вход по аккаунту


GasLiquid Mass Transfer in Hot Sparged Systems.

код для вставкиСкачать
Dev. Chem. Eng. Mineral Process. 12(3/4), pp. 323-332, 2004.
Gas/Liquid Mass Transfer in Hot Sparged
Y. Zhu* and J. Wu
Energy & Thermofluids Engineering, CSIRO Manufacturing &
Infrastructure Technology, PO Box 56, Highett, Victoria 3190,
Understanding the effect of temperature on the volumetric mass transfer rate is of
vital importance for the design of chemical reactors operated at elevated
temperatures. The present study is motivated by a poor understanding of this efiect
due to previous limited investigations. Such an effect has been studied at the CSIRO
Energy & Thermofluids Engineering Laboratory in a mechanically agitated air-water
mixing tank with operating temperatures in the range of 20-80°C at atmospheric
pressure. The oxygen concentration of the liquid is measured by a dissolved oxygen
probe. The volumetric mass transfer rate is estimated using a dynamic method. It has
been found that the saturated oxygen concentration of the liquid does not change
significantly with the gas sparging rate and impeller speed. The change of
temperature aflects the saturated oxygen concentration significantly. The volumetric
mass transfer rate increases with the increase of liquid temperature in the range of
18-60 “c. Beyond 60 “c, the transfer rate decreases. Such a decrease is mainly due to
the dramatic drop in the gas void fraction.
Many chemical processes such as oxidation and leaching are operated at elevated
temperature, and the performance of these processes is dictated by the volumetric
mass transfer rate, kra, where kL is the mass transfer coefficient and a is the interfacial
area. Understanding gas dispersion and mass transfer behaviour is of vital importance
to improve the efficiency of these processes. However, there is little information
available regarding the effect of temperature on kLa, since most previous studies were
carried out at room temperature (-20°C, e.g. [ 1, 21).
The temperature effect on the mass transfer coefficient (kL)was studied by Suresh
et al. [3] (see also [4, 51). In their study, the variation of diffusivity (i.e. the diffusion
coefficient of solute in a solution) was brought about by the change of temperature
* Author f o r correspondence (
Y. Zhu andJ. Wu
(293-435K, 18-163°C). Although the value of kL obtained does not seem to change
significantly over the diffusivity range, the authors claimed that the variation of
diffusivity has some effect on the value of kL,the effect being larger than those due to
the change in viscosity, surface tension and density. The increase of kL with
temperature was also studied by Calderbank and Moo-Young [6], who found that kL is
a function of the viscosity and density of the liquid and the solute diffisivity. An
increase in temperature results in an increase in kL.
Although the value of kL increases with temperature, the volumetric mass transfer
rate (kLa)may not necessarily increase with temperature. The studies by Smith and
Gao [7] and Gao et al. [8] revealed that the gas void fraction was significantly
reduced at the boiling points. This indicates that the interfacial area may be
significantly reduced. Therefore, the value of kLa may not necessarily increase
monotonically with temperature.
The dependence of kLa on temperature, at low temperatures, was studied by Panja
and Phaneswara Rao [9]. They investigated the effect of temperature on kLa at
temperatures of 291, 308 and 318K (i.e. 18-45OC) at constant gas velocity and
various stirring rates, The results indicated a substantial increase in kLa with
increasing temperature. It was found that kLa increased by about I7-21% for every
10°C rise in temperature. However, the dependence of kLa on temperature in hot or
boiling systems has not been examined.
It appears from the previous studies that the effect of temperature on k ~ has
a not
been fully understood. The difference in the behaviour of kL and the interfacial area a
with temperature makes it very hard to predict the overall value of kLa, especially at
high temperatures. The aim of the present study is to examine ths effect.
Experimental Details
The experiments were carried out in the Energy & Thermofluids Engineering
Laboratory of CSIRO Manufacturing & Infrastructure Technology at Highett,
Victoria. A flat-bottomed cylindrical mixing vessel with a height of 1.0 m and an
internal diameter of T = 390 mm was chosen (see Figure 1). Four baffles were
installed in the tank, with each baffle having a width of TI12 and a wall clearance of
Ti64. This tank was filled with deionised water to a depth of 390 mm. A Rushton
impeller (DT6) of diameter 130 mm (D/T = 1/3) was used, and it was located at a
height of 130 mm from the tank bottom (C/T = 1/3).
