ENHANCING THE SIGNAL-TO-NOISE RATIO OF SONIC LOGGING WAVEFORMS BY SUPER-VIRTUAL INTERFEROMETRIC STACKING

Sonic logs are essential tools for reliably identifying interval velocities, which, in turn, are used in many seismic processes. Borehole irregularities and washout zones along the borehole surface can cause the signal-to-noise ratio of recorded sonic waveforms to be quite low. Noisy borehole conditions can decrease the signal-to-noise ratio and mask the signal recorded by the receiver stations. To mitigate this problem, we have extended the theory of super-virtual refraction interferometric stacking to enhance the signal-to-noise ratio of sonic waveforms. This theory is composed of two redatuming steps followed by a stacking operation. The first redatuming procedure is of correlation type, where sonic waveforms are correlated with each other to obtain virtual waveforms with the sources datumed to the refractor. The second datuming step is of convolution type, where virtual sonic waveforms are convolved with the recorded waveforms to dedatum the sources back to their original positions. The stacking procedure following each step enhances the signal-to-noise ratio of the refracted P-wave first arrivals. Datuming with correlation and convolution of traces introduces spurious events known as correlation artifacts in the super-virtual dataset. To overcome this problem, we replace the correlation-type datuming step by a deconvolution-type datuming step. Although the cross-correlation-based datuming method is more robust, the deconvolution-based datuming method significantly suppresses the spurious artifacts. To limit the noise amplification effect caused by the deconvolution step, we add a non-zero regularization parameter to stabilize the deconvolution of the virtual sonic waveforms. Our tests on synthetic and real data examples show remarkable signal-to-noise ratio enhancement of refracted P-wave arrivals in the sonic waveforms. These tests further demonstrate how the use of deconvolution-type datuming instead of the conventional correlation-type datuming may significantly suppress the correlation artifacts. The semblance analysis demonstrated that the super-virtual sonic waveforms yields more robust formation velocities compared to the raw waveforms.