Integration of 3D Seismic Attributes and Well Logs for Electrofacies Mapping and Prediction of Reliable Petrophysical Properties

Abstract

Abstract Accurate distribution of geological and petrophysical properties such as facies, porosity and water saturation in carbonate reservoir is an essential part of building robust static and dynamic models for proper reservoir management and making reliable decisions. An integration based approach is applied for the prediction of the essential reservoir properties using well logs and 3D seismic data. The method is based on the derivation of electrofacies classes using Artificial Neural Networks (ANN) method and Geostatisticsl Conditional Simulation (GCS) Approaches. The electrofacies of the carbonate reservoir of late cretaceous understudy, which is located in the south west of Iran, were classified using the unsupervised ANN method. Based on the verification of cross plots of input logs (GR, DT, RHOB, NPHI, and LLD) versus electrofacies classes, seven rock types including one class as shale rock and others sex classes as argillaceous, dense, poor, moderate, good and best limestone rocks were identified. The unsupervised approach has provided unbiased classes of electrofacies that cover the vertical variations of the well logs very well and also provided a very fine correlation for the entire reservoir. Based on the 3D seismic data, nine seismic volume attributes were calculated. The most correlative attributes for prediction each reservoir properties were selected based on the correlation coefficient investigation. At this stage, the ANN was used to create 3D model of electrofacies, effective porosity and water saturation based on the associated 3D seismic derived attributes. These seismic derived reservoir properties were used as secondary variable in collocated cokriging equation through the GCS. The 3D simulation of electrofacies represented similar seismic attribute responses, which was interpreted geologically, resulted in good consistency with sedimentary directions and conditions of the understudy area. In conclusion, the seismic based 3D simulations of electrofacies, effective porosity and water saturation generated using ANN's and GCS have provided geologically more meaningful information about the lateral facies variations and reliable properties distributions in this carbonate reservoir.