深层火山岩储集层定量预测方法的探讨

Discussion on quantitative prediction of deep valcanics reservoir

在探井资料有限的研究区,如何充分利用邻区的钻井、地震资料,对深层火山岩储集层作定量预测,本文结合实际对比做了有益的探讨,并形成了一套定量预测方法。在经过地质、地震可行性综合分析的前提下,先将有效趋势分析与机理分析相结合,选取特征地震参数,然后通过相关分析、校正等手段,将相邻测区地震资料相关信息作动态合并处理,再应用BP神经网络进行储层厚度预测。该方法在FS-9井区侏罗纪末期火山岩储集层厚度定量预测中,取得了令人满意的效果,经后验探井检验,符合很好。

In a research area limited in prospect well data,how to fully use drilling data and seismic data in an adjacent area to do quantitative prediction of deep volcanics reservoir, combining the practice we made an useful approach in the paper and formed a sort of quantitative prediction method. In a precondition of that geologic and seismic integrated analysis is feasible, using combination of effective trend analysis and mechanism analysis to select characteristic seismic data and through such tools as correlation analysis and correction,etc. ,the dynamic combination processing of relative seismic data in an adjacent area has been done,then using related BP neural network to carry out reservoir prediction thickness. The method has been used in quantitative thickness prediction of late Jurassic valcanics reservoir in well Fangshen 9 and achieved a good results which are consistent with post-examination drilling data.