D.Ing. (Mechanical Engineering) / Most coal fired power stations in South Africa are equipped with Electrostatic Precipitators (ESP's). With the ongoing reduction of allowable emissions, as negotiated with the Chief Air . Pollution Control Officer (CAPCO) of the Department of Environmental Affairs and Tourism (DEAT), ways to reduce emissions are sought. In the case of emission levels exceeding the values set by the controlling authority load losses are required for compliance. This however has the effect of plant operating inefficiently and a loss of revenue will result. Specifically in times of growing demand, when more and more of the currently installed generation capacity is required to satisfy the demand, forced load reductions are not desirable. Performance enhancement of ESP's can be achieved by means of system optimisation. Research was initiated to achieve the capability of modelling important dynamic aspects of ESP performance using Computational Fluid Dynamics (CFD). This modelling capability would create the opportunity to investigate the different influencing factors which govern the dust collection efficiency. In the past ESP flow has been modelled by means of mathematical modelling with various degrees of success world-wide. It was found that the accuracy of flow modelling as presently carried outby researchers world wide, is not sufficient to represent the complex inlet flow. Commercially available performance simulation software is based on empirical modelling principles and do not include the complexity of flow fields and re-entrainment and thus results have been limited in accuracy. Computational fluid dynamics software is commercially available and widely used to simulate industrial flow for plant design and optimisation. This technology has been applied with increasing confidence and success in the past. However, often the physical phenomena relevant for the performance simulation of the plant is not integrated into the code and specialised user routines are created to achieve a valid performance model. This research study introduces a unique integrated simulation methodology based on a commercial CFD code. The work focuses on the accurate modelling of fluid flow and collection dynamics in an ESP. User subroutines have been created to simulate particle charging, collection and re-entrainment. The results of the simulations are compared to measurement at actual plant.