The tank was constructed of stainless steel to withstand the high temperatures, and
was fitted with a sight glass mounted on the outside tank wall so that the level of fluid
could be monitored (see Figure 1). Heating was provided by a 2 kW immersion heater
and a 750 W capacity ribbon heater wound around the outside of the tank under the
insulation. A draw-off system was designed to sample the hot liquid for oxygen
concentration measurement. A counterflow condenser was fitted before the dissolved
oxygen (DO) probe to cool the water to a temperature suitable for the DO probe. The
temperature differential and residence time of the fluid in the condenser ensured that
the maximum temperature experienced by the probe never exceeded 27OC, even for
the maximum operating temperature of 80°C.
The gas was sparged through a 110 mm diameter sparging ring and 53 evenly
spaced holes located underneath the turbine and 65 mm from the tank bottom. The
GadLiquid Mass Transfer in Hot Sparged Systems
sparged system was operated at rotor speeds of 300 and 450 rpm for comparison with
data obtained previously at the same speeds. The gas sparging rates were 15.3, 23.9
and 35.3 Llminute for both rotor speeds. The system was operated at atmospheric
pressure and at room temperature, 40, 60 and 80°C, respectively.
sleel tank
Electrical heating
tape on outsde of
lank and under
Dacron lnsulatlon
Figure 1. Set-up of the aerated mixing tank with temperature control.
A variable-speed, three-phase electric motor in conjunction with a belt-drive speed
reducer was used to drive the agitator via an O N 0 S O W in-line torque transducer.
The analogue torque signal was digitised into an IBM PC via an A/D board for
computing the power consumption of the system. The liquid level in the vessel was
measured by an ultrasonic level meter.
Measurement of the mass transfer rate was carried out using a dynamic gassingout method [e.g. 2, 10-123. The dissolved oxygen concentration (DOC) of the water
was measured by a polarographic DO probe (ORION PCM800). This probe has a
response time of about 4 seconds, which is appropriate to resolve the present mass
transfer rate. A small volume of liquid was withdrawn from the tank using a
peristaltic pump and cooled to below 45°C before being passed through the DO
probe. The sampled water was passed over the sensing membrane of the probe at a
speed of approximately 0.02 m d . The bubbles in the withdrawn liquid were
removed from the loop by a slammer before the liquid was passed through the DO
probe. The temperature of the sample was measured using the in-built thermistor
sensor of the probe. The signals from the probe were sampled using an ORION data
acquisition package at a sampling frequency of 1 Hz.
Estimation of Mass Transfer Rate
The procedure for calculating the mass transfer rate has been described by Zhu et al.
[ 131 and therefore only a brief description is given here. The DOC data was analysed
to determine the mass transfer coefficient (kLa) based on the liquid-phase model:
Y. Zhu andJ. Wu
dC/dt = kLa(C, - C)
where C is the DOC at time t; and C, is the equilibrium concentration of oxygen in
the liquid phase. The integration of Equation (1) yields:
where C* = (C - Co)/(C,- Co)is the normalised DOC at time t; and C,, is the initial
DOC at time to,i.e. the starting time of air sparging.
Figure 2 shows a typical variation of DOC as a fhction of time. The oxygen in
the water was initially purged by sparging nitrogen gas through the system. When the
oxygen content is almost zero in the system, agitation was stopped so that the bubbles
could escape fiom the free surface. The agitation was then resumed and the sparging
gas was changed to compressed air. The DOC at the sampling point fluctuates (with a
standard deviation within k2% of the mean value) due to the time-dependent
behaviour of the local DOC and flow velocity fluctuations. However, the mean value
shows a clear exponential behaviour and the starting time (to) can be accurately
determined fiom the curve. Afier applying a least-squares fit using Equation (2), the
kLavalue can be determined from the response curve with an error of less than 2 4 % .
Results and Discussion
Figure 3 shows the saturated oxygen concentration as a function of gas sparging rate
at different temperatures. The saturated oxygen concentration is independent of
sparging rate, however temperature has a pronounced inverse effect on the
concentration. Figure 3 also indicates that the agitation speed affects the saturation
level of oxygen in water, although this is much less than the effect of temperature.
t (Second)
Figure 2. Normalised oxygen concentration C* = (C- C(s/(C, - CQ)as a function of
time. The solid line is from measurements with the DO probe placed outside the tank;
the ragged line is from measurements with the probe placed directly in the tank.
