the column passes from one stagnant zone to the next and remains unchanged in composition. At the top of the column, this heavy vapor contaminates the purified vapor rising from the central portion of the trays, causing a major drop in apparent efficiency. Thorogood (172) modeled this scenario as two parallel columns, with the reflux ratio in one column lower than in the other. An early pinch takes place in the column simulating the stagnant regions (lower L/V), and this reduces the purity of the overhead stream. This model is identical to the model for maldistribution in packed columns (Sec. 9.2.2).
Counteracting this highly detrimental stacking effect is mixing of both vapor and liquid. Mixing of vapor between trays is strongest at high ratios of tray spacing to column diameters, i.e., in smaller columns, but weakens as column diameter increases. Mixing of liquid occurs in the downcomers and is analogous to packed-tower redistribution (Sec. 9.2.3), but occurs far more frequently. Thorogood (172» shows that only a small amount of remixing in the downcomer—as little as 2 to 5 percent—is sufficient to mitigate the stacking effect. Weiler et al. (174) shows that at least such a limited amount of remixing occurs in the downcomer.
Modeling. Perhaps the most popular model for evaluating the effects of maldistribution on tray efficiency is the Lockett et al. stagnant region model (SRM) (12,148,167,171). This model postulates liquid plug flow at the center of the tray with stagnant zones near the wall. The stagnant zones occupy the areas enclosed by the chords joining the ends of the inlet and outlet downcomers (Fig. 7.76). An alternative model by Bell and Solari (158,175) represents liquid channeling by a nonuniform velocity profile across the tray. The model can include the effect of liquid recirculation in the stagnant regions. Sohlo and Kouri i l42) initially used a similar model, and later (176) attempted to account for the way nonideal flow develops. Kler and Lavin (166) and Estevez and Arreaza (177) proposed models based on two-dimensional velocity distributions.
Test data. The detrimental effects of stagnant regions on tray efficiency have never been confirmed by experiment. The only published data are by Yanagi and Scott (160,164), who measured tray efficiency in an 8-ft column with and without a flow-straightening device. They observed the severely channeled profile of Fig. 7.86 without their flow straighteners and plug flow with them. Despite that, they measured the same tray efficiency with and without the flow straighteners. Lockett and Safekourdi (12,169) and Bell and Solari (158) argued, based on their models, that the Scott and Yanagi test conditions were such that liquid flow maldistribution should not have had a noticeable effect on efficiency.
Case studies were reported (170,174) of large-diameter (> 15-ft) towers with sieve trays not reaching the expected efficiency. Maldistribution was cited as the culprit or at least one of the causes. Improving liquid flow patterns, often among other modifications, was the fix. The only other evidence that channeling adversely affects tray efficiency comes from the above-mentioned theoretical models.
Improving liquid flow patterns. A number of special tray designs have been developed to improve liquid velocity distribution on large-diam-eter trays. Their main applications are vacuum distillation. In pressure distillation, liquid flows are usually high and multipass trays are used, so that stagnant zones are seldom a problem. Some means of improving the liquid flow patterns are
1. Using valve trays. Their sideways gas movement alleviates liquid channeling (Sec. 7.3.2, item 9). On the basis of eliminating the stagnant regions, Biddulph (168) expects valve tray efficiencies in large-diameter columns to be 15 to 20 percent higher than sieve trays, but this has not been verified by experiment.
2. Using suitably directed slots and baffles to avoid channeling. Such devices (170,178) have improved efficiency in large-diameter vacuum towers with long flow paths. Use of these devices in large-diameter towers has been advocated for vacuum services (24,161) and for pressure services with tall weirs (161).
3. Using arc-shaped downcomers. Special tray designs (165,169) have arc-shaped downcomers, which direct liquid toward the column walls. These devices improve flow patterns, but their effect on efficiency has not been tested.
4. The Fractionation Research Inc. unique design for inlet downcomer baffle and outlet weir (160,164) converted a highly channeled liquid flow into plug flow. Despite that it did not improve efficiency.
5. Increasing the number of passes (167). This reduces the tendency to form stagnant zones (Sec. 7.3.2, item 7).
Vapor maldistribution. Most popular theoretical models (such as the AIChE and the Chan and Fair models, Sec. 7.2.1) postulate perfectly mixed vapor flow. In large-diameter columns, vapor is more likely to rise in plug flow. Modeling work showed (143,179,180) that in the absence of stagnant zones on the tray, vapor flow pattern has generally little effect on tray efficiency. When column efficiency exceeds 80 percent (143), or when stagnant liquid zones exist (171,173,180), vapor plug flow reduces tray efficiency.
Some theoretical work (181,182) has been done on vapor velocity maldistribution due to hydraulic gradients (Fig. 7.9). It has been
Figur« "".9 Possible nonuniform vapor flow pattern.
shown (182) that even if it occurs, it has only a minor effect on tray efficiency. A recent study suggests a more significant effect (183), but this study is based on some doubtful extensions of the AIChE correlation.
