Use of AVOA data to estimate fluid indicator in a vertically fractured medium

TLDR

Microstructural attributes of vertical fractures, such as crack density, aspect ratio, and fluid infill, control the elastic response of the medium, with normal fracture weakness strongly dependent on fluid content. The study aims to quantify fracture fluid content by defining a fluid indicator based on the ratio of S‑ to P‑wave velocities in the unfractured medium. Using linear‑slip theory and a Born‑formalism derivation, the authors relate PP and PS reflection coefficients to fracture weakness and develop a method to compute the fluid indicator from wide‑azimuth PP‑AVOA data. The derived formulas show that PP reflections are insensitive to fracture surface roughness while PS reflections are not, and synthetic‑data inversion demonstrates that the fluid indicator can be accurately recovered for liquid‑filled or partially saturated fractures but degrades with increasing gas saturation and is robust to moderate errors in g except when fractures are gas‑saturated.

Abstract

Microstructural attributes of cracks and fractures, such as crack density, aspect ratio, and fluid infill, determine the elastic properties of a medium containing a set of parallel, vertical fractures. Although the tangential weakness [Formula: see text] of the fractures does not vary with the fluid content, the normal weakness [Formula: see text] exhibits significant dependence on fluid infill. Based on linear-slip theory, we used the ratio [Formula: see text] — termed the fluid indicator — as a quantitative measure of the fluid content in the fractures, with g representing the square of the ratio of S- and P-wave velocity in the unfractured medium. We used a Born formalism to derive the sensitivity to fracture weakness of PP- and PS-reflection coefficients for an interface separating an unfractured medium from a vertically fractured medium. Our formulae reveal that the PP-reflection coefficient does not depend on the 2D microcorrugation/surface roughness with ridges and valleys parallel to the fracture strike, whereas the PS-reflection coefficient is sensitive to this microstructural property of the fractures. Based on this formulation, we developed a method to compute the fluid indicator from wide-azimuth PP-AVOA data. Inversion of synthetic data corrupted with 10% random noise reliably estimates the normal and tangential fracture weaknesses and hence the fluid indicator can be determined accurately when the fractures are liquid-filled or partially saturated. As the gas saturation in the fractures increases, the quality of inversion becomes poorer. Errors of 15%–20% in g do not affect the estimation of fluid indicator significantly in case of liquid infill or partial saturation. However, for gas-saturated fractures, incorrect values of g may have a significant effect on fluid-indicator estimates.