CFD Modelling of Chemical Reaction Systems in Jacketed Stirred Tank Reactors

Modelling of jacketed stirred tank reactors (STRs) generally assumes perfect mixing within the reactor leading to uniform concentration/temperature distributions and uniform flow of the heat transfer fluid through the jacket. Our previous computational fluid dynamics (CFD) study of hydrodynamics and conjugate heat transfer between the jacket fluid and the reactor contents in a pilot-scale STR revealed that perfect mixing and uniform flow are unrealistic assumptions. The main objective of this project is to develop an organic synthesis reaction process model for relevant reaction systems in the fine chemicals and pharmaceutical industries via integration of reaction kinetics into the CFD-heat transfer model. Chemical synthesis of organic compounds is often complex involving multiple competitive reactions and conversion/selectivity. The product distributions are critically dependent on the hydrodynamics, mixing and heat transfer characteristics of the reactor and particularly for large-scale STRs. Any inhomogeneous and transient hydrodynamic conditions prevailing will result in spatial and temporal variations in reaction conditions which can generate undesired by-products and reduce yield. These variations are quite different depending upon the size of the STRs from laboratory to pilot plant to manufacturing scales. Finally, in many instances, the active ingredients are isolated from the reactor contents via crystallisation and the presence of by-products can adversely affect the quality of crystals. Even the crystallisation steps can be influenced adversely by any inhomogeneous and transient hydrodynamic conditions prevailing within vessel and the presence of by-products. It is therefore of paramount importance to minimise by-products formation through improved design and efficient operation of laboratory, pilot plant and industrial reactors. Application of first principles models of the physical phenomena occurring in the reactor and during the crystallisation can facilitate this.
In addition to the modelling, experiments will be carried out in laboratory scale STRs (0.5 - 5 L) to determine reaction kinetics and for collection of process data for model validation. This comprehensive model will be an extremely powerful tool for the "design by modelling" approach, which can predict accurately reaction product distributions and facilitate reactor design, scale up and optimision of operating conditions, and determining a safe operational envelope via exploring process conditions leading to thermal runaways. The model can also be used to conduct computational experiments to minimise the number of real experiments in large scale sizes which are inherently unsafe, costly and time consuming.