ANALYSIS OF SEISMIC SPECTRAL ATTENUATION BASED ON WIGNER-VILLE DISTRIBUTION FOR SANDSTONE RESERVOIR CHARACTERIZATION - A CASE STUDY FROM THE WEST SICHUAN DEPRESSION, CHINA

摘要

A mass of previous work has shown the application of seismic spectral decomposition for low frequency shadows detection and stratigraphic mapping. In this paper, we further analyze the characteristics of spectral attenuation in fractured sandstone reservoirs saturated with different fluids by Wigner-Ville distribution (WVD) based method. We give a description on the geometric structure of cross-terms due to bilinear nature of WVD with a multicomponent signal composed of three Gaussian time-frequency atoms. Then cross-terms are eliminated using improved WVD with time and frequency independent Gaussian kernels as smoothing windows, what is called the smoothed pseudo WVD (SPWVD). At last the SPWVD is applied to spectral decomposition of seismic data from West Sichuan depression of China. We focus our study on the comparison of characteristics of spectral attenuation in order to verify whether SPWVD can differentiate different fluids content. It shows that prolific gas reservoirs show higher peak spectral amplitude at higher peak frequency, which attenuate faster than low quality gas reservoirs and dry or wet reservoirs. SPWVD is able to separate prolific gas reservoirs from low-quality gas reservoirs nnd dry or wet reservoirs. This can be regarded as a spectral attenuation signature for future exploration in the study area.