Artificial Neural Networks as a Valuable Tool for Well Log Interpretation

Abstract Artificial neural networks (ANNs) are rapidly gaining popularity in the area of oil exploration. This article discusses the importance of ANNs to petroleum engineers and geoscientists and its advantages over other conventional methods of computing. ANNs can assist geoscientists in solving some fundamental problems such as formation, permeability prediction, and well data interpretation from geophysical well log responses with a greater degree of confidence comparable to actual well test interpretation. The main goal of the present article is to use the artificial neural network from a petroleum geoscientist's point of view and encourage geoscientists and researchers to consider it as a valuable alternative tool in the petroleum industry. A three-layer feed-forward back-propagation network has been used to predict neutron log (NPHI) and density log (RHOB) values using gamma ray (CGR), resistivity log (IDPH), and sonic log (DTCO) input parameters. The results are also compared by analysis performed by multivariate regression analysis (MVRA).