ICCPGE 2016, 1, 201 - 209

Study of Mass Transfer Coefficient in a Wetted Wall Column

1,*
Abdualnaser M. Almagrbi , Moammer M. Elzwail1, Walid A. Alaswad1, Saleh O. Handi2
1 Chemical and Petroleum Engineering Department, Al-Mergib University, Alkhoms, Libya
2 Chemical Engineering Department, High Institute for Comperhensive Professions, Alkhoms, Libya
* Corresponding Author: [email protected]

Abstract
Gas absorption is the unit operation in which one or more soluble components of gas mixture are dissolved
in a liquid. The absorption may be a purely physical phenomenon or may involve chemical reaction with
one or more constituents in the liquid solution. In order to obtain the highest rate of absorption, gas and
liquid streams flow in opposite directions in counter-current flow. The unit in this work has been designed
to help grasp the basic principles of the chemical and physical aspects involved in absorption. This study
unit is made of borosilicate transparent glass in order to show the water spread in the column and get
the visual distribution of fluids behavior which helps to fully understand the phenomenon. In the spray
wetted wall columns, an efficient contact between the phases is achieved by atomizing through fixed or
rotary nozzles. The nozzles are arranged along the periphery and the fluids are fed in a counter-current
flow. This obstruction-free operation makes this type of process suitable to treat flow rate containing
a great number of solid particles. In this work, the wetted wall column is used to determine gas/liquid
mass transfer coefficients, which is essential to design absorption towers. This study investigates the
absorption of oxygen from air into deoxygenated water (prepared by nitrogen sparging) in liquid film-
controlled absorption experiment. The liquid film mass transfer coefficient is calculated at various mass
flow rates of water and air. This work also studies the effect of water flow and air flow on oxygen
concentration in the oxygenation and de-oxygenation process.
Keywords: Absorption; wetted wall; mass transfer coefficient; oxygenation; deoxygenation.

1. Introduction fecting the process. In considering the design of

equipment to achieve gas absorption, the main re-
The removal of one or more selected components quirement is that the gas should be brought into
from a mixture of gases by absorption into a suit- intimate contact with the liquid, and the effecting
able liquid is the second major operation of chem- of the equipment will mainly be determined by the
ical engineering that is based on interphase mass contact degree reached between the two phases. In
transfer controlled largely by rates of diffusion. The absorption, usually the gas is fed at the bottom of
gas-liquid absorption process may be treated as a the column, and the solvent is fed at the top. The
physical process; the chemical reaction doesn’t hold solvent and absorbed gas leave at the bottom, and
considerable effect. However when nitrogen oxides the unabsorbed components leave as gas from the
are absorbed in water to produce nitric acid, or top. The essential difference between distillation
when carbon dioxide is absorbed in a solution of and absorption is that in the former the vapour has
sodium hydroxide, a chemical reaction occurs, which to be produced in each stage by partial vaporization
effects the actual rate of absorption. Absorption of the liquid at its boiling point, whereas in absorp-
processes are therefore conveniently divided into tion the liquid is well below its boiling point. In
two groups, those in which the process is solely distillation there is a diffusion of molecules in both
physical and those where a chemical reaction is af-

201
directions, so that for an ideal system equimolecular amount of solid particles.
counter diffusion takes place, though in absorption Some of the spray columns advantages are Contin-
gas molecules are diffusing into the liquid, with neg- uous contact, Low pressure drop, Low cost, Gas
ligible transfer in the reverse direction, in general, phase controlling and Considered as one stage.
the ratio of the liquid to the gas flow rate is con- The main goal in this experimental work is to study
siderably greater in absorption than in distillation of mass transfer coefficient in wetted wall columns,
with the result that layout of the trays is different and to investigate the effects of water and air flow
in the two cases [1]. rates on the mass transfer coefficient [2].

2. Mythology and Experimental work

Wetted wall gas absorption columns can be used
to estimate the mass transfer coefficients between
gases and liquids, essential in the design of absorp-
tion towers. This aspect is the most important one
to determine the size of such towers, since they
serve as a basis from which correlations have been
developed for absorption towers.

Figure 1.1: Spray column used in gas absorption

[2]
.

