ACS PRF | ACS
All e-Annual Reports
45799-G9
Estimation of Macroscopic Fracture Transport Properties from Micro-Scale Flow Structures and Fracture Geometry
During the first year of this project, we investigated single-phase flow dynamics in a rough fracture model constructed from the scanned CT images, using computational fluid dynamics. Flow simulations gave a detailed description of pressure and velocity fields and showed that this approach can be used for the deterministic analysis of single-phase flow through fractures. Our results demonstrate the formation of preferential flow channels and their correlation with local structural characteristic of the fracture. We also identified the need to implement a different modeling approach for two-phase flow simulations, due to limitations in the computational fluid dynamics method for the simulation of discrete fluid interfaces under capillary effects. A modified invasion percolation is proposed, which allows the incorporation of contact angles, surface curvatures, and tracking of fluid interfaces for direct comparison with experimental findings.
During single-phase simulations, our results demonstrate the formation of preferential flow channels and their correlation with local structural characteristic of the fracture. When roughness and tortuosity are present in a real fracture, the mismatch of pressure drop predictions from the real fracture and the parallel plate model reached two orders of magnitude. The underestimation of pressure drop at a fixed fracture flow velocity by using the parallel plate model is inherently caused by the assumption of a smooth parabolic profile, excluding the effects of roughness and channel tortuosity. In addition, the effect of roughness and tortuosity can be accounted for with a scaling factor applied to the hydraulic conductivity term in the cubic law.
For two-phase simulations, we implemented a modified invasion percolation (MIP) method to model capillary drainage, map immiscible fluid location on the discretized, CT-scanned fracture, and determine fracture capillary pressures. The MIP technique is an algorithm seeking the least resistant pathway for the advancement of the invading fluid using the Young Laplace equation. Preliminary results, showing fluid distributions generated by the MIP model, yield fair agreement with experimental observation of phase structures. Comparison of simulation results against the experiments suggest that the accessibility of fluids into the fractures as well as a hierarchy of sequence play a major role in the modeling of primary drainage.
The importance of the effect of fracture morphology on the flow of fluids through fractures has been observed both experimentally and numerically. Nevertheless, understanding the correlation between the aperture field of a fracture and its macroscopic transport properties remains limited. During the progress of the second part of this project, we will continue with the description of transport properties from pore-scale measurements, through modeling and experimentation.
In the modeling part, we will examine the relationship between geostatistical characteristics of the fracture aperture field and a fracture's capillary pressure curve. The study will use our Modified Invasion Percolation approach with the added improvement of normal and in-plane curvature terms of the fluid interface for calculating the local capillary pressure. A satisfactory agreement with the experimental results will enable us to further study the characterization of capillary pressure for a wide range of the fracture structures having different geostatistical characteristics. The generation of new fracture structures will be based on a matrix decomposition technique. Mean, spatial correlation length and variance of the aperture distribution will be the geostatistical parameters used in the generator to characterize each fracture aperture field. The results will allow us to quantify the impact of fracture aperture connectivity on the shape of the capillary pressure curve, and to predict capillary pressures from summarized geostatistical properties describing fracture apertures.
In addition to the described modeling effort, core flood experiments will be conducted to study the effect of wettability on the distribution and transmissibility of fluids in porous media. Current findings provide a link between pore channel structure and the distribution of two immiscible fluids inside a rough open fracture, and demonstrate the importance of capillary forces in the mobility of fluids in fractured rocks. These findings contribute to the understanding of micro-scale fluid dynamics in fractures, and provide the basis for the study of fluid flow in interconnected pore channels. Experimental results will allow the identification of flowing conditions favoring displacement mechanisms such as snap-off and trapping as well as the investigation of the effect of wettability and pore structure on fluid distribution at the pore scale. High-resolution X-ray CT scanning will be used to accomplish a detailed characterization of the system, as function of space and time, during the core-flood experiments.
