Generalized regression and feed-forward back propagation neural networks in modelling porosity from geophysical well logs

Geophysical formation evaluation plays a fundamental role in hydrocarbon exploration and production processes. It is a process which describes different reservoir parameters using well field data. Porosity is one of the parameters that determines the amount of oil present in a rock formation and research in this area is mainly carried out by engineers and geoscientists in the petroleum industry. Accurate prediction of porosity is a difficult problem. This is mostly due to the failure in the understanding of spatial porosity parameter distribution. Artificial neural networks have proved to be a powerful tool for mapping complicated and non-linear relationships in petroleum studies. In this study, we analyze and compare generalized regression neural network (GRNN) and feed-forward back propagation neural network (FFBP) in modeling porosity in Zhenjing oilfield data. This study is calibrated on four wells of Zhenjing oilfield data. One well was used to find an empirical relationship between the well logs and porosity, while the other three wells were used to test the model's predictive ability in the field, respectively. The findings proved that the GRN network can make more accurate and credible porosity parameter estimation than the commonly used FFBP network. Artificial intelligence can be exploited as a powerful instrument for predicting reservoir properties in geophysical formation evaluation and reservoir engineering in petroleum industry.