Executive Summary
This report summarizes the main achievements of the three years of development of new crystallization technologies for improved crystal size and shape control during the crystallization process. The successful crystallization process and system design requires an interdisciplinary effort, which ranges from population balance model (PBM) development of the system concept, through efficient implementation of model equations to soft-sensor development, which is required for the model predictive control (MPC) design as well. This report gives a deeper insight into these interdisciplinary development efforts, which also highlights the achievable improvements enabled by the combination of process modeling, high performance process simulation and optimization.
In the first year we showed that the application of wet-milling during crystallization can significantly improve the process flexibility and attainable crystal size domain. The GPU acceleration halved the simulation time, which enabled faster optimization. In the same year we also developed a portable PBM solver for batch crystallization of 1D and high aspect ratio 2D particles with optional GPU acceleration, involving primary and secondary nucleation, crystal growth and dissolution. For model based control applications, a fast, approximate, geometrical model based CSD to CLD and aspect ratio distribution (ARD) transformation was also developed.
In the second year the concept of integrated crystallizer-wet mill process simulation and optimizing design has been extended to 2D high aspect ratio crystal case. Furthermore, using the portable PBM solvers developed in the first year, we designed and implemented a full PBM based nonlinear model predictive control (NMPC) system for the batch cooling crystallization of L-ascorbic acid. This NMPC, in addition to solved the barrier of computational time on an average PCs also used the FBRM CLD as a feedback device, for the first time in the literature.
In the third year, the optimal operation of integrated crystallizer-wet milling systems was generalized by discovering a repeating crystallization mechanism scheduling pattern in the optimal operating profiles regardless of the applied process model (1D or 2D), initial conditions (seeded/unseeded and seed loading) and the objective (smaller or larger desired crystal size). Furthermore, by improving the calculation performance of the portable GPU accelerated PBM solver, it was showed that the real time model based shape- control became real-time feasible. In addition, the iterative model based experimental design (IMED) was applied for a spherical agglomeration process, which work demonstrates how the experimental conditions can be design to minimize the number of experiments for precise parameter estimation.
Achieved deliverables
- Development of a generic Matlab based software package for the simulation of 1D and 2D batch and continuous MSMSP crystallizers.
- Quick design method for integrated batch crystallizer-wet mill systems.
- Fast and approximate 2D CSD to CLD and ARD transformation development.
- 1D NMPC development and implementation for L-ascorbic acid crystallization.
- Real-time feasible 2D NMPC simulation.
- Application of IMED for a spherical agglomeration process.