New models for particle embedment during micro-particle injection into naturally fractured reservoirs are developed. The proposed models aim to predict production benefit from the application of micro-particle injection during coal seam gas (CSG) stimulation with broader applications to other naturally fractured reservoirs. The elastoplastic finite element modelling is applied to coal sample from Surat basin (Australia), to predict micro-particle embedment and fracture deformation under various packing densities and closure stresses. The coupled lattice Boltzmann-discrete element model (LBM-DEM) is then used for permeability prediction. These results are combined in a radial Darcy flow analytical solution to predict the productivity index of CSG wells. Modelling results indicate that considering elastoplastic fracture surface deformation leads to smaller permeability increase and less production enhancement, if compared with the linear elastic deformation of fracture implemented in traditional models. Although focused on Australian coals, the developed workflow is more broadly applicable in other unconventional resources. Modelling of particle transport and leak-off in coal fracture intersected with a cleat using LBM-DEM approach demonstrates the effects of particle and cleat sizes, particle concentration and sedimentation on the leak-off process. The leak-off is significantly affected if the particle-cleat size ratio is higher than 0.5. Particle sedimentation increases leak-off into vertical cleat substantially, but has no effect on horizontal cleat. Suspensions of higher concentration result in higher leak-off for cleats with different apertures.