Geomechanics Rock Property Modeling using Statistical Concepts

Geomechanicalrock properties, including elastic moduli, unconfined compressive strength, and friction angle, play a crucial role in decisions related to drilling, completions, production, and reservoir management. While laboratory tests for geomechanical properties are typically conducted for each field, the data are often sparse and may not encompass the full range of the variability and anisotropy of the formations. Consequently, empirical correlations based on limited core test data from individual fields may lack robustness for various applications. In absence of core measurements, correlations from publications are often used to determine rock properties. However, these correlations may have been developed for formations very different from the field and lithotypes in question, making it difficult to a scertain their applicability. This paper introduces several statistical tools and methods for developing more robust rock property correlations. These correlations are based on core test results from approximately 200 depths from multiple unconventional shale plays. By employing these statistical methods, multivariate correlations of rock properties with wireline logs were developed. This approach facilitates the development of both global and region-specific correlations. Additionally, the resulting empirical correlations can identify outliers when the input logs fall outside of the data range used to build the model.