Prediction of free flowing porosity and permeability based on conventional well logging data using artificial neural networks optimized by Imperialist competitive algorithm - A case study in the South Pars gas field

Nuclear Magnetic Resonance (NMR) logging is one of the most effective tools in determining permeability and porosity of a formation layer, so it is a reliable method to characterize a reservoir. Since NMR logging is not applicable in certain circumstances, its parameters are usually correlated with conventional logging parameters. In this research, an Artificial Neural Network (ANN) with Multi Linear Perceptron (MLP) structure and a feed-forward back-propagation algorithm is employed to predict the NMR logging parameters from their conventional counterparts. The ANN is optimized using the Imperialist Competitive Algorithm (ICA). The ANN-ICA model is applied to two data sets measured in two separate gas wells of south Pars gas field. The data obtained from one well are used as the training data set and the other well data are utilized to test the proposed model. Finally, a sensitivity analysis (based on the parameters of the ICA algorithm and the number of neurons in the ANN) is applied to explore their effects on the ANN's performance. According to the results, the accuracy and efficiency of the proposed model are more desirable than the traditional neural network. It is found that the optimization method is not sensitive to the ICA parameters.