A new multidisciplinary workflow is suggested to re-characterize the Hamra Quartzite (QH) formation using artificial neural networks. This approach involves core description, routine core analysis, special core analysis and raw logs of fourteen wells. An efficient electrofacies clustering neural network technology based on a self-organizing map is performed. The inputs in the model computation are: neutron porosity, gamma ray and bulk density logs. According to the self-organizing map results, the reservoir is composed of five electrofacies (EF1 to EF5): EF1, EF2 and EF3 with good reservoir quality, EF4 with moderate quality, and EF5 with bad quality. Hydraulic flow units are determined from well logs and core data using the flow zone indicator (FZI) approach and the multilayer perception (MLP) method. Obtained results indicate eight optimal hydraulic flow units. Hydraulic flow units for un-cored well are determined using the MLP, the used inputs to train the neural system are: neutron porosity, gamma ray, bulk density and predefined electrofacies. A dynamic rock typing is achieved using the FZI approach and combining special core data analysis to better characterize the hydraulic reservoir behavior. A best-fit relationship between water saturation and J-function is established and a good saturation match is obtained between capillary pressure and interpreted log results.
Dynamic Rock Type Characterization Using Artificial Neural Networks in Hamra Quartzites Reservoir: A Multidisciplinary Approach
