Fracture characterization and estimation of fracture porosity of naturally fractured reservoirs with no matrix porosity using stochastic fractal models

摘要

Determining fracture characteristics at the laboratory scale is a major challenge. It isknown that fracture characteristics are scale dependent; as such, the minimum samplesize should be deduced in order to scale to reservoir dimensions. The main factoraffecting mechanical and hydrological characteristics of natural fractures is aperturedistribution, which is a function of scale and confining pressure, rather than roughnessof one fracture surface. Scale and pressure dependencies of artificial and naturalfractures were investigated in this study using an X-Ray CT Scanner. Fractal dimension,D, and amplitude parameter, A, of fracture aperture approaches a constant value withincreased sampling area, similar to the behavior of fracture roughness. In addition, bothparameters differ under different confining pressures for a reference sampling area.Mechanical properties of fracture-fracture deformation behavior and fracture normalstiffness were obtained from CT scan data as well.Matrix porosity is relatively easy to measure and estimate compared to fractureporosity. On the other hand, fracture porosity is highly heterogeneous and very difficult to measure and estimate. When matrix porosity of naturally fractured reservoirs (NFR)is negligible, it is very important to know fracture porosity to evaluate reservoirperformance. Since fracture porosity is highly uncertain, fractal discrete fractal network(FDFN) generation codes were developed to estimate fracture porosity. To reflect scaledependent characteristics of fracture networks, fractal theories are adopted. FDFNmodeling technique enables the systematic use of data obtained from image log andcore analysis for estimating fracture porosity. As a result, each fracture has its ownfracture aperture distribution, so that generated FDFN are similar to actual fracturesystems. The results of this research will contribute to properly evaluating the fractureporosity of NFR where matrix porosity is negligible.