多子波分解与重构法砂岩储层预测

鄂尔多斯盆地北部大牛地气田的天然气藏是大型岩性圈闭气藏，储层具有岩性致密、非均质性强的特点，由于地震资料分辨率较低、砂泥岩波阻抗差微弱、煤层反射振幅过强，造成砂体的地震响应特征不明显，储层预测难度极大。通过对叠后地震数据的多子波分解与重构，也就是将地震道分解成不同主频的子波集合，将反映储层频率成分段的子波进行叠加，用于储层含气性预测，在含气砂岩储层预测中取得了良好的效果。

Prediction of Sand Reservoir with Multi-Wavelet Seismic TraceDecomposition and Reconstruction

Daniudi Gasfield in Ordos Basin is typically a lithologic gas reservoir with tight and anisotropic characteristics.There is no obvious seismic response of sand body due to low resolution，small difference of acoustic impedance between sandand shale，and strong reflection events from coal. Therefore，it is very difficult to predict the reservoirs in the area. Using themulti-wavelet seismic trace decomposition and reconstruction，each post-stack seismic trace is decomposed into a set of Rickerwavelets of different dominant frequencies. A new seismic volume is then reconstructed with wavelets within a selected rangeof dominant frequencies that reflect the reservoirs and used to predict the reservoirs and gas-bearing property.