Under some conditions, vapor velocity maldistribution induced by hydraulic gradient pt tray tilt can lead to excessive nonuniform weeping (183a; Also, see Sees. 6.2.12, 6.2.13). Such excessive weeping can be detrimental to tray efficiency.
Downcomer mixing. The AIChE model assumes that liquid along the downcomer length is perfectly mixed. A recent study using a tracer injection technique (174) showed that in large trays, the downcomer liquid is unmixed. A computational analysis (174) showed that efficiency reduction due to an unmixed downcomer is relatively small, but will intensify in the presence of liquid stagnant regions on the tray.
Weir height. Taller weirs raise liquid level on the tray in the froth and emulsion regimes. This increases interfacial area (136,137) and vapor contact time. Larger interfacial areas and contact times enhance efficiency, especially when the mass transfer resistance is concentrated in the vapor (most distillation systems). In the spray regime, weir height affects neither liquid level nor efficiency (Sec. 6.4.4).
In distillation systems, the improvement of tray efficiency due to taller weirs is small (5). Koch Engineering (8), Kreis and Raab (28), and Kalbassi et al. (184) observed little effect of weir height on distillation tray efficiency for weirs 1.5 to 3 in, 1 to 2 in, and 0.5 to 1 in tall, respectively. Finch and Van Winkle (185) reported an efficiency increase of the order of 5 to 10 percent as weir height is raised from 1 to 3 in; a similar increase was reported by Prado and Fair (110,144) in humidification and stripping tests.
A significant drop in tray efficiency was observed (184-186) when changing from a weir even as short as 0.5 in to no weir. Raising weir heights from 2 to 4 in was shown to raise froth regime efficiency by 20 to 50 percent for caustic absorption of acid gases (28), for absorption of ammonia in water (117), and for absorption of oxygen into low-viscos-ity water/glycerol solutions (186). In the spray regime, weir height was shown to have little effect on absorption efficiency (117,186).
Length of liquid flow path. Longer liquid flow paths enhance the liquid-vapor contact time, the significance of liquid plug flow, and, therefore, raise efficiency. However, flow-path increases are coupled with col umn diameter increases, and a point is reached where the tray may suffer from channeling (Sec. 7.3.3).
One experience was reported (74) where a refinery reboiled absorber was revamped by replacing single-pass trays by two-pass trays. Although the expected capacity gain was achieved, the reduction in flow-path length from 36 to 18 in caused a loss in efficiency large enough to justify reinstalling the original trays. Finch and Van Winkle (185) report an efficiency increase of the order of 0.5 percent per inch of flow-path length for flow paths 11- to 22-in long.
Fractional hole area. Efficiency increases with a reduction in fractional hole area (23,28,110,144,186). Yanagi and Sakata (23), experimenting with commercial-scale towers, show a 10 to 15 percent increase in tray efficiency when fractional hole area was lowered from 14 to 8 percent of the bubbling area (Fig. 7.10a). Kreis and Raab (28) show an identical increase for N2/02 separation, and an even larger increase (20 to 25 percent) when fractional hole area was lowered from 8 to 5 percent of the bubbling area. Prado and Fair (110,144) showed an efficiency increase of the order of 5 percent as fractional hole area was reduced from 11 to 6 percent of the bubbling area in humidi-fication and stripping tests. The above data were collected both in the froth and spray regimes.
Hoi® diameter. In the froth regime, small holes increase the interfa-cial area, possibly also froth height (24,62). The enhancement in efficiency is small, often negligible (5,28,136,185). Experiments by Kal-bassi et al. (184) and Lopez-Bonillo et al. (187) showed that a fourfold reduction in hole diameter increased efficiency by about 15 percent or less. Prado and Fair (110,144) showed a tray efficiency rise of less than 2 percent when hole diameter was halved in humidification and stripping tests. Absorption experiments with low-viscosity systems (117,186) show a negligible to small enhancement in efficiency due to smaller holes. For high-viscosity absorption systems, efficiency significantly improved when hole diameter was reduced (186). In spray regime absorption, a threefold hole diameter reduction was shown to drop efficiency by 15 to 25 percent (117,186).
Vapor-llquid loads. A higher vapor load reduces the vapor contact time but also increases the interfacial area (136,137). These two factors have counteracting effects on tray efficiency. Usually, the contact time dominates, and efficiency decreases with higher vapor rates (185). A higher liquid load increases tray efficiency (185) because it increases tray liquid holdup, and therefore vapor contact time.
Most distillation operations are performed at constant LfV ratios.
A. IS9BIITAHE-H-9UTtME - llSlkPa
. 9 A ClfCLMflAM-JWTUE --3D Ifl
A. IS9BIITAHE-H-9UTtME - llSlkPa
. 9 A ClfCLMflAM-JWTUE --3D Ifl
Was this article helpful?