Heat transfer in packed-bed bioreactors has received a significant amount of
modeling attention. In wide packed beds lacking internal heat transfer plates
the design and operational variables available are the bioreactor height, and the
aeration rate and the temperature of the inlet air [97]. Using unsaturated air to
promote evaporation is not an option because the bed will dry out quickly. Of
the operating variables, increasing the superficial velocity is the most effective
in reducing axial temperature gradients [97]. Using low inlet air temperatures is
not practical since, in order to achieve reasonable control of the temperature at
the air outlet end, the air inlet temperature must be of the order of 10–15 °C
below the optimum temperature for growth, and such low air temperatures cool
the inlet region of the bed to temperatures which restrict growth [97].
Regarding bioreactor height, if it is possible to identify a temperature that
should not be exceeded at any point in the bed during the fermentation, then,
due to the steady rise in temperature with bed height, the bed can be no higher
than the height at which this critical temperature is reached at the time of peak
heat production [142]. This height can be denoted the critical height. It depends
on the growth kinetics and the operating conditions used. For growth rates and
operating conditions which have been reported in the literature, this height
ranges from as low as 0.1 m to as high as 2.5 m [142].
The heat transfer dynamics are quite different in the case where a thin packed
bed is used. The contribution to overall heat removal by radial heat conduction
to the walls and convective removal by cooling water can be significant. This was
observed experimentally by Saucedo-Castaneda et al. [96], who used a waterjacketed column of 6 cm internal diameter. There was a significant axial temperature gradient in the first 5 cm at the air inlet end of their 35 cm high column.
Above this, the axial gradient was negligible but there was a significant radial
gradient. An explanation for this behavior was provided by the modeling work
of Mitchell and von Meien [143]. In the lower regions of the column the temperature is kept low enough by the incoming air that the organism grows rapidly.
It grows so rapidly that the heat transfer to the cooling water cannot remove all
the heat and therefore the air temperature rises as it moves up through the column. The less effective cooling by the warmer air means that there is an increase
in bed temperature with height. As the bed temperature rises above the
optimum, the growth rate slows. Once the growth rate slows to a value which
gives a rate of heat production equal to the rate of heat removal to the cooling
water, then there is no further increase in bed temperature with height. In those
regions where the axial temperature gradient is negligible the evaporation will
be negligible, so this design also reduces the importance of the water balance.
A Zymotis-type bioreactor with closely spaced internal heat transfer plates
will perform similarly to the thin column of Saucedo-Castaneda et al. [96], in
that, after a short distance near the air inlet in which there is a significant axial
temperature gradient, there will be a negligible axial gradient with height [143].
This suggests that a Zymotis-type bioreactor can be of infinite height. However,
as pointed out below, pressure drop considerations may limit the height of
Zymotis-type bioreactors.
Thin packed bed columns of less than 10 cm are obviously impractical for
large-scale applications because of the small capacity of individual columns,
108
D.A. Mitchell et al.
modeling attention. In wide packed beds lacking internal heat transfer plates
the design and operational variables available are the bioreactor height, and the
aeration rate and the temperature of the inlet air [97]. Using unsaturated air to
promote evaporation is not an option because the bed will dry out quickly. Of
the operating variables, increasing the superficial velocity is the most effective
in reducing axial temperature gradients [97]. Using low inlet air temperatures is
not practical since, in order to achieve reasonable control of the temperature at
the air outlet end, the air inlet temperature must be of the order of 10–15 °C
below the optimum temperature for growth, and such low air temperatures cool
the inlet region of the bed to temperatures which restrict growth [97].
Regarding bioreactor height, if it is possible to identify a temperature that
should not be exceeded at any point in the bed during the fermentation, then,
due to the steady rise in temperature with bed height, the bed can be no higher
than the height at which this critical temperature is reached at the time of peak
heat production [142]. This height can be denoted the critical height. It depends
on the growth kinetics and the operating conditions used. For growth rates and
operating conditions which have been reported in the literature, this height
ranges from as low as 0.1 m to as high as 2.5 m [142].
The heat transfer dynamics are quite different in the case where a thin packed
bed is used. The contribution to overall heat removal by radial heat conduction
to the walls and convective removal by cooling water can be significant. This was
observed experimentally by Saucedo-Castaneda et al. [96], who used a waterjacketed column of 6 cm internal diameter. There was a significant axial temperature gradient in the first 5 cm at the air inlet end of their 35 cm high column.
Above this, the axial gradient was negligible but there was a significant radial
gradient. An explanation for this behavior was provided by the modeling work
of Mitchell and von Meien [143]. In the lower regions of the column the temperature is kept low enough by the incoming air that the organism grows rapidly.
It grows so rapidly that the heat transfer to the cooling water cannot remove all
the heat and therefore the air temperature rises as it moves up through the column. The less effective cooling by the warmer air means that there is an increase
in bed temperature with height. As the bed temperature rises above the
optimum, the growth rate slows. Once the growth rate slows to a value which
gives a rate of heat production equal to the rate of heat removal to the cooling
water, then there is no further increase in bed temperature with height. In those
regions where the axial temperature gradient is negligible the evaporation will
be negligible, so this design also reduces the importance of the water balance.
A Zymotis-type bioreactor with closely spaced internal heat transfer plates
will perform similarly to the thin column of Saucedo-Castaneda et al. [96], in
that, after a short distance near the air inlet in which there is a significant axial
temperature gradient, there will be a negligible axial gradient with height [143].
This suggests that a Zymotis-type bioreactor can be of infinite height. However,
as pointed out below, pressure drop considerations may limit the height of
Zymotis-type bioreactors.
Thin packed bed columns of less than 10 cm are obviously impractical for
large-scale applications because of the small capacity of individual columns,
108
D.A. Mitchell et al.