Multiphase Bioreactor Design

RESULTS AND DISCUSSION

Preliminary Tests

The solid phase distribution in the reactor volume was virtually uniform within the whole
range of experimental conditions. The suspension of ZnO particles was therefore further
treated as a pseudohomogeneous continuum, characterized for each particular solid
concentration by appropriate values of slurry density and viscosity, evaluated from
relations proposed by Brinkman (1952) and Roscoe (1952):

(2)

(3)

Results of pressure loss measurements in the sectionalised column (N=3 and 6) showed
its “autoregulation” behaviour. The total pressure drop of the column, defined as the sum
of wetted plates pressure drop and of the hydrostatic pressure of the liquid phase (slurry)
in the column, was constant for given values of liquid flow rate and solid phase
concentration within the whole region of gas flow rates. Obviously, this implies that the
increase of plates pressure drop with increasing gas flow rate was appropriately
compensated by the simultaneous decrease of overall liquid (slurry) holdup in the column
(see Table 1.1).

Longitudinal Liquid Phase (Slurry) Mixing

Evaluation of RTD characteristics proved that the axial mixing of the liquid phase or
suspension in the sectionalised reactor was appropriately described by the two-parameter
model of perfectly mixed tanks in series with backflow between the adjacent tanks, for
the number of series terms equal to the real number of column stages. Values of the
second parameter, Ei+1,i, obtained from independent measurements proved further that the
extent of liquid backflow was appreciably suppressed by the proper design of internal
plates e.g. in comparison with the data obtained in our previous study (Kaštánek et al.,
1974) in a three-stage column of identical diameter for free plate area ratios φ=4 and 8%
respectively (Table 1.2). Formation of gas cushions below the internal sectionalising
plates was observed within the regions of gas and liquid superficial velocities, UoG=
0.025–0.050 m s−^1 , uoL=0.013–0.029 m s−^1 , and consequently, liquid or suspension
backflow was virtually negligible at such flow conditions (Eav≤0.1). In this operation
region, liquid mixing was satisfactorily described by the simple tanks-in-series model.
Incidentally, the same result was reported by Poncin et al. (1990) from experiments in a
six-stage countercurrent-flow bubble column sectionalised by dual-flow plates. An
increase of backflow coefficients observed at uoG>0.050 m s−^1 can be ascribed to the
partial conversion of the plates working regime at such conditions. Due to the sharp
increase of plates pressure drop at these uoG values, a certain portion of gas passed
upwards through the downcomer tubes causing excessive liquid entrainment. Based upon
such results, the optimum operating region of internal plates can be defined by the gas
velocity range uoG=1–2 (uoG)crit, where (uoG)Crit is the critical value of superficial gas

Multiphase bioreactor design 6