随机模拟和随机反演是两种比较常用的较精细的储层预测方法,而变差函数是随机模拟、随机反演的核心技术.以喇嘛甸油田北北二区块SⅢ层段为研究对象,选用300 m×300 m、600 m×600 m、800 m×800 m、1200m×1 200m不同变程变差函数,对随机模拟和随机反演结果对比分析,探讨不同变程对预测结果的影响.结果显示:随着变程的增大,随机模拟砂体结果由离散发展成连续片状分布,而随机反演砂体分布整体趋势一致,只有局部薄层砂体受到较小程度的影响；通过盲井验证,随机反演较随机模拟,其预测的砂体与井点的符合率更高.分析认为,随机反演较随机模拟多了井间地震的约束,减少了砂体分布的不确定性和多解性,提高了井间砂体的预测精度,便于在油田开发阶段推广应用.%Stochastic modeling and stochastic inversion are two commonly used reservoir prediction methods, and variogram is the core technology of stochastic simulation and stochastic inversion. With SIH Formation of North-North second block in the Lamadian Oilfield as the study object, the authors made a comparative analysis of the stochastic simulation and stochastic inversion on different variograms of 300 m ×300 m, 600 m ×600 m, 800 m x800 m and 1 200 m × 1 200 m, so as to probe into the effects of ranges of different variables on prediction results. The results show that, with the increasing range, the stochastic simulation results vary from the sand body to a continuous sheet of discrete distribution, and stochastic inversion sand distribution are consistent with the overall trend, with only a thin layer ofsand being affected in an insignificant degree. Blind shaft verification shows that stochastic inversion has higher consistency between predicted sand bodies and well points than stochastic simulation. An analysis reveals that relative to stochastic simulation, stochastic inversion is subjected to seismic constraints, thus reducing the uncertainty of sand distribution and multiple solutions and improving the prediction precision of inter-well sand bodies. Therefore, it deserves extension and application at the oilfield development stage.
展开▼
