【摘要】 为了充分挖潜窄小河道砂体剩余油,以大庆油田北二西为例,在精细等时地层格架建立的基础上,用波形聚类、相干体、分频、属性分析等地震预测技术对窄小河道砂体预测适应性进行了分析。通过分析发现,不同技术对窄小河道砂体边界识别程度是不同的,波形聚类可以大致判断河流方向,对于窄小河道边界识别不明显;相干体对于窄小河道边界的识别好于波形聚类;分频可以清楚地看到窄小河道砂体边界细微变化。通过对RMS属性分析,建立了薄互层砂体预测厚度与砂体实际厚度的关系,并利用后验井进行了验证。对于厚度>2 m的砂体,预测精度高,对于厚度为1～2 m的砂体,在预测相对误差<35%时,预测精度可以达到60%,对于厚度<1 m的砂体,属性预测不可靠。更多还原

【Abstract】 Seismic prediction technologies for narrow channel sand body are analyzed such as waveform clustering,coherence cube,frequency division and attribute analysis in order to produce remaining oil of narrow channel sand body,taking Beierxi area of Daqing oilfield as an example,on the basis of isochronous stratigraphic framework. It is found that different prediction technologies have different results to narrow channel sand bodies. Waveform clustering can approximately determine direction of narrow channel,but it can not distinguish narrow channel boundary obviously. Coherence cube is more superior than waveform clustering in narrow channel boundary recognition. Frequency division can clearly show narrow channel boundary subtle change. The relationship between the predicted sand body thickness and the actual sand body thickness is established through the analysis of RMS attribute,and it is verified that prediction accuracy of sand body is different. The prediction accuracy is high when sand body thickness is more than 2 m and the prediction accuracy can reach 60% on the condition that relative prediction error is less than 35% when thickness of sand body is between 12 m. The prediction accuracy of attribute is not reliable when thickness of sand body is less than 1 m. This prediction method has important significance in the production of remaining oil of narrow channel sand body.更多还原