Artificial Neural Network Approaches for Permeability Estimation

Permeability should be dispersed conveniently to control the aquifer's type and quality. Permeability in a variety of porous media can be determined using different methods depending on the environment and the scope of the porosity media. These days, permeability of core samples and well logging data with greater aquifer heterogeneity, artificial intelligence algorithms are well-known for estimating permeability. Machine learning and artificial intelligence have gained popularity and credibility across all scientific fields. To address the dearth of resources in geosciences generally and hydrology specifically.

As soft computing techniques, Artificial Neural Networks (ANNs) have demonstrated the capacity to estimate acceptable outputs with tolerable outcomes. The ANN model uses basic processing units, which are networks of interconnected neurons. The simplest approach is the Feed-Forward Artificial Neural Network (FF-ANN). The Middle Jurassic Hugin Formation may have been deposited as a mouth bar setting during the period of general transgression, as evidenced by fluctuating permeability values brought on by changes in the sediment supply, which varying porosity values brought on by variations in the amount of clay and size of grains.

Keywords: Artificial Neural Network, Feed-Forward Artificial Neural Network, Volve oilfield, Hugin Formation, Permeability estimation.