phenomena for which it is not practical to make experimental measurements,
and therefore these sections will also comment on what insights the mathematical models have given us as to what controls the performance of SSF systems.
Mathematical models can be applied to SSF bioreactors in two slightly different ways. First, predictive models can be proposed. These models predict
system performance based on the initial conditions and the operating parameters of the bioreactor. Such models can be used to explore the likely performance of the bioreactor under conditions which have not yet been tried
experimentally, and therefore can be useful tools in guiding the scale-up
process. A model developed for a smaller scale bioreactor can be used to
simulate performance at larger scales before the larger scale bioreactor is built.
This increases the chances of identifying and avoiding operating problems on
the large scale.
Second, interpretive models can be proposed. These models take as input
operating variables of the bioreactor and measurements of those state variables
which it is practical to measure, and give as output estimates of other state
variables, including state variables which it may be impossible or impractical to
measure during the fermentation. Such models are quite useful in SSF, because
it is not practical to obtain direct measurements of the biomass, and parameters
which can be measured on-line such as oxygen or carbon dioxide concentrations are only indirectly related to the amount of biomass. The accuracy of such
models can be checked by predicting state variables which are easily measured
experimentally, such as bed temperature. These models are interpretive and not
predictive because they rely on the constant input of fermentation data; they
cannot predict bioreactor performance simply on the basis of initial conditions.
However, they are still quite useful since, if the measured variables can be
measured on-line, the model can be used quite successfully in control schemes.
The application of interpretative models has been hampered by the technical
challenges in collecting adequate data on-line, and as a result, to date bioreactor
models have been of the predictive type. Some general comments can be made
about such models First, due to the heterogeneity of SSF systems, a spatial
variable is often involved, which leads to partial differential equations and
therefore makes solution of the equations more difficult than would occur in
perfectly mixed systems. Second, the sophistication of the model and the detail
with which it describes the system depend on the complexity of the system and
the motivation behind the modeling work.
4
Microscale Phenomena Occurring Within SSF Bioreactors
The microscale processes demonstrated in Fig. 3 are intrinsic to SSF due to the
particulate nature of the substrate. They occur in all bioreactors and relatively
little can be done to influence them in the way the bioreactor is designed and
operated since they occur at the surface and inside the individual substrate
particles. The most that can be achieved through bioreactor design and operational strategies is to promote exchange between the particle and air phases
and to ensure that the transport processes within the air phase are not limiting.
Biochemical Engineering Aspects of Solid State Bioprocessing
81
and therefore these sections will also comment on what insights the mathematical models have given us as to what controls the performance of SSF systems.
Mathematical models can be applied to SSF bioreactors in two slightly different ways. First, predictive models can be proposed. These models predict
system performance based on the initial conditions and the operating parameters of the bioreactor. Such models can be used to explore the likely performance of the bioreactor under conditions which have not yet been tried
experimentally, and therefore can be useful tools in guiding the scale-up
process. A model developed for a smaller scale bioreactor can be used to
simulate performance at larger scales before the larger scale bioreactor is built.
This increases the chances of identifying and avoiding operating problems on
the large scale.
Second, interpretive models can be proposed. These models take as input
operating variables of the bioreactor and measurements of those state variables
which it is practical to measure, and give as output estimates of other state
variables, including state variables which it may be impossible or impractical to
measure during the fermentation. Such models are quite useful in SSF, because
it is not practical to obtain direct measurements of the biomass, and parameters
which can be measured on-line such as oxygen or carbon dioxide concentrations are only indirectly related to the amount of biomass. The accuracy of such
models can be checked by predicting state variables which are easily measured
experimentally, such as bed temperature. These models are interpretive and not
predictive because they rely on the constant input of fermentation data; they
cannot predict bioreactor performance simply on the basis of initial conditions.
However, they are still quite useful since, if the measured variables can be
measured on-line, the model can be used quite successfully in control schemes.
The application of interpretative models has been hampered by the technical
challenges in collecting adequate data on-line, and as a result, to date bioreactor
models have been of the predictive type. Some general comments can be made
about such models First, due to the heterogeneity of SSF systems, a spatial
variable is often involved, which leads to partial differential equations and
therefore makes solution of the equations more difficult than would occur in
perfectly mixed systems. Second, the sophistication of the model and the detail
with which it describes the system depend on the complexity of the system and
the motivation behind the modeling work.
4
Microscale Phenomena Occurring Within SSF Bioreactors
The microscale processes demonstrated in Fig. 3 are intrinsic to SSF due to the
particulate nature of the substrate. They occur in all bioreactors and relatively
little can be done to influence them in the way the bioreactor is designed and
operated since they occur at the surface and inside the individual substrate
particles. The most that can be achieved through bioreactor design and operational strategies is to promote exchange between the particle and air phases
and to ensure that the transport processes within the air phase are not limiting.
Biochemical Engineering Aspects of Solid State Bioprocessing
81