文章检索
首页» 过刊浏览» 2024» Vol.9» lssue(2) 282-296     DOI : 10.3969/j.issn.2096-1693.2024.02.020
最新目录| | 过刊浏览| 高级检索
油气钻采数字孪生模型构建方法及应用案例
林伯韬, 朱海涛, 金衍, 张家豪, 韩雪银.
1 中国石油大学( 北京) 信息科学与工程学院/ 人工智能学院,北京 102249 2 中国石油大学( 北京) 石油工程学院,北京 102249 3 中海油能源发展股份有限公司工程技术分公司,天津 300452
Modeling approach and case studies of digital twin in drilling and production of oil and gas fields
LIN Botao, ZHU Haitao, JIN Yan, ZHANG Jiahao, HAN Xueyin.
1 College of Information Science and Engineering/College of Artificial Intelligence, China University of Petroleum-Beijing, Beijing 102249, China 2 College of Petroleum Engineering, China University of Petroleum-Beijing, Beijing 102249, China 3 CNOOC EnerTech-Drilling & Production Co., Tianjin, 300452

全文:   HTML (1 KB) 
文章导读  
摘要  油气钻采过程中地质的不确定性、井下实时工况的不可见性、工程仿真的复杂性阻碍了其科学高效的设计及施工。数字孪生技术能够提供实时智能且可视化的方案设计和工程决策,但缺乏针对油气钻采的系统建模方法。对此,本文首先剖析油气钻采数字孪生的国内外研究及应用现状,进而应用成熟度指标定量评价该技术的发展程度;其次,逐次提出油气钻采数字孪生模型的建模方法,包括建模流程、拆分策略、装配及融合架构、建模工具,并以钻井井壁稳定和海上生产系统为例,介绍数字孪生在钻井与开采方面的应用案例;最后,分析困难与挑战并提出发展建议。研究发现,相对制造业,钻采孪生多处于可视化阶段,整体成熟度偏低。油气钻采系统的复杂需求被拆分为若干清晰且较容易实现的子需求;基于需求分析将建模对象在粒度、维度、生命周期上拆分为不同的子模型,通过模型层、功能层、需求层逐层装配子模型,进而实现多维度、多领域模型间的融合。同时,需要在模型管理、数据管理和工程仿真方面完善方法和提高效率。此外,钻采孪生面临多源异构数据选择与融合困难、子模型定义模糊、模型验证不清的问题,以及复杂动力学过程、多部门多任务协同、自主软件工具开发方面的挑战。综上,本文提出的数字孪生模型构建方法和案例能为油气钻采工程提供方法指导和应用参考。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
关键词 : 油气,钻井,开采,数字孪生,数据科学,人工智能
Abstract

The uncertainty of geological composition, the invisibility of the under-well real-time working conditions, and the complexity of the engineering simulation in the oil and gas field drilling and production process have hindered its scientific and efficient design and construction. The digital twin technology can bring up real-time, intelligent, and visualized project design and decision-making but has yet to lack a systematic method for modeling oil and gas field drilling and production. In this regard, the article first explored the current levels of investigation and implementation both domestically and abroad, based on that the level of development by applying the maturity index was quantified. It then proposed the digital twin modeling approach for drilling and production in the oil and gas field, which encompassed the modeling workflow, model division strategies, architecture for model assembly and integration, and modeling tools for constructing the digital twin. Also, two case were studied for drilling and production, using wellbore stability while drilling and offshore gas well production system as two examples, respectively. Finally, the difficulties and challenges related to the digital twin deployment in the field were analyzed, based on which the suggestions for its future development are proposed. It is found that the digital twin for drilling and production has stayed at the visualization level and at a relatively low degree of maturity compared to the manufacturing field on digital twin. The complex demand for oil and gas drilling and production systems can be divided into several clear and easy realized sub-demands. Based on requirement analysis, the modeled object can be separated to be various sub-models based on the granularity, dimension, and lifecycle. The sub-models are then assembled layer by layer across the model, function, and demand layers so that the multi-dimension and multi-field models can be integrated. Meanwhile, an improvement of their methods and an increase in efficiency for the model administration, data management, and engineering simulation ae desired. Moreover, the digital twin faces the problems such as difficulty in selection and fusion of multi-source heterogeneous data, vagueness in the sub-model definition, and ambiguity in the model validation, as well as the challenges such as the complicated kinetics processes, multi-division and multi-task collaboration, and development of domestic software tools. In summary, the digital twin modeling approach and the case studies in this article can provide a methodological guidance and practical reference for oil and gas drilling and production practices.


Key words: oil and gas; drilling; production; digital twin; data science; artificial intelligence
收稿日期: 2024-04-30     
PACS:    
基金资助:国家自然科学基金面上项目“砾岩储层砾石—交界面—基质合压水力裂缝非平面扩展机制研究”(42277122) 资助
通讯作者: linb_cupb@163.com
引用本文:   
林伯韬, 朱海涛, 金衍, 张家豪, 韩雪银. 油气钻采数字孪生模型构建方法及应用案例. 石油科学通报, 2024, 02: 282-296 LIN Botao, ZHU Haitao, JIN Yan, ZHANG Jiahao, HAN Xueyin. Modeling approach and case studies of digital twin in drilling and production of oil and gas fields. Petroleum Science Bulletin, 2024, 02: 282-296.
链接本文:  
版权所有 2016 《石油科学通报》杂志社