【摘要】 在对准噶尔盆地腹部石南21井区岩性地层油气藏研究中,将层序地层学的研究成果与地球物理中的速度分析相结合,以高精度层序地层格架为指导,对层序地层格架内的地层进行速度研究,目的在于预测出储层内部有利砂岩发育区及储层埋深,为井位论证提供重要的基础资料.首先以岩心资料、测井资料和地球化学资料为基础,确定层序界面识别标志,通过单井相划分、合成地震记录制作,在地震剖面上识别层序界面,通过层序对比和追踪,建立起研究区等时地层格架,并以此作为速度研究的模型层位,利用模型迭代层速度反演法,有针对性地进行层序地层格架内部的层速度分布规律研究.结果表明,研究目标层序白垩系清水河组(CX0-CX3)的层速度分布规律是:南及东南部砂体发育区地层速度高,北部泥岩相对发育区地层速度低,速度的高低反映了砂岩含量的多少;速度图上高速区和低速区的平面分布形态,与属性图的平面分布范围和形态基本吻合,振幅属性图上的异常体是高速异常体,速度与振幅类地震属性在反映储层非均质性和砂体横向变化方面是相互印证、相互补充的.此外,在层序格架下的层速度反演基础上得到的目标层序CX3界面的平均速度图,其系统误差和随机噪音的影响小,时深转换所得CX3界面构造图等值线走势合理,即使井资料较少,也能得到形态正确精度较高的构造图.模拟勘探评价5个阶段的误差分析表明,钻前预测误差不超过0.33%,有效降低了岩性地层油气藏勘探的深度设计误差和风险.更多还原

【Abstract】 In this paper,we present a case study where we combine sequence stratigraphiy with seismic velocity study for a 3D seismic survey landing at the SN21 well area in Junggar Basin,China.The main trap in this area is subtle composite trap of structure and lithology.Our purpose is to provide essential maps for demonstrating well location by predicting reservoir’s lateral distribution and its depth.Firstly,according to the principle of sequence stratigraphy and the correlation analysis of well logs and seismic data,we constructed high-resolution stratigraphic frameworks in this subtle-reservoir area.Each framework has its given geological meaning.The three main frameworks are CX0(reflection of QSH top),CX3(reflection of TTH top),and CX5(reflection of TTH bottom).Secondly,by using ray tracing techniques,we merged these sequence stratigraphic frameworks with the seismic rms velocities,as well as controlled logging and geological constraints,to estimate seismic interval velocity model of these stratigraphic frameworks.Study results demonstrate that,within target strata(CX0-CX3),the south and southeastern sand distribution area shows high velocity values,while the north mudstone low velocity values.The magnitude of velocity reflects the sand content.Study results also prove that the seismic velocity and the seismic amplitude attribute corroborates each other in predicting sand distribution.Furthermore,the estimated final velocity model for time to depth conversion has few systemic and random errors,and the predicted structure maps with this velocity model turn out to be of relatively high precision and coincide with the subsequent drilling data.更多还原