Mass Transfer Coefficients Between Gas and Liquid Phases in Packed Columns

enhances the probability

of the bubble

coalescence

and

makes bubble larger. On the other hand, in the present

study the dimention of the apparatus is far larger than that of the distributor and the depth of the liquids is

shallow. Therefore, the probability of the bubble coalescence over the distributor is less than that of the Houghton et al.'s

case, because

the circulation

of the liquid

in the

apparatus makes bubbles easy to apart each other. This fact is recognized when the porous plate distributor is used for liquid of group A and C in the bubble column with continuous

liquid

flow.

are obtained. Acknowledgment The authors are grateful for valuable advices.

Prof

T. Sakurai,

Tokyo

Inst.

Nomenclature

[cm] [cm] [cm/sec2] [g/cm- sec2] [cm/sec] [-] [-] [g/cm3] [cm] [-] [dyne/cm]

d = volume equivalent bubble diameter d = average bubble diameter g = gravitational acceleration z/Po = excess pressure required to generate bubble ug = gas flow rate per unit area of porous plate Fr = ug2/e2gd, Froude numer We = ug2dp/e2o, Weber number p -density of liquid

Summary

Bubbles which have been just generated from the porous

plate are small and have an equal size, but sometime coalescence of these small bubbles occurs at a location

slightly removed from the distributor, where the gas holdup is very large. Therefore, large and wide size distri-

bution of bubbles are observed. This occurs easily in pure water and pure solvents. The surface active substances in water and solvents obstruct this coalescence of bubbles. In concentrated inorganic salt solutions, this obstruction is also recognized. For the extreme cases

when no coalescence is observed and the coalescence occurs at the maximumrate, the correlations of the average bubble diameter and the conditions of bubble generation

Tech.,

8 = average pore diameter denned by Eq.(l) e =porosity of porous plate a - surface tension of liquid Literature

cited

1) 2)

Foulk, C.G.: Kolloid. Z., 60,115 (1932) Gleim, V. G. and Shelomov, I. K.: J. Appl. Chem. USSR, 32, 799 (1959) 3) Gleim, V.G., Shelomov, I.K. and Shidlovski, B.R.: J. Appl. Chem. USSR, 32, 1069 (1959) 4) Houghton, G., Mcleam, A.M. and Ritchie, P. D.: Chem. Eng. Sd.f 7, 40 (1957) 5) Koide, K. Hirahara, T. and Kubota, H.: Kagaku Kogaku, 30, 712 (1966)

6) Vershoor,

H. : Trans. Inst.

Chem. Eng.(London),

28,52(1950)

MASS TRANSFER COEFFICIENTS BETWEEN GAS AND LIQUID PHASES IN PACKED COLUMNS* KAKUSABURO ONDA, HIROSHI

Dept. of Chem. Eng., University

x CFr)"0-05

It has

Introduction Mass transfer

coefficients

for gas absorption,

desorption

and vaporization in packed columns have been studied by many mvestigators3'5>11>22>26l30>31). Assuming that the wetted surface on packing pieces is identical with the gas-liquid interface, Onda et al. presented the empirical equations of the gas and liquid-side mass transfer coefficients, kG and kL, for the gas absorption and desorption12~18).

Recently, a newequation for the wetted surface area,

aw, taking into account the liquid surface tension and the surface

energy

follows150

of

packing

was derived

= l-exp{-1.45(W18'2()>24l27'28) are divided by aw of Eq. (1). The kL thus obtained

are

correlated as well as that in our previous paper12>18) by JOURNAL

OF

CHEMICAL

ENGINEERl|NG

OF

JAPAN

Fig.