The equipment used for continuous gas/liquid con-

tact can be a tower filled with solid packing mate-
rial, an empty tower into which the liquid is sprayed
and through which the gas flows, or a tower that
contains a number of bubble-cap, sieve or plates.
Other typical absorption operations are carried out
in spray columns, wetted wall columns, and stirred
vessels. In general the gas and the liquid stream
counter current to each other in order to obtain
the greatest rate of absorption.
In this work an absorption spray column is used,
where the contact surface between the phases is Figure 2.1: Wetted wall gas absorption column.
obtained by atomizing through fixed nozzles. The
nozzles are arranged along the periphery and the As seen in Figure 2.1, a wet film absorption col-
fluids are fed in counter-current flow. Obstruction umn (or wetted wall gas) contains, essentially, in a
free operation makes this type of columns particu-
larly suitable to treat flow rate containing a great

202
vertical tube through which a liquid inclines; tem- are small and by saturating the inlet air with wa-
porarily a gas rises through it. The gas, when con- ter, humidification effects are eliminated. Thus it
tacting the liquid, is partially spread in it by means is possible to maintain reasonably isothermal con-
of the mass transfer. This process is usually used ditions throughout the column [5]. Experimental
to study the mass transfer between two components procedure allows a power-law relationship to be cal-
in two different phases due to the fact that there is culated and a comparison is made between this and
an interfacial film between such phases in the wet published relationships such as:
film column, dissimilar other processes. It will be
0.5 0.17 n
assumed that such superficial area corresponds to Sh = (Sc) (Ga) (Re) (3.1)
the inner surface of such tube, so that the liquid
That is, for a given gas/liquid system:
can be in the shape of a thin laminar film [5].
n
(Sh)L α (Re)L
3. Experimental Description
KL Z ρud
The operating principle of the unit is very simple. α
DL µ
The feeding pump (2) draws water from tank (1) to
the oxygenation column (3), where it is sprayed by a Thus, construction of a logarithmic graph:
special nozzle onto the internal wall of the cylinder KL Z
in order to obtain a wetted wall. The air pump (9) DL
vs (Re)L
pumps air in a counter-current at flow to the flow-
ing water in the column. Oxygen is continuously for various flow rates of oxygen (in the form of air)
absorbed from the flowing air by the seeping water a power law can be determined and compared to
on the wall. The water is drawn from the bottom published values.
of the column by the pump P2 (4) and sent to flow Inter diffusion coefficient in dilute solutions can be
meter (5) and to the sampling vessel 1 (6) where calculated as follows:
probe 1 measures the saturation of the treated wa- 0.5
ter. Afterwards, the oxygenated water is sent to 1.17 × 10−13 (ζB M WB ) T
DAB = (3.2)
the de-oxygenation column (7), where it is sprayed V A0.6 µ
on the internal wall of the cylinder. Nitrogen is Recommended values for the association factor are:
pumped inside the column in counter- current flow 2.6 of water. Oxygen → Water
to extract the oxygen from the water. solvent B = H2 O , solute A=O2
VA =2×7.4×10−3
m3
= 0.0148 kgmol

0.5
1.17 × 10−13 (2.6 × 18) × 298.15
DAB = 0.6
(0.0148) × 8.96 × 10−4

m2
= 3.336 × 10−6
s
Reynolds Number can be estimated by following
equation:
ρL ud
ReL = (3.3)
µL
Column surface area
Figure 3.1: Flow chart of the plant [2].
2
A = π (r) (3.4)
The system chosen for the experiment is the ab-
sorption of oxygen into oxygen free water. In this 2
system the solubility and enthalpy of the solution π (0.016) = 0.804 × 10−3 m2
Calculation of velocity

203
 u = Q/A (3.5) KL Z
Sh = (3.9)
DL
1.389 × 10−6 m 0.01 × 0.9
u= = 1.727 × 10−3 = = 2697.84
0.804 × 10−3 s 3.336 × 10−6
996.28 × 1.727 × 10−3 × 0.032
Re = Table 3.2: Data of Reynolds number
8.96 × 10−4
m3
   
m

kg m

KL Z
u s Q s Γ m.s BF KL s DL
= 61.449 −3 −6 −3 −4
1.72×10 1.4×10 1.6×10 9.4×10 0.01 2698
3.5×10−3 2.8×10−6 3.1×10−3 1.5×10−4 0.014 3777
Table 3.1: Data of Reynolds number 5.2×10−3 4.2×10−6 4.6×10−3 2.0×10−4 0.017 4856
6.90×10−3 5.6×10−6 6.2×10−3 2.4×10−4 0.02 5396
m3
     
