Quantitative characterization and monitoring of reservoir properties, pressures, fluids and fractures with multicomponent and quasi-continuous full-waveform seismology

全波形(FWI)地震方法在复杂近海油藏成像方面取得了令人瞩目的工业和学术成就，但作为监测陆上常规和非常规生产、二氧化碳和废水注入以及提高采收率方法的实用、常规工具，基础和应用科学仍需取得进展。幸运的是，为了帮助取得这样的进展，地球物理学家现在手头有以高性能计算和人工智能技术为形式的强大的新地球计算工具，以及以永久可控震源、宽带地震检波器和分布式声学传感(DAS或光纤)地震传感器和钻柱声学技术为形式的强大的新仪器和地震采集工具。我们建议创建下一代实用的FWI油藏描述和监控工具，包括通过分析复杂的现代油藏环境的弹性、粘度和各向异性来确定岩石物理属性(压力、流体、裂缝和粘度)的高分辨率地图。我们小组在宽带和光纤领域和实验室采集、实用的多参数弹性、粘弹性和各向异性FWI方法开发、岩石物理地震反演、近地表表征、钻柱成像、机器学习和用于大计算/大数据问题的HPC方法方面进行了重要的研究，但仍是初步的。我们计划扩大和扩大这些早期的成功，创建一个实用的储集层波形包。涉及多分量和DAS数据的FWI、弹性FWI-岩石物理敏感性分析、面波分析和处理、作为地震反演框架的机器学习以及新的混合采集和去混合方法，是这项研究的关键预期成果。这些努力将带来高分辨率地质填图方面的知识和技术创新。这使加拿大通过HQP技术培训和研究成果(高分辨率监测能力)受益，这些成果直接影响资源开采效率、减轻地质灾害，并有助于减少土地、水和能源的使用。

Full waveform (FWI) seismic methods have achieved spectacular industrial and academic successes in image-forming of complex offshore reservoirs, but as practical, regular-use tools for monitoring of onshore conventional and unconventional production, CO2 and wastewater injection, and EOR methods, basic and applied scientific progress is still required. Fortunately, in aid of making such progress, geophysicists now have at hand powerful new geo-computational tools, in the form of HPC and artificial intelligence technology, and powerful new instrumentation and seismic acquisition tools, in the form of permanent controllable sources, broadband geophones and distributed acoustic sensing (DAS, or fibre-optic) seismic sensors, and drillstring acoustic technology. We propose to create the next generation of practical, FWI reservoir characterization and monitoring tools, involving the determination of high resolution maps of rock physics properties - pressures, fluids, fractures and viscosities - through analysis of the elasticity, viscosity, and anisotropy of the complex modern reservoir environment. Our group has carried out significant, though initial, research in broadband and fibre-optic field and laboratory acquisition, practical multi-parameter elastic, viscoelastic and anisotropic FWI method development, rock-physics seismic inversion, near surface characterization, drillstring imaging, machine learning, and HPC methods for large computation / large data problems. We propose to grow and expand these early successes, creating a practical reservoir waveform package. FWI involving multicomponent and DAS data, elastic FWI-rock physics sensitivity analysis, surface wave analysis and processing, machine learning as a framework for seismic inversion, and new blended acquisition and deblending methodologies, are key expected outcomes of the research. These efforts will bring about both knowledge and technology creation in high-resolution geological mapping. This benefits Canada through technical HQP training, and research with outcomes (high resolution surveillance capabilities) that directly affect resource extraction efficiency, geohazard mitigation, and which contribute to reduction of land, water, and energy use.