The terms in Eqs. (14-123) to (14-126) are in English units and are explained in the Nomenclature. For sieve trays, m = 1.94 and Cw = 0.79. Note that the constants are a slight revision of those presented in the original paper (C. L. Hsieh, private communication, 1991). Clear liquid height is calculated from Colwell's correlation [Eqs. (14-115) to (14-122)]. The Hsieh and McNulty correlation applies to trays with 9 percent and larger fractional hole area. For trays with smaller hole area, Hsieh and McNulty expect the weeping rate to be smaller than predicted.
Weeping from Valve Trays An analysis of weeping from valve trays [Bolles, Chem. Eng. Progs. 72(9), 43 (1976)] showed that in a well-designed valve tray, the weep point is below the gas load at which the valves open; and throughout the valve opening process, the operating point remains above the weep point. In contrast, if the tray contains too many valves, or the valves are too light, excessive valve opening occurs before the gas pressure drop is high enough to counter weeping. In this case, weeping could be troublesome.
Weep point correlations for valve trays were presented by Bolles (loc. cit.) and by Klein (Chem. Eng., Sept. 17, 1984, p. 128). Hsieh and McNulty (loc. cit.) gave a complex extension of their weep rate correlation to valve trays.
Dumping As gas velocity is lowered below the weep point, the fraction of liquid weeping increases until all the liquid fed to the tray weeps through the holes and none reaches the downcomer. This is the dump point, or the seal point. The dump point is well below the range of acceptable operation of distillation trays. Below the dump point, tray efficiency is slashed, and mass transfer is extremely poor. Operation below the dump point can be accompanied by severe hydraulic instability due to unsealing of downcomers.
Extensive studies on dumping were reported by Prince and Chan [Trans. Inst. Chem. Engr. 43, T49 (1965)]. The Chan and Prince dump-point correlation was recommended and is presented in detail elsewhere (Kister, Distillation Design, McGraw-Hill, 1992). Alternatively, the dump point can be predicted by setting the weep rate equal to 100 percent of the liquid entering the tray in the appropriate weep correlation.
Turndown The turndown ratio is the ratio of the normal operating (or design) gas throughput to the minimum allowable gas throughout. The minimum allowable throughput is usually set by excessive weeping, while normal operating throughput is a safe margin away from the relevant flooding limit.
Sieve and fixed valve trays have a poor turndown ratio (about 2:1). Their turndown can be improved by blanking some rows of tray holes, which reduces the tendency to weep, but will also reduce the tray's maximum capacity. Turndown of moving valve trays is normally between about 4:1 to 5:1. Special valve designs can achieve even better turndown ratios, between 6: 1 and 10:1, and even more. Turndown can also be enhanced by blanking strips (which require valve removal) or valve leg crimping. Sloley and Fleming (Chem. Eng. Progr., March 1994, p. 39) stress that correct implementation of turndown enhancement is central to achieving a desired turndown. When poorly implemented, turndown may be restricted by poor vapor-liquid contact rather than by weeping.
Vapor Channeling All the correlations in this section assume an evenly distributed tray vapor. When the vapor preferentially channels through a tray region, premature entrainment flood and excessive entrainment take place due to a high vapor velocity in that region. At the same time, other regions become vapor-deficient and tend to weep, which lowers tray efficiency.
Work by Davies [Pet. Ref. 29(8), p. 93, and 29(9), p. 121 (1950)] based on bubble-cap tray studies suggests that the vapor pressure drop of the tray (the dry pressure drop) counteracts channeling. The higher the dry tray pressure drop, the greater the tendency for vapor to spread uniformly over the bubbling area. If the dry tray pressure drop is too small compared with the channeling potential, channeling prevails.
Perhaps the most common vapor channeling mechanism is vapor crossflow channeling (VCFC, Fig. 14-39). The hydraulic gradient on the tray induces preferential vapor rise at the outlet and middle of the tray, and a vapor-deficient region near the tray inlet. The resulting high vapor velocities near the tray outlet step up entrainment, while the low vapor velocities near the tray inlet induce weeping. Interaction between adjacent trays (Fig. 14-39) accelerates both the outlet entrainment and the inlet weeping. The net result is excessive entrainment and premature flooding at the tray middle and outlet, simultaneous with weeping from the tray inlet, accompanied by a loss of efficiency and turndown.
VCFC takes place when the following four conditions exist simultaneously [Kister, Larson, and Madsen, Chem. Eng. Progr., p. 86 (Nov. 1992); Kister, The Chemical Engineer, 544, p. 18 (June 10, 1993)]:
FIG. 14-39 Vapor crossflow channeling. Note entrainment near the tray middle and outlet, and weep near the tray inlet. (Kister, H. Z., K. F. Larson, and P. Madsen, Chem. Eng. Prog., Nov. 1992, p. 86; reproduced with permission.)
1. Absolute pressure < 500 kPa (70 psia).
2. High liquid rates [>50 m3/(mh) or 6 gpm/in of outlet weir].
3. High ratio (>2:1) of flow path length to tray spacing.
4. Low dry tray pressure drop. On sieve and fixed valve trays, this means high (>11 percent) fractional hole area. On moving valve trays, this means venturi valves (smooth orifices) or long-legged valves (>15 percent slot area). On all trays, the channeling tendency and severity escalate rapidly as the dry pressure drop diminishes (e.g., as fractional hole area increases).
Hartman (Distillation 2001: Topical Conference Proceedings, AIChE Spring National Meeting, p. 108, Houston, Tex. (April 22-26, 2001)] reports VCFC even with conventional valve trays (14 percent slot area) at very high ratio (3.6:1) of flow path length to tray spacing and tray truss obstruction.
VCFC is usually avoided by limiting fractional hole areas, avoiding venturi valves, and using forward-push devices. Resitarits and Pap-pademos [Paper presented at the AIChE Annual Meeting, Reno, Nev. (November 2001)] cited tray inlet inactivity as a contributor to VCFC, and advocate inlet forward-push devices to counter it.
Was this article helpful?