石油学报 ›› 1987, Vol. 8 ›› Issue (1): 11-19.DOI: 10.7623/syxb198701002

• 地质勘探 • 上一篇    下一篇

生油门限统计预测新模型

张启锐   

  1. 中国科学院地质研究所
  • 收稿日期:1985-06-05 出版日期:1987-01-25 发布日期:2013-07-08

New Statistical Prediction Model for Hydrocarbon Genesis Threshold

Zhang Qirui   

  1. Institute of Geology, Academia Sinica, Beijing, China
  • Received:1985-06-05 Online:1987-01-25 Published:2013-07-08

摘要: 在大量生油门限的预测方法中,统计预测法还末受到广泛的重视,它是以一定数量的已知数据为基础,通过一定的办法建立经脸表达式,然后应用于实际。Connan(1974)咨次建立了生油门限温度与生油岩地质年龄之间的函数关系,礁在此基础上,于志钧(1981,1983)做了一些补充和改进,井通过逐步回归分析,得到了一个新的表达式,其多重相关系数为0.931,比Connan的方程高。在实际应用中,Connan和于志钧的公式部不人方便,因为共因变量都是生油岩年龄值的对数。事实表明,若以生油岩的现今地层温度为因变量,则所得摸型将更为有效合理。对于新的模型,笔者采用了较先进的叠代变换回归法(Bo和Tidwell,1962)进行拟合,产生了比于志钧公式更高的拟合度。用门限已知的其他一些盆地数据对方程进行检验,结果是令人鼓舞的。预测的门限地温与实测值最大误差只有4℃。

Abstract: Among the methods of hydrocarbon generation threshold prediction, the statistical method is the one not yet been fully reciated.Based on certain number of known data, an empia-ical expression is developed and applied to interested basins to predict their threshald values, which are important information at the early stage of hydrocarbon exploa atian and development.The functional relationships between temperature of hydrocarbon generation threshold and age values of source rocks were first suggested by Connan(1970.Based on Connan's equation and data, Yu Zhijun (1981,1983) established a new expression by using stepwise-regression analysis method to enlarged data set. The coefficient of multiple correlation equals 0.931, which is better than that of Connan's equation.