This paper consists of two parts. The first part gives introduction on the Indonesian model as the most reliable model for prediction of water saturation in this reservoir type because it has the least uncertainty if the least-square error method (LSE) among six frequently used conventional water saturation models is deployed (i.e., total shale, modified total shale, Indonesian, dispersed clay, simandoux and modified simandoux). The authors innovate a new method in the second section to present an equation that could improve the Indonesian model.
This equation is presented through resistivity, gamma ray, neutron and density logs, which are the common available logs in most of the wells and formations under study. In this study, fourteen wells (in 39 different zones) from five basins of Mesaverde tight gas sandstones across western US basins were chosen in order to demonstrate the reliability of this method. The equation that is introduced in this study has shown good performance in improving the estimation of water saturation in 85 percent of the total zones (i.e., 33 zones out of 39).