Gas/Liquid Mass Transfer in Hot Sparged Systems
Qg (I/m)
Figure 3. Temperature eflect on DOC
The effect of temperature on kLa at the two impeller speeds is shown in Figures 4
and 5. The kLa value increases with sparging rate and impeller speed (or power input),
in agreement with the general trend observed at room temperature for many rotors
under a wide variety of operating conditions [12, 13, 15-17]. For temperatures of
18-60°C, kLa increases with temperature. This finding is consistent with Panja and
Phaneswara Rao [9] results, they found that kLa increased when the temperature was
increased from 17 to 45°C. However, at about 6 0 T , kLa reaches a maximum for all
sparging rates. Further increase of temperature causes a decrease of mass transfer rate.
+Qg=23.93, I/rnin
Temperature ("C)
Figure 4, Effect of temperature on kLaat N
= 300 rpm.
Y. Zhu andL Wu
Temperature (“C)
Figure 5. Effect of temperature on kLa at N
= 450
The measured power consumption and gassing rate allow the plot of kLa in the
form suggested by van’T Riet [ 11, that is:
kLa = a(
where a is a constant (van’T Riet found the value of LY is 0.026 for coalescence
systems); P / V is the specific power demand; and UJ is the superficial gas velocity.
Figure 6 shows the measured kLa as a hnction of ( P l Vy4ZJ:’ . It is revealed that at
different temperatures, kLa is a linear function of ( P l V)”‘UPs, in spite of the impeller
speed and gassing rate.
A least-squares fit to the data points allows the value of a to be estimated for each
temperature. The present data at room temperature (-20°C) suggest a value of 0.03,
which agree to within 20% with the value of van’T Riet [l]. The average value of a at
each temperature is plotted in Figure 7. The value of a increases with temperature and
peaks at about 6OoC, beyond which the value decreases.
The diffisivity coefficient DL of the liquid phase varies with temperature. As
temperature increases, the value of DL increases (e.g. [14]). The mass transfer
coefficient (kL)is related to DLvia k, oc
(e.g. [5, 61). Therefore, kL increases as
temperature increases. On the other hand, as temperature increases, the interfacial
area a decreases since:
0 the void fraction is significantly reduced (e.g. [7]); and
0 the gas bubbles become bigger due to the increase of vapour content, and
therefore the bubbles rise faster than at room temperature. The overall gas holdup is significantly reduced at higher temperatures.
Gas/Liquid Mass Transfer in Hot Sparged Systems
0.06 1
020 0 4 0 A60 080 0 2 0
I ’
B40 A60 e80
m u
(PN)O 4U,0.5
Figure 6. Volumetric gasRiquid mass transfer coefficient at diferent temperatures.
Open symbols: N = 300 rpm; closed symbols: N = 4.50 rpm. The solid line is
Equation (3); the dotted lines mark the HO% boundaries of the correlation.
0.02 -
DT6 Impeller
0.01 -
Figure 7. Value of constant a as a function of temperature. a is the constant in the
correlation k,.a = a ( P l V)”‘Up’.
Y. Zhu andJ. Wu
The trend of kL and a with increasing temperature is depicted in Figure 8. The
trend of kL at large temperatures was estimated by extrapolating the diffusivity data of
Gowing [14] using a power law and the model of Calderbank and Moo-Young [6],
i.e. k, cc D:' . The interfacial area variation is estimated only through the variation of
vapour pressure inside the bubble. Although the exact values of these two quantities
are not estimated in Figure 8, the trend with variation of temperature is reasonable.
The product of kL and a will have a peak at intermediate temperatures. This explains
the behaviour of kLa with increasing temperature.
Mass Transfer
Coefficient k L
- - - ._- - - - -
Volumetric Mass Transfer
Temperature ("C)
Figure 8. Variation of the interfacial area a and mass transfer coeficient kL with
increasing temperature.
