Spectral Attributes for Reservoir Parameter Detection

Reservoir parameters e.g. lithology, thickness and porosity can be detected and classified via multiple attributes. These derived attributes are either time based or frequency based. The advantage of any spectral decomposition technique is that it adds an extra dimension, the frequency, to study the sensitivity of bright or dim amplitude anomalies to reservoir parameters. Such anomalies are often hidden among many frequencies in a broad seismic spectrum. We propose to derive attributes from a high resolution spectral decomposition technique called Continuous Amplitude and Phase Spectral (CAPS) Transform. Many spectral attributes can be derived from the time-frequency volumes. We have focused on two significant parameters namely: amplitude gradient and amplitude variance. The performances of the spectral parameters are tested on two public seismic fields. Results indicate that spectral parameters are effective in reservoir description.