mass transfer rate equal to that achieved at comparable working conditions in gas-liquid
reactors.
The functional dependence of kLaL on gas holdup was within the whole range of
experimental conditions virtually independent of phases flow rates and solid phase
concentration and varied only with the number of column stages. In cases where the
suspension of solid particles can be treated as a pseudohomogeneous continuum (i.e. for
small non-sedimenting particles), gas holdup ratio can thus be obviously regarded, in
analogy with the gas-liquid systems, as a lumped parameter representing, for a particular
reactor configuration, the effect of individual operating variables on the rate of gas-liquid
mass transfer. Experimental data proved a favourable effect of secondary gas dispersion
in sectionalised bubble columns on kLaL values corresponding to given gas holdup in the
reactor. Comparison with the dual-flow plates proved the suitability of the plate design
tested in our work for the countercurrent flow arrangement, regarding both higher values
of bubble bed voidage achieved and lower plates pressure drop.
RECOMMENDATIONS FOR FUTURE WORK
While there seems to be no doubt that sectionalised bubble columns may offer distinct
advantages to reaction processes in g-l-s systems, as compared with other types of
multiphase contactors, further experimental work is still needed, aimed at clarifying
particular aspects of their performance with slurry systems and ultimately at providing
reliable, generally applicable design and scale-up recommendations. An outline of
suggested research topics is given in the following.
Performance of multistage slurry reactors in the two basic flow arrangements,
concurrent upflow and countercurrent flow, should be compared under different working
conditions (gas and liquid flow rates, solids loading, particle size and density). Both dual
flow plates and plates with downcomers should be tested in the countercurrent flow
regime with the purpose of formulating general criteria for objective selection of flow
arrangement and plate type for specific process demands. In comparison with the single-
stage bubble columns, the multistage units exhibit significantly smaller operating
flexibility, i.e. a much narrower region of gas and liquid (slurry) flow rates in which these
reactors can be operated for a specific plate geometry. The selection of the optimum plate
parameters for given working conditions thus represents an essential step of the
multistage reactor design. The design rules proposed for dual flow plates in gas-liquid
reactors (see e.g. Deckwer, 1985 or Kaštánek et al., 1993) should be re-examined in the
view of possible solid phase effect and, similarly, additional experiments should be
devoted to testing plates with downcomers in a wide region of phases flow rates.
In analogy with the single-stage bubble columns, conditions of the homogeneous
bubbling regime formation in multistage slurry reactors should be investigated.
Generally, parameters governing the formation and stability of the homogeneous
bubbling regime in g-l-s systems include plate and column geometry, liquid properties,
solids loading and particle size and density (Pandit and Joshi, 1986; Gavroy et al., 1995).
The effect of such parameters and their mutual, possibly synergistic, coupling should be
examined with due regard to the specific features of multistage units (e.g. gas
redispersion in individual stages, lower bed height and Hoi/Dr ratio etc.), which could,
Multiphase bioreactor design 20