复杂碳酸盐岩缝洞型储层的地震精细预测方法

利用地震属性能够对非常规储层中的裂缝等不连续异常进行识别, 但碳酸盐岩缝洞型储层非均质性强, 地震信号响应复杂, 使得储层精细结构描述面临重大挑战。为此, 提出针对目标储层的裂缝精细预测技术流程, 首先使用构造导向滤波(SOF)方法去除地震数据中的噪声, 在保留断层构造特征的基础上, 增强地震数据中的有效信息; 其次, 充分挖掘地震信号的多尺度特征, 开展多方法技术的组合应用, 包括基于三维多尺度体曲率属性(3D MSVC)、蚂蚁追踪、最大似然属性以及基于二维希尔伯特变换的储层裂缝多尺度体边缘检测; 最后, 将这些方法获得的缝洞储集体的多源特征信息进行降维优化和融合, 从而降低地震精细预测的多解性和不确定性, 实现碳酸盐岩缝洞储集体不同尺度特征要素信息的同步检测。实际地震资料应用结果表明, 该方法对裂缝、溶洞、古河道和隐伏断裂有很好的检测效果, 能够为碳酸盐岩缝洞型储层勘探开发提供有效的技术支撑。

Seismic attributes can be used to predict discontinuous anomalies (e.g.fractures) in unconventional reservoirs.For carbonate fractured-vuggy reservoirs, however, the strong heterogeneity leads to complex seismic responses, making the fine description of reservoir structure very challenging.In this paper, a method for fine prediction of fractures in target reservoirs was proposed.Firstly, the structure-orientated filter (SOF) is used to remove the noise from seismic data, so that the effective information in seismic data is improved while the characteristics of fault structures are preserved.Secondly, the multi-scale characteristics of seismic signals are adequately identified with the support of multiple methods/techniques including 3D multi-scale volumetric curvature (3D-MSVC) attribute, ant tracking, maximum likelihood attribute, and 2D Hilbert transform based multi-scale reservoir fracture edge detection.Finally, the multi-source characteristic information of fractured-vuggy reservoir obtained by these methods/techniques is optimized and fused to mitigate the ambiguity and uncertainty in fine seismic prediction and realize the synchronous detection of featured information of carbonate fractured-vuggy reservoirs at different scales.The application results of actual seismic data showed that the method proposed in this paper can finely predict the fractures, vugs, ancient channels and subtle faults, so as to efficiently support the exploration and development of carbonate fractured-vuggy reservoirs.