Catalisys

Fig. 15. Boundaries of the hydrodynamic regimes ¡n a RAGL F for the air-water system: cylindrical particles, size or 1.9 * 1.9 cm [26j; spherical particles with a diameter of 1.22 cm [264]. 1 — emulsification regime; 2 — wave (oscillation) regime: 3 — pulsation, or piston-flow, regime; 4 — bubble regime.

Fíg. 16. Boundaries of the hydrodynamic regimes in aRAGLF ¡rom [60]: 1 — bubble regime; 2 — emulsification regime: 3 — pulsation. regime.

Fíg. 17. Boundaries of the hydrodynamic regimes in a
RAGLF from [263]: 1 — bubble regime; 2 — piston-flow
regime: 3 — emulsification regime

as the gas velocity is increased (at a constant liquid velocity) and rises as the liquid velocity is increased (at a constantgas velocity). The influence of the first effect is exhibited less dramatically at high liquid velocities, and that of the second effect at low gas velocities.
By using the definition of the drag coefficient for two-phase flow fp , the following equation for computing the pressure drop was obtained in References [61,265]:

where gc is theacceleration due to gravity, the Reynolds numbers for the gas and liquid are computed for the nominal diameter of the granular bed, and the mass flow rate is normalized to the total cross-sectional area of the empty column; the gas density is the density at the equipment inlet, and the 4ra g coefficient fp is determined fromthe following relationship:

A t high gas velocities and large liquid loadings, the Lockhart-Martinelli parameters ᶲL and x given in Figure 18 for descending flow [264]. are more convenient, A number of correlating relationships for the pressure drop i n the regimes of single-phase and two-phase flow are presented i n the monograph by Shah [10].
A large number ofexperimental studies of the pressure drop in a RAGLF have been carried out at the Pittsburgh Energy Research Center [2621. These experiments were performed in a column with a diameter of 10.2 cm and particles with sizes of 0.635 x 0.635 cm and 0.32 x 0.32 cm in the regimes of emulsification and pulsation. The total number of experimental results, which exceed300 in number, are approximated well by the correlation of Tallmadge [266]. Sato and coworkers
[264] represented these data [262] i n the form of an equation of the Lockhart-Martinelli type.
A n important parameter for RAGLF calculations in the average gas holdup EG. In the general case, the gas holdup i n packed reactor s a t low gasvelocities is much greater than that i n empty sparged columns ; the opposite i s true a t high gas velocities .
The average gas holdup can be divide d into two components : a dynamic EGd and a static EGs, but not always inactive component , which is deter mine d b y experimental mean s a t zero velocity of the gas flow (a nimmobile gas phase) . The porosity of the granular bed €, the overall liquid holdup hL and the gas holdup Ta are relate d t o one another as follows:

s = hL -}-f 5 = hL + e0j -fi-j; (162)

Sometimes , the e liquid d an d ga s holdups are e represented i n the e form of fractions B o f thefree volume o f the e be d [21:

1 ="i3¿ -bSo = 1 • + - H a s (163;

A s indicate d by published data a [260,264-269] , the overall gas holdup increases with increasing velocity of the gas stream and is dependent upon the physical properties of bot h moving phases , the geometry of the stationary bed ,...