有机地球化学参数总有机碳含量预测

Prediction of Total Organic Carbon Content in Organic Geochemical Parameters

页岩气储层中总有机碳含量(TOC)反映了页岩的生烃潜力,连续准确获取页岩气储层TOC含量对页岩气的勘探开发具有重要意义。本文结合威远地区龙马溪组页岩的实际情况,建立了多元回归曲线法、ΔlogR法、支持向量机回归算法三种预测模型,并对该地区W1井页岩进行了TOC含量评定。结果表明:在威远地区采用支持向量机回归算法模型其精度最高,可以为页岩的TOC含量预测提供指导。

The total organic carbon content (TOC) in shale gas reservoir reflects the hydrocarbon generation potential of shale. Continuous and accurate acquisition of TOC content in shale gas reservoir is of great significance for shale gas exploration and development. Combined with the actual situation of Longmaxi formation shale in Weiyuan area, this paper establishes multiple regression curve method, ΔLogR method and support vector machine regression algorithm that are used to evaluate the TOC content of shale in well W1 in this area. The results show that the precision of support vector machine regression model is the highest in Weiyuan area, which can provide guidance for the prediction of TOC content of shale.