非正交小波变换的频谱分解技术在大牛地气田储层预测中的应用

鄂尔多斯盆地北部上古生界是陆相致密的隐蔽性气藏，具有低孔、低渗、低压、多层叠置的特点。由于储层与围岩的波阻抗差异小和薄互层多，储层预测成为目前该盆地北部地区勘探开发的主要难题之一。为提高储层预测的精度，开展了基于地震频谱分解技术的储层预测技术研究。Gabor Morlet小波变换的谱分解技术直接使用频率参数控制，克服了常规频谱分解技术的局限，适用非平稳地震信号的分析，能更有效地突出信号的局部特征，具有可逆性和非正交性。应用Gabor Morlet小波变换的频谱分解技术，对该盆地北部大牛地气田D16井区下石盒子组2+3段储层进行地震频谱成像，其成果展示了河道砂体的非均质空间展布特征，与实钻情况吻合，证实了该方法的有效性。

APPLICATION OF SPECTRUM DECOMPOSITION OF NONORTHOGONAL WAVELET TRANSFORMS TO RESERVOIR PREDICTION IN DANIUDI GAS FIELD

The reservoir in Upper Palaeozoic of North Ordos Basin is compacted continental hidden reservoir characterized by lower porosity, lower permeability, lower pressure and multi layer overlapping. Because little difference of acoustic impedance exists between the reservoir and surrounding rocks, along with many thin interbeddings, reservoir prediction is one of the major difficulties for the exploration and exploitation at this area presently. In order to improve the reservoir prediction precision, the paper developed the research on reservoir prediction techniques based on seismic spectrum decomposition. The spectrum decomposition by the Gabor Morlet wavelet transform is directly controlled by frequency parameter, which overcomes the limits of the conventional spectral decompositions. This method, being reversible nonorthogonal wavelet transforms, can not only be applicable for analyzing unstable seismic signal, but also effectively display the local characteristics of the signal. By using spectrum decomposition technique of the Gabor Morlet wavelet transform, the paper performed spectrum imaging for the reservoir in sublayer 2+3 in Upper Palaeozoic Shihezi formation at Daniudi gas field. The imaging results which displayed heterogeneous spatial distribution properties of the channel sandstone were found out to be well consistent with the actual drilling data, and this verified the effectiveness of this method.