2

m

kg kg
Q A m u ρL m3
µL ReL
s s m.s
8.6×10−3 6.9×10−6 7.7×10−3 2.7×10−4 0.023 6205
−6 −3 −3
1.4×10 0.8×10 1.72×10 996.3 8.96×10−4 61
2.8×10−6 0.8×10−3 3.5×10−3 996.3 8.96×10−4 123
−6 −3 −3
4.2×10 ×10 5.2×10 996.3 8.96×10−4 184
5.6×10−6 0.8×10−3 6.90×10−3 996.3 8.96×10−4 246
Sh = 0.32Sc0.5 Ga0.17 Re0.5 (3.10)
6.9×10 −6
0.8×10 −3
8.6×10 −3
996.3 8.96×10 −4
307 Calculation of Schmidt number
µ
Calculation of liquid flow rate ScL = (3.11)
ρDL
Q
Γ= ×ρ (3.6) 8.96 × 10−4
h = = 0.27
996.28 × 3.336 × 10−6
1.389 × 10−6 Calculation of Galileo number
Γ= × 996.28
0.9
dp3 gρ2
GaL = (3.12)
kg µ2
= 1.537 × 10−3
m.s 3 2
(0.032) × 9.81 × (996.28)
  31 = 2
3ur (8.96 × 10−4 )
BF = (3.7)
ρ2 g
= 0.397 × 109
! 13
3 × 1.727 × 10−3 × 1.537−3
BF = 2
(996.28) × 9.81 0.5 0.5
0.17
Sh = 0.32 × (61.679) (0.27) 0.397 × 109
= 9.351 × 10−5
Calculation of mass transfer coefficient = 37.618
s
DL Γ Table 3.3: Data of Sherwood and Reynolds num-
KL = 0.422 (3.8)
ρBF 2 ber
Sh
s Sh Re Sc.Ga
3.336 × 10−6 × 1.537 × 10−3 2.50 62 37.6
KL = 0.422
996.28 × (9.351 × 10−5 )
2
3.54 123 53.2
4.33 185 65.1
m
= 0.01 5.00 247 75.1
s 5.60 308 84.0
Calculation of Sherwood number
Calculation of flooding gas velocity

204
  0.5 4. Results and Discussion
σ
Uf = F1 F2 (3.13) Theoretical and experimental data of mass trans-
ρg
fer coefficient at different Reynolds numbers or air
The data covered column sizes up to 50-mm (2- velocities were compared. The plotted data in Fig-
in) diameter; the correlation should be use l with ures 4.1 and 4.2 show the comparison at various
caution for larger columns. Reynolds numbers. A close fit between the pre-
 0.25 dicted and the experimental data was found in Fig-
G
F2 = (3.14) ures 4.1 and 4.2. These findings indicate that the
L mass transfer coefficient equation developed in this
work has a sufficient degree of accuracy. Generally,
L = Qρ (3.15)
it was found that the discrepancy was lower at low
Reynolds but it increases as Reynolds number in-
L = 6.944 × 10−6 × 996.28 creases. Also the work shows that Sherwood num-
ber increases proportionally with Reynolds number.
kg
= 6.9 × 10−3 n 0.5
Sh = 0.32 (Re) (Sc) (Ga)
0.17
(4.1)
s
In Figure 4.1, it is observed that the oxygen con-
G = Qρ (3.16)
centration is highly dependent on Reynolds num-
ber and it increases linearly as Reynolds number
G = 8.33 × 10−4 × 1.186 increases. Generally, all curve trends show Sher-
wood number is increasing linearly with Reynolds
kg number, and they can be correlated as follows:
= 9.88 × 10−4
s 0.5 0.5 0.17
0.25 Sh = 0.32 (Re) (Sc) (Ga) (4.2)
−4

9.88 × 10
F2 = In Figure 4.2 clearly shows that, the deviation be-
= 0.615
6.918 × 10−3 tween the calculated and experimental values is in-
Calculation of Surface Tension creasing as Reynolds number increases. At low
2 3
Reynolds number the deviation is around 35%, while
σ = A (1 − Tr )(
B+CTr +DTr +ETr )
(3.17) at high Reynolds number it is about 45%, because
A = 0.18548 , B = 2.717 , C = -3.554 , D = 2.047 of the effect of air high velocities on oxygen concen-
,E=0 tration.

T
Tr = (3.18)
Tc
298.15
= = 0.46
647.13
 
2.7+(−3.5)(0.46)+(2.04)(0.46)2 +(0)(0.46)3
σ = 0.185 (1 − 0.46)

N
= 0.0729 Figure 4.1: Comparison between the experimen-
m tal data & published data
di
F1 : 1.22 when 3.2 > 1.0
σ
As 3.2 0.0729
0.032
= 1.404
F1 =1.22

 0.5
0.0729 m
Uf = 1.22 × 0.615 = 0.186
1.186 s

205
Figure 4.2: Relationship between Sherwood num-
ber & Reynolds number

Figure 4.3 to 4.6 illustrate the experimental corre-

Figure 4.4: Relation between air flow rate & oxy-
lation between the air flow rate and oxygen con-
gen concentration rate in Probe 1 at flow rate water
centration with a constant counter water flow rate
(2.778×10−6 m3 /s).
of (1.389×10−6 m3 /s) within a time interval of (2
min). In general, these figures show a gradual in-
crease in oxygen concentration as the air flow rate
increases, which is due to the oxygen absorption in
the water. The oxygen absorption rate, itself, grad-
ually decreases till the water reaches its saturation
state, and the oxygen oxygen absorption stops.

