Simulation of agglomeration reactors via a coupled CFD/direct Monte-Carlo method

Mechanistic modelling of fluid bed granulation, Part I: Agglomeration in pilot scale process

The present study aims to develop a mechanistic model to predict the performance of a fluid bed granulation process. Therefore, the behavior of the bed was investigated experimentally for various operating conditions. It was observed that the granule Loss on Drying (LoD) and granule size are strongly interrelated. In detail, the maximum final granule size was observed at an intermediate final LoD. Consequently, there is an optimum spray rate and inlet temperature with respect to the granule size. Besides, it was demonstrated that the experiments delivering lower LoD result in a more elongated final granule. Aimed at enabling the prediction of the bed performance numerically, a single-compartment, population-balance-based model was developed and validated against experimental data. The model parameters associated with the growth rate of granule were estimated and mechanistically correlated to the relevant operating conditions. Detailed analysis of the experimental results suggested that these model parameters may be partially connected to the granule LoD. Subsequently, in order to examine the accuracy of the developed model, a simulation was performed for a new set of operating conditions not previously accounted for in the correlations. The comparison of the simulated bed performance, when compared to the experimental results, proved with reasonable accuracy the reliability of the developed model in predicting the temporal evolution of granule size. Therefore, this study can be a step forward in developing a stand-alone granulation model, via modeling heat and mass transfer, to simulate evaporation and drying in a fluid bed granulator.

Mechanistic modelling of fluid bed granulation, Part II: Eased process development via degree of wetness

The performance of a fluid bed granulator was investigated through experimental and numerical study to develop a stand-alone fluid bed granulation model. The single-compartment model proposed in part I (for agglomeration modeling) was extended to account for i) evaporation of freely-flowing droplets, and ii) particle drying. This model enables us to predict the granule liquid content and temperature besides the granule size. Accurately, the equations of heat and mass conservation were solved in parallel to the population balance calculation of the agglomeration. In the same manner as for the agglomeration model, the model parameters associated with the drying model were estimated and correlated to the relevant quantities. The analysis of the experimental results revealed the significant contribution of the system "degree of wetness" to the bed performance, i.e., granule size and loss on drying (LoD). As the agglomeration model parameters were partially correlated to LoD in Part I, the presented model was revisited by inclusion of the degree of wetness. The reliability of the developed model in predicting the temporal evolution of granule size, liquid content, and temperature was proven through comparing the bed performance between simulation and experiment. Subsequently, to lowering the costs associated with experimental run, an approach was proposed based on the degree of wetness, aimed at reducing the number of experiments required for the design of experiment (DoE). The results of our simulation using reduced experiments demonstrated that the degree of wetness can be a promising indicator for the performance of the fluid bed granulator as well as for more efficient design of experiment.