用数据挖掘技术优选侧钻井井位

DATA EXCAVATING IN DETERMINING OF SIDETRACKING WELL LOCATION

张勇斌 梁荣华 马玉书

石油大学人工智能研究中心, 北京昌平 102200

Artificial Intelligence of Petroleum University

运用数据挖掘技术对油井实施开窗侧钻效果进行预测,以辅助侧钻井井位优选决策,为科学、客观地进行报废井侧钻提供依据,从而达到油田高效挖潜目的。介绍了数据挖掘技术的有关理论及算法,选取已实施侧钻的油井相关数据作为数据源训练神经网络。同时,预测了5口井的侧钻效果,经现场实施后,预测和实际效果基本吻合。通过实例及其结果的分析,证实了该方法新颖、简便、费用低、准确度高,易于推广。

The data excavating technology can be used for predicting the efficiency of oil well sidetracking, and assisting the determination of a proper sidetracking well locating, thus provides a scientific and reliable basis for sidetracking the abandoned well. The related theories and calculating methods are presented, and the data from the sidetracked wells is used for training the nerve network. It is used to predict the sidetracking of five wells, and the results are identical with the and field operations. This technology proves to be a novel, simple and accurate method.

10.3969/j.issn.1000-7393.2001.04.003