A fully triaxial multi-array logging tool has been successfully field tested. A forward model inversion of the fully triaxial multi-array tool response provided vertical resistivity (Rv, and horizontal resistivity (Rh in a deep water turbidite sequence. From the vertical and horizontal resistivity, a laminated sand-shale reservoir model was built in conjunction with Nuclear logs and NMR logs. This process enhanced the hydrocarbon estimation in this low resistivity pay, and reduced the uncertainty of the gas in place assessment. The analysis of the data suggests that shale anisotropy plays an important role in the laminated sand-shale analysis.
The interval in question is a deep water turbidite levee/overbank deposit, consisting of highly organized thin layers of high quality gas-bearing sand. Layer thicknesses range from almost a meter to less than a centimeter, with most of the layers in the centimeter range. The original analysis of the data (based on conventional induction tool and a dual-water shaly-sand approach lead to unreasonably high water saturations.
A laminated sand-shale model where the shales are anisotropic has been built based on Rv, Rh values from the triaxial induction tool. This petrophysical model leads to much lower water saturations. The Sw values computed with this second model agree very well with NMR-derived values using diffusion-editing techniques and with analogous core Sw measurements and capillary pressure data. This field example demonstrates the use of fully rriaxial multi-array tool data, and the importance of including shale anisotropy, for the interpretation of low resistivity pay in laminated sand-shale formations.
Subsequent core description and core analysis have shown that the interpreted section was indeed a high quality thinly laminated reservoir. Downhole formation tester data indicated the layers to be in communication, and a drill stem test confirmed the producibility and gave a minimum indication of the reservoir extent. Mini-permeameter measurements have shown a qualitative similarity between resistivity and permeability anisotropy.
It also shows the beginning of a paradigm-shift where we are not focused on detecting the thin beds by increased resolution, but evaluating them quantitatively with the measured anisotropy. Multi-component induction will become a necessary part of an evaluation program in these environments.
