One of the goals of seismic prospecting has been to determine petrophysical properties of the reservoir (such as lithology, porosity and pore fluid type) using remote measurements, yet little has been done to analyze the ambiguity and sensitivity of the seismic measurements to the petrophysical properties of interest. A likely reason for this is that AVO attributes, commonly used to reduce the risk in qualitatively determining petrophysical properties, cannot be easily related to physical properties of rocks given that the attributes’ amplitudes give information about changes across interfaces with no significant information about the intervals above and below these interfaces. Furthermore, common practice has been to estimate two term AVO which results in two attributes related to changes of three physical properties of rocks (P- and S-wave velocity and density) across interfaces. Unambiguous estimation of the three properties (Vp, Vs and r) or their reflectivities is not possible with only two attributes.

P- and S-wave velocities and density determine reflection amplitude as a function of offset, and their estimation (or attributes related to them) from seismic data is important given that reservoir properties in clastic reservoirs are related to these rock properties through effective media relations.The reconstruction of P- and S-wave velocity and density logs for different reservoir conditions through the effective media relationships allows for the ambiguity and sensitivity analysis of rock properties to different reservoir conditions. The same rock properties used to analyze sensitivity and ambiguity through well log reconstruction can be obtained from seismic data by post-stack inversion of AVO attributes. The resultant seismic data is a measure of rock properties of subsurface formations (not changes across interfaces) and can be related directly to well log data.

In this presentation examples of ambiguity and sensitivity analyses of rock properties are presented at both well log and seismic resolution and for the case of two and three term AVO analyses followed by post-stack inversion.

End_of_Record - Last_Page 17---------------