Flooding point

The most important boundary condition for the application of the proposed power characteristic Ne(Q, Fr, D/d, h/d) is however given by the upper boundary value of the gas throughput number Q. Each stirrer can, namely, at a particular stirrer speed only disperse a particular maximum gas throughput in the liquid. Upon exceeding this value the stirrer is flooded by gas i.e. it is completely enveloped by gas and is incapable of dispersing it. 

In dispersion processes in gas/liquid (G/i) systems gas cushions are developed in so-called suction regions behind the flowed against surfaces of the stirrer, whereby the pumping ability of the stirrer for liquid (circulation flow) ceases. 

Thereby the discharge of the gas dispersion is also diminished, the gas cushions becomes ever larger and at a particular value of the gas throughput number Qmax the stirrer is flooded with gas. 

In order to be able to determine reliably the upper value of the gas throughput number, experiments have been carried out at different but constant stirrer speeds such that the air throughput was initially slowly increased until flooding occurred. Then the gas throughput was slowly reduced and the value at which the dispersing effect of the stirrer was restored noted [612]. 

The relationship between these numbers is shown in Fig. 2.13 and proves, that turbine stirrers performed best even as regards the efficiency of gas dispersion. In this regard the 12-blade turbine stirrer is clearly superior to the usual 6-blade turbine stirrer. 

---

As the throughput in a contactor represented by the superficial velocities and is increased, the holdup / increases in a nonlinear fashion. A flooding point is reached at which the countercurrent flow of the two Hquid phases cannot be maintained. The flow rates at which flooding occurs depend on system properties, in particular density difference and interfacial tension, and on the equipment design and the amount of agitation suppHed (40,65). 

The nonuniformity of drop dispersions can often be important in extraction. This nonuniformity can lead to axial variation of holdup in a column even though the flow rates and other conditions are held constant. For example, there is a tendency for the smallest drops to remain in a column longer than the larger ones, and thereby to accumulate and lead to a locali2ed increase in holdup. This phenomenon has been studied in reciprocating-plate columns (74). In the process of drop breakup, extremely small secondary drops are often formed (64). These drops, which may be only a few micrometers in diameter, can become entrained in the continuous phase when leaving the contactor. Entrainment can occur weU below the flooding point. 

Plate-Column Capacity The maximum allowable capacity of a plate for handling gas and liquid flow is of primaiy importance because it fixes the minimum possible diameter of the column. For a constant hquid rate, increasing the gas rate results eventually in excessive entrainment and flooding. At the flood point it is difficult to obtain net downward flow of hquid, and any liquid fed to the column is carried out with the overheaa gas. Furthermore, the column inven-toiy of hquid increases, pressure drop across the column becomes quite large, and control becomes difficult. Rational design caUs for operation at a safe margin below this maximum aUowable condition. 

An alternate method for predicting the flood point of sieve and valve plates has been reported by Kister and Haas [Chem. Eng. Progi , 86(9), 63 (1990)] and is said to reproduce a large data base of measured flood points to within 30 percent. It applies to entrainment flooding only (values of Flc less than about 0.5). The general predictive equation is... 

For distillations, it is often of more interest to ascertain the effect of entrainment on efficiency than to predic t the quantitative amount of liquid entrained. For this purpose, the correlation shown in Fig. 14-26 is useful. The parametric curves in the figure represent approach to the entrainment flood point as measured or as predicted by Fig. 14-25 or some other flood correlation. The abscissa values are those of the flow parameter discussed earher. The ordinate values y are fractions of gross hquid downflow, defined as follows ... 

