3D Modeling of Electrofacies from Seismic and Well-log Data Using a Geostatistical Algorithm

Abstract

Facies analysis represents a major part of reservoir characterization studies. The present study examines reservoir electrofacies (EFs) based on seismic and well-log data using several intelligent clustering methods. Results of the clustering were then evaluated using geostatistical algorithms for static modeling of the reservoir facies. The facies were classified based on logs by the multi-resolution graph-based clustering (MRGC) and self-organizing map (SOM) methods to obtain the porosity, as a petrophysical parameter, for 3D modeling. By means of a geostatistical approach to facies modeling via the sequential indicator simulation (SIS), well and seismic data were combined to come up with an accurate 3D model, which was then generalized to the whole reservoir. Application of the SOM and MRGC methods led to identification of 7 facies. In both methods, facies 1 exhibited the best reservoir properties. Upon the modeling, a 3D facies model was established for the depth interval ranging from the Frontier (second well creek) to the Crow Mountain horizons in the Teapot Dome. This model is of help in well planning and nomination of new well locations for drilling.