Fed-Batch Bioproduction of Spectinomycin
33
oxygen demand varies according to the metabolic state of the microorganism.
Consequently, the air flow rate is linked to the rate of antibiotic production.
The oxygen balance contains the same activation-reaction-inhibition terms as
the model but with the yield coefficients defined for oxygen utilization. Since the
air flow rate reflects the oxygen demand of the process, it should be possible to
predict the airflow rate from the oxygen balance equation. For this prediction of
the air flow rate to be meaningful, the parameters evaluated from the data by
nonlinear regression should be used. If this prediction is possible, it would
demonstrate that the energetic relationship which exists between spectinomycin
and oxygen utilization is also described by the model. Hence, if the air flow rate
profile for maximum spectinomycin productivity is known a priori, then the
maximum spectinomycin productivity can be obtained for every fed-batch run
by controlling the air flow rate profile.
It is indeed possible to reconstruct the air flow rate profile using the parameters evaluated (Table 2) by nonlinear regression from the spectinomycin and
glucose data. The air flow rate data points used for demonstrating the prediction
are each the mean values of 20 on-line sample points. The yield coefficients
YCA, YI"IA, and YP2A were calculated from the data for the substrate uptake for
growth and product formation. Here kta is treated as a parameter. Once kta is
evaluated, the air flow rate can be computed using a suitable expression relating
air flow rate to the mass transfer coefficient k~,. Figures 16-20 show the
predicted air flow rate profiles along with the actual air flow rate data.
12
-i
0
144
9
in
i
I
i
I
a
I
i
l
i
I
i
24
48
72
96
120
Time (h)
Fig. 16. Reconstruction of the air flow rate profile for spectinomycin bioproduction with glucose
feed concentration of 100 gl - 1 9 air flow rate data; -- predicted air flow rate
33
oxygen demand varies according to the metabolic state of the microorganism.
Consequently, the air flow rate is linked to the rate of antibiotic production.
The oxygen balance contains the same activation-reaction-inhibition terms as
the model but with the yield coefficients defined for oxygen utilization. Since the
air flow rate reflects the oxygen demand of the process, it should be possible to
predict the airflow rate from the oxygen balance equation. For this prediction of
the air flow rate to be meaningful, the parameters evaluated from the data by
nonlinear regression should be used. If this prediction is possible, it would
demonstrate that the energetic relationship which exists between spectinomycin
and oxygen utilization is also described by the model. Hence, if the air flow rate
profile for maximum spectinomycin productivity is known a priori, then the
maximum spectinomycin productivity can be obtained for every fed-batch run
by controlling the air flow rate profile.
It is indeed possible to reconstruct the air flow rate profile using the parameters evaluated (Table 2) by nonlinear regression from the spectinomycin and
glucose data. The air flow rate data points used for demonstrating the prediction
are each the mean values of 20 on-line sample points. The yield coefficients
YCA, YI"IA, and YP2A were calculated from the data for the substrate uptake for
growth and product formation. Here kta is treated as a parameter. Once kta is
evaluated, the air flow rate can be computed using a suitable expression relating
air flow rate to the mass transfer coefficient k~,. Figures 16-20 show the
predicted air flow rate profiles along with the actual air flow rate data.
12
-i
0
144
9
in
i
I
i
I
a
I
i
l
i
I
i
24
48
72
96
120
Time (h)
Fig. 16. Reconstruction of the air flow rate profile for spectinomycin bioproduction with glucose
feed concentration of 100 gl - 1 9 air flow rate data; -- predicted air flow rate