Quantitative Prediction of 3-D Multiple Parameters of Tectonic Fractures in Tight Sandstone Reservoirs Based on Geomechanical Method

Conventional fracture characterization model for low-permeability tight reservoirs cannot be directly built through data statistical method and simplified strain energy density method under high confining pressures. In this research, we make an attempt to establish a series of geomechanical models for prediction of fracture distribution in brittle reservoirs, especially for tight sandstone reservoirs. First, we emphasize that energy generated by tectonic stress on brittle sandstone can be distinguished into accumulating elastic strain energy, fracture surface energy, and residual strain energy and natural fractures can be interpreted or inferred from geomechanical-model-derived strains. For this analysis, we confirm 0.85 σ_c as the key threshold for mass release of elastic strain energy and bursting of macrofractures, then deduce and build a physical relation model between fracture volume density and strain energy density under uniaxial stress state based on theory of geomechanics and CT scanning. Then on this basis, using combined Mohr-Coulomb criterion and Griffith's criterion and considering the effect of filling degree in fractures, we continue to modify and deduce the mechanical models of fracture parameters under complex stress states. Finally, all the geomechanical equations are planted into the FE platform to quantitatively simulate the present-day 3-D distributions of fracture density, aperture, porosity and occurrence based on paleostructural restoration of Keshen anticline. It is concluded that the evolutionary model, the Keshen structure, present to us is a distinct and intuitive top-pop-up fold with low amplitude and relatively low-density fractures in the top, has only experienced pre-folding stage and strong compression stage, and has not yet been developed to rapid uplift stage. Its predictions also agreed well with presentin-situ core observations and FMI interpretations.