--> --> Seismic Reservoir Characterization with Geostatistical Approach to Delineate Sand Bodies in a Complex Fluvial/Lacustrine Rift Basin (Melut Bain), Sudan

International Conference & Exhibition

Datapages, Inc.Print this page

Seismic Reservoir Characterization with Geostatistical Approach to Delineate Sand Bodies in a Complex Fluvial/Lacustrine Rift Basin (Melut Bain), Sudan

Abstract

Fluvial/lacustrine rift basins are among the most complex and heterogeneous sedimentary systems with potential reservoir units below seismic resolution. However, they are of great interest forming potential hydrocarbon plays. Complexity of the depositional staked pattern, structural framework and barriers are poorly understood which necessities further detailed reservoir characterization efforts. Melut basin is one of the well-known interior rift basins in Sudan. It is regionally linked to the Mesozoic-Cenozoic rift system in central and western Africa. The Eocene Yabus Formation is the main oil producing reservoir in the basin. It is dominated by channel sandstone and shale that deposited in fluvial/lacustrine environment during the third phase of rifting which demonstrates significant degree of heterogeneities in different scales. In this paper, two acoustic impedance inversion approaches (deterministic and stochastic) for post-stack seismic data have been applied to enhance seismic resolution and delineate detailed stratigraphic units. Geostatistical analysis/integration of the higher (vertical) resolution obtained from well logs was successfully utilized to constrain the higher (lateral) resolution from seismic data. Therefore, not only the acoustic properties are used but also the detailed layering characteristics. The optimized workflow begins with seismic data conditioning, petrophysical evaluation and rock physics analysis, then well to seismic tie and wavelet extraction, followed by low frequency analysis and deterministic AI inversion. For stochastic inversion, a structural (3D gridding) model was constructed based on 10 subzones identified within Yabus Formation, and using geostatistical analysis for each sub-layer including vertical and horizontal semi-variogram. A total of 50 multi-realizations of AI and porosity were generated with high resolution to predict sand bodies and reveal lateral extends. These realizations were also used for uncertainty analysis by producing several combinations of porosity and probability cut-off values to provide better understanding of the reservoir quality. The results validated with available well data and showed good match.