Publications

Typically, facies modeling is based on a geostatistical approach for controlling the geometry and the spatial distribution of the facies. The geostatistical modeling allows the stochastic simulation of the reservoir properties, providing alternative, equi-probable, solutions that can be used for capturing the complexity of the reservoir. The most common stochastic simulation techniques for facies modeling are object-based modeling, pixel-based modeling, and multi-point statistics (MPS) modeling. This poster presentation compares the methodology of these three methods applied to a turbidite system. As a guidance for facies modeling, a conceptual model of the submarine fan deposits is created based on the interpretation of the facies log for 6 wells. From this model, azimuth maps for guiding the facies and depositional regions are derived which are input to all three modeling techniques. Object-based modeling makes it possible to create submarine fan geometries by combining different geometrical facies bodies. The pixel-based technique uses the Gaussian indicator simulation (GIS) method to produce a spatial distribution of the facies by employing variograms. The newest technique is MPS which uses training images as the main driver for the facies modeling. By comparing the results, the advantage and disadvantage of these three methods are discussed. Based on streamline simulation the connectivity of the facies bodies and their influence on the flow behavior is analyzed.