在油藏勘探开发中,利用地震波阻抗反演技术预测储层空间分布特征,自应用以来发挥着非常重要的作用.特别是在油藏开发中后期,需要对薄层、薄互层等油层进一步开发时,该项技术显得尤为重要.尽管该方法以测井和地质等资料为约束,但依据传统的构造解释方法建立的初始构造模型不够准确,仍然无法降低地震反演的多解性,井间储层分布与后验井存在较大差别,预测出的薄层、薄互层及储层岩性的可信度较低.本文利用测井资料进行高分辨层序地层划分,依据单井合成地震记录将划分结果标定在地震剖面上,并完成各层序界面的地震层位解释,井震结合建立高分辨率层序地层格架.结果表明,将该地层格架作为初始模型进行测井约束的地震资料波阻抗反演,对不同旋回发育的薄层或薄互层储层,横向预测不会以一套厚储层呈现,提高了储层反演预测储层分布的精度.通过大庆长垣高台子油田中的具体应用,展示了高分辨率层序地层格架约束下的波阻抗反演技术在薄层、薄互层储层预测中的广阔应用前景.
During the reservoirs exploration and development, the impedance inversion method based on seismic data could be adopted to predict the reservoir distribution. Especially in the stage of reservoirs development, it is necessary to know the thin layer and thin interbed reservoirs distribution and to take measures in the further development. In order to acquire accurate inversion results, it is important to build a refined initial model. Although the impedance inversion is based on log and geological data, the initial model built on the traditional structure interpretation results is not accurate enough, and is unable to reduce multi-solution problem of seismic inversion. The impedance inversion based on high-resolution sequence stratigraphy framework has been introduced. First the high-resolution sequence stratigraphy classification results could be calibrated in the seismic profile by the synthetic seismogram. Then every sequence interface must be interpreted in the seismic profile. As to the sequence interface that does not have the corresponding seismic wave in seismic profile, it could be traced according to the logging data. The initial model for the log-constrained impedance inversion has been constructed on the basis of the interpretation of high-resolution sequence framework. Finally the reservoirs distribution is able to be required from the inversion results. The method is successfully applied to the Gaotaizi Oil field of Daqing Placanticline. The predictive reservoirs have been confirmed by the later real drilling data. The case study indicates the wide application prospects for the inversion method constrained by high-resolution sequence framework in the prediction of the thin interbed reservoir.