叠前地震Q值反演在储层流体识别中的应用研究

Application of pre-stack seismic Q value inversion in reservoir fluid recognition

相较于常规的叠后资料Q值估算方法,本文的Q值反演法充分利用了叠前地震资料所具有的丰富振幅、频率和走时信息.首先简化黏弹性介质纵波反射系数近似,利用贝叶斯理论,建立反演方程,然后通过加入柯西约束和模型约束对算法进行优化,增加反演细节和横向连续性,提高反演精度和可靠性,最后将算法应用胜利油田东部探区的复杂岩性油气藏中,同时反演纵横波速度、密度和地层吸收衰减参数Q,成功地区分了强振幅引起的"真假亮点",识别了储层的含油气性,弥补了传统方法的不足.

Compared to conventional Q estimation methods based on post-stack data,the Q value inversion method makes full use of abundant amplitude,frequency and travel information in pre-stack seismic data in this article. First simplified viscoelastic medium longitudinal wave reflection coefficient approximation, and constructed inverse equation based on Bayesian theory. Then,by adding Cauchy constraints and model constraints optimized algorithm, which increased the inversion details and lateral continuity and improved retrieval accuracy and reliability. Finally,the algorithm was used to complex lithological reservoirs in eastern exploration region in Shengli oilfield, which, simultaneously inversed longitudinal velocity,shear velocity,density and formation attenuation parameter Q, successfully distinguished the strong amplitude caused by the ＂genuine bright spots ＂,identified the shortcomings of hydrocarbon potential,and maked up the traditional methods of the reservoir.