469279 A Phenomenological Model for the Prediction of Granule Drying Times in a Fluid Bed Dryer

Monday, November 14, 2016: 12:30 PM
Continental 4 (Hilton San Francisco Union Square)
Ariel R. Muliadi and Joseph Bullard, Chemical and Materials Engineering, Vertex Pharmaceuticals, Boston, MA

A computational model designed to predict the drying times of pharmaceutical granules in a fluid bed dryer is described. The model is comprised of two major components, the first of which utilizes a well-established two-stage phenomenological drying model, and assumes that granule drying consists of evaporation of liquid ‘domains’ from the granules surfaces followed by liquid evaporation from within the intra-granular porous network. The transient evolution of the granule moisture content is solved by numerically integrating the one-dimensional heat equation. The moving liquid/air interface is solved in a stationary computational domain via a Landau-style mapping. The second component numerically simulates the stochastic nature of the fluidization behavior of granules inside the fluid bed dryer. The combination of this stochastic simulation with the existing phenomenological approach allows for the study of mechanistic correlations between the granule physical characteristics (size, porosity, permeability, and initial moisture content) and the process conditions (e.g., drying air flow rate and temperature, and fluidization “efficiency”). We demonstrate that center-point conditions from an extensive series of process development runs can be used to derive the empirical inputs to the model, and that these inputs can be scaled according to the aforementioned process conditions—ultimately allowing for the prediction of drying times across a broad range of processing conditions.

Extended Abstract: File Not Uploaded