摘要
川西新场气田侏罗系沙溪庙组气藏是典型的低渗透致密碎屑岩气藏,由于地质条件复杂,非均质性极强,储层单井产能预测已成为难点之一。通过分析测井信息和储层物性,提取声波时差、裂缝张开度、裂缝渗透率等11个与产能相关的参数,并运用相关系数法确定裂缝张开度、裂缝渗透率、产能系数、裂缝孔隙度、裂缝发育指数和综合评价指数6个指标作为产能控制特征参数;然后选取33个已测试层位作为已知样本,利用支持向量机构建适合该区的产能预测方程,预测标准误差为0.0661,平均绝对误差仅为0.0182,预测精度较高;最后,利用所构建的产能方程对24个未测试层位的产能进行预测,完成沙溪庙组气藏单井产能评价。
The gas reservoirs of Jurassic Shaximiao are typical low-permeability tight clastic ones. Formation in the Xinchang Gas Field ot West bichuan Due to complex geological conditions and strongly heterogeneous characteristics in this area, it is difficult to predict single well productivity. By analyzing logging information and reservoir physical property, we first extract 11 parameters related to productivity, such as interval transit time, fracture aperture, fracture permeability, etc. Second, with correlation coefficient method, 6 productivity controlling characteristic parameters are determined, including fracture aperture, fracture permeability, productivity coefficient, fracture porosity, fracture developing index and composite evaluation index. Third, we select 33 tested wells as known samples, construct productivity forecast equation with the theory of support vector machine, and then get the standard error of 0. 066 1 and the mean absolute error of 0. 018 2. In the end, with the equation, we predict 24 untested wells to complete single well productivity evaluation of Shaximiao Formation.
出处
《石油实验地质》
CAS
CSCD
北大核心
2009年第4期338-342,共5页
Petroleum Geology & Experiment
基金
四川省学术和技术带头人培养基金(川人办发[2008]24号)
四川省教育厅人文社科重点项目(2008ZB026)
成都理工大学商学院自立项目(Sxyzc08-09)
关键词
产能评价与预测
支持向量机
致密碎屑岩
沙溪庙组
新场气田
productivity evaluation and prediction
support vectors machine
tight clastic rock
Shaximiao Formation
Xinchang Gas Field
作者简介
石磊（1958-），男，高级工程师，主要从事石油和天然气勘探工作。E-mail：shil@edut．edu．cn。
