摘要
页岩油水平井生产动态变化规律复杂,常规方法难以实现全生命周期的生产动态预测。为此,以吉木萨尔页岩油水平井为例,通过分析一口典型水平井自喷期生产动态曲线,将生产过程划分为焖井阶段、快速上产阶段、产量快速递减阶段和低产稳产阶段等4个阶段,并以此为基础,引入见油时刻、峰值产油量、峰值时间及平均递减率4个特征参数,推导建立页岩油水平井定量化表征模型,结合SPSA算法,实现全生命周期生产动态曲线的拟合、预测与参数解释反演。研究结果表明,峰值产油量和平均递减率对生产动态曲线影响较大,而见油时刻和峰值时间影响较小。运用提出的定量化表征模型及曲线拟合优化算法,进行了实例应用,证明了模型的适用性,也表明该模型可作为一种切实可靠的方法为现场提供页岩油水平井生产动态预测。
The production performance of shale oil horizontal wells is complex,and it is difficult to predict the production performance of the whole life cycle through conventional methods.Therefore,taking the shale oil horizontal well of Lucaogou Formation in Jimsar as an example,the production process is divided into four stages by analyzing the production performance curve of a typical horizontal well in the blowout period,including soaking stage,up production stage,production decline stage and low and stable production stage.The quantitative characterization model of shale oil horizontal well is deduced and established.Combined with SPSA algorithm,the whole life cycle production performance curve fitting,prediction and parameter interpretation inversion are realized.The results show that the peak oil production and average decline rate have a great influence on the production performance curve,while the oil breakthrough time and peak time have little influence.Finally,the application of the quantitative characterization model and curve fitting optimization algorithm is carried out to prove the applicability of the model;it also shows that the model can be used as a practical and reliable method to predict the production performance of shale oil horizontal wells.
作者
陈依伟
周玉辉
梁成钢
徐田录
何永清
CHEN Yiwei;ZHOU Yuhui;LIANG Chenggang;XU Tianlu;HE Yongqing(Jiqing Oilfield Operation Area,Xinjiang Oilfield Company,PetroChina,Karamay,Xinjiang 834000,China;School of Petroleum Engineering,Yangtze University,Wuhan,Hubei 430100,China)
出处
《西南石油大学学报（自然科学版）》
CAS
CSCD
北大核心
2021年第5期97-103,共7页
Journal of Southwest Petroleum University(Science & Technology Edition)
关键词
页岩油
水平井
生产动态
表征模型
产量预测
曲线拟合
shale oil
horizontal well
production performance
characterization model
production prediction
curve fitting
