The Investigation, Development, and Analysis of Models of Electrodialysis

Electrodialysis is an emerging electromembrane salt separation technology, industrial implementation of which is currently inhibited by the risk-averse nature of industry and poor optimisation of units. The work presented in this thesis involves understanding how electrodialysis is modelled, rigorously testing one of the most basic assumptions, and developing more advanced modelling strategies. A review of existing published literature revealed there is a wide variety of models both in their fundamental construction and the assumptions taken to neglect phenomena. Despite this, good agreement with empirical data is near ubiquitous but attributed to an over-reliance on fitting parameters. One assumption taken by all researchers is channel uniformity whereby the repeating geometry is leveraged to significantly reduce model complexity. In Part I, this assumption was investigated through computational fluid dynamics simulations through which significant flow maldistribution between channels was revealed. Further, an analytical model was found to capture the degree of maldistribution well in a dimensionless number and revealed how it can be influenced. A study on the impact of maldistribution revealed that while the electric resistance is marginally affected, the limiting current density is significantly reduced. Particle image velocimetry experiments empirically demonstrated the presence of maldistribution and transport experiments validated its impact. In Part II, an advanced process model of electrodialysis was developed using the analogy of an electric circuit. This model was designed to be adaptable and avoid the use of empirical fitting parameters. A membrane transport number and resistance model proved vital in ensuring predictive accuracy over a range of concentrations and voltages. To demonstrate the generality of the model, it was extended to describe bipolar membrane electrodialysis which similarly showed good agreement with experimental data. Overall, this work has contributed significant advancements to how electrodialysis is analysed, designed, and modelled.