A case study of azimuthal AVO analysis with anisotropic spreading correction

Seismic methods play an important role in the development of tight frac-
tured reservoirs with substantial spatial variability of fracture density. Here,
we apply a comprehensive processing sequence designed for layered azimuthally
anisotropic media to wide-azimuth P-wave data acquired over a fractured gas
sand formation in the Rulison field, Colorado. The main processing steps
include nonhyperbolic moveout inversion based on an orthorhombic veloc-
ity model, estimation of effective and interval normal-moveout (NMO) el-
lipses, anisotropic geometrical-spreading correction, and azimuthal amplitude-
variation-with-offset (AVO) analysis.
The azimuthal AVO response, obtained by computing the azimuthally
varying AVO gradient on common-midpoint supergathers, proved to be the most
sensitive fracture-detection attribute. In particular, we have identified two areas
of extremely high AVO ellipticity at the bottom of the reservoir. One of these
AVO anomalies becomes much more pronounced and spatially coherent after
application of the moveout-based anisotropic geometrical-spreading correction
(MASC). In contrast, the interval NMO ellipticity in the reservoir is relatively
weak, most likely because of the influence of vertical and lateral heterogeneity.
The two major AVO-gradient anomalies at the bottom of the reservoir
coincide with the intersections of two wrenching fault systems, which suggests
that the azimuthal AVO analysis helped to detect “soft spots” of high fracture
density. Also, the dominant fracture direction (N70W) estimated from the AVO
ellipses is aligned with one of the fault systems and is in good agreement with
EMI logs.