Estimation of Shear Wave Velocity Using Empirical, MLR, and GEP Techniques-Case Study: Kharg Island Offshore Oilfield

Determination of the shear wave velocity, Vs, is an integral part in creation of reservoir geomechanical models. This parameter together with the compressional wave velocity and rock density are utilized to calculate the dynamic elastic moduli of the subsurface formations. In well logging, the Vs can be directly measured through the Dipole Shear Sonic Imager (DSI) logs which need special requirements and technical considerations. Therefore, many researchers have strived to develop cost-effective accurate methods for Vs estimation in the oil/gas fields. The Kharg Island offshore oilfields, located in the Persian Gulf, consist of a giant limestone reservoir called Asmari formation. In the past, numerous studies were conducted to develop mathematical relations for Vs prediction in the Asmari reservoir; however those relations were not capable of estimating the Vs values correctly. In this research, the well logging data related to a vertical offshore well was utilized to develop three mathematical relations for Vs estimation in the Asmari formation. To do this, linear regression (LR), Multivariate Regression (MLR), and Gene Expression Programing (GEP) methods were applied. Moreover, the accuracy of those relations was compared with some available empirical correlations for Vs prediction in limestone rocks. Comparing the results of those data-driven equations with the empirical equations illustrated that the results of the GEP method are more accurate than other equations. Moreover, the Pickett empirical correlation was found to be more suitable than other empirical correlations for Vs estimation in the Asmari reservoir. The methodology applied in this research is a reliable procedure to estimate the Vs in the study area as well as other geologically similar oil reservoirs. Such an application leads to generation of robust geomechanical models increasing the project success and oilfield development progression.