A variable aperture model, including the random variation of fracture aperture as against the conventional parallel plate model, has been developed to adequately examine the transport of colloids/suspended particles in a single coupled fracturematrix system. Rather than relying on a complex geostatistical method for an accurate representation of fracture aperture, which requires an enormous field data and resource for its validation, a simple statistical method (linear congruential generator) is implemented in the present article. The random variation of fracture aperture is an honest representation of the unpredictable geometry/ morphology of fracture aperture in comparison with widely applied the conventional parallel plate model or the simple mathematical functions based on fractal theory (self-affine structures). A considerable number of parameters are involved in investigating the extent of penetration of colloids into the rock matrix, which creates complexity and ambiguity in the analysis. To overcome this problem, a single parameter “Maximum Penetration Factor” has been introduced for simple and reliable assessment of diffusion of colloids within the rock matrix. Additionally, a non-dimensional parameter ‘Matrix Mitigation Factor’ has been introduced in the present study, which can provide a means of evaluating the diffusion of suspended particles within the rock matrix when it comes to real time applications like microbial enhanced recovery (MEOR) and chemical enhanced recovery (CEOR) in the petroleum industry (nanoparticles and nanofluids). A semi-implicit finite difference model has been adopted for solving the coupled partial differential equations in the present numerical study. Finally, Neumann and Robinson boundary conditions as a function of time have been applied at the fracture inlet to better represent the field scenario as against the conventional constant source condition (Dirichlet). The model results indicate that there is a difference in concentration between the parallel plate model and random fracture model when it comes to colloidal concentration in the fracture and rock matrix. The variance in concentration is due to the inclusion of variation of the aperture in the variable aperture model, which is absent in the parallel plate model. Additionally, the results suggest that the variable source boundary condition has a significant influence on the transport of colloids in fracture-matrix system. Finally, from the evaluation of the extent of diffusion of colloids into rock matrix, it was concluded that that variable aperture model is associated with more mitigation of colloids compared to the parallel plate model, especially in the case of random fracture. © 2017, The Association of Korean Geoscience Societies and Springer-Verlag Berlin Heidelberg.