changing driving force r
Fig. 8.13. The importance of the headspace flow patterns in affecting bed-to-headspace heat and mass transfer. Two extreme cases are shown. In both cases it is assumed that the bed is well mixed. (a) If the headspace is well mixed, then the driving force for heat transfer is equal at all axial positions (b) If the flow through the headspace occurs by plug flow, then the driving force for heat transfer decreases as the air flows past the bed surface
In some cases the curves were consistent with a flow regime consisting of several well-stirred regions in series (Fig. 8.14(a)). In other cases they were consistent with plug-flow with axial dispersion (Fig. 8.14(b)). The rotational speed did not affect the type of headspace flow regime. Drums without substrate gave patterns at both gas flow rates that were consistent with the presence of 1 to 2 well-mixed regions in series within the headspace. However, in the presence of substrate there was a difference between the flow patterns at the two different gas flow rates. At both substrate loadings, the response curves obtained with the gas flow rate of 2.7 L min-1 were consistent with the presence of 1 to 3 well-mixed regions in series within the headspace, whereas the response curves obtained with the gas flow rate of 2.7 L min-1 were consistent with plug-flow with axial dispersion.
Hardin et al. (2001) used CO as a tracer to study flow patterns in a 200-L drum. The patterns were consistent with those that would be expected for a central plug-flow region surrounded by a dead region (Fig. 8.15(a)). The dead region includes a part of the headspace gases and all of gas in the inter-particle spaces in the bed. The dead region is well mixed in the radial direction but there is no axial transport.
The presence or absence of baffles and the superficial velocity had the greatest effects on the fraction of the drum occupied by the dead region and the rate of transfer between the plug-flow and dead regions (Fig. 8.15(b)). With an increase in the superficial velocity of the air (defined as the volumetric air flow rate divided by the cross-sectional area of the empty drum) there was less mixing between the plug-flow and dead regions and the dead region occupied a greater proportion of the gas volume in the drum. Compared to the absence of lifters, the presence of lifters led to a greater degree of exchange between the plug-flow and dead regions and meant that the dead region represented a smaller proportion of the drum.
Unfortunately, it is not possible to make generalizations from these studies. Flow patterns within the headspace of rotating-drum bioreactors will be greatly influenced by the design and positioning of the air inlet and outlet. One thing is clear, however: If end-to-end aeration is used, it is not reasonable to assume that the headspace is well mixed.
Was this article helpful?