Estimation of effective porosity in the space between wells using classification of seismic attributes by Adaboost algorithm

In this paper, by combination of seismic data and well log data, the effective porosity in the space between wells is estimated. One of the important petroleum reservoir features is effective porosity that engineers are always looking to find an appropriate model for distribution of this parameter in the reservoir. The petrophysical properties of petroleum reservoirs are very complex. In the last few decades, the effective porosity estimation procedures have become one of the hot topics in the industry to evaluate these procedures or methods. In the current research, by integration of petrophysical, seismic data and seismic attributes classification using Adaboost algorithm, it is tried to estimate the effective porosity in a twodimensional seismic cross section of the block F3 Dutch sector of the North Sea. In the first step, seismic attributes of two-dimensional seismic section has been extracted. Based on the feature selection methods, six seismic attributes have positive impact on the process of the classification. The six attributes of seismic data and effective porosity have been classified and labeled in the positions of wells. Adaboost algorithm has been used for training, and training has been led to the accuracy of 76.6 percent. The next step is the accuracy of Adaboost algorithm or validation, which has been led to accuracy of 71.7 percent. Adaboost algorithm can also be used for classification of effective porosity in other parts of the reservoir where data are not available. The algorithm outputs have shown a good effective porosity in the lower layers, where there may be an economic horizon for the oil production.