Table

I

Correlation

Experimental results organic solvents

Packings Raschig ring Berl saddle Sphere " " Rod

Size 10mm l/2-in l/2-in " 1-in 14mm

replacing at the relation vs. modified line represents kL (pL/fiLg)

I

Absorbent CC14 » " CH3OH » CC14

of liquid-phase

of gas absorption Temp. 25°C » " 20°C " 25°C

data

tration

into a 0.120 0.0862 0.0227 0.0735 0.0389 0.093

n 0.70 0.73 0.86 0.76 0.83 0.70

in Reynolds number by aw. Fig. 1 shows of kL {pL/^LgYn I 36). This effect of addition of surfactant may result from the two phenomena à"the reduction of liquid mixing

at the junction of packing pieces as pointed out by Hikita7) and the inter facial resistance with increase in concenT VOL.1

Sanyo-Kasei

Co., NO.1

Ltd. 1968

for gas absorption

and desorption

by using

water

of surfactant.

The kL calculated from these data are compared with those obtained by water in Fig,1 in which the datain for a=47 dynes/cm, in this work and 42 dynes/cm, the

literature7) deviate pretty from Eq. (2). 1.2 Gas absorption by organic solvent Many investigation

on the gas

absorption

in

packed

column have been carried out by using water as an absorbent. However, there are so far only a few data5>13>28) on the gas absorption by organic solvent. In the present work, the gas absorption

of pure

CO2

into methanol and carbon tetrachloride were carried out. The columns used were 6-and 12cm I. D. and packed with 10~25mmRaschig rings, Berl saddles, spheres and rods for 20~30cm height. The mass transfer results are given in Table 1 as a

relation of kLa-aLn. Applying Eq. (l) to kLa data obtainded in this work and reported in the literature5>13>28)

for organic solvents, the same plottings are shown in Fig. 2 in which the agreement of the observed values and Eq. (2) is also satisfactory. Thus, the liquid-side mass transfer coefficients, kz, for gas absorption

and desorption

been correlated

by Eq. (2) within an error of ±20% for

in packed columns, have

organic solvents as well as water.

2. Gas-side Mass Transfer Coefficient

: kG

2.1 Absorption The Izgci data for gas absorption reported in the literature1>8ll5>16'17'30'32) are divided by aw calculated from Eq. (l). The ko thus obtained are shown in Fig. 3 as a plot of {kGRT/atDG)/{^G/pGDGyn UA,) "2 0 vs. modified 57

Fig. 2 Correlation of absorption "data by using organic solvent with Eq. (2)

Fig. 3 Correlation of phase data for absorption

Reynolds number. The equation for the best line passing through the points in the follows à" koRT/atDo = 5.23(G/WG)0-7

higher

group in Fig.3

(^g/PgDg)U3

is as

feD,)"2'° (3)

(3)

into

2.00.

This

difference

gas-

comes from the fact that

kaa data for packing smaller than 15mmtend to decrease monotonously with the increase of at as reported in the literatures1 16). However,this cause is not clear at present. The jD-factor

for mass transfer

can be obtained

In Fig. 3, data for Raschig rings and Berl saddles smaller than 15mmare situated on the lower group and are best correlated by merely changing the constant, 5.23, in Eq.

rearranging

58

JOURNAL OF CHEMICAL ENGINEERING

3.4

by

Eq. (3). For example, since atDp is 6(1-e) =

for spheres, Eq. (3) becomes > = 0.771[GDP7Ml-s)r°:30

(4) OFJAPAN

Fig. 4 investigators

Comparison of /cgO data for vaporization at L=78OOkg/m2 hr

Fig.

by various

Fig. 5 Schematic diagram of experimental apparatus for vaporization

6-a

in water-air

Fig.

system

6-b

25mm Raschig

rings

15mm Raschig

rings

I in Berl

saddles

for

data10'21>23'30) for vaporization because of the difficulties in the experimental techniques, as shown in Fig.4, in which kGa for air-water system are plotted against the gas massvelocity, G, for 1-in. Raschig rings. To ascertain their results, the rates of vaporization were measured for air-water system under the condition of adiabatic process-i. e, constant temperature of water.

Fig. 6 Vaporization data in this work

(Operational temperature of water was about 25°C and the differ ence of the temperature between top and bottom of the column was within 0.1°C.)

Fig.

Shulman et al.22) reported the following sublimation of dry naphthalene packings jD = 1.195[GDPV^(l-s)]"0-36

6-c

equation (5)

The agreement between Eqs. (4) and (5) is fairly good within the region 2 à"2 Vaporization

of 100