In this contribution, we propose an alternative way to calibrate and estimate S-wave velocities using a regression analysis methodology. Prediction of S-wave velocities is critical in locations where sonic logs only have P-wave velocities available or for some reason only in discrete intervals, generally in slow formations. The proposed method applies two robust quadratic models for estimating S-wave velocities, assuming V p , clay volume and, fractional effective porosity as parameters affecting V s . Considering S-wave velocities varying according to quadratic models, the estimation process confirms previous results showing V p as the most important dependence parameter in the regressions. The preliminary results corroborate the idea that the use of the proposed quadratic model by regression analysis can represent a suitable procedure for predicting shear wave velocities in slow siliciclastic rocks.

Viscoelastic waves are modeled using a 3D time-domain finite-difference scheme with three classical rheological models -Maxwell (M), Kelvin-Voigt (KV), Standard Linear Solid (SLS) -each of them with a single relaxation mechanism. Attenuation and dispersion effects are introduced in terms of the quality factor Q for a certain angular frequency ω. Initially, seismic attenuation of the 2D poststack seismic data from Brazilian pre-salt is evaluated through a frequency analysis. After that, numerical results are validated with analytical solutions based on a homogeneous model with constant seismic properties. The behavior of each rheological model are discussed based on phase, wave amplitude and spectrum. A simplified multilayered geological model is proposed considering regions of the poststack seismic data from the Brazilian pre-salt, Búzios field. Synthetic data are then generated for each rheological model considering two approaches in terms of Q distribution: a constant Q and a varying Q per layer. Both waveform and spectra are extracted from seismograms and compared with selected regions of Búzios poststack data. The amplitude spectra confirmed that the shallowest post-salt package is more dissipative than deeper regions.Viscoelastic Models Applied to the Brazilian Pre-salt Moreover, based on the analysis of synthetic data, KV and SLS models proved to be effective to mimic seismic wave propagation at Búzios field. Synthetic results also show that a Q distribution per layer can yield a seismic attenuation behavior similar to the one observed within the data.

Viscoelastic waves are modeled using a 3D time-domain finite-difference scheme with three classical rheological models – Maxwell (M), Kelvin-Voigt (KV), Standard Linear Solid (SLS) – each of them with a single relaxation mechanism. Attenuation and dispersion effects are introduced in terms of the quality factor \textit{Q} for a certain angular frequency $\omega$. Initially, seismic attenuation of the 2D poststack seismic data from Brazilian pre-salt is evaluated through a frequency analysis. After that, numerical results are validated with analytical solutions based on a homogeneous model with constant seismic properties. The behavior of each rheological model are discussed based on phase, wave amplitude and spectrum. A simplified multilayered geological model is proposed considering regions of the poststack seismic data from the Brazilian pre-salt, Búzios field. Synthetic data are then generated for each rheological model considering two approaches in terms of \textit{Q} distribution: a constant \textit{Q} and a varying \textit{Q} per layer. Both waveform and spectra are extracted from seismograms and compared with selected regions of Búzios poststack data. The amplitude spectra confirmed that the shallowest post-salt package is more dissipative than deeper regions. Moreover, based on the analysis of synthetic data, KV and SLS models proved to be effective to mimic seismic wave propagation at Búzios field. Synthetic results also show that a \textit{Q} distribution per layer can yield a seismic attenuation behavior similar to the one observed within the data.

No abstract

No abstract