Application of artificial neural networks in lithofacies interpretation used for 3D geological modelling

This paper represents a study using Artificial Neural Networks (ANN) to perform automatic interpretation of lithofacies in a reservoir scale. This technique having been used successfully to interpret lithofacies automatically in the Sha20 Block, Shanian oilfield. Description and interpretation from a cored section in the key well was used to train the Supervised neural network. Having trained the network, it was then used to recognise and interpret the units vertically and laterally in the studied reservoir. The unsupervised neural network was run to classify the cored interval into 2 and 6 classes respectively and the results were then compared with the supervised network output. The results were observed to be over 87% accurate. Then a 3D geological model was built using the sequential indicator simulation method, the excellent results obtained from the developed model shows that the method is quite effective and gets satisfying prediction precision for the lithofacies in reservoir modeling.