高分辨率非线性地震波阻抗反演方法和应用
|
摘要
在岩性高度非均质的复杂储层横向预测中,由于可观测信息的重叠、缺失以及噪声的干扰。反演问题总是对应着不唯一、不稳定和不确定的解,因此基于线性假定的储层横向预测技术已不适应这类地区的岩性勘探。为此。提出了一种多井约束分频非线性地震波阻抗反演方法。给出了反演方法的基本原理和算法结构,详细讨论了方法中的关键技术——井旁多级地震反子波的提取、直接反演初始地震波阻抗、井旁多级地震子波的提取、间接反演最终地震波阻抗和大尺度地质模型约束等。在琼东南盆地和南黄海北部盆地。利用实际资料对多井约束分频非线性地震波阻抗反演方法进行了验证。首先对井的波阻抗曲线进行了多尺度分解,并进行了频谱分析；在此基础上,通过井约束波阻抗反演提取了多级统计地震反子波序列；然后分别利用井的不同尺度的频率分量作为约束条件,进行了地震波阻抗反演,获得了低频波阻抗剖面、中高频波阻抗剖面以及由低频和中高频波阻抗剖面合成的全频波阻抗剖面。对这些波阻抗剖面的分析表明,低频波阻抗剖面反映的是大套地层的岩性结构,中高频波阻抗剖面则很好地刻画了地层岩性的细节。因此,可以利用高分辨率的中高频波阻抗剖面识别砂体的边界以及进行含油气性判别。
In predicting complex reservoir with highly heterogeneous li- thology,the integrated reservoir inversion and prediction often have multi-solutions which are instable and uncertain because of the noise interference,and the overlapping and deficiency of geophysi- cal information.So the reservoir prediction technologies based on the linear assumption are not applicable to the exploration for this kind of lithology.Aiming at this problem,a nonlinear multi-well constrained frequency-divided impedance inversion method is pro- posed,given its basic principles and algorithm,and discussed the key technologies such as the extraction of multi-stage seismic in- verse wavelets beside the well,the initial acoustic impedance of di- rect inversion,the extraction of multistage seismic wavelets beside the well,and the final acoustic impedance of indirect inversion and the restraint of large-scale geological model,and so on.In the southeast Qiong basin and the northern basin of south Yellow sea, the actual data was used to confirm the method proposed by us. First,multi-scale decompositions are carried out for impedance curves and the spectra of the data is analyzed.On the basis of this, the multi-stage statistical inverse-wavelet series are extracted by well-constrained impedance inversion.Finally,the seismic imped- ance inversion is performed with the constraint of different-scale frequency component to achieve the low-frequency impedance sec- tion,the middle/high-frequeney impedance section,and the overall frequency impedance section synthesized by the former two sec- tions.The analysis of these impedance sections shows that the low- frequency impedance section reflects the macro lithologic structure of the strata,while the middle/high-frequency impedance section reflects the detailed lithologic structure of the strata.Therefore, the high-resolution middle/high-frequency impedance section can be used to recognize the boundary of sand body and to predict the hy- drocarbon zones.
In predicting complex reservoir with highly heterogeneous li- thology,the integrated reservoir inversion and prediction often have multi-solutions which are instable and uncertain because of the noise interference,and the overlapping and deficiency of geophysi- cal information.So the reservoir prediction technologies based on the linear assumption are not applicable to the exploration for this kind of lithology.Aiming at this problem,a nonlinear multi-well constrained frequency-divided impedance inversion method is pro- posed,given its basic principles and algorithm,and discussed the key technologies such as the extraction of multi-stage seismic in- verse wavelets beside the well,the initial acoustic impedance of di- rect inversion,the extraction of multistage seismic wavelets beside the well,and the final acoustic impedance of indirect inversion and the restraint of large-scale geological model,and so on.In the southeast Qiong basin and the northern basin of south Yellow sea, the actual data was used to confirm the method proposed by us. First,multi-scale decompositions are carried out for impedance curves and the spectra of the data is analyzed.On the basis of this, the multi-stage statistical inverse-wavelet series are extracted by well-constrained impedance inversion.Finally,the seismic imped- ance inversion is performed with the constraint of different-scale frequency component to achieve the low-frequency impedance sec- tion,the middle/high-frequeney impedance section,and the overall frequency impedance section synthesized by the former two sec- tions.The analysis of these impedance sections shows that the low- frequency impedance section reflects the macro lithologic structure of the strata,while the middle/high-frequency impedance section reflects the detailed lithologic structure of the strata.Therefore, the high-resolution middle/high-frequency impedance section can be used to recognize the boundary of sand body and to predict the hy- drocarbon zones.
