摘要
In hydrocarbon reservoirs, seismic waveforms become complex and the correlation dimension becomes smaller. Seismic waves are signals with a definite frequency bandwidth and the waveform is affected by all the frequency components in the band. The results will not define the reservoir well if we calculate correlation dimension directly. In this paper, we present a method that integrates empirical mode decomposition （EMD） and correlation dimension. EMD is used to decompose the seismic waves and calculate the correlation dimension of every intrinsic mode function （IMF） component of the decomposed wave. Comparing the results with reservoirs identified by known wells, the most effective IMF is chosen and used to predict the reservoir. The method is applied in the Triassic Zhongyou group in the XX area of the Tahe oil field with quite good results.
在储层发育带,通常地震波波形会变复杂,所以关联维会变小。但由于地震波是有一定带宽的信号,波形受这个带宽内所有频率成分的影响,如果直接进行关联维的计算,其结果往往不能很好地反映储层。鉴于这种情况,提出了经验模态分解(EMD)与关联维相结合的方法,利用EMD对地震波进行分解,对分解后的每一个平稳的固有模函数(Intrinsic Mode Function,IMF)分量进行关联维计算,将计算结果与有效的IMF分量进行储层预测。利用该方法对XX区三叠系中油组进行了试算,取得了较好的效果。
基金
sponsored by the National Nature Science Foundation of china(Grant No.40774064)
National Hi-tech Research and Development Program of China(863 Program)(Grant No.2006AA0AA102-12)
