Improving the 3D facies model with the seismic-derived log volumes: a case study from the Asmari Formation in the Hendijan Field, southwest Iran

Abstract

Abstract The prolific Oligocene-Miocene Asmari Formation is the primary reservoir rock hosting important hydrocarbon resources in Iran and consists of fluvial/deltaic siliciclastic and shallow marine carbonate rocks in the Hendijan Field. Due to the significant facies variability of the formation, the presence and quality of the reservoir pose a significant uncertainty in the characterization of the reservoir. This study compares two facies models, one based on well logs only and the second based on estimated facies volumes as a secondary variable in facies modelling. The petrophysical evaluation with microscopic thin sections and electrofacies analysis were used to classify the facies and determine the reservoir quality. As a result, the Ghar Member was identified as a highly porous interval, while the lower part of the Asmari Formation is characterized by tight facies. A sequential Gaussian simulation (SIS) algorithm was used to build the 3D facies model on the basis of the well logs. Acoustic impedance, shear impedance and density (derived from pre-stack inversion) were used as inputs to an artificial neural network to generate acoustic and density log volumes. Using electrofacies cut-offs, facies volume was constructed and used as a secondary variable to improve the initial facies model. The final facies model was compared with the blind well to check the validity of the prediction and satisfactory results were obtained. Since the values are present in all the cells of the reservoir (the traditional facies model only has values for the well location), the estimated facies volume is an accurate variable in the prediction of the facies model for the Asmari reservoir and for this reason the secondary facies model is more reliable than the primary one.