Autor(es):
Ferreira, António
; Ferreira, Cecília
; Teixeira, J. A.
; Rocha, F. A.
Data: 2010
Identificador Persistente: http://hdl.handle.net/1822/11399
Origem: RepositóriUM - Universidade do Minho
Assunto(s): Bubble column; Mass transfer; Absorption; Multiphase reactors; Particles; Temperature
Descrição
The knowledge about the effect of solid phase properties, at different temperatures, on gas–liquid transfer
and the respective physical mechanisms has been poorly studied. In the present work, the temperature
and solid properties (size and density) effects on the gas–liquid mass transfer characteristics in a bubble
column were experimentally evaluated. Gas–liquid–solid systems formed by air/water/polyvinyl chloride
(PVC) beads and air/water/expandable polystyrene (EPS) beads were used. For each system, volumetric
liquid side mass transfer coefficient, kLa, was determined under different temperatures (20–35 ◦C),
superficial gas velocities (up to 7.2 mm/s), solids sizes (210, 549 and 591 m) and concentration (up to
5 vol.%). The results show that the temperature plays an important role on mass transfer phenomena as
kLa increases as temperature is increased. However, temperature does not have, significantly, influence
on the solids effect on kLa. In what concerns the effect of the solids on kLa, a negative effect of their
presence is observed being this effect more pronounced for the largest particles (for PVC). In addition,
for the same solid size, a decrease in kLa occurs when the solid loading increases (observed in both cases,
PVC and EPS). Complementary, bubble/particle interaction studies, using an image analysis technique,
were also done aiming a better understanding of the effects of the two types of particles – PVC and EPS
– on kLa values, being clearly confirmed that physical (e.g. density) and chemical properties of the solids
are important parameters to be taken into account on mass transfer and hydrodynamic studies.
Based on theoretical models for kL and a, an empirical correlation for kLa dependence on the experimental
variables was developed. The experimental results were fitted with a mean deviation of 5%, that
is similar to the experimental error (5%).