The purpose of the present study is focused at investigating the application of a multilayer perception neural - network to estimate permeability in non-cored intervals using this technique which is performed on the reservoir of Trias Argileux Grèseux Inférieur in Sif Fatima oil field. To ensure robustness of models, the following well logging data has been used for their developing: density, sonic, resistivity, porosity, gamma ray, porosity neutron, water saturation. The performance of models was evaluated using Root Mean Square Error, Mean Absolute Error and correlation coefficient. The application of model allows recovering permeability in 151.63 m of six (6) non-cored wells. Missing permeability have been restored and validated with a good performance. This approach has a significant economic benefit allowing the gain of the time, the money and the energy compared to laboratory determination.



 Télécharger l'article : Development of Artificial Neural Network models for predicting permeability: case study of Sif Fatima oil field, Berkine basin (Southern of Algeria)