Optimism drives our business. From Spindletop to today, it only takes one spectacular well to start a boom, but possibly dozens of dry holes or, in today’s shale plays, sub-economic wells to turn that optimism into realism.

CGGVeritas has developed an integrated workflow, which includes high-quality 3-D seismic survey design and acquisition through reservoir characterization, to derive detailed geomechanical and lithological models that can guide more targeted and effective drilling and completion programs.

The company has published results for the Haynesville shale using the Tri-Parish CGGVeritas multiclient data (SEG 2011). The Hampson-Russell multi-attribute analysis of various reservoir quality indicators was used to predict actual gas production at the wells covered by the study, and excellent correlation was achieved. With this strong link established between seismic attributes and production measurements, several areas of high potential (sweet spots) for future development were identified.

cggveritas_prod

The first six months of production predicted from seismic and correlated to actual values are shown. Note potentially high production areas (circled) that have not been drilled. (Image courtesy of CGGVeritas)


The first step in identifying potentially productive zones is to understand the lithological and geomechanical properties that control reservoir quality. In the Haynesville shale, prestack simultaneous inversion was used to extract acoustic and shear impedances to predict properties such as density, porosity, brittleness, mineral composition, and total organic content, which were calibrated against existing well information. This showed that rock properties of the lithology above and below the Haynesville are very different from those within it (even though post-stack seismic reflection amplitudes may look similar). Having identified the potential sweet spots, the effectiveness of hydraulic stimulation was assessed by using stress estimates derived from anisotropic analysis of the wide-azimuth seismic data calibrated against available core measurements.

The strength of this technique is the number of different attributes used to produce the models. No single attribute produces a reliable correlation with production, but by applying multi-attribute analysis to a wide range of lithological and geomechanical properties, areas of potentially high production can be predicted. Partners in this study have commissioned further proprietary studies over their licenses.

This workflow is as applicable to liquids as to gas and is being extended to shale oil reservoirs. The models can be used to predict reservoir drainage geometry and induced fracture behavior so that the most productive well locations and fracture stages can be determined, reducing exploration risk. They also can mitigate the risks associated with drilling hazards and hydraulic stimulation by locating ductile areas, which form boundaries to the fracture zones, and identifying faults that might allow leakage of gas or proppant out of the desired zone. This type of study will enable optimal drilling, completion, and recompletion of the Haynesville to power US energy for decades.

Comments