Biomedical Engineering Reference
In-Depth Information
(a)
Term
Scaled estimate
Prob>|t|
Intercept
Temperature (induction)
Ramp time
Total induction duration
(Temperature (induction) -40.5)*(Ramp time -40)
(Temperature (induction) -40.5)*(Total induction duration -5.04348)
(Ramp time -40)*(Total induction duration -5.04348)
90.37115
-8.595555
-2.349906
8.2248835
-1.774924
-2.872896
1.2456973
<.0001*
0.0001*
0.1801
<.0001*
0.3050
0.1056
0.4678
(b)
Term
Scaled estimate
Prob>|t|
Intercept
Temperature (induction)
Ramp time
Total induction duration
(Temperature (induction) -40.5)*(Ramp time -40)
(Temperature (induction) -40.5)*(Total induction duration -5.04348)
(Ramp time -40)*(Total induction duration -5.04348)
112.06084
10.553237
-0.565352
4.5083364
-0.812693
-2.941176
-3.212074
<.0001*
0.0161*
0.8872
0.2286
0.8384
0.4639
0.4245
(c)
Term
Scaled estimate
Prob>|t|
Intercept
Temperature (induction)
Ramp time
Total induction duration
(Temperature (induction) -40.5)*(ramp time -40)
(Temperature (induction) -40.5)*(Total induction duration -5.04348)
(Ramp time -40)*(Total induction duration -5.04348)
108.55072
5.5253623
-1.068841
3.7691131
-5.625
-2.291667
-0.625
<.0001*
0.3959
0.8681
0.5260
0.3872
0.7220
0.9225
Figure 5.10. Scaled estimates of main effects of induction-phase operational parameters on
product yield (a), product quality 1 (b), andproduct quality 2 (c) performance froma full factorial
DOE study. Note: Data have been normalized against the average small-scale model perfor-
mance at set point operating conditions.
earlier and illustrated inFig. 5.12, implementationof thedesign spaceconcept shouldallow
the manufacturers to make changes in operating conditions without getting preapproval
from the regulatory bodies as far as the changes are still within the approved design space.
Classification of performance parameters is primarily based on process develop-
ment studies from which it is determined whether a performance parameter is a useful
measure of process consistency (key) and/or product quality (critical) for a particular unit
operation. Of all the variability in operational parameters that was examined, none
resulted in unacceptable product quality as ascertained by the results from the worst case
study shown in Table 5.4. Hence, for the fermentation step in the case study it was
determined that there were no critical performance parameters. Step yield, product
quality 1, and product quality 2 were identified as key performance parameters (KPPs).
Process characterization data are also used to classify operational parameters. This
is done based on the effect of an operational parameter on the key and critical
performance parameters of the process. As illustrated in Fig. 5.13, if variability of an
operational parameter will result in a critical performance parameter (CPP) to deviate
outside the respective PVAC, then it would be identified as a critical operational
parameter. If the variability in the operational parameter would not result in a CPP to
fail its PVAC but does have a significant impact on the CPP or KPP, the parameter is
identified as a key operational parameter. Parameters that do not significantly impact
CPP or KPP are identified as nonkey operational parameters. For the case study, and as
Search WWH ::




Custom Search