Abstract: 3D Facies Modeling Using Seismic Inversion Data as Soft Constraint

BASHORE, WILLIAM M.

Acoustic impedance data derived from seismic trace inversion have been shown to contribute substantially to modeling of interwell reservoir property distributions by using geostatistical integration techniques. A method is presented which extends this integration beyond continuous properties (e.g., porosity, sand-shale ratio) to include categorical variables such as facies.

The conversion of acoustic impedance to a facies classification is rarely a direct one-to-one mapping. Typically, overlaps exist among the facies-dependent distributions of impedance. This uncertainty is predicting facies from impedance should be included in the model building process. Relative proportion curves for each facies can be generated from the facies logs and the seismic-based impedance at the well-seismic tie points. The impedance traces can then be converted into proportion traces for each facies.

Categorical indicator kriging provides the basis on which to integrate the "hard" facies logs with the "soft" facies proportion data. Several techniques are available to conditionally simulate facies models: probability-fields, sequential, and truncated Gaussian. The soft data are included using the proportions as local means or using indicator cokriging. Porosity and permeability models may then be generated from the resulting facies models where the facies-dependent univariate and bivariate relationships are honored. The incorporation of facies dependencies are always preferred.

Examples of facies modeling using seismic-based impedance are given. Comparisons between facies-based porosity and permeability models with models built by directly integrating porosity and impedance illustrate the added-value of 3D facies modeling with impedance as a soft constraint.

AAPG Search and Discovery Article #90942©1997 AAPG International Conference and Exhibition, Vienna, Austria