基于大数据的躺井预警及原因诊断方法研究  

Research on the Method of Well Lying Early Warning and Cause Diagnosis based on Big Data

作　　者：焦钰嘉 张利军 JIAO Yu-jia;ZHANG Li-jun(CNOOC Research Institute Co.,Ltd.,Beijing 100000,China)

机构地区：[1]中海油研究总院有限责任公司,北京100000

出　　处：《化工管理》2023年第2期159-161,共3页Chemical Engineering Management

摘　　要：实际开发生产过程中,躺井不仅影响产量,还增加作业成本,严重影响油田正常生产。特别是进入高含水期的老油田,躺井率居高不下,躺井原因分析工作量逐渐增加。伴随信息化技术及人工智能技术的发展,文章旨在建立躺井自动识别及躺井原因自动诊断体系,通过构建躺井识别业务模型,实现躺井自动判断,为躺井发现提供一种新途径,而不再只是依赖于现场巡检人员发现。应用躺井原因自动诊断大数据模型能够对已躺井进行原因自动判识,自动推送躺井原因,有效提高技术人员躺井原因分析效率,辅助油田实际工作。In the actual development and production process, lying well not only affects the production, but also increases the operation cost, which seriously affects the normal production of the oilfield. Especially in the old oil fields entering the high water cut period, the well lying rate remains high, and the workload of well lying cause analysis gradually increases. With the development ofinformation technology and artificial intelligence technology, this paper aims to establish an automatic identification system of lying wells and automatic diagnosis system of lying well causes. By constructing a business model of lying well identification, it can realize the automatic judgment of lying wells, and provide a new way for lying well discovery, which is no longer just dependent on the discovery of on-site inspectors. The application of big data model for automatic diagnosis of well lying causes can automatically identify the causes of lying wells, automatically push the causes of lying wells, effectively improve the efficiency of technicians’ analysis of well lying causes, and assist the actual work of the oilfield.

关 键 词：躺井识别 神经网络学习 躺井原因 

分 类 号：TE345[石油与天然气工程—油气田开发工程]