Rock mechanical properties are critical to reduce drilling risk and maximize well and reservoir productivity. This paper present a methodology of predicting mechanical properties (Young's modulus and Poisson's ratio) from 3D rock models generated using laboratory measurements or downhole logging data. The 3D rock model provides the microstructure and boundary to simulate rock elastic properties. Mechanical properties are computed from rock models using the finite element method.
Laboratory measurements were conducted on four Berea sandstone samples to determine the mechanical properties for comparison and other properties as input in rock modeling, such as bulk and grain density, porosity, mineralogy, and grain-size distributions. Numerical results from rock models generally match the core measurements of the corresponding samples. The methodology proposed in this study could potentially be applied downhole for predicting the mechanical property profile along the wellbore, as all input parameters to generate rock models can be derived from logging measurements.
The application of geomechanics has a significant impact on all aspects of field development - from the exploration and appraisal phases through development and harvest of the field, to the final abandonment. One optimized field development plan using geomechanics can result in very large cost savings over the life of the field. Geomechanical properties of reservoir rocks (Young's modulus and Poisson's ratio) are always required in various geomechanics applications (Yale et al., 1995). Their determination are usually from laboratory triaxial compression testing on a limited number of core samples. Test data of cores are commonly considered as the standard and used as a reference and calibrated value to build the geomechanical model. Because of cost, time and limited core availability, triaxial testing is difficult to conduct for each interval and each well. Alternatively, the static mechanical properties can be derived from the dynamic properties obtained from well logs using some correlation built on core data.