Several numerical simulations of fluid flow were performed using the Lattice Boltzmann method and image processing techniques to estimate the effective properties of 2-D porous rocks. The effective properties evaluated were the physical characteristics that allow fluid flow including the effective porosity, permeability, tortuosity, and average throat size to determine the storage and transport of fluids in porous rocks. The permeability was compared using the Darcy model simulation and the empirical Kozeny–Carman Equation. The results showed that the Lattice Boltzmann method and image processing techniques can estimate the effective parameters of porous rocks. Furthermore, there was a good correlation between permeability and parameters such as effective porosity, tortuosity, and average throat size. The Darcy model simulation revealed a gamma distribution in the permeability, while the empirical Kozeny–Carman Equation showed a log-normal distribution.

Hydrocarbon production in the reservoir depends on fluid flow through its porous media, such as fractures and their physical parameters, which affect the analysis of the reservoir’s physical properties. The fracture’s physical parameters can be measured conventionally by laboratory analysis or using numerical approaches such as simulations with the Lattice Boltzmann method. However, these methods are time-consuming and resource-intensive; therefore, this research explores the application of machine learning as an alternative method to predict the physical parameters of fractures such as permeability, surface roughness, and mean aperture. Synthetic three-dimensional digital fracture data that resemble real rock fractures were used to train the machine learning models. These included two convolutional neural networks (CNNs) designed and implemented in this research—which are referred to as CNN-1 and CNN-2—as well as three pre-trained models—including DenseNet201, VGG16, and Xception. The models were then evaluated using the ${\mathrm{R}}^2$ and mean absolute percentage error (MAPE). CNN-2 was the best model for accurately predicting the three fracture physical parameters but experienced a drop in performance when tested on real rock fractures.