Seismic Inversion for Azimuthally Anisotropic Models of Fractured Formations

This invited overview paper discusses recent advances in seismic data processing and inversion for azimuthally anisotropic media and application of kinematic and amplitude signatures of reflected waves in fracture characterization. Because of the complexity of fractured reservoir models, it is beneficial to acquire wide-azimuth, multicomponent data and combine the results of moveout and AVO (amplitude-variation-with-offset) analysis. Still, quantitative fracture characterization faces significant challenges, especially for reservoirs with multiple fracture networks. The transition from the effective parameters that control seismic signatures to fracture orientations and compliances requires using additional information (e.g., well logs, core measurements, and geologic data). The potential and pitfalls of using long-spread, wide-azimuth P-wave data in fracture detection is illustrated on a case study from Rulison field in Colorado. A comprehensive processing sequence designed for layered azimuthally anisotropic media helped to identify pronounced azimuthal AVO anomalies at the top and bottom of the tight gas sand reservoir. Geologic and well-log information indicates that these anomalies correspond to areas of intense fracturing. Estimation of the fracture parameters, however, is complicated by a poor correlation between the AVO and NMO (normal-moveout) ellipses.