A more accurate estimation of fracture weaknesses constrained using fracture facies

Huaizhen Chen, Jian Han, Kristopher A. Innanen

Establishment of model constraints is crucial for seismic inversion. Well-logging data is usually used to generate initial models in the case of estimating elastic parameters (e.g. P- and S-wave velocities, impedances and moduli, etc.). However, in the case of seismic inversion for fracture indicators (e.g. weakly anisotropy parameters, fracture weaknesses), conventional well logging data cannot be directly used to build initial models of fracture indicators. To implement a better estimation of fracture weaknesses, we propose a two-stage inversion method, which is implemented as: 1) estimating azimuthal elastic impedance (AEI) and fracture facies; and 2) predicting fracture weaknesses using the estimated AEI as the input and model constraints constructed using the predicted fracture facies. In the first stage, we use Gaussian mixture prior distribution to obtain the AEI of different incidence angles and azimuths, and predict the fracture facies according to Bayesian classification; and in the second-stage, we implement a model-constrained Bayesian inversion for fracture weaknesses. We apply the proposed two-stage inversion method to synthetic seismic data, which illustrates the inversion method is robust even in the case of employing noisy seismic data of signal-to-noise ratio (S/N) of 1 as the input. Example of real data reveals that reliable results of fracture weaknesses are obtained using the proposed inversion method, which verifies that our method is a valuable tool for generating reliable fracture indicators from azimuthal seismic data.