引文
1李庆忠．走向精确勘探的道路——高分辨率地震勘探系统工程剖析[M]．北京：石油工业出版社，1993．154～174
2周竹生，周熙襄．宽带约束反演方法[J]．石油地球物理勘探，1993，28(5)：523～536
3杨谦．地震资料宽带约束反演方法及应用[J]．石油物探，1996，35(4)：37～42
4张繁昌，印兴耀，吴国忱等．用模拟退火神经网络技术进行波阻抗反演[J]．石油大学学报(自然科版)，1997，21(6)：16～18
5杨文采．地球物理反演的遗传算法[J]．石油物探，1995，34(1)：116～122
6杨文采．神经网络算法在地球物理反演中的应用[J]．石油物探，1995，34(2)：116～120
7 Robinson E A.Predictive decomposition of time series with application to seismic exploration[J].Geophysics,1967,32(3):418～484
8 Robinson E A.Predictive decomposition of seismic traces[J].Geophysics,1957,22(4):767～778
9 Robinson E A,Treitel S.Geophysical signal analysis[M].New Jersey:Prentice-Hall,Inc,1980.56～84
10 Fu L Y.A neuron filtering model and its neural network for space-and time-varying signal processing[R].Third International Conference on Cognitive and Neural systems.Boston University,1999,174～178
11 Fu L Y.Caianiello neural network method for geophysical inverse problems[A].In Poulton M.Computational neural networks for geophysical data processing[C].Amsterdam:Pergamon Press,2001.187～215
12 Fu L Y.Joint inversions of seismic data for acoustic impedance[J].Geophysics,2004,69(4):994～1004
13 Lines L R,Treitel S.A review of least square inversion and its application to geophysical problems[J].Geophysical Prospecting,1984,32(2):159～186
14 Brillouin L.Science and Information Theory[M].New York:Academic Pr,1962.286～324
15 Saggaf M M,Robinson E A.A unified framework for the deconvolution of traces of nonwhite reflectivity[J].Geophysics,2000,65(5):1660～1676
16 Caianiello E R.Outline of a theory of thought-processes and thinking machines[J].J Theoret Biol,1961,2(3):204～235
17 Rumelhart D E,Hinton G E,Williams R J.Learning representations by error propagation[A].In Rumelhart D E,McClelland J L.Parallel distributed processing:Cambridge[C].MA:MIT Press,1986.318～362
18 Fu L Y.Joint Lithologic Inversion[A].In Wong P,Aminzadeh F,Nikravesh M.Soft Computing for Reservoir Characterization and Modeling[C].Heidelberg:Springer-Verlag Publishers,2002.511～530
2周竹生，周熙襄．宽带约束反演方法[J]．石油地球物理勘探，1993，28(5)：523～536
3杨谦．地震资料宽带约束反演方法及应用[J]．石油物探，1996，35(4)：37～42
4张繁昌，印兴耀，吴国忱等．用模拟退火神经网络技术进行波阻抗反演[J]．石油大学学报(自然科版)，1997，21(6)：16～18
5杨文采．地球物理反演的遗传算法[J]．石油物探，1995，34(1)：116～122
6杨文采．神经网络算法在地球物理反演中的应用[J]．石油物探，1995，34(2)：116～120
7 Robinson E A.Predictive decomposition of time series with application to seismic exploration[J].Geophysics,1967,32(3):418～484
8 Robinson E A.Predictive decomposition of seismic traces[J].Geophysics,1957,22(4):767～778
9 Robinson E A,Treitel S.Geophysical signal analysis[M].New Jersey:Prentice-Hall,Inc,1980.56～84
10 Fu L Y.A neuron filtering model and its neural network for space-and time-varying signal processing[R].Third International Conference on Cognitive and Neural systems.Boston University,1999,174～178
11 Fu L Y.Caianiello neural network method for geophysical inverse problems[A].In Poulton M.Computational neural networks for geophysical data processing[C].Amsterdam:Pergamon Press,2001.187～215
12 Fu L Y.Joint inversions of seismic data for acoustic impedance[J].Geophysics,2004,69(4):994～1004
13 Lines L R,Treitel S.A review of least square inversion and its application to geophysical problems[J].Geophysical Prospecting,1984,32(2):159～186
14 Brillouin L.Science and Information Theory[M].New York:Academic Pr,1962.286～324
15 Saggaf M M,Robinson E A.A unified framework for the deconvolution of traces of nonwhite reflectivity[J].Geophysics,2000,65(5):1660～1676
16 Caianiello E R.Outline of a theory of thought-processes and thinking machines[J].J Theoret Biol,1961,2(3):204～235
17 Rumelhart D E,Hinton G E,Williams R J.Learning representations by error propagation[A].In Rumelhart D E,McClelland J L.Parallel distributed processing:Cambridge[C].MA:MIT Press,1986.318～362
18 Fu L Y.Joint Lithologic Inversion[A].In Wong P,Aminzadeh F,Nikravesh M.Soft Computing for Reservoir Characterization and Modeling[C].Heidelberg:Springer-Verlag Publishers,2002.511～530
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心