237 7. Seismic inversion and characterization applied to geothermal energy Easy to put in place, these methods always output results which will, of course, highly depend on the selection of attributes used as inputs. The main challenge of such methods is to properly interpret the obtained features. Coupled with supervised techniques, this is a good methodology to prove that the supervised training is a typical response, well identified in the seismic data. Recent advances in this methodology, suitable for exploration, suggest afterward calibration with well data, or with conceptual section from the proposed by geologists. 7.4 Example: identification of lithology, good porosity and fractured areas through a seismic inversion study In the following example, a seismic inversion and characterization study has been conducted to highlight the most prospective area in carbonates, either in terms of rock properties (lithology, porosity/permeability) and fault/fracture presence over a 3D survey. For this characterization case, the fluid is stored in good matrix properties, while the permeability is ensured by the fractures. In geothermal activities, these parameters are key to ensure the targeted flow rate. The described work can be performed in 2D. This recent case (Baillet et al., 2024) has been scenarized as a geothermal project, and these methodologies have already been applied successfully in this context in Paris basin or in the North of France. A stratigraphic joint inversion, using angle-stacks, has been performed. Both P- and S- impedance are optimized. The zone is covered by 6 wells, with DT, RHOB and partial DTS completed by empirical laws when needed. A seismic characterization has been performed to predict lithology (Figure 7.11). A petro-elastic model, built at wells, upscaled at seismic scale, has been used as a training sample for discriminant analysis. Shaly carbonates could have been discriminated against dolomites with a satisfactory rate. A particularly interesting layer, in terms of porosity/permeability, called unit-C, has been identified, at the limit of the seismic resolution. A connectivity analysis, in 3D, has been undertaken, and proves its connection/extension, invisible in section only. In Figure 7.11, this layer is plotted in orange and has been used to propose an update of the horizon, in red dots. In each lithology, a law estimating the porosity has been derived from well logs and applied, based on P-impedance. The porosity is very heterogeneous vertically, as observable in another section, Figure 7.12. If the fluid is present in the most porous matrix lithologies, the provided flow rate might not be enough to sustain a geothermal project. Faults and fractures, identified, can greatly increase the prospectivity of an area. A fracture characterization, processing and mixing key attributes (as dip-steered similarity, spectral decomposition, energy,
RkJQdWJsaXNoZXIy NjA3NzQ=