应用地震多属性反演与随机模拟技术预测砂体展布——以川西丰谷构造须四段为例
|
摘要
针对丰谷构造须四段储层具有埋深大、物性差、砂泥岩速度差异小等特点,综合采用基于神经网络的地震多属性反演与基于地质统计学的随机模拟进行砂体展布预测。该方法在目标曲线预分析和井震标定的基础上,首先根据地震属性与目标曲线的相关性,对地震属性进行排序与优选;然后通过概率神经网络训练,寻找目标曲线与多种地震属性的非线性关系,得到目标曲线的反演数据体;最后以反演体为约束条件,通过随机模拟的方法建立砂岩百分含量模型,完成砂体的井间预测。应用结果表明,该方法可以有效地减少砂体预测的多解性、提高砂体预测的精度,较为客观地展现各期砂体的空间分布以及相互叠置关系,具有一定应用价值。
The reservoir of Xu4 formation of Fenggu Structure is characterized by high buried depth,poor reservoir property,and little velocity's difference between sand and mud,etc.In view of these characteristics,using seismic multiple attributes inversion based on neural network and stochastic simulation based on geological statistics to predict sandbody distribution.The concrete step is that firstly the efficient pre-analysis on target curve and the well seismic calibration are made.And seismic attributes are optimally sorted and selected based on the correlation between single seismic attribute and target curve.Secondly,the non-linear relation between multiple attributes and target curve is searched and the three-dimensional inversion volume is obtained through the training of Probabilistic Neural Network.Finally,a model of sandstone content is built using this volume as the constrained condition,and then the inter-well distribution of sandbody is predicted.The practical application shows that using this method can effectively reduce the mistiness of prediction and improve the precision of prediction.Besides,spatial distribution and superimposition relationship of different sandbodys can be accurately displayed using the method.It has a certain value for the sandodys prediction.
The reservoir of Xu4 formation of Fenggu Structure is characterized by high buried depth,poor reservoir property,and little velocity's difference between sand and mud,etc.In view of these characteristics,using seismic multiple attributes inversion based on neural network and stochastic simulation based on geological statistics to predict sandbody distribution.The concrete step is that firstly the efficient pre-analysis on target curve and the well seismic calibration are made.And seismic attributes are optimally sorted and selected based on the correlation between single seismic attribute and target curve.Secondly,the non-linear relation between multiple attributes and target curve is searched and the three-dimensional inversion volume is obtained through the training of Probabilistic Neural Network.Finally,a model of sandstone content is built using this volume as the constrained condition,and then the inter-well distribution of sandbody is predicted.The practical application shows that using this method can effectively reduce the mistiness of prediction and improve the precision of prediction.Besides,spatial distribution and superimposition relationship of different sandbodys can be accurately displayed using the method.It has a certain value for the sandodys prediction.
引文
[1]Chen Q,Sidney S.Seismic attribute technology for eservoir forecastingand monitoring[J].The Leading Edge,1997,16(5):445-456.
[2]Hampson D P,Schuelke J S,Quirein J A.Use of multiattributetransforms to predict log properties from seismic data[J].Geophysics,2001,66(1):2035-2041.
[3]Deutsch C V,Journel A.Geostatistical software library and users guide[M].Oxford:Oxford University Press,1992.
[4]Wu S H,Zhang Y W,Ringas J E.Reservior stochastic modelingcontrained by quantitative geological conceptual pattern[J].PetroleumScience,3(1):27-33.
[5]赵政璋,赵贤正,王英民,等.储层地震预测理论与实践[M].北京:科学出版社,2004.Zhao Zhengzhang,Zhao Xianzheng,Wang Yingmin,et al.Theory andpractice of reservior's seismic prediction[M].Beijing:Science Press,2004.
[6]陈遵德,陈富贵.地震数据油气预测中的属性优化方法[J].石油物探,1998,37(4):32-40.Chen Zunde,Chen Fugui.Geophysical Prospecting for Petroleum,1998,37(4):32-40.
[7]郑晓芳.地震多属性反演方法的研究及应用[D].北京:中国地质大学,2005.Zheng Xiaofang.The study&application of seismic multi-attributeinversion[D].Beijing:China University of Geoscience,2005.
[8]李东安,宁俊瑞,刘振峰.用神经网络和地质统计学综合多元信息进行储层预测[J].石油与天然气地质,2010,31(4):493-498.Li Dongan,Ning Junrui,Liu Zhenfeng.Oil&Gas Geology,2010,31(4):493-498.
[9]Liu Z P,Liu J Q.Seismic-controlled nonlinear extrapolation of wellparameters using neural networks[J].Geophysics,1998,63:2035-2041.
[10]吴胜和.储层表征与建模[M].北京:石油工业出版社,2009.Wu Shenghe.Reservior characterization&modeling[M].Beijing:Petroleum Industry Press,2009.
[11]叶泰然,曾小英,刘兴艳.丰谷、玉皇观地区须家河组四段井位论证报告[R].成都:中国石化西南分公司勘探开发研究院,2006.Ye Tairan,Zeng Xiaoying,Liu Xingyan.Report of well locationdemonstration of Xu4 Member in the area of Fenggu and YuHuangguan[R].Chengdu:Research Institute of Exploration andDevelopment of Southwest Oil&Gas Campany,Sinopec Group,2006.
[2]Hampson D P,Schuelke J S,Quirein J A.Use of multiattributetransforms to predict log properties from seismic data[J].Geophysics,2001,66(1):2035-2041.
[3]Deutsch C V,Journel A.Geostatistical software library and users guide[M].Oxford:Oxford University Press,1992.
[4]Wu S H,Zhang Y W,Ringas J E.Reservior stochastic modelingcontrained by quantitative geological conceptual pattern[J].PetroleumScience,3(1):27-33.
[5]赵政璋,赵贤正,王英民,等.储层地震预测理论与实践[M].北京:科学出版社,2004.Zhao Zhengzhang,Zhao Xianzheng,Wang Yingmin,et al.Theory andpractice of reservior's seismic prediction[M].Beijing:Science Press,2004.
[6]陈遵德,陈富贵.地震数据油气预测中的属性优化方法[J].石油物探,1998,37(4):32-40.Chen Zunde,Chen Fugui.Geophysical Prospecting for Petroleum,1998,37(4):32-40.
[7]郑晓芳.地震多属性反演方法的研究及应用[D].北京:中国地质大学,2005.Zheng Xiaofang.The study&application of seismic multi-attributeinversion[D].Beijing:China University of Geoscience,2005.
[8]李东安,宁俊瑞,刘振峰.用神经网络和地质统计学综合多元信息进行储层预测[J].石油与天然气地质,2010,31(4):493-498.Li Dongan,Ning Junrui,Liu Zhenfeng.Oil&Gas Geology,2010,31(4):493-498.
[9]Liu Z P,Liu J Q.Seismic-controlled nonlinear extrapolation of wellparameters using neural networks[J].Geophysics,1998,63:2035-2041.
[10]吴胜和.储层表征与建模[M].北京:石油工业出版社,2009.Wu Shenghe.Reservior characterization&modeling[M].Beijing:Petroleum Industry Press,2009.
[11]叶泰然,曾小英,刘兴艳.丰谷、玉皇观地区须家河组四段井位论证报告[R].成都:中国石化西南分公司勘探开发研究院,2006.Ye Tairan,Zeng Xiaoying,Liu Xingyan.Report of well locationdemonstration of Xu4 Member in the area of Fenggu and YuHuangguan[R].Chengdu:Research Institute of Exploration andDevelopment of Southwest Oil&Gas Campany,Sinopec Group,2006.
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