层状TI介质中微地震定位和各向异性速度结构同时反演

微地震监测被广泛应用于非常规油气资源的水力压裂作业、油藏描绘和水驱前缘监测工程中.微地震定位采用的初始速度模型一般是基于地震测井记录和射孔数据建立,该速度模型的不准确性易引起定位误差.为降低这种定位误差,本文发展了一种微地震定位和各向异性速度结构同时反演的方法.研究对象为1-D的层状TI介质,其中对称轴方向任意.利用改进的分区多步最短路径算法计算qP、qSV和qSH波的到达时间和射线路径,结合共轭梯度法求解带约束的阻尼最小二乘问题.数值模拟结果表明,该算法能同时进行各向异性速度结构模型(每层的Thomsen参数和界面深度)和微震震源参数(空间坐标和发震时刻)的反演,并且对随机噪声不敏感,有利于实际工程应用.

Simultaneous inversion for anisotropic velocity structure and microseismic location in layered TI media

Microseismic monitoring is widely applied in the engineering of fracturing operations, reservoir delineation and water front monitoring in unconventional oil/gas resources. Microseismic inversion generally makes use of an initial velocity model based on well logs and seismic data from perforation shots. The problem of precision in the model will cause errors in microseismic sources locating. To reduce errors, this paper has developed a method that simultaneously updates the velocity model and locates events in space and time. The 1-D layered anisotropy with a transversely-isotropic (TI) symmetry is assumed, where the symmetry axis can be arbitrarily oriented. The arrival times and ray paths of the qP, qSV and qSH waves are calculated with a modified multistage shortest-path algorithm. Combined with the conjugate gradient method, a damped, minimum-norm, least-squares and constrained problem is solved. The numerical examples show that the proposed algorithm can be used to invert the anisotropic velocity model (Thomsen parameters and interface depth for each layer) and locate the microseismic sources and their onset times, simultaneously. And it is not so sensitive to random noise, which is beneficial for the actual engineering applications.