(182b) Case Studies Involving Mechanistic and Empirical Models Used to Define Design Space Ranges and Corresponding Large Scale Verification | AIChE

(182b) Case Studies Involving Mechanistic and Empirical Models Used to Define Design Space Ranges and Corresponding Large Scale Verification

Authors 

Tabora, J. E. - Presenter, Bristol Myers Squibb
Hallow, D., Bristol-Myers Squibb Company


Placeholder only follows.  Abstract to be added.

Abstract Text guidelines: A minimum of 150 words is required for the abstract text, and no more than 1500 words may be submitted. Each abstract is allowed to include up to 20 figures.
From http://www.pharmaqbd.com/bms_tabora_aiche_award_2011/
There is some resistance to the adoption of MVA modeling stemming from the prolific reliance on one-factor-at-a-time experimentation. While this approach is often appropriate and acceptable in early stages of process development, it frequently continues to be utilized in later stages of development although increased efficiencies may be attained through the implementation of combined MVA modeling/experimental approaches to improve process understanding and overcome key development challenges.
In addition, the discussion of which parameter(s) to explore simultaneously can be nuanced and complex, thereby compounding the reluctance to incorporate multivariate experimentation into process development and optimization workflows. However, I am very optimistic about the potential for increased acceptance and uptake of such approaches and have seen the adoption of these techniques increase substantially over the past five years. It is not only a regulatory expectation, but also leads to substantial gains in efficiency, which our scientists are recognizing quite rapidly.
It is widely acknowledged in the business intelligence sector that efficient multivariate analysis and modeling generally require significant efforts in data aggregation. The same holds true in chemical process development, and although tools to support this requirement are not currently readily available, there is an increasing appreciation of the importance of data consolidation. Hence, the industry is aggressively moving toward solutions to address this gap through the use of local databases and datamarts that aggregate data from multiple sources.