Abstract

Fracture detection and fluid discrimination play significant roles in the field of geothermics, hydrogeology, and exploration geophysics. However, it is still a challenging task how to reliably and robustly characterize fracture and fluid distribution from seismic data. Therefore, this paper focuses on fracture detection and fluid discrimination from reflected seismic data in the saturated rock with the tilted parallel fractures. And we adopt the dry fracture weakness parameters related with the fracture density and the fluid bulk modulus mainly affected by fluid as indicators of fractures and fluid, respectively. Firstly, using the anisotropic fluid substitution equation and the linear-slip model, the analytical approximate expressions of the stiffness parameters are derived for the gas-bearing rock. Numerical examples illustrate that these approximations possess satisfactory accuracies for the case with the small dry weakness parameters and small fluid bulk modulus. Further, following the connection between the reflection coefficient and the scattering function, and coupling rock physical relationships, we derive a novel linearized PP-wave reflectivity equation parameterized only by six unknown parameters, and rewrite it in terms of the Fourier series to decouple isotropic and anisotropic components of seismic data. Based on the analysis of the contribution of each unknown parameter to the reflection coefficient, a two-step Bayesian inversion method is proposed for estimations of the shear modulus, fluid bulk modulus, porosity, dry weakness parameters, and dip angle, in which the Karhunen–Loève transform is introduced to enhance robustness of inversion. Finally, synthetic and field data are utilized to verify feasibility and stability of the proposed method.

Article Highlights

• Reflectivity is parameterized by only six model parameters for the gas-bearing rock with tilted aligned fractures.

• A two-step inversion incorporating the Fourier series and Karhunen–Loève transform can increase the prediction stability.

• This paper may be useful for the exploration and exploitation of gas-bearing fractured tight sandstones.

摘要

裂缝检测和流体判别在地热学、水文地质学和勘探地球物理学领域发挥着重要作用。然而，如何从地震数据中可靠且稳健地表征裂缝和流体分布仍然是一项具有挑战性的任务。因此，本文重点研究具有倾斜平行裂缝的饱和岩石中反射地震数据的裂缝检测和流体识别。并采用与裂缝密度相关的干裂缝弱化参数和主要受流体影响的流体体积模量作为裂缝和流体的指标。首先，利用各向异性流体置换方程和线性滑移模型，推导出含气岩石刚度参数的解析近似表达式。数值例子表明，这些近似对于具有较小干弱参数和较小流体体积模量的情况具有令人满意的精度。进一步，根据反射系数和散射函数之间的联系，耦合岩石物理关系，我们推导了一个新的线性化PP波反射率方程，仅由六个未知参数参数化，并用傅里叶级数重写，以解耦各向同性和地震数据的各向异性分量。在分析各未知参数对反射系数贡献的基础上，提出了一种两步贝叶斯反演方法，用于估计剪切模量、流体体积模量、孔隙度、干弱参数和倾角，其中引入了 Karhunen-Loève 变换以增强反演的鲁棒性。最后，利用合成数据和现场数据验证了所提方法的可行性和稳定性。

文章重点

• 对于具有倾斜对齐裂缝的含气岩石，反射率仅由六个模型参数参数化。

• 结合傅立叶级数和 Karhunen-Loève 变换的两步反演可以提高预测稳定性。

• 本文可为含气裂缝性致密砂岩的勘探开发提供参考。