基于经验模态分解的碳酸盐岩储层含气性检测方法研究

Deep carbonate reservoirs whose resources are rich in China have the low proved rate. The main difficulties existed in the exploration are that resources always have characteristics of poor reflection continuity and weak gas-bearing response and weak difference between the reservoir and the non-reservoir, which limit the methods and techniques applied in high accuracy gas reservoir prediction and fine description. For the problems in deep carbonate reservoirs, weak signal detection methods based on Ensemble Empirical Mode Decomposition (EEMD) and Complete Ensemble Empirical Mode Decomposition (CEEMD) which have high resolution are used. Spectral decomposition methods with high resolution using EEMD or CEEMD and conventional complex seismic trace analysis and TK energy separation method and wavelet transform method and so on are developed and analyzed based on the detection of weak signals and frequency attenuation characteristics of the seismic waves. Attenuation gradient analysis methods and formation absorption section estimation methods with high resolution based on EEMD or CEEMD which can give a quantitative estimate of the attenuation information are further studied. The study is expected to show clearer statistical interpretations of hydrocarbons and improve the accuracy of gas detection and further establish the identification method and technical system of quantitative attenuation parameters which are suitable for identifying gas and water in carbonate reservoir. Using ChuanXi marine carbonate rocks formation as a typical test area, the application theoretical models of EEMD- and CEEMD-based gas-bearing detection methods are established. The analysis and evaluation of practical application effects are carried out to form a technology system which can be suitable to the gas-bearing detection in carbonate reservoir. Finally, we want to extend the technology to other areas to provide critical technical support and services for carbonate reservoirs exploration.

深层碳酸盐岩储层储量丰富而探明率低。勘探难点主要在于储层反射连续性差、含气响应微弱、储层与非储层区别不明显，导致能用于高精度气藏储层预测和精细描述的方法技术有限。针对这些问题，本项目从弱信号检测和地震波频域衰减特性出发，采用具有高分辨率的基于聚合经验模态分解（EEMD）和完备聚合经验模态分解（CEEMD）的弱信号检测方法，发展基于EEMD或CEEMD联合常规地震道分析法、TK能量分离法和小波变换等的高分辨率谱分解方法；研究基于EEMD或CEEMD的高分辨率定量估计衰减信息的衰减梯度分析法和地层吸收剖面方法，提高烃类信息的统计性解释结果和含气预测准确度，建立气、水识别的方法技术体系。以川西海相碳酸盐岩地层为典型研究试验区，建立EEMD和CEEMD在含气性检测中应用的理论模式并对实际应用效果分析和评价，形成一套适合碳酸盐岩含气性检测的方法技术体系。为我国碳酸盐岩气藏勘探提供关键技术支持和服务。

利用地震资料进行深层碳酸盐岩储层含气性检测的重点是识别弱信号背景下的弱含气响应差别。本项目针对叠后高保真地震数据，研究了利用基于聚合经验模态分解（EEMD）和完备聚合经验模态分解（CEEMD）的自适应随机信号处理算法提取弱含气储层响应情况下的地球物理信息，并在此基础上发展了基于EEMD和CEEMD的碳酸盐岩含气储层地震预测和含气性检测方法技术。分析了模态混叠对基于经验模态分解的含气性检测方法的影响，提出了基于EEMD和CEEMD的高亮体属性提取算法；发展了结合EEMD和小波变换的频率衰减梯度估计方法； 提出了基于CEEMD和希尔伯特变换、最小二乘法的高精度地震衰减估计算法；发展并优化了基于CEEMD的地层吸收剖面算法。将这些研究成果应用到川西海相碳酸盐岩储层含气性检测中，储层预测结果表明，利用本项目所发展方法可以有效地识别宽带地震响应中特定频率的强振幅异常，给出的含烃类统计性解释结果与已知井的含气测试结果相吻合。与常规地震属性分析方法对比显示，本项目发展的基于经验模态分解的含气性检测方法在碳酸盐岩储层含气性检测中具有较好的实用性，为深层碳酸盐岩储层烃类检测提供了新的技术支撑。