基于GPU的可控源电磁三维各向异性储层参数反演

The lag of interpretation technique of has becoming the bottleneck of development of three dimensional (3D) controlled-source electromagnetic (CSEM) method. Due to complexity of real earth strata, 3D inversion is urgently needed for real data interpretation. Reservoir parameters play important roles in the stage of oil production. Recognizing the space distribution of reservoir parameters such as porosity, permeability, fluid saturation is not only important for the precision and stability of reservoir assessment, but also crucial for guiding the well design as well as production deployment. Nowadays, the research of rapid numerical calculation of EM responses of 3D CSEM method and 3D anisotropic reservoir parameter inversion is the key issue to be solved for the application of electromagnetic method. In this research, 3D parallel algorithm of CSEM based on GPU will be developed. The multi-scale and multi-source information model integrated with rock physics, numerical simulation, regional geology，well logging as well as reservior production data is used to carry out the precise 3D anisotropic reservoir parameter inversion. By the transformation of geoelectrical and reservoir parameters, the scientific purpose of acquiring the space distribution of 3D anisotropic reservoir parameters will be realized.

三维可控源电磁法的解释技术一直相对落后，成为了制约三维可控源电磁法发展的瓶颈。地下实际地质情况复杂，迫切需要三维反演作为资料解释的重要工具。储层参数在油气田开发阶段具有非常重要的作用，掌握储层的电阻率、孔隙度、渗透率，流体饱和度等参数的空间分布，对油藏储量估算的精确度和可靠性以及指导井位设计、部署开发方案至关重要。研究可控源电磁法快速三维正演计算和各向异性储层参数反演方法是当前提高电磁法应用水平和应用前景迫切需要解决的问题。本研究将开发基于GPU的三维可控源电磁法快速并行正演计算，实现岩石物理、数值模拟、区域地质、测井、油藏开发等资料的多尺度和多源信息的联合建模，实现多种信息约束下的各向异性介质的高精度联合反演成像。通过电性参数与储层参数的转换关系，达到获取三维各向异性储层参数空间分布规律的科学目的。

以三维可控源电磁法的理论为基础，首先详细开展了基于各向异性可控源电磁法的理论，详细推导了含源Maxwell方程的场源关系表达式，研究并提取了不同视电阻率定义的敏感性参数。以砂岩和页岩为例，分别研究了不同岩性储层的电阻率、孔隙度、渗透率，声波速度、TOC含量等参数的转换关系，对油藏评价以及指导井位设计、部署开发方案提供了重要的基础资料。研究了基于GPU和有限元法的三维可控源电磁法快速正演计算，实现了岩石物理、数值模拟、区域地质、测井等资料的多尺度和多源信息的联合建模，初步实现了多信息约束下的各向异性介质的联合反演成像。通过电性参数与储层参数的转换关系，达到了获取三维各向异性储层参数空间分布规律的科学目的。