地震波形分类技术在季节性河流沉积研究中的应用

Application of Seismic Waveform Classification in Research on Ephemeral Stream Deposits

针对前人对阿根廷库约盆地Cacheuta坳陷PCo区块Barrancas组季节性河流相砂泥岩薄互层沉积的物源方向、沉积相带展布特征等描述不清的问题,在分析该地区满足地震波形分类技术应用条件的基础上,通过确定合适的地震时窗类型和大小、波形分类数等,采用神经网络算法对反映不同沉积相的地震波形进行多次试验和分类,最终快速有效地预测出了季节性河流的多个物源方向和复杂沉积相带展布特征。预测结果与该区区域地质背景吻合,且与单井测井相匹配较好。在此基础上建立了季节性河流相沉积及演化模式,对油田下一步开发部署调整以及剩余油挖潜深入研究提供了重要指导。

As the unclear description of source direction and sedi- mentary facies belt distribution in the previous study of the ephem- eral stream sandstone and mudstone thin interbed deposits in Bar- rancas Formation of PCo Block in Cuyo Basin, Argentina, with the precondition that seismic waveform classification is analyzed to be feasible in this area, firstly, appropriate time window type and size, waveform classification numbers and etc. are determined, then dif- ferent waveforms which can reflect different sedimentary facies are classified by using neural network algorithm, after many times＇ trials, and eventually, the ephemeral stream with multiple source di- rections and complicated characteristics of sedimentary facies dis- tribution is rapidly and effectively predicted, which coincides with the regional geological setting. And furthermore, the logging facies of single well matches well with the predicted seismic facies. Bas- ing on the result, the ephemeral stream sedimentary evolution mod- els of this studied area have been built, which will provide signifi- cant guidance for the future oilfield development planning and re- maining oil tapping.