【摘要】 概率密度函数随机地震反演不需繁杂的模型初始化,能充分考虑地下地质的随机特性,使反演结果更符合实际地质情况,对于钻井较少的工区及储层非均质性较强的储层更具优势。其引用测井资料参与反演约束,同时通过对测井资料概率密度函数和波阻抗可能分布范围的求取,再在这两个条件约束下,求取满足以上随机特征又满足地震响应的波阻抗序列来实现反演;解决了随机模拟地震反演的变差函数在井资料较少、甚至在工区内井分布不均匀的情况下变得极不稳定,而使反演结果可靠性不高的问题,能较为准确地预测出储层低速异常带的分布。通过在蜀南嘉陵江组储层预测中的成功应用,表明该方法在薄储层且勘探初期钻井资料较少、现今流行的叠后反演方法不能准确预测储层的条件下,取得了较好的应用效果。更多还原

【Abstract】 Without complicated model initialization, the stochastic inversion of probability density function (PDF) takes fully geologic features into consideration and makes the inversion results be in close agreement with actual geological situation. In addition, the inversion is more superior in some studied area with a little drilling and some reservoirs with strong heterogeneisty. In this inversion, well-logging data is involved in the inversion constraint, the PDF for well-logging data and the possible distribution scope of wave impedance are obtained simultaneously; after that,some results, which are satisfactory with both stochastic features and wave-impedance sequence of seismic response, are gained to realize an inversion. Moreover, the stochastic seismic inversion makes variogram become stable under the conditions of a little well data and an irregular well distribution, and achieves a higher reliable inversion result, and may predict more accurately the distribution of low-velocity abnormal zone. It has not only been applied to the reservoir prediction in Jialingjiang Formation of southern Sichuan Basin very successfully, but also obtained a better result in such domains as thin reservoirs with a little data during an initial exploration, which may not be predicted by post-stack inversion.更多还原