Semi-quantitative effective stress prediction using seismic amplitude variation with offset inversion

Abstract

The accurate prediction for in-situ stress or pressure is highly relevant to many practical applications in hydrocarbon exploration and development such as safe production, geofluid discrimination and tectonic characterisation. The stress prediction using seismic data is currently based on the empirical (or fitting) relationship between effective stress and several estimated elastic attributes of formations, such as P-wave velocity and bulk modulus. The empirical approaches dominantly rely on the well-logging data constraints and cannot reveal the theoretical relationship between effective stress and seismic signatures. Thus, they may generate unexpected misleading interpretations in the frontier areas without sufficient prior well-logging data and in some fields with strong lateral formation discontinuity. To address this problem theoretically, we derive a PP-wave reflection coefficient equation in terms of the defined decoupled bulk and shear moduli, fluid factor, density and effective stress based on the elastic piezosensitivity model, Mavko-Jizba modified moduli model and Gassmann fluid substitution model. The new reflection coefficient equation is consistent with the exact Zoeppritz equation within the moderate incident angles. Further, a seismic amplitude-variation-with-offset (AVO) inversion method incorporating our reflection coefficient equation is developed to predict effective stress using an iterative least-squares algorithm. Two synthetic tests and a field application confirm the feasibility and robustness of the proposed method for stress prediction.