Electrofacies Determination Using SOM Neural Network in Bangestan Reservoir, SW Iran

The reservoir electrofacies study is one of important subjects in hydrocarbon reservoirs scope now. Determination of the high reservoir quality zones can play an important role in production view of the hydrocarbon reservoir and their development. Electrofacies is defined on the basis of clustering which is grouping all similar log data in unique set and distinguished it from other sets. In the present research at first using SOM, MRGC and DC methods, primary model of electrofacies in a number of field’s wells has been determined. Electrofacies have determined by different methods correlated with identified flow unit’s derived Core storage capacity (phie*h)-flow capacity(K*h) data, and SOM method has been chosen among them for clustering which had the highest accordance. 9 created Initial electrofacies reduced to 4 electrofacies according to the analogy of some parameter; such as, porosity and gamma logs. This electrofacies have been generalized for entire filed resulting in creation of a model with separation capability of the deferent reservoir zones. This model shows a decrease in reservoir quality from the upper part to the bottom of the reservoir also depicts reservoir quality changes whole the studied field.