The extraction and processing of minerals from surface mines and quarries can produce significant fugitive emissions as a result of site activities such as blasting, unpaved road haulage, loading, primary crushing and stockpiling. Uncontrolled fugitive dust emissions can present serious environmental, health, safety and operational issues impacting both site personnel and the wider community. Simulation technology is finding increasing use for the purposes of advanced warning of potential problem emissions in addition to providing a basis for future planning applications where demonstrable compliance with regulatory requirements are necessary. The initial re-entrainment and subsequent dispersion of fugitive dust presents a process complicated by the combination of the in pit topography, the surrounding natural topography and the dynamic nature of emissions from these sites. These factors impact upon the accuracy and reliability of the conventional Gaussian plume based computational prediction methods employed for regulatory compliance and IPPC applications. This paper proposes that optimal modelling of open pit emissions may be more accurately achieved by the use of a multi-scale predictive modelling approach utilising computational fluid dynamic (CFD) methods for high resolution near source dispersion and conventional Gaussian based methods for far field dispersion modelling. This paper presents a numerical based flow and dispersion analysis of a typical UK based open pit utilising CFD in conjunction with a conventional Gaussian plume based methods. Typical operating emissions and meteorological conditions are obtained from long te

rm data records collected at a large operating quarry extraction operation in the UK. Emissions are modelled using a Lagrangian framework within conventional atmospheric boundary layer (ABL) profiles expressed as functions of turbulence and velocity parameters under assumed neutral conditions. Results are presented in terms of the impact of site topography on in pit retention as compared to the Gaussian based method.