摘要
压裂施工参数优化是水力压裂过程中的一个重要环节。为了精准设计压裂施工参数,达到节约开发成本和增产的目的,首先初选出影响水力压裂效果的参数,再用灰色关联分析法对参数进行排序,然后讨论参数个数对BP(back propogation)神经网络模型效果的影响,最终确定参数个数为10个。利用粒子群算法构建日产气量最优化模型,进而反演出最优压裂施工参数。应用于X区块的97口致密气井,BP神经网络模型准确率为86.52%。对7口井进行压裂施工参数优化后,每口井所有层的总平均增产率为5.57%。该方法具有投入成本低和操作简单等优点,为现场压裂设计提供借鉴和参考,具有一定的实际意义。
The optimization of fracturing parameters is an important link in the process of hydraulic fracturing.In order to accurately design fracturing operation parameters and increase production while saving development cost,the parameters affecting hydraulic fracturing effect were preliminarily selected.Then the grey correlation analysis method was used to sort the parameters.Afterwards the number of parameters was discussed by BP neural network model,result showed that the optimal number is 10.The particle swarm optimization algorithm was used to operate the optimization model of daily gas production,and then the optimal fracturing parameters were inversely obtained.Applied to 97 tight gas wells in block X,the accuracy of BP neural network model reached 86.52%.The total average production increase rate of 7 target wells was 5.57%after the optimization of fracturing parameters.The proposed method has the advantages of low investment cost and simple operation.It provides reference for field fracturing design and has certain practical significance.
作者
郭大立
唐乙芳
李曙光
张天翔
康芸玮
GUO Da-li;TANG Yi-fang;LI Shu-guang;ZHANG Tian-xiang;KANG Yun-wei(School of Sciences, Southwest Petroleum University, Chengdu 610500, China;China United Coal Bed Methane National Engineering Research Center Co., Ltd., Beijing 100095, China;CNPC Coal Bed Methane Co., Ltd., Beijing 100028, China)
出处
《科学技术与工程》
北大核心
2022年第19期8304-8312,共9页
Science Technology and Engineering
基金
国家科技重大专项(2016ZX05065,2016ZX05042-003)。
关键词
致密气
参数优选
BP神经网络
粒子群算法
压裂参数优化
tight gas
parameter optimization
BP neural network
particle swarm algorithm
optimization of fracturing parameters
