Traditionally, fluid substitutions are often conducted on log data for calculating reservoir elastic properties with different pore fluids. Their corresponding seismic responses are computed by seismic forward modeling for direct gas reservoir identification. The workflow provides us with the information about reservoir and seismic but just at the well. For real reservoirs, the reservoir parameters such as porosity, clay content, and thickness vary with location. So the information from traditional fluid substitution just at the well is limited. By assuming a rock physics model linking the elastic properties to porosity and mineralogy, we conducted seismic forward modeling and AVO attributes computation on a three-layer earth model with varying porosity, clay content, and formation thickness. Then we analyzed the relations between AVO attributes at wet reservoirs and those at the same but gas reservoirs. We arrived at their linear relations within the assumption framework used in the forward modeling. Their linear relations make it possible to directly conduct fluid substitution on seismic AVO attributes. Finally, we applied these linear relations for fluid substitution on seismic data and identified gas reservoirs by the cross-plot between the AVO attributes from seismic data and those from seismic data after direct fluid substitution.
