Dust explosions are a major hazard frequently encountered in vital sectors like food, energy, defense (propellants and explosives) and pharmaceuticals. These explosions emanate from rapid combustion of clouds of suspended fine particles in the micron range. Quantitative estimation of dust explosion propagation is crucial to their mitigation, and for providing estimates of physical quantities that determine the explosion behavior for design of safety systems. In this contribution, a multi-scale reaction engineering approach for modeling dust explosions has been presented. In the model, two scales are considered. At the particle scale, a detailed model involving various transport steps is written and solved for a variety of different boundary conditions. This model is used to build an effective reaction rate term which is incorporated in the dust cloud-scale CFD model through an appropriate source term. The CFD model for the effective dust cloud mixture is then executed to model the propagation of the dust explosion. Validation of the developed CFD model has been carried out for two different kinds of dust particles, aluminum (metallic dust) and starch (organic dust), for experimental data published in literature. Finally, a case study has been presented which shows the applicability of present approach in modeling real situations. It is demonstrated that using the same physics and estimated kinetic parameters from particle scale model or smaller scale experiments, fairly satisfactory prediction for dust explosions in real geometries could be obtained. (C) 2012 Elsevier B.V. All rights reserved.