AAPG ANNUAL CONFERENCE AND EXHIBITION
Making the Next Giant Leap in Geosciences
April 10-13, 2011, Houston, Texas, USA
Identifying Lithology Variation and Gas Occurrence in Deep Carbonate Reservoirs Using S Transform
(1) Geology and Geophysics Department, Texas A&M University, College Station, TX.
Deeply buried gas reservoirs in Sichuan Basin provide an important future energy resource for China. However, at depths of about 7km, reservoir rocks are highly pressured and could have porosity as low as 3-5%. One of the greatest difficulties encountered is thus to identify accurately commercially viable targets before drilling. Standard 3D seismic methods for direct hydrocarbon indication have often proven problematic, if not impossible, for detecting deep gas targets at such great depths. We attempt a recently developed spectral decomposition method via S-transform to identify the responses of gas-saturated zones in time-frequency domain from those due to lithology variations.
Different subsurface structures respond differently to each frequency component of the incident wave. Amplitude and phase spectra of seismic pulse change during propagation due to attenuation, dispersion and scattering. Combing the advantages of short-time Fourier analysis and wavelet analysis, S-transform can preserve the phase information in the decomposition and present time-dependent isofrequency profiles. Spectral decomposition using S transform thus allows a continuous time-varying analysis of the effects of lithology and fluid changes on frequency content. Our results show that superimposed on the overall trend of frequency decay with travel time, low-frequency anomalies of about 10-17 Hz usually associated with gas-saturated zones or lithology tuning whereas high-frequency variations are due to lithology change only, confirmed by log analysis. In normalized time-frequency domain, sharp transitions usually occur between the high-frequency anomaly (seal) and low-frequency anomaly (gas-saturated zone). Further separation of gas-saturated zones from lithology tuning needs integrated interpretation of both time-frequency analysis and seismic waveform data in time domain. The apparent frequency of lithology-tuning layers is usually low on seismic section whereas that of a gas zone is not. The former is due to the tuning of structural beddings as appeared on seismic section whereas the low-frequency anomaly of the latter is rather due to intrinsic attenuation and resonance caused by the presence of gas. These findings enable us to improve rock-physics-based quantitative seismic interpretation for deep gas exploration.