Alimadadi - Islamic Azad University, North Tehran Branch, Tehran, Iran,
Behroozi - Iran University of Science and Technology, Tehran, Iran
Sadati - Mechanical Engineering, K.N.Toosi University of Technology, Tehran, Iran

Obtaining physical reservoir characteristics is extremely important and necessary to determine the correlations, productions, and field development. Reservoir characteristics include porosity, permeability, cementation, and the like which are obtained from petrophysic and petrographic analyses. From these properties porosity is the most important static property of petroleum reservoirs that can be used to perceive permeability, fluid behaviors, capillary pressure, and sedimentological interpretations. One of the goals of prediction, accomplished in this paper, is to find out the missed porosity logs to interpret a gas reservoir in the well due to available and suitable petrophysical logs gathered from near wells. In some wells, we cannot measure a number of petrophysical properties whereas wells are maybe washed out or the borehole tools are not available for old wells. Therefore, petroleum geologist should pursue some methods to transfer accessible data into faulty wells. It means that they predict missed data using information which is available in its near wells. For prediction purposes of this property, “Resistivity Logs”, “Gamma Ray Log”, and “Sonic Log” will have to be used as input information. The relationships of porosity logs versus the logs mentioned above are absolutely nonlinear. Soft computing methods are one of the powerful approaches used to identify lost data.

Density and neutron logs, Neural Network, Petrographic data, Petrophysic data, Prediction, well logging