【摘要】 准噶尔盆地车排子地区超覆在石炭系之上的古近系,聚集丰富的岩性油藏资源,由于受石炭系强振幅的影响和地震资料分辨率的制约,常规的地震属性无法有效刻画储层的纵横向分布。针对此问题,利用改进的广义S变换的谱分解技术对地震资料进行时频分析,结合已有钻井、测井资料精细标定层位、录井显示和测井解释等综合分析,优选出70Hz的分频剖面识别该地区的岩性油藏薄储集层。储层预测结果与现有钻井资料都非常吻合。据此研究成果又部署了一口评价井,并获得了工业油流,进一步表明谱分解技术适合该地区古近系的油气储层预测,也可为其他地区类似的岩性油藏储层预测提供思路。更多还原

【Abstract】 Lithologic reservoir resources are rich in Carboniferous near the top of the Paleogene Chepaizi area,Junggar Basin.But due to the impact of strong amplitude in Carboniferous and limited resolution of seismic data,the conventional seismic attributes cannot effectively portray the vertical and horizontal distribution of the reservoirs.For this issue,spectral decomposition techniques are used to identify thin lithologic reservoirs in the study area.Improved general S transform spectral decomposition techniques are used to perform frequency division of seismic data.Combined with results of a comprehensive analysis,such as the fine horizon calibration of existing drilling data,logging presentation and log interpretation,the 70 Hz frequency division section is preferred for reservoir prediction.Results of reservoir prediction are consistent very well with the existing drilling data.They are used to deploy the evaluation well C55-1,which has obtained industrial oil.This further shows that spectral decomposition techniques are effective reservoir exploration techniques in the region with Paleogene Formations and can also be used for thin lithologic reservoirs in other areas.更多还原