Numerical study of fracture aperture characteristics and their impact on single-phase flow and capillary-dominated displacement

The importance of the effect of fracture morphology on fluid flow through fractures has been extensively recognized. Nevertheless, understanding the relationship between fractures' void structure and the associated macroscopic transport properties remains limited. The objective of this study is to examine both single-phase flow and capillary-dominated displacement in a rough-walled fracture. A commercially available computational fluid dynamics software was used to investigate single-phase flow. A physically-based percolation model was developed to study capillary-dominated displacement, and validated by using previous experimental data as a modeling reference.;Knowledge of detailed descriptions of a fracture from Computed Tomography (CT) measurement was used for generating a realistic fracture replica. Single-phase flow simulations were carried out on the CT-scanned fracture by means of Computational Fluid Dynamics (CFD) simulation. The effects of fracture roughness and path tortuosity were explored by visualizations of pressure contours and vector velocity profiles of flows thorough the realistic fracture. The results of CFD simulations were compared against predictions using idealized smooth fractures or parallel plate models. Noticeable discrepancies showed that the cubic law was inadequate in describing flows in fractures because of lack of a valid physical representation of a real fracture. The deviations of the cubic law predicting flow in a real fracture can be corrected by applying a scaling factor to its hydraulic conductivity.;A modified invasion percolation (MIP) approach is proposed to model primary drainage. A detailed fracture structure and fluid distribution maps were obtained from previous experiments using X-ray computed tomography; these results were used as reference for validating the proposed model. Parameters representing fracture characteristics are mean aperture, standard deviation and spatial correlation length. The effect of these three parameters on fracture capillary pressure curves were quantified relative to the entry pressure and the irreducible water saturation. The results show that the geostatistical parameters are strongly related to the magnitudes of the entry pressure and the irreducible water saturation. Geostatistical analysis elaborates on the correspondence between structure characteristics of fractures, preferential flow channeling and fracture capillary pressure curves, thus providing a methodology to predict fracture capillary pressure from fracture aperture parameterization.