Seismic dynamic monitoring in CO_2 flooding based on characterization of frequency-dependent velocity factor  

基于速度频散因子表征的CO_2驱地震监测方法（英文）

作　　者：张军华[1,2] 李军 肖文[1,2] 谭明友 张云银[3] 崔世凌 曲志鹏[3] 

机构地区：[1]中国石油大学(华东)地学院,青岛266580 [2]海洋国家实验室海洋矿产资源评价与探测技术功能实验室,青岛266071 [3]胜利油田物探研究院,东营257015

出　　处：《Applied Geophysics》2016年第2期307-314,418,共9页应用地球物理（英文版）

基　　金：supported by the Pilot Project of Sinopec(P14085)

摘　　要：The phase velocity of seismic waves varies with the propagation frequency, and thus frequency-dependent phenomena appear when CO2 gas is injected into a reservoir. By dynamically considering these phenomena with reservoir conditions it is thus feasible to extract the frequency-dependent velocity factor with the aim of monitoring changes in the reservoir both before and after CO2 injection. In the paper, we derive a quantitative expression for the frequency-dependent factor based on the Robinson seismic convolution model. In addition, an inversion equation with a frequency-dependent velocity factor is constructed, and a procedure is implemented using the following four processing steps： decomposition of the spectrum by generalized S transform, wavelet extraction of cross-well seismic traces, spectrum equalization processing, and an extraction method for frequency-dependent velocity factor based on the damped least-square algorithm. An attenuation layered model is then established based on changes in the Q value of the viscoelastic medium, and spectra of migration profiles from forward modeling are obtained and analyzed. Frequency-dependent factors are extracted and compared, and the effectiveness of the method is then verified using a synthetic data. The frequency-dependent velocity factor is finally applied to target processing and oil displacement monitoring based on real seismic data obtained before and after CO2 injection in the G89 well block within Shengli oilfield. Profiles and slices of the frequency-dependent factor determine its ability to indicate differences in CO2 flooding, and the predicting results are highly consistent with those of practical investigations within the well block.二氧化碳注入地层驱油过程中会引起地震波速度的频散现象,这是地震波传播过程中相速度随频率发生变化所造成的。所以综合考虑二氧化碳气驱情况下的速度频散效应,提取频散因子,进而动态监测气驱前后的油藏变化,具有方法的可行性。本文从Robinson频散褶积模型出发,首先推导了速度频散因子的定量表征式,构建了含有速度频散因子的反演方程,给出了四个处理步骤的实现过程,其中包括广义S变换谱分解、井旁道子波提取、谱均衡处理和阻尼最小二乘频散因子提取。根据粘弹性介质理论,建立了基于Q值变化的衰减地层模型,对止演后的偏移剖面进行了时频谱分析,在此基础上进行了频散因子提取与对比,验证了方法的有效性。最后,将本文方法应用到胜利油田高89井区注气前后两期资料的目标处理与驱油监测中。从提取的两期地震频散因子剖面和切片来看,方法能较好地指示二氧化碳驱前后的差异性,研究成果与井场实际考察结果非常吻合。

关 键 词：CO2 flooding Frequency-dependent velocity factor G89 well block Reservoir dynamic monitoring 

分 类 号：TE357.7[石油与天然气工程—油气田开发工程] P631.4[天文地球—地质矿产勘探]