Use this URL to cite or link to this record in EThOS:
Title: Multi-scale population balance modelling and controllability of granulation processes
Author: Ramachandran, Rohit
ISNI:       0000 0004 2671 7149
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
Access from Institution:
Many continuous granulation plants operate below their design capacity, suffering from high recycle rates and instabilities. Thus, there is an immediate economic incentive for effective operation and control of granulation units. The overall granulation process is integrated and interacting, with limited manipulated variables available (e. g. binder addition, nozzle locations and mixing rate). Hence. the complex process dynamics and operational challenges presented warrant a fundamental model-based strategy for design, operation, control and scale-up that is well supported by experimental analyses. A realistic model of the granulation process has to account for the granule size, the binder content, and the porosity (or related parameter bulk density), thereby necessitating a three-dimensional population balance model to yield a good representation of the process. While this multidimensional population balance model is warranted by the physics of the problem. it is a bigger challenge to derive kernels (rate laws) for the key granulation mechanisms. Most kernels in the literature are empirical and/or semi-empirical and provide little insight into the intricacies of the granulation mechanisms. This effectively results in an inability to make the necessary engineering decisions to improve control of the granulation process. Hence, this thesis is concerned with a more systems-centric approach to enhance the design, control and scale-up of granulation processes. Experimental studies on a lab-scale batch drum granulator for a Calcite/PVOH-H20 system were performed to assess granulation kinetics and model development of the granulation process. Effects of process /material properties and liquid binder distribution on granule properties, illustrating the non-homogeneity of key particle attributes and which justify the need for multi-dimensional population balances, were studied. Process sensitivities, manipulations and potential disturbances were identified, formulating a comprehensive control configuration for granulation processes, with application seen in a continuous drum granulation of limestone. While carrying out experiments, multiple granule attributes were characterised and this presents a challenge, which this research addresses accordingly. A population balance model incorporating nucleation, aggregation, breakage and consolidation was developed in this research. Novel aspects are the mechanistic formulations of the nucleation, aggregation and breakage kernels which are derived from first-principles. Such mechanistic descriptions of the rate processes lend themselves to a more in-depth understanding of the granulation process, contributing fundamental knowledge to the design, control and scale-up of these processes. A sensitivity analysis of the model was then performed to ascertain the influence of model parameters on the particle density distribution. Continuing from this, a compartmentalised version of the combined population balance model was developed, for the purpose of controllability analysis. Results obtained were used to identify suitable control-loop pairings to facilitate enhanced control-loop performance. Experimental validation of the population balance model is an integral part of this research. The model was quantitatively validated using lab-scale experimental data for granule size, binder content and porosity. The tuned model was then able to predict evolutions and distributions of granule attributes for different operating conditions and formulations. The model was also validated for different granulation systems. This illustrates the robustness and flexibility of the model and these results are promising toward the longer-term step of a first-principles based predictive model for the granulation process that can help alleviate the need for large number of experiments. As an alternative to deriving the above-mentioned mechanistic kernels, a discrete element modelling (DEM) approach was also undertaken in this thesis. Based on a Volume of Fluid (VOF) method, the analysis carried out provided useful information to help understand the effect of primary particle morphology on granulation kinetics making it possible to establish relationships between material and process/design properties and granulation process behaviour
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available