Determination and Estimation of Hydraulic Flow Units from Core and Well Log Data by Neuro-Fuzzy Approach

Porosity-permeability relationships in the framework of hydraulic flow units can be used to characterize heterogeneous reservoir rocks. AHydraulic Flow Unit (HFU) with identical hydraulicproperties shows Flow Zone Indicator (FZI) with close values. In the present study, we attempt to makea quantitative correlation between flow units and well log responses using Neuro-fuzzy method in the Shaly sandstone Balakhani Formation at the Shahdenizoilfield, South Caspian. First, HFUs are determined to improve the prediction of flow units ininterval/wells. These measures are based on porosityand permeability of cores. A popular Neuro-fuzzy method, i.e., Adaptive Network-based Fuzzy Inference Systems (ANFIS), that is based oncombination of adaptive neural networks and fuzzy inference systems (FIS) is then used to offer a powerful tool for improving the flow units prediction.The results of this study demonstrate that there is a significant agreement between the core-derived and ANFIS-based logs derived flow units. The conductedexperimental results show that ANFIS method is efficient for modeling the flow units from well logs at well locations of which no core data was available