ABSTRACT
The pharmaceutical and biopharmaceutical industry is changing. Terms that reflect the current state are quality by design (QbD), design space, control strategy, process analytical technology (PAT), and process signature. Process modeling is an integral part of the QbD framework. Models for process understanding, statistical process monitoring, and process control are required during the life cycle of the product. Batch or semi-batch processes are dynamic, nonlinear, and of finite duration. Batch process variables are both autocorrelated and cross correlated. These characteristics (due to the nature of the batch process operation itself) should be considered when modeling batch operations for process understanding, statistical process control (SPC), real-time process control, and optimization. Such models may be mechanistic (based on first principles), or empirical (based on appropriate data), or hybrid. Batch unit operations may be described with first principles models when the chemical, biochemical, and physical processes that are taking place in the batch vessel are well understood (e.g., Cinar et al. 2003). However, when this is not the case or it is time-consuming and appropriate data are available, empirical models can be developed.