Direct Scale-Up of Laboratory Distillation Ljficiency Measurements It has been found by Fair, Null, and Bolles [Ind. Eng. Chem. Process Des. Dev., 22, 53 (1983)] that efficiency measurements in 25- and 50-mm (1- and 2-in-) diameter laboratory Oldersbaw columns closely approach tbe point efficiencies [Eq. (14-129)] measured in large sieve-plate columns. A representative comparison of scales of operation is shown in Fig. 14-37. Note that in order to achieve agreement between efficiencies it is necessaiy to ensure that (1) tbe systems being distilled are tbe same, (2) comparison is made at tbe same relative approach to tbe flood point, (3) operation is at total reflux, and (4) a standard Oldersbaw device (a small perforated-plate column with downcomers) is used in tbe laboratoiy experimentation. Fair et al. made careful comparisons for several systems, utibzing as large-scale information tbe published efficiency studies of Fractionation Research, Inc. 

Flooding and Loading Since flooding or phase inversion normally represents the maximum capacity condition for a packed column, it is desirable to predict its value for new designs. The first generalized correlation of packed-column flood points was developed by Sherwood, Shipley, and Holloway [Ind. Eng. Chem., 30, 768 (1938)] on the basis of laboratory measurements primarily on the air-water system. 

This model apphes in the region belowthe loading point, and it cannot predict the flood point because it does not include the effects of gas velocity on liquici holdup. The model of Stichlmair et al. [Gas... 

Sepn. Purif., 3, 19 (1989)] takes holdup into account and applies to random as well as structured packings. It is somewhat cumbersome to use and requires three constants for each packing type and size. Such constants have been evaluated, however, For a number of commonly used packings. A more recent pressure drop and holdup model, suitable for extension to the flood point, has been pubhshed by Rocha et al. [Jnd. Eng. Chem. Research, 35, 1660 (1996)]. This model takes into account variations in surface texturing of the different brands of packing. 

The concepts of shp velocity and characteristic velocity are useful in defining the Flooding point and operational regions of different types of column contactors. The shp (or relative) velocity is given by the equation ... 

As flooding is approached, the slip velocity continues to decrease until at the flood point is zero and the following relationship apphes ... 

Estimate the flood point from Figure 8-137, which accounts for liquid flow effects and is a ratio of liq-uid/vapor kinetic effects [79]. Flooding velocity is obtained from... 

Use with Figure 8-137 to estimate both flood point and entrainment. 

Figure 8-137 is used for estimating the entrainment-flood point. Liquid particle entrainment is generally considered as reducing tray efficiency (performance). 

Higher efficiencies are obtained for operating conditions within 85-95% of the tray flood point. 

At the flood point, liquid continues to flow down the column, but builds up at a greater rate from tray to tray. Sutherland [69] demonstrated that flooding moves up the column from the point of origin. For this reason it is important to design perforated trays without downcomers with extra care, as changing internal rates are quickly reflected in performance if the proper hole requirements are not met. They are a usefiil tray for steady state operations. 

Kister [93] has developed a new approach at establishing the flood point that appears to suit the available data and is obviously more accurate than reading the upper curve on Figure 9-21C. 

Strigle [82] identifies a regime 20% above point F on Figure 9-22 as the maximum hydraulic capacity and is termed the flooding point for atmospheric operations. 

Strigle [94] proposed this term to better describe the performance of a packed column at or near the previously described loading point. Kister [93] evaluated the limited published data and proposed using the MOC at 95% of the flood point. The flood point can be estimated by Equation 9-20 or from the plots in References 90 and 93. The data are reported to be within 15-20% of the prediction [93]. See Figure 9-22 for the identification of MOC on the HETP vs. Cg chart For more accurate information... 

Figure 9-34D. Pressure drop correlation at flood point for use with Table 9-33. Used by permission of Zenz, F. A., Chemical Engineering, Aug. (1953) p. 176 all rights reserved.

As a comparison or alternate procedure, the pressure drop at the flooding point as indicated by the upper break in the pressure drop curve can be estimated from Table 9-33 and Figure 9-34D for rings and saddles [81]. The values in the table multiplied by the correction ratio gives the pressure drop for the liquid in question, expressed as inches of water. 

---