Prediction effective-porosity of carbonate Reservoir Using Artificial Neural Network

In this paper we propose Artificial Neural Network (ANN) as a new approach for prediction effective-porosity () of carbonate formations from petrophysical log data. Advantages and disadvantages of Artificial Neural Network (ANN) have been discussed by several authors [1]. Although many interpretations for determining total porosity in carbonate formation exist, finding the effective porosity is a challenge in reservoirs with no coring sample and well test data. In this paper, 767 data sets were used from five wells of a reservoir in Iran. Depth, NPHI, PHOB and SGR were used as the input data and porosity obtained by coring was as target data. 60% of these data points were used for training and the remaining for predicting the effective porosity (validation and test).An ANN was developed and a correlation coefficient (R) of 0.901 was obtained by comparing effective porosity predictions and the actual measurements. Data sets are well log and core data of a reservoir