Compartment based population balance modeling of a high shear wet granulation process using data analytics

Abstract

Population balance models (PBMs) are used extensively to model various particulate processes such as granulation. A high shear granulation process is often assumed to be well mixed and is represented using a single compartment PBM. However, the inhomogeneities existent within the granulator are not effectively addressed using the single PBM representation for the process. Thus, a multi-compartment model is needed to account for the inhomogeneities within the granulator. In this study, the multiple compartments are identified from data mining methods (e.g. clustering) and their average values are thereby obtained. Using regression analysis, a general expression is obtained for the size of the compartments and the average values for different operating conditions. These expressions are then used within a multi-compartment PBM formulation to describe the process dynamics. Validation for the regression expressions also showed good agreement against the varying operating conditions. Also, the multi-compartment model is able to account for mechanical dispersion behaviors that is typically associated with a high-shear process. This study shows that the assumption of a single compartment for a high-shear granulator is often inadequate and the multi-compartment based approach can offer a better physical representation of the process.