Wave-induced fluid flow is the main cause of seismic attenuation and dispersion. So the estimated velocity dispersion information can be used to identify reservoir fluid and effectively reduce the risk of reservoir drilling. Using equivalence of dispersion and attenuation between poroelastic and viscoelastic media, we developed the method of FAVO (frequency-dependent amplitude variation with offset) dispersion quantitative estimation based on the analytical solution of 1D viscoelastic wave equation. Compared with the current single-interface velocity dispersion estimation method, the new nonlinear approach uses the analytical solution of 1D viscoelastic wave equation as the forward modeling engine. This method can conveniently handle the attenuation and generate the full-wave field response of a layered medium. First, the compound matrix method (CMM) was applied to rapidly obtain the analytical solution by vectorization. Further, we analyzed the seismic response characteristics through the model data to clarify the effectiveness of the forward modeling method. Then, the more reliable P-wave velocity, S-wave velocity, and density were recovered based on prestack viscoelastic waveform inversion (PVWI). Combining with the inversion results, the derivative matrix was calculated to perform nonlinear velocity dispersion estimation. Finally, the new estimation method was tested with the model and actual data. The experiments show that the developed method is clearly superior to the single-interface dispersion estimation method in accuracy and resolution. This approach can be used as a new choice reservoir fluid identification.

中文翻译：

---

基于粘弹性波方程解析解的非线性FAVO色散量化

波浪引起的流体流动是地震衰减和扩散的主要原因。因此估计的速度弥散信息可用于识别储层流体，有效降低储层钻井风险。利用多孔弹性介质和粘弹性介质之间色散和衰减的等价性，我们开发了基于一维粘弹性波动方程解析解的FAVO（频率相关振幅随偏移量变化）色散定量估计方法。与目前的单界面速度频散估计方法相比，新的非线性方法使用一维粘弹性波动方程的解析解作为正演建模引擎。这种方法可以方便地处理衰减并生成层状介质的全波场响应。第一的，应用复合矩阵法(CMM)通过矢量化快速得到解析解。此外，我们通过模型数据分析了地震响应特征，以阐明正演建模方法的有效性。然后，基于叠前粘弹性波形反演（PVWI）恢复了更可靠的纵波速度、横波速度和密度。结合反演结果，计算导数矩阵进行非线性速度频散估计。最后，用模型和实际数据对新的估计方法进行了测试。实验表明，所开发的方法在精度和分辨率上明显优于单界面色散估计方法。该方法可作为储层流体识别的新选择。