致密碎屑岩储层地震反演技术方案及应用
|
摘要
致密碎屑岩储层与围岩地震波阻抗差异微弱,应用常规储层地震反演方法的有效性较差,精度较低。针对致密碎屑岩储层的地质、地球物理特点,本文提出了将神经网络和地质统计学结合起来的致密碎屑岩储层地震反演技术方案。在此方案中,通过神经网络地震反演获得地质涵义较为明确的但垂向精度较低的反演结果,以此结果为约束,以测井数据作为条件数据(硬数据)进行储层参数地质统计学随机反演/模拟,进而得到较为精细的、同时横向分布较为符合地质规律的储层参数反演成果。通过在D气田致密碎屑岩储层地震反演中的应用,提高了储层预测精度,达到了良好的应用效果。
The feeble difference in acoustic impedance between reservoir and its adjacent rocks is unfavorable to use seismic inversion to characterize reservoirs,and the corresponding results are always invalid and coarse.Based on the knowledge of the geology and geophysics of tight clastic reservoir,this paper outlines a new program used for seismic inversion in tight clastic reservoir characterization.Firstly a coarse inversion result in accordance with geological pattern is obtained through artificial neural network.Then geo-statistics inversion is applied to integrate well data and the inversion data mentioned above,and to get the high-resolution inversion data honored with lateral geological pattern.This seismic inversion program is employed in D gas field,in which tight gas reservoir is the main type,and the high-resolution results are obtained which can be used for reservoir characterization.
The feeble difference in acoustic impedance between reservoir and its adjacent rocks is unfavorable to use seismic inversion to characterize reservoirs,and the corresponding results are always invalid and coarse.Based on the knowledge of the geology and geophysics of tight clastic reservoir,this paper outlines a new program used for seismic inversion in tight clastic reservoir characterization.Firstly a coarse inversion result in accordance with geological pattern is obtained through artificial neural network.Then geo-statistics inversion is applied to integrate well data and the inversion data mentioned above,and to get the high-resolution inversion data honored with lateral geological pattern.This seismic inversion program is employed in D gas field,in which tight gas reservoir is the main type,and the high-resolution results are obtained which can be used for reservoir characterization.
引文
[1]丁晓琪,张哨楠,周文等.鄂尔多斯盆地北部上古生界致密砂岩储层特征及其成因探讨.石油与天然气地质,2007,28(4):491～496Ding Xiaoqi,Zhang Shaonan,Zhou Wen et al.Char-acteristics and genesis of the Upper Paleozoic tightsandstone reservoirs in the northern Ordos basin.Oil&Gas Geology,2007,28(4):491～496
[2]于兴河,李胜利,赵舒等.河流相油气储层的井震结合相控随机建模约束方法.地学前缘,2008,15(4):33～41Yu Xinghe,Li Shengli,Zhao Shu et al.The constrai-ning method on stochastic modeling for fluvial petro-leum reservoir controlled by depositional facies in-tegrating wells with seismic data.Earth ScienceFrontiers,2008,15(4):33～41
[3]McCormack M D.Neural computing in geophysics.The Leading Edge,1991,10(1):11～15
[4]孙翠娟,张文.概率神经网络技术在油气藏物性参数预测中的应用.中国海上油气(地质),2002,16(5):349～354Sun Cuijuan,Zhang Wen.An application of probabi-listic neural network technique to parameter predic-tion for oil and gas reservoir.China Offshore Oiland Gas(Geology),2002,16(5):349～354
[5]张绍红.概率神经网络技术在非均质地层岩性反演中的应用.石油学报,2008,29(4):549～552Zhang Shaohong.Application of probabilistic neuralnetwork technique to lithology inversion of heteroge-neous stratum.Acta Petrolei Sinica,2008,29(4):549～552
