基于物理模型的碳酸盐岩孔洞地震属性量化分析

发布时间：2018-02-09 18:41

本文关键词： 碳酸盐岩定量分析稀疏约束反演谱分解多孔洞组合 Q值估算　出处：《中国石油大学(北京)》2016年博士论文　论文类型：学位论文

【摘要】：我国的碳酸盐岩储层拥有巨大的勘探和开发潜力,但是油气藏埋藏深,缝洞的分布规律复杂,勘探开发和定量预测难度大,因此展开基于物理模型的碳酸盐岩孔洞地震属性定量化分析,对后续的储层预测和储量估算奠定基础。以塔里木哈拉哈塘地区碳酸盐岩储层三维地震物理模型为数据基础,研究了潜山岩溶区和层间岩溶区孔洞的地震响应特征,并通过地震属性技术和三维可视化技术,建立了潜山岩溶区和层间岩溶区孔洞储集体雕刻体积和真实体积之间的定量关系曲线,着重分析了潜山岩溶区和层间岩溶区孔洞储集体在不同上覆层结构、不同孔洞充填物以及不同孔洞尺度因素下的雕刻体积和真实体积之间的关系曲线的差异,取得了一些认识。基于稀疏约束反演谱分解的理论,通过合成地震记录的例子,认识到其时频能量谱比传统的谱分解方法如STFT、S变换及CWT等具有较高的时频分辨率、较好的时频聚焦性和较好的抗噪性,能够有效反映信号的局部特征,不受时窗长度和形状的影响;其时频相位谱不但清楚反映了时频能量分布,还带有有效可识别的相位信息。将稀疏约束反演谱分解的时频能量谱与时频相位谱结合,用于碳酸盐岩多孔洞储集体组合的识别中,通过物理模型数据中的规则洞组合、陷落柱洞和层状垂直组合的例子以及一个实际地震剖面的例子,说明该方法有利于判断多孔洞储集体的组合形式和孔洞组成个数,为后续的储层预测和孔洞储集体体积的定量估算提供有利依据。此外,根据稀疏约束反演谱分解的稀疏性和地震信号的可重构性,对碳酸盐岩物理模型数据和实际数据的叠后地震剖面进行了去噪,有效去除随机噪声。以频谱比值法为算法基础,认为频率域的Q值估算方法对于谱估计的结果非常敏感,而精确的谱估计更有利于Q值的估算,使用传统谱分解方法估算Q值,存在时频分辨率和抗噪性差的问题,而稀疏约束反演谱分解恰好弥补了这一缺陷,因此提出了基于稀疏约束反演谱分解的Q值估算方法。通过合成地震数据的例子,证明该方法的优势以及有效性,并将其应用到不同含气量和不同孔隙度的碳酸盐岩孔洞中,得到了一些认识,为判断孔洞含油气特性提供了依据。
[Abstract]:Carbonate reservoirs in China have great potential for exploration and development, but the depth of oil and gas reservoirs is deep, the distribution of fractures and cavities is complex, and the exploration and development and quantitative prediction are difficult. Therefore, the quantitative analysis of seismic attributes of carbonate pores based on physical model is carried out, which lays a foundation for the subsequent reservoir prediction and reserve estimation, and takes the three-dimensional seismic physical model of carbonate reservoir in Hala Hattang area of Tarim as the data base. The seismic response characteristics of holes in buried hill karst area and interlayer karst area are studied, and the seismic attribute technology and 3D visualization technology are used to study the seismic response characteristics of holes in buried hill karst area and interbedded karst area. The quantitative relationship curve between the volume of hole reservoir and the real volume in buried hill karst area and interlayer karst area is established, and the different overburden structure of hole reservoir in buried hill karst area and interlayer karst area is analyzed emphatically. Based on the theory of spectral decomposition of sparse constrained inversion, the difference of relationship curves between engraved volume and real volume under different pore filling and different pore size factors is obtained. Based on the theory of spectral decomposition of sparse constrained inversion, an example of synthetic seismic records is obtained. Recognizing that the time-frequency energy spectrum has higher time-frequency resolution, better time-frequency focusing and better anti-noise than the traditional spectral decomposition methods such as STFTs transform and CWT, and can effectively reflect the local characteristics of the signal, The time-frequency phase spectrum not only clearly reflects the time-frequency energy distribution, but also has effectively identifiable phase information. The time-frequency energy spectrum decomposed by sparse constrained inversion spectrum is combined with the time-frequency phase spectrum. Used in the identification of multi-porosity reservoir assemblages in carbonate rocks, through regular hole combinations in physical model data, examples of cave-hole and layered vertical combinations, and an example of an actual seismic profile. It is shown that this method is helpful to judge the combination form and the number of pore components of the multi-porosity reservoir, and to provide a favorable basis for the subsequent reservoir prediction and quantitative estimation of the volume of the pore body. According to the sparsity of spectral decomposition of sparse constrained inversion and the reconfiguration of seismic signal, the post-stack seismic profile of carbonate physical model data and actual data is de-noised, and the random noise is effectively removed. The spectral ratio method is used as the basis of the algorithm. It is considered that the estimation method of Q value in frequency domain is very sensitive to the result of spectrum estimation, and accurate spectrum estimation is more favorable to the estimation of Q value. The problem of time-frequency resolution and poor noise resistance exists when the traditional spectral decomposition method is used to estimate Q value. The sparse constraint inversion spectral decomposition makes up for this defect. Therefore, a Q value estimation method based on sparse constraint inversion spectral decomposition is proposed. The advantages and effectiveness of this method are proved by an example of synthetic seismic data. It has been applied to carbonate pores with different gas content and different porosity, and some understanding has been obtained, which provides a basis for judging the oil-gas characteristics of pores.
【学位授予单位】：中国石油大学(北京)
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P618.13;P631.4

【相似文献】