A predictive model framework for DTB crystallisers is presented. The overall model contains relat... more A predictive model framework for DTB crystallisers is presented. The overall model contains relations for secondary nucleation, growth, attrition, fines classification and dissolution, but only contains two parameters that need to be estimated. These parameters can be estimated using experimental data from a different crystalliser than the one that is to be simulated. In this paper the influence of fines removal is studied for a 1.1 m 3 DTB crystalliser. Experimentally observed and simulated product median crystal sizes are compared for fines withdrawal rates of 2 and 3 l's 1 and dissolution residence times of 10 and 100 s. The predictive capabilities of the framework are good with respect to the average product median size, but unsatisfactory with respect to its oscillations. The results are promising considering the absence of parameters that are dependent of crystalliser scale, type and operating conditions.
A general, systematic procedure is presented to support the development and statistical verificat... more A general, systematic procedure is presented to support the development and statistical verification of dynamic crystallization process models. Within this procedure, methods are presented to address the optimal design of dynamic experiments for model discrimination among several crystallization kinetics models. The problem of designing dynamic experiments is cast as an optimal control problem that enables the calculation of optimal discrete sampling points, experiment duration, fixed and variable external control profiles, and initial conditions of a dynamic experiment subject to general constraints on inputs and outputs. Within this framework, methods are presented to provide experiment design robustness, accounting for parametric and thus prediction uncertainty.
Useful components and deficiencies of conventional crystallisation design procedures and models a... more Useful components and deficiencies of conventional crystallisation design procedures and models are identified. A systematic procedure consisting of a hierarchical decomposition into design of the product, task, flowsheet and individual crystalliser is presented. At each level the relevant specifications, variables and necessary knowledge are identified. For the flowsheet and crystalliser level a predictive crystallisation model is presented to estimate kinetic parameters from experimental CSD data and to analyse the behaviour of design alternatives. Currently, the procedure is applicable to cooling, flashcooling and evaporative crystallisation. The developed model is restricted to systems with little agglomeration. Validation of the proposed design procedure and models requires industrial case studies.
A predictive model framework for DTB crystallisers is presented. The overall model contains relat... more A predictive model framework for DTB crystallisers is presented. The overall model contains relations for secondary nucleation, growth, attrition, fines classification and dissolution, but only contains two parameters that need to be estimated. These parameters can be estimated using experimental data from a different crystalliser than the one that is to be simulated. In this paper the influence of fines removal is studied for a 1.1 m 3 DTB crystalliser. Experimentally observed and simulated product median crystal sizes are compared for fines withdrawal rates of 2 and 3 l's 1 and dissolution residence times of 10 and 100 s. The predictive capabilities of the framework are good with respect to the average product median size, but unsatisfactory with respect to its oscillations. The results are promising considering the absence of parameters that are dependent of crystalliser scale, type and operating conditions.
A general, systematic procedure is presented to support the development and statistical verificat... more A general, systematic procedure is presented to support the development and statistical verification of dynamic crystallization process models. Within this procedure, methods are presented to address the optimal design of dynamic experiments for model discrimination among several crystallization kinetics models. The problem of designing dynamic experiments is cast as an optimal control problem that enables the calculation of optimal discrete sampling points, experiment duration, fixed and variable external control profiles, and initial conditions of a dynamic experiment subject to general constraints on inputs and outputs. Within this framework, methods are presented to provide experiment design robustness, accounting for parametric and thus prediction uncertainty.
Useful components and deficiencies of conventional crystallisation design procedures and models a... more Useful components and deficiencies of conventional crystallisation design procedures and models are identified. A systematic procedure consisting of a hierarchical decomposition into design of the product, task, flowsheet and individual crystalliser is presented. At each level the relevant specifications, variables and necessary knowledge are identified. For the flowsheet and crystalliser level a predictive crystallisation model is presented to estimate kinetic parameters from experimental CSD data and to analyse the behaviour of design alternatives. Currently, the procedure is applicable to cooling, flashcooling and evaporative crystallisation. The developed model is restricted to systems with little agglomeration. Validation of the proposed design procedure and models requires industrial case studies.
Uploads
Papers by S. Bermingham