Abstract: Application of Adaptive Lattice Filtering in Sonic Well Log

Z. Zhou, T. W. Spencer

Adaptive lattice digital filtering techniques have been used in seismic data deconvolution and ground roll suppression and linear prediction. However, there has not been much application on sonic well logging. Because of their head wave nature and frequency response to compressional waves and shear waves in fluid-filled borehole, it is difficult to use conventional deconvolution to compress these types of waves. Also, due to the dispersion of pseudo-Rayleigh waves and low frequency response of Stoneley waves, it has not been a very successful method to extract compressional, shear, and Stoneley waves which are very useful to local geology and structural interpretation. In this paper we present a new application of adaptive lattice filtering on borehole acoustic waveforms. Although, there are a number of different algorithms for calculating the reflection coefficient of the lattice filter, the method used in this paper basically follows Burg's algorithm (harmonic mean) modified by Makhoul with weighting factor (attenuation). An experiment on synthetic borehole full acoustic waveforms shows that adaptive lattice filtering technique is very effective in the enhancement and compression of different arrivals. This is a potentially important application of this type of filtering because it decouples successive stages of the filtering and automatically estimates the reflection coefficient, thereby enabling the optimization of the output.

AAPG Search and Discovery Article #90955©1995 GCAGS 45th Annual Meeting and Gulf Section SEPM, Baton Rouge, Louisiana