[6]孙思敏,彭仕宓.基于模拟退火算法的地质统计学反演方法.石油地球物理勘探,2007,42(1):38～43Sun Simin,Peng Shimi.Inversion of geostatisticsbased on simulated annealing algorithm.OGP,2007,42(1):38～43
[7]印兴耀,刘永社.储层建模中地质统计学整合地震数据的方法及研究进展.石油地球物理勘探,2002,37(4):423～430Yin Xingyao,Liu Yongshe.Methods and develop-ment of integrating seismic data in reservoir model-building.OGP,2002,37(4):423～430
[8]王霞,张延庆,于志龙等.叠前反演结合地质统计模拟预测薄储层.石油地球物理勘探,2011,46(5):744～748Wang Xia,Zhang Yanqing,Yu Zhilong et al.Thinreservoir prediction by prestack inversion combinedwith geostatistic simulation.OGP,2011,46(5):744～748
[9]李建雄,谷跃民,党虎强等.高密度井网开发区井震联合储层预测方法——以SL盆地JN2区块为例.石油地球物理勘探,2011,46(3):457～462Li Jianxiong,Gu Yuemin,Dang Hu-qiang et al.Wellseismic and surface seismic joint prediction for reser-voir rock in a development area with denser produc-ers—Case study of Block JN2in SL Basin.OGP,2011,46(3):457～462
[10]刘振峰,郝天珧,杨长春.沉积模型和储层随机建模.地球物理学进展,2003,18(3):519～523Liu Zhenfeng,Hao Tianyao,Yang Changchun.Sedi-mentation models and stochastic reservoir modeling.Progress in Geophysics,2003,18(3):519～523
[11]董宁,刘振峰,周小鹰等.鄂尔多斯盆地致密碎屑岩储层地震识别及预测.石油与天然气地质,2008,29(5):668～675Dong Ning,Liu Zhenfeng,Zhou Xiaoying et al.Seis-mic detection and prediction of tight gas reservoirs inthe Ordos Basin.Oil&Gas Geology,2008,29(5):668～675
[12]Caers J,Ma X.Modeling conditional distributions offacies from seismic using neural network.Mathema-tical Geology,2002,34(2):143～167
[13]Caers J,Avseth P,Mukerji T.Geostatistical integra-tion of rock physics seismic amplitudes and geologicalmodels in North-Sea turbidite systems.The LeadingEdge,2001,20(3):308～312
[14]姚逢昌,甘利灯.地震反演的应用与限制.石油勘探与开发,2000,27(2):53～56Yao Fengchang,Gan Lideng.Application and re-striction of seismic inversion.Petroleum Explorationand Development,2000,27(2):53～56
[15]凌云,郭向宇,高军等.油藏地球物理面临的技术挑战与发展方向.石油地球物理勘探,2010,49(4):319～335Ling Yun,Guo Xiangyu,Gao Jun et al.The techni-cal challenges and the development trend of reservoirgeophysics.OGP,2010,49(4):319～335
[16]Humpson D P,Schuelke J S,Quirein J A.Use of multiattribute transforms to predict log propertiesfrom seismic data.Geophysics,2001,66(1):220～236
[17]Masters T.Signaland Image Processing with Neural Networks.New York:John Wiley&Sons Incorpora-tion,1994:1～200
[18]Pan Guocheng著;李钟山译;夏立显校.地质统计学中的区域化变量理论.世界地质,1997,16(2):85～93
[19]PanGuocheng著;李钟山,陈永良译;夏立显校.地质统计学中结构分析的理论与方法.世界地质,1997,16(3):70～82
[20]Pan Guocheng著;何凯涛,辜华良译;夏立显校.地质统计学中的估值技术和条件模拟.世界地质,1997,16(3):83～100
[21]陈建阳,田昌炳,周新茂等.融合多种地震属性的沉积微相研究与储层建模.石油地球物理勘探,2011,46(1):98～102Chen Jianyang,Tian Changbing,Zhou Xinmao et al.Sedimentary microfacies studies and reservoir model-ing by integration of multiple seismic attributes.OGP,2011,46(1):98～102
[2]于兴河,李胜利,赵舒等.河流相油气储层的井震结合相控随机建模约束方法.地学前缘,2008,15(4):33～41Yu Xinghe,Li Shengli,Zhao Shu et al.The constrai-ning method on stochastic modeling for fluvial petro-leum reservoir controlled by depositional facies in-tegrating wells with seismic data.Earth ScienceFrontiers,2008,15(4):33～41
[3]McCormack M D.Neural computing in geophysics.The Leading Edge,1991,10(1):11～15
