基于遗传算法的CO_2驱最小混相压力预测  

CO_2-oil Minimum Miscible Pressure Prediction by Genetic Algorithm Method

作　　者：苏坤 廖新维[1] 赵晓亮[1] 卢宁[1] 窦祥骥[1] 

机构地区：[1]中国石油大学(北京)石油工程学院,北京102299

出　　处：《科学技术与工程》2013年第16期4524-4528,共5页Science Technology and Engineering

基　　金：国家基础研究项目(2011CB707302);国家科技重大专项项目(2011ZX05016-006)资助

摘　　要：最小混相压力是CO2驱替过程中判断混相与否的关键参数;该值的准确预测对CO2驱替区块的筛选具有重要意义。以7组细管实验数据为研究样本,结合国外对最小混相压力影响因素的研究进展,通过分析各因素对最小混相压力的影响规律,筛选了影响CO2驱最小混相压力的主要参数;并在确定了经验公式形式的基础上,采用遗传算法优化了公式参数。通过结果比较,证明新公式具有理想的预测精度,其对最小混相压力的预测值可以作为相应CO2驱区块筛选过程中的参考依据。Minimum Miscible Pressure （MMP） is among key parameters to predict CO2 flooding performance during candidate reservoirs screening. Key factors determining MMP are screened after investigation over previous researches, and then by seven samples of oil component, a reasonable empirical function is developed to predict MMP for reservoirs in China, of which coefficients are calculated by genetic algorithm （GA） to improve accuracy for further application. After comparison, it proves better which can be widely used for further MMP prediction for prediction by new function than the ones already existed, pilot reservoir in China.

关 键 词：遗传算法 因素分析 最小混相压力 经验公式 混相驱 

分 类 号：TE341[石油与天然气工程—油气田开发工程]