论文部分内容阅读
地震反射层深度偏移的准确成像要求有精确的速度模型。在几乎没有井约束的边缘地区,速度估算一般运用了一些象正常时差分析、地震波传播时间层析或迭代叠前深度偏移等常规方法。这些方法很有效,但同时有可能费用较高或较为费时。我们从一系列与地震反射面相交的油井中获取了有关地层顶部的资料,在这些情况下,我们运用最小平方最优化方法来估算速度模型。这种方法产生的速度模型能够使井约束方面的深度偏移通过最小平方反演得到最优化并使深度偏移成像与地层顶部相匹配。测井资料用于叠后偏移最优化上,从而缩减了速度分析的时间和费用。除了应用使地层顶部深度偏移最优化的反演方法之外,我们还能运用“最高平方反演”的灵敏度分析法来确定一系列的速度模型使之能够提供数学上可以接受的解决方法。灵敏度分析对估算厚地层速度要优于估算薄地层速度这一预期结果进行了量化。我们所提出的最优化方法在纽芬兰海岸的Hibernia油田的合成和实际资料中得到了成功的验证。
Accurate imaging of the depth of seismic reflection layer requires an accurate velocity model. In the marginal areas where there is almost no well constraint, velocity estimation generally uses some conventional methods such as normal time difference analysis, seismic wave propagation time tomography or iterative prestack depth migration. These methods are effective, but at the same time may be more costly or time-consuming. We obtained data about the top of the formation from a series of wells that intersected the seismic reflection surface. In these cases, we applied the least squares optimization method to estimate the velocity model. The velocity model produced by this method optimizes the depth offset in well constraints by least-square inversion and maps the depth migration to the top of the formation. Logging data is used to optimize poststack migration, reducing the time and expense of velocity analysis. In addition to using inversion methods that optimize depth drift at the top of the formation, we can use the “highest square inversion” sensitivity analysis to determine a range of velocity models that will provide a mathematically acceptable solution. Sensitivity analysis quantifies the expected result of estimating thick formation velocity over estimating thin formation velocity. The optimization method we proposed has been successfully validated in the synthesis and actual data of the Hibernia field off the coast of Newfoundland.