微地震弹性波模式解耦矩张量成像与反演方法

Microseismic monitoring plays an important role in commercial development of unconventional resources such as shale gas, because it is the key technology to determine the structure of rock damage and stimulated reservoir volume. For surface or near-surface microseismic monitoring, induced events are always hided in noise, and thus make challenges for source location based on picking of P and S wave arrivals. As an alternative technology, migration-type location approach is just at the initial stage of application and has limited capability to characterize the source property using the multi-component data. Source mechanism inversion often uses the approaches introduced from earthquake seismology. The existing approaches have low reliability due to the effects of multi-parameter trade-offs and sensitivity to velocity errors. In this project, we will extend our recent study on active source seismic wave propagation, imaging and waveform inversion for anisotropic media to passive source seismic. Based upon the investigation of focal mechanisms of seismic sources and the characteristics of near-field radiation, far-field polarization and propagation in anisotropic media, we plan to apply mode decoupling to the adjoint elastic wavefields, which are the essential wavefields for both time-reversal event location and moment tensor full-waveform inversion. The proposed moment tensor imaging approach can simultaneously locate the events and characterize their source properties, while the mode decoupling based moment tensor inversion approach will mitigate the parameter trade-offs and be more robust to the errors of velocity models. We believe that this project will produce new technology with promising potential for domestic unconventional petroleum industry.

微地震监测是水力压裂过程中确定岩石破损结构和储层改造体积的核心技术，在页岩气等低渗透性油气资源商业开发中发挥关键作用。地面或近地表接收的微地震信号常常被噪声掩盖，给基于P和S波初至拾取的定位方法带来挑战；视为替代技术的偏移型定位方法刚进入应用领域，尚未充分挖掘多分量波场对震源机制的描述潜力；从天然地震学领域借鉴过来的震源机制反演方法受参数折衷效应、速度误差和观测孔径限制等因素困扰，在实际应用中可靠性不高。本项目拟将申请人在各向异性介质主动源地震波传播、成像与波形反演方面的研究向被动源地震领域延伸。以各向异性介质水力压裂微地震物理机制、近场辐射与远场极化和传播特征分析为基础，对逆时定位和矩张量全波形反演中最核心的弹性波伴随场引入模式解耦，研究可同时进行震源定位和属性刻画的矩张量成像方法，降低矩张量全波形反演的参数折衷效应和速度误差敏感性，为我国非常规油气产业发展提供有良好前景的创新技术。

无论是非常规油气储层水力压裂动态监测，还是各种类型天然地震的震源问题研究，迫切需要发展新型震源定位与矩张量成像方法，建立适合震源与弹性介质多参数建模的波形反演新方法。本项目在弹性波动力学理论基础上，调查了典型震源机制的能量辐射模式和弹性体波在各向异性非常规储层中的极化或偏振特征，提出适应任意各向异性介质的弹性波高精度模拟算子，开发出基于偏振投影的各向异性弹性波模式解耦高精度、高效率算法，建立基于P/S波模式解耦的弹性波矢量成像与逆时聚焦定位方法，为微地震数据震源定位、矩张量成像与震源机制描述提供了优秀的弹性波传播引擎和伴随场成像技术。针对多分量地震数据震源与介质参数建模与结构成像问题，提出P/S模式解耦预条件波形反演方法，为解决包括震源矩张量、纵-横波速度等多参数反演面临的参数耦合难题提供了创新思路。调查了主动源和被动源多分量地震数据P/S分离的人工智能解决方案，初步建立基于深度学习方法的地面与海底多分量地震记录P/S分离技术，为基于P波和S波震相走时与波形数据的震源定位、机制反演以及纵、横波速度建模提供了有前景的技术储备。此外，将微地震伴随场成像方法拓展到了天然地震远震阵列数据，剖析了地壳和上地幔各向异性对岩石圈弹性界面成像的重要影响。项目研究成果在Geophysical Journal International、Geophysics、Geophysical Prospecting、IEEE Geoscience and Remote Sensing Letters以及《地球物理学报》、《石油地球物理勘探》等国际与国内核心期刊正式发表，也有部分成果投稿后处于录用或修改过程中。负责人应邀赴新加坡国立大学与沙特阿卜杜拉国王科技大学进行短期访问，在国际与国内学术会议做邀请报告或专题报告，并通过与工业界的交流合作推动我国非常规油气资源地震勘探开发技术进步。