Determining the electrical facies utilizing multi-resolution graph-based and artificial neural network clustering methods in an Oilfield, SW Iran

Abstract

Asmari Formation is the most significant reservoir of the Mansouri oil field in SW Iran. This Formation is generally composed of carbonate lithologic but has sandstone layers in some zones. It is also the youngest reservoir rock in Iran. Cross-lithology diagrams determined the lithology of the Asmari Formation by applying 280 core samples from one drilled well in the studied reservoir, and a combination of lime, dolomite, anhydrite, and sandstone with shaley layers was estimated. Also, five zones were identified with different geological/reservoir conditions, with the best reservoir quality in zones 3 and 5 with sandstone/shale lithology. Furthermore, this study determines multi-resolution graph-based clustering (MRGC) and artificial neural network (ANN) facies clustering methods. For this purpose, the combined log readings of DT, RHOB, GR, Sw, NPHI, and PHIE were used. Lithology was evaluated and estimated in each sequence using corrected and edited logs and lithology cross-sections. After applying the Geolog software, an optimal model with 8 clusters (facies) with better separation of rock species was obtained. Finally, five facies with different lithological compositions and reservoir conditions were identified. The information obtained from determining electrical facies by clustering method defines the reservoir zone from non-reservoir and qualitatively (good, medium, and poor).