Delineation of Reservoir Compartments in Fluvial Sand Systems by using Spectral Decomposition and Seismic Attributes: Case Study from the Gulf of Thailand

Ahmad, Mirza N.; Sriburee, Suchada; Rowell, Philip

Stacked fluvial sand systems can be complex and generally contain a series of sandy point-bar deposits interspersed with mud-filled abandonment channels. The implications for reservoir characterization of these depositional systems due to this complex geometry is that the point-bar sand accumulations which form the main reservoir can be highly compartmentalized as a result of cross-cutting mud-filled channels. Therefore, delineation of point bar and channel geometries has important consequences for hydrocarbon exploration and development. The objective of this study is to develop a geophysical workflow to delineate these depositional geometries and define different reservoir compartments This study attempts to define these compartments by applying various imaging techniques such as RMS attributes, coherency cube data and discrete frequency amplitude volumes using the Continuous Wavelet Transform (CWT) technique on seismic data within the Pattani Basin of the Gulf of Thailand. Spectral decomposition outputs through CWT reveal that thick sands (>30m) can be detected by amplitude spectra at 20~25 Hz, while relatively thin sands can be detected by higher frequencies. Bright amplitudes were observed at 20 Hz for gas sands. At shallow stratigraphic levels (down to 1800m) the combination of RMS and coherency volumes successfully identified sand bodies and mud-filled channels associated with meander belts. On the other hand at deeper stratigraphic levels amplitude spectra of CWT at frequencies 20~25Hz along with coherency cube data better image the sands and mud-filled channels. These mud-filled channels may act as barriers and could potentially compartmentalize the reservoirs. This effect can be seen in the changes in reservoir pressure data in different wells across the mud-filled channels. Some sands are distributed within small area cross-cut by low amplitude mud-filled channels. In this case closely spaced wells would be required to effectively drain the hydrocarbons from these sands. Sand distribution model was prepared within the zone of interest and the mapped sands based on seismic data are in match with GR log at the drilled well locations. This suggested workflow of sand prediction might help to reduce exploration risk and to identify the optimum well spacing for field development.

AAPG Search and Discovery Article #90163©2013AAPG 2013 Annual Convention and Exhibition, Pittsburgh, Pennsylvania, May 19-22, 2013