overflow vessel, 6—pump, 7—
rotameters, 8—oxygen electrodes; A—
amplifier, ChS—channel selector,
CC—conductivity cell, ADC—analog-
digital converter, PC—personal
computer, DLS—data logging system,
CB—compensating bridge.
<0.05 (Kaštánek et al., 1993). The tracer concentration in the exit stream was measured
continuously by means of a conductivity cell installed at the column outlet (see Figure
1.1). The amplified electrode signal was sampled and stored in a PC and the experimental
response curves were subsequently transformed to a dimensionless form, corresponding
to the system response to a unit pulse signal (Dirac function). On the basis of previous
analysis (Kaštánek and Zahradník, 1973), values of the abscissa of curve maximum
(θYmax) and ordinale at θ=1 (Yθ=1) were selected as suitable criteria for a simple
comparison of experimental RTD curves with those calculated from respective mixing
models. Values of the liquid backflow coefficient, Ei+l, i, defined as the ratio of the liquid
backflow between the adjacent column stages to the net liquid throughput, QL, were
measured by the stationary colorimetric method. Solution of methylene blue dye was
continuously dosed to the lower column stage and its concentration in samples taken
from individual column stages under steady state conditions was determined by the
photometer Spekol-2 (Carl Zeiss Jena). Values of the backflow coefficient were then
calculated from the relation
(1)derived from the steady state tracer balance (column stages are numbered from the top).
Comparison of experimental RTD curves proved, in agreement with the literature (Kato
and co-workers, 1972; Vlaev and Zahradník, 1987), that the presence of the solid phase
had negligible effect on the liquid phase residence time distribution within the whole
range of experimental conditions. Values of the backflow coefficient were therefore
measured only in the air-water system.
The stationary method was employed for determining values of the volumetric gas-
liquid mass transfer coefficient, kLaL. Deoxygenised liquid or slurry phase was
continuously fed to the upper column stage and steady state concentration of dissolved
oxygen in individual column stages was measured by polarographic electrodes of the
Clark type. Respective kLaL values were then calculated from the oxygen balance
equations assuming plug flow of the gas phase. Liquid (slurry) mixing in the column was
described , on the basis of our RTD measurements, by the model of perfectly mixed tanks
in series with backflow, for values of the backflow coefficients determined at respective
flow conditions.
The average values of gas holdup ratio (bubble bed voidage) in individual column
stages were determined by the method of pressure differences modified for the
application in slurry systems. Values of wetted plate pressure drop were calculated from
readings of pressure taps situated below and above the respective distributing plates.
New methodologies for multiphase bioreactors 1 5