Characterization of hydraulic flow units from seismic attributes and well data based on a new fuzzy procedure using ANFIS and FCM algorithms, example from an Iranian carbonate reservoir

A robust methodology to determine geological facies in oil and gas fields is the integration of seismic attributes and well data to estimate flow zone indicator (FZI). Recently, the application of fuzzy and Neuro-Fuzzy approach regarding this purpose has enjoyed an increasing attention. The current study was carried out in the Surmeh (Arab) formation at Persian Gulf basin, Southern Iran. A Nero-fuzzy system was applied to estimate FZI cube from seismic attributes. To do so, core data and seismic data from four wells were imported to ANFIS system. Subsequently, the outcomes were compared with those of probabilistic neural network (PNN). Finally, a fuzzy C-Means clustering (FCM) technique was applied to characterize different hydraulic flow units (HFUs). The results of this study demonstrate that adaptive neuro-fuzzy inference systems (ANFIS) turn out to be successful in modeling FZI from seismic attributes and well data for a faraway well location. Moreover, the results achieved suggest that using the FCM technique is an efficient methodology to determine different HFUs from FZI cube.