Fluidized Bed Spray Granulation – From process understanding to modeling of nucleation and dust integration

Fluidized bed spray granulation is a process to obtain coarse-grained particles directly from suspensions, solutions or melts, ready for sale without any need for post-treatment. The process combines spray drying with granulation and classification, thus allowing the product particle size distribution to be adjusted according to customer demands. Well-known applications include pharmaceutical products as well as mass products like detergents or fertilizer. Because of the process intensification combining several normally consecutive unit operations in one apparatus, it is imaginable, that design and specification of process parameters is mainly decided on the basis of intuition or experience, if available. This is partly due to a lack of knowledge concerning the occurring mechanisms, but also due to the complex interaction of these mechanisms which makes it difficult to pre-estimate process behavior without simulation tools. There are models published in literature helping to understand the process but still neglecting kinetics so far unknown like e.g. nucleation and dust integration. Both are important for a continuously running process deciding on whether it is possible to achieve steady state or not. In this work these mechanisms have been examined experimentally with the help of tracer methods in the case of dust integration and by analysis of continuous experiments in the case of nucleation (cf. (Grünewald et al., 2006)). On the basis of the findings, the kinetics of dust integration and seed formation have been modeled by combining strongly simplified fluid dynamics with deposition models and heat and mass transfer. The kinetics along with a population balance approach allows the dynamic modeling of the process including transient effects. The dust integration mechanism mainly controls the dust content in the plant and thus influences strongly the seed formation rate. The long-term experiments show a damped oscillating behavior of the particle size distribution and the production rate until achieving steady-state. The simulation results are able to reflect this behavior.