Publications

This study focuses on BSJ gas field located in Central Sulawesi, Indonesia. The reservoir in this field is Limestone with irregular porosity and because of that, porosity is an important consideration. The objective of this study is to generate porosity using seismic data for reservoir characterization. Seismic 2D pre-stack depth migration (PSDM) was inverted to generate acoustic impedance data using the model based inversion method. Then, the acoustic impedance attribute along with other seismic attributes were analyzed using multiattribute transforms and Probabilistic Neural Networks (PNN). Cross validation is used to estimate the reliability of the derived multiattribute transforms. It was observed that there are only 2 numbers of attributes with a good combination to predict porosity. Based on the results, the cross correlation between actual porosity from well log data and predicted porosity was about 72% using multiattribute transforms and increased significantly to 92% using Probabilistic Neural Network (PNN). Therefore, PNN’s porosity section was used to map porosity distribution around the study area. This study, may serve as a guide for further field development, especially for this study area.