A Fuzzy Logic Model for Predicting Dipole Shear Sonic Imager Parameters From Conventional Well Logs

摘要

Sonic wave velocities, including compressional, shear, and Stoneley, are functional and practical parameters for various branches of reservoir characterization. In the last few years, several tools such as the dipole shear sonic imager (DSI) were introduced for measuring sonic wave velocities. Most of these instruments are expensive and drilling companies do not run them in all wells of a field because of economical restrictions. In this study, an accurate, intelligent, and indirect method was presented for prediction of sonic wave velocities, which are directly obtained from dipole shear sonic log, utilizing conventional well log data and fuzzy logic technique. Furthermore, the proposed methodology was applied on a carbonate reservoir of Iran. The results of the case study demonstrated the capabilities of fuzzy logic for estimation of sonic wave velocities, where DSI may not be run. The MSEs of the predicted V_p, V_s, and V_(st) in the test data calculated 1.47 (US/F), 4.96 (US/F), and 2.219 (US/F), respectively, which correspond to the R2 values of 91.8%, 89.3%, and 90.4%, respectively.