Numerical simulation and interpretation of sonic measurements in vertical and highly deviated wells

Borehole sonic measurements are widely used to estimate formation elastic properties and to construct synthetic seismograms. However, presence of noise compromises the accuracy of sonic logs. Sonic logs are prone to errors originating from near wellbore damage or mud-filtrate invasion. Moreover, sonic logs are calculated from the numerical processing of waveforms over a wide range of receivers. Numerical processing induces errors in the sonic slowness because the slowness value is averaged over the length of the receiver array. I apply a fast modeling method using spatial sensitivity functions to calculate sonic logs. First, I define the spatial sensitivity function for the compressional and flexural modes. Then, I apply the fast modeling in a joint inversion of shear and compressional slowness logs to mitigate noise contaminating sonic logs. Joint inversion is performed in vertical and slightly-dipping wells, to estimate layer-by-layer formation elastic and mechanical properties for isotropic and anisotropic formations. Finally, I introduce a fast modeling procedure for compressional and flexural modes in deviated and horizontal wells. Results of the fast modeling are compared to finite-difference numerical simulations. The fast modeling of sonic borehole measurements in deviated wells can be applied in a joint inversion to estimate formation elastic and geometrical properties.Petroleum and Geosystems Engineerin