| Popis: |
Prior to this work, several amplitude anomalies (bright spots) related to the Upper Cretaceous Winters sands have been identified and drilled based on 2D interpretation. After 3D seismic was shot, other techniques, including AVO analysis, time slices of coherency, and fluid factor volume were applied to further enhance reservoir characterization (Lanning & Cambois, 1998), but the direct analysis of the variation of seismic trace shape had not been considered. This paper illustrates a new approach to the development and reservoir characterization of the Winters sands. This original methodology is based on seismic facies analysis (using Neural Network Technology applied to seismic trace shape) combined with litho-seismic modeling from well curves. Seismic facies maps were used to characterize regional depositional environment and analyze gas sands associated with a channel system. Seismic responses (“trace shapes”) at well C (gas producer) and D (dry hole) were then integrated into litho-seismic models for calibration purpose. Modifying the sand thickness, reservoir porosity and fluid content, several seismic responses were generated and compared with model traces generated by the Neural Network. This methodology quickly and accurately characterized gas sand reservoirs and built an initial lithoseismic model of the Winters sands that agrees with current data. This technique can be directly used in future field development (without any pre-processing of seismic data or use of wells), or in conjunction with other techniques (such as seismic attribute analysis or seismic inversion). |
