Hold-up and pressure drop for water irrigating “non-wettable” coke

Abstract

Simultaneous measurements of the liquid hold-up and of gas pressure drop, for beds of “non-wettable” coke (contact angle about 90°) irrigated by water counter-current to an uprising gas stream, were made for water rates from 0·8 to 12ft. 3/hr. ft. 2 of column cross-section and for pressure drops up to 4″ W.G./ft. height of column. Coke size ranges were 1 4 ″ − 1 2 ″, 1 2 ″ − 3 4 ″ and 3 4 ″ −1″. It is shown that the hold-up can be separated into two quantities — that assumed to be in a “virtual state of rest” and that flowing over the packing surface. The first part consists of a quantity H S , the “static” hold-up, which will not drain from the packing when liquor and gas flow are stopped; a quantity H F , which must be added to H S when liquor flows, but which is independent of the liquor rate, and a quantity K 1 h n which indicates that some liquid is held in a state of rest by the pressure drop h. The contribution to the hold-up from liquid flowing over the packing surface is expressed by means of a flow equation for the case of zero gas velocity, but the equation must be modified to account for the retardative or dispersive effect of the gas on the liquid when the gas flows. The contribution of flowing liquid to the hold-up is shown by the last term in the complete hold-up equation for the experimental system. ▪ The exponent r is constant for all packing sizes and equals 7·4. H S , H F , K 1, K 2, K 3 and n all vary with packing size but, with the exception of H S , none can be correlated with that variable, as each shows a reversal of trend in passing from the smallest coke size to the largest. This anomaly may not be experienced for wetting systems. The pressure drop is correlated in terms of the true gas velocity. It is found that there are two gas velocity ranges, in the lower of which the liquor rate has no direct effect upon the pressure drop and in the upper one it has only a slight but definite effect. ▪ ▪ k 1 and k 2 are empirically correlated in terms of packing size. In conclusion, it is emphasized that the correlations in this work were made possible by using the true gas velocity, based on the free space in the packing, rather than the superficial velocity, and it is suggested that the mechanics of packed columns will never be fully understood until this variable finds greater use in theoretical investigations.