| Summary: | Rock mechanical properties including wave velocities, Poisson's ratio, Young's, shear and bulk modulus have important applications in geomechanical analysis of petroleum reservoirs. Direct measurement of these parameters is usually not viable due to the high cost of testing or a lack of available data, particularly in old wells. Therefore, indirect methods are often used to predict these parameters from available data. The simplest, and still extremely common, method is empirical equations. However, these relationships are highly sensitive to different types of fluids or lithologies and are often not relevant for local geology. In recent years, intelligent systems have been used in different branches of sciences and technologies and have often been demonstrated to be useful in prediction and optimization problems. Herein, we attempt to predict a range of geomechanical parameters for different rock types using intelligent systems. For this purpose, wave velocities and rock mechanical properties of different lithologies were predicted from conventional petrophysical data that are available in most of petroleum wells. The results showed that the used methodologies are fast and reliable (around 95% accuracy) in estimation of geomechanical properties in our case studies and can be used in geomechanical/petrophysical modelling of coal seam gas and petroleum reservoirs.
|
|---|