Rigorous model-based control strategy for batch crystallizers (CrysCODE)

Project members: A.N. Kalbasenka, A.E.M. Huesman, O.H. Bosgra

Sponsored by: AKZO-Nobel, BP-Amoco, DSM, Purac BioChem, Solvay

This research is a part of the multi-disciplinary research project CrysCODE (Crystallizer COntrol and DEsign).

The ultimate goal of the present research is to design a model-based control system for batch crystallization processes. This ambiguous research title can be elaborated further resulting in the following subgoals:

• validation of the models for a crystallizer (draft-tube and draft-tube-baffle crystallizers;
• validation of the models for process actuators (conventional and ultrasonic seeding techniques, impeller frequency);
• validation of the model for a measurement system (crystal size distribution measurement and supersaturation measurement devices);
• model reduction;
• design and validation of a dynamic observer (extended Kalman filter);
• design and validation of a model predictive controller.