This study compares the calculated fracture apertures in a fragmented rock layer under different stress scenarios using two different approaches. Approach 1 is a simplified method using a two-dimensional (2D) mapping of the fracture network and projects the far-field stresses to individual fractures, and calculates the dilation, normal and shear displacements using experimental stiffnesses available in the literature. Approach 2 employs a three-dimensional (3D) finite element method (FEM) for the mechanical analysis of the fragmented rock layer considering the interaction with the neighbouring rock layers, frictional interfaces between the rock blocks, stress variations within the fragmented rock layer, and displacements, rotations and deformations of rock blocks. After calculating the fracture apertures using either of the approaches, the permeability of the fragmented rock layer is calculated by running flow simulations using the updated fracture apertures. The comparison between the results demonstrates an example of the inaccuracies that may exist in methods that use simplified assumptions such as 2D modelling, ignoring the block rotations and displacements, projected far-field stresses on fractures, and the stress variations within the rock layer. It is found that for the cases considered here, the permeability results based on apertures obtained from the simplified approach could be 40 times different from the results from apertures calculated using a full mechanical approach. Hence, 3D mechanical modelling implementing realistic boundary conditions, while considering the displacements and rotations of rock blocks, is suggested for the calculation of apertures in fragmented rocks.