歧北斜坡滩坝砂体地震预测技术  

Seismic prediction technology for beach bar sand-body in Qibei slope

作　　者：郭淑文[1] 吴雪松[1] 祝文亮[2] 邢兴[1] 国春香[1] 

机构地区：[1]中国石油大港油田勘探开发研究院,天津300280 [2]中国石油大港油田勘探事业部,天津300280

出　　处：《物探化探计算技术》2015年第5期628-633,共6页Computing Techniques For Geophysical and Geochemical Exploration

基　　金：国家重点基础研究发展计划项目(2013CB228606)

摘　　要：歧北斜坡沙二段滩坝砂体储层预测面临薄互层地层地震分辨率低,灰质砂岩与常规砂岩地震反射特征相似难以预测两个难点。通过试验优选,形成了两项针对性的地震储层预测方法。针对地震资料分辨率低,薄砂层预测困难的问题,采用谱分解和地震属性相结合方法,在分频体上提取地震属性,提高薄砂层识别能力。针对高灰质含量的滩坝砂体难以预测的问题,探索了利用地震属性降维和数据融合技术剔除灰质的方法。实际应用表明,谱分解技术和多地震属性降维技术,能够较好地预测歧北斜坡滩坝砂体的薄砂层,并在滩坝砂主砂体发育区的预测中,剔除了灰质的影响。Seismic reservoir prediction method was studied for beach bar sand-body in Qibei slope aiming at low seismic resolution of thin sand bed and similar reflection character for conventional sand and lime-sand.Two corresponding seismic reservoir prediction methods were formed through the experimental optimization.For the problem of thin bed and low resolution in the conventional seismic profile,combining the spectral decomposition and seismic attributes to improve the resolution ability of beach bar sand-body though deriving seismic attributes from separate frequency data basing seismic modeling.Attribute dimension reduction and data fusion techniques were tested to strip the lime sand to resolve the problem of prediction for high lime content sand.The field test showed that spectral decomposition and seismic attribute reduction techniques can be used to predict thin beach bar sand in Qibei slope and strip the influence of lime during prediction of main beach bar sand.

关 键 词：滩坝砂体 谱分解 灰质剔除 奇异值分解 属性降维 数据融合 

分 类 号：P631.4[天文地球—地质矿产勘探]