Bureau of Economic Geology, The University of Texas at Austin (www.beg.utexas.edu).
AAPG 2002 Meeting, Houston, Texas, March 1013, 2002
Quantitative Seismic Sedimentology to Predict
Lesli. J. Wood and Hongliu Zeng
Three-dimensional seismic data have revolutionized the understanding of reservoir character by providing spatially dense, high-resolution imaging of planform and improved discrimination of petrophysical variability. Classic seismic facies analysis can be supplemented through the study of horizontal imaging of depositional forms and the spatial relationships between reflection patterns to quantify the morphology of the system. Width-to-depth ratios, sinuosity, meander wavelength, and amplitude can be calculated on these forms and related to container fill using basic geomorphic principles. Lithology prediction can be achieved by integrating well log-calibrated reflection amplitude and facies patterns, and further improved by Neural Network-assisted lithology inversion. We present an example of this quantitative approach in the Miocene-Pliocene succession, offshore Louisiana, where a series of stratal slices were generated from key sequence stratigraphic horizons picked within the seismic data volume. Depositional morphology on stratal slices shows distribution and evolution of fluvial-deltaic systems in coastal plain-shallow marine deposits of the highstand systems tract and incised-valley and distributary channel deposits of the lowstand systems tract. Calculations show the relationship between width and fill type in incised features, as well as sinuosity differences among morphic groups. Analysis of seismic image texture and asymmetry reveals important depositional processes, such as progradation, lateral accretion and channel avulsion, and their effect on reservoir-level architecture.