The effect of temperature on the volumetric mass transfer rate has been studied in a
mechanically agitated mixing tank at temperatures in the range of 20-80°C. The
Volumetric mass #transferrate is measured using a dynamic method. The saturated
oxygen concentration of the liquid does not change significantly with gas sparging
rate and impeller speed. The change of temperature affects the saturated oxygen
concentration significantly. The volumetric mass transfer rate increases with the
increase of liquid temperature in the range of 18-60°C. Beyond 60"C, the value
decreases, mainly due to the dramatic drop in the gas void fraction.
GasiLiquid Mass Transfer in Hot Sparged Systems
The authors would like to acknowledge the support from AMIRA International
through the AMIRA Project P419A sponsored by BillittodQueensland Nickel (QNI),
Iluka Resources Limited and Rio Tinto/Comalco.
1I .
van’T Riet, K. 1979. Review of measuring methods and results in nonviscous gas-liquid
mass transfer in stirred vessels. Ind. Eng. Chem. Process Des. Dev., 18(3), 357-364.
Linek, V., and Sinkule, J. 1991. The Influence of gas and liquid axial dispersion on
determination of kLa by dynamic method. Trans IChemE, 69(A), 308-312.
Suresh, A.K., Sridhar, T., and Potter, O.E. 1988. AIChEJ., 34,55.
Lee, J.H., and Foster, N.R. 1990. Measurement of gas-liquid mass transfer in multi-phase
reactors. Appl. Catalysis, 63, 1-36.
Kawase, Y., and Moo-Young, M. 1992. Mathematical models for design of bioreactors:
Applications of Kolmogorov’s theory of isotropic turbulence. Chem. Eng. J., 43, B 19B41.
Calderbank, P.H., and Moo-Young, M.B. 1961. The continuous phase heat and mass
transfer properties of dispersions. Chem. Eng. Sci., 16,39-54.
Smith, J.M., and Gao, Z. 2000. Vertical void distribution in gas-liquid reactors.
Proceedings of the 3rd International Symposium on Mixing in Industrial Processes, pp.
1 89- 196.
Gao, Z., Smith, J.M., Zhao, D., and Muller-Steinhagen, H. 2000. Void fraction and mixing
in sparged and boiling reactors. Proceedings of the 10th European Conference on Mixing,
pp. 2 13-220.
Panja, N.C., and Phaneswara Rao, D. 1991. Experimental studies on kLa in a mechanically
agitated contactor from transient electrical conductivity response data. Trans IChernE,
69(A), 302-307.
Linek, V., Moucha, T., and Sinkule, J. 1996. Gas-liquid mass transfer in vessels stirred
with multiple impellers: I1 Modelling of gas-liquid mass transfer. Chem. Eng. Sci., 51( 15),
Linek, V., Vacek, V., and Benes, P. 1987. A critical review of experimental verification of
the correct use of the dynamic method for the determination of oxygen transfer in aerated
agitated vessels to water, electrolyte solutions and viscous liquids. Chem. Eng. J., 34, 1 I34.
Chen, Z.D., and Chen, J.J.J. 1999. Comparison of mass transfer performance for various
single and twin impellers. Trans IChemE, 77(A), 104-109.
Zhu, Y., Bandopadhayay, P.C., and Wu, J. 2001. Measurement of gas-liquid mass transfer
in an agitated vessel: A comparison between different impellers. J. Chem. Eng. Japan,
34(5), 579-584.
Gowing, S. 1987. Dissolving of Bubbles in Liquids. David W. Taylor Naval Ship Research
and Development Centre, Report DTNSRDC-87/006.
Hickman, A.D. 1988. Gas-liquid oxygen transfer and scale-up: A novel experimental
technique with results for mass transfer in aerated agitated vessels. Proceedings of the 6th
European Conference on Mixing, Pavia, Italy, pp. 369-374.
33 I
Y. Zhu andf. Wu
16. Warrnoeskerken, M.M.C.G., and Smith, J.M. 1989. The hollow blade agitator for
dispersion and mass transfer. Chern. Eng. Res. Des., 67, 193-198.
17. Martin, T., McFarlane, C.M., and Nienow, A.W. 1994. The influence of liquid properties
and impeller type on bubble coalescence behaviour and mass transfer in sparged, agitated
reactors.Proceedings of the 8th European Conference on Mixing, pp. 57-64.
Без категории
Размер файла
441 Кб
mass, hot, transfer, sparged, gasliquid, system
Пожаловаться на содержимое документа