Figure 4.5: Relation between air flow rate & oxy-

gen concentration rate in Probe 1 at water flow rate
(4.167×10−6 m3 /s ).

Figure 4.3: . Relation between air flow rate &

oxygen concentration rate in Probe 1 at water flow
rate (1.389×10−6 m3 /s ).

206
Figure 4.6: Relation between air flow rate & oxy- Figure 4.8: Relation between air flow rate & oxy-
gen concentration rate in Probe 1 at water flow rate gen concentration rate in Probe 2 at flow rate water
(5.555×10−6 m3 /s). (2.778×10−6 m3 /s ).

The curves trends in Figures 4.7 and 4.8, show that, Figures 4.9 and 4.10 show the effect of water flow
at same constant water flow rate, the oxygen con- rate on the oxygen concentration. The oxygen con-
centration in (probe 2) are slightly increasing with centration decreases as the water flow rate increases.
the increase of air flow rates. This hardly notice- This is due to the water turbulence flow in the col-
able effect is due to the fact that the water is either umn. During the turbulence flow, the oxygen ab-
at, or is approaching its oxygen saturation state. sorption rate drops significantly because of smaller
mass transfer areas and less times of exposure as
opposed to the smooth laminar flow.

Figure 4.7: Relation between air flow rate & oxy-

gen concentration rate in Probe 2 at water flow rate
(1.389×10−6 m3 /s ).
Figure 4.9: . Relation between water flow rate
& oxygen concentration rate in Probe 1 at air flow
rate (2.778×10−4 m3 /s ).

207
 Figure 4.13: Relation between water flow rate &
oxygen concentration rate in Probe 1 at nitrogen
flow rate (250 m3 ⁄s).
Figure 4.10: Relation between water flow rate &
oxygen concentration rate in Probe 1 at air flow
rate (4.166×10−4 m3 /s).

Figures 4.11, and 4.12 show that the oxygen con-

centration in (probe 2) declines tiredly with the in-
crease of water flow rate at some constant air flow
rate which is caused by the gradual increase of oxy-
gen concentration.

Figure 4.12: Relation between water flow rate &

oxygen concentration rate in Probe2 at air flow rate
(6.944×10−4 m3 /s).

In De-oxidation process, the air pump has to be

switched off. The oxygen concentrations are ob-
tained at different water flow rates, and then plot-
ted in Figure 4.13. The oxygen concentrations at
probe1 drop significantly with the increase of water
Figure 4.11: Relation between water flow rate & flow rate at the beginning of the trend. However,
oxygen concentration rate in Probe2 at air flow rate the trend declines at a smaller gradient in the rest
(2.778×10−4 m3 /s). of the curve.

5. Conclusion and recommendation

• Good agreement between the previously published
and the experimental data was found indicating
that the models used in this work to estimate the

208
 mass transfer coefficient have a sufficient degree
of accuracy.

• Some of the plotted correlations indicate that the

oxygen concentration at probe2 increases slightly
with the increase of air flow rate at constant wa-
ter flow.

• The oxygen concentrations at probe1 drop signif-

icantly with the increase of water flow rate at the
beginning of the trend.
• Nitrogen has a significant influence as a reduction
factor to extract the oxygen from the oxygenated
water in the de-oxidation process.
• For future studies, the temperature effect on oxy-
gen concentration can be investigated in wetted
wall gas absorption. In addition, graphical corre-
lations of mass transfer coefficients of some other
substances can be thoroughly studied, instead of
O2

References
[1] 1. J. R. Backhurst, J. H. Harker and J. F.
Richardson; Chemical Engineering, Solutions to
Problems in Volume 2, 5th Ed, 1993.
[2] 2. http:// www.didacta.it

[3] 3. Christian, H. E. Nielsen., Søren Kill, Henrik

W. et al. Chem Eng Sci, 53 (3), (1998) :495-503.
[4] 4. Bird R. Stewart W. and Lightfoot E., “Trans-
port Phenomena”, Wiley "2004".

[5] 5. http:// www.armfield.co.uk

[6] 6. McCabe, W. L., Smith J. C., and Harriott, P.,
Unit Operations of Chemical Engineering, 6th
Ed, McGraw Hill Company, New York, 2001.

209