[4]孙翠娟,张文.概率神经网络技术在油气藏物性参数预测中的应用.中国海上油气(地质),2002,16(5):349～354Sun Cuijuan,Zhang Wen.An application of probabi-listic neural network technique to parameter predic-tion for oil and gas reservoir.China Offshore Oiland Gas(Geology),2002,16(5):349～354
[5]张绍红.概率神经网络技术在非均质地层岩性反演中的应用.石油学报,2008,29(4):549～552Zhang Shaohong.Application of probabilistic neuralnetwork technique to lithology inversion of heteroge-neous stratum.Acta Petrolei Sinica,2008,29(4):549～552
[6]孙思敏,彭仕宓.基于模拟退火算法的地质统计学反演方法.石油地球物理勘探,2007,42(1):38～43Sun Simin,Peng Shimi.Inversion of geostatisticsbased on simulated annealing algorithm.OGP,2007,42(1):38～43
[7]印兴耀,刘永社.储层建模中地质统计学整合地震数据的方法及研究进展.石油地球物理勘探,2002,37(4):423～430Yin Xingyao,Liu Yongshe.Methods and develop-ment of integrating seismic data in reservoir model-building.OGP,2002,37(4):423～430
[8]王霞,张延庆,于志龙等.叠前反演结合地质统计模拟预测薄储层.石油地球物理勘探,2011,46(5):744～748Wang Xia,Zhang Yanqing,Yu Zhilong et al.Thinreservoir prediction by prestack inversion combinedwith geostatistic simulation.OGP,2011,46(5):744～748
[9]李建雄,谷跃民,党虎强等.高密度井网开发区井震联合储层预测方法——以SL盆地JN2区块为例.石油地球物理勘探,2011,46(3):457～462Li Jianxiong,Gu Yuemin,Dang Hu-qiang et al.Wellseismic and surface seismic joint prediction for reser-voir rock in a development area with denser produc-ers—Case study of Block JN2in SL Basin.OGP,2011,46(3):457～462
[10]刘振峰,郝天珧,杨长春.沉积模型和储层随机建模.地球物理学进展,2003,18(3):519～523Liu Zhenfeng,Hao Tianyao,Yang Changchun.Sedi-mentation models and stochastic reservoir modeling.Progress in Geophysics,2003,18(3):519～523
[11]董宁,刘振峰,周小鹰等.鄂尔多斯盆地致密碎屑岩储层地震识别及预测.石油与天然气地质,2008,29(5):668～675Dong Ning,Liu Zhenfeng,Zhou Xiaoying et al.Seis-mic detection and prediction of tight gas reservoirs inthe Ordos Basin.Oil&Gas Geology,2008,29(5):668～675
[12]Caers J,Ma X.Modeling conditional distributions offacies from seismic using neural network.Mathema-tical Geology,2002,34(2):143～167
[13]Caers J,Avseth P,Mukerji T.Geostatistical integra-tion of rock physics seismic amplitudes and geologicalmodels in North-Sea turbidite systems.The LeadingEdge,2001,20(3):308～312
[14]姚逢昌,甘利灯.地震反演的应用与限制.石油勘探与开发,2000,27(2):53～56Yao Fengchang,Gan Lideng.Application and re-striction of seismic inversion.Petroleum Explorationand Development,2000,27(2):53～56
[15]凌云,郭向宇,高军等.油藏地球物理面临的技术挑战与发展方向.石油地球物理勘探,2010,49(4):319～335Ling Yun,Guo Xiangyu,Gao Jun et al.The techni-cal challenges and the development trend of reservoirgeophysics.OGP,2010,49(4):319～335
[16]Humpson D P,Schuelke J S,Quirein J A.Use of multiattribute transforms to predict log propertiesfrom seismic data.Geophysics,2001,66(1):220～236
[17]Masters T.Signaland Image Processing with Neural Networks.New York:John Wiley&Sons Incorpora-tion,1994:1～200
[18]Pan Guocheng著;李钟山译;夏立显校.地质统计学中的区域化变量理论.世界地质,1997,16(2):85～93
[19]PanGuocheng著;李钟山,陈永良译;夏立显校.地质统计学中结构分析的理论与方法.世界地质,1997,16(3):70～82
[20]Pan Guocheng著;何凯涛,辜华良译;夏立显校.地质统计学中的估值技术和条件模拟.世界地质,1997,16(3):83～100
[21]陈建阳,田昌炳,周新茂等.融合多种地震属性的沉积微相研究与储层建模.石油地球物理勘探,2011,46(1):98～102Chen Jianyang,Tian Changbing,Zhou Xinmao et al.Sedimentary microfacies studies and reservoir model-ing by integration of multiple seismic attributes.OGP,2011,46(1):98～102
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