摘要：Virtual Flow Meters (VFMs) have garnered considerable attention due to their substantial benefits in curtailing the expenses associated with multiphase flow measurement. To overcome the constraints of current technologies, we introduce the Distributed Hybrid Virtual Flow Meter (HVFM-Fleet) method. This innovative approach leverages a data-driven model that integrates multi-scale features, significantly enhancing the accuracy of flow characteristic identification. Moreover, a mechanical model is crafted to translate arbitrary well networks into a virtual layout, thereby amplifying the system's adaptability to a variety of scenarios. The HVFM-Fleet capitalizes on distributed computing to reduce computational costs in large-scale simulations, establishing a database that surpasses a thousand multi-well cases鈥攖he most extensive in the transient multiphase flow simulation sector. Case studies integrate simulation data with actual measurement data from oil fields in the North Sea of the UK, with a detailed analysis of key elements such as individual wells, multiple wells, and flow allocation. The HVFM-Fleet has notably reduced the mean absolute percentage error and achieved a remarkable 140% improvement in performance compared to existing models. Furthermore, this research pioneers new metrics for assessing the volatility of predictions, setting a new standard in the evaluation of flow measurement reliability. 漏 2024 Elsevier B.V.

ISSN号：2949-8910

卷、期、页：卷243

发表日期：2024-12-01

影响因子：0.000000

期刊分区（SCI为中科院分区）：二区

收录情况：SCI（科学引文索引印刷版）,EI（工程索引）,SCIE（科学引文索引网络版）

发表期刊名称：Geoenergy Science and Engineering

参与作者：王彪

通讯作者：马赫,朱志勇,项小龙

第一作者：韩国庆,梁星原

论文类型：期刊论文

论文概要：马赫,韩国庆,朱志勇,王彪,项小龙,梁星原，Hybrid virtual flow metering on arbitrary well patterns for transient multiphase prediction driven by mechanistic and data model，Geoenergy Science and Engineering，2024，卷243

论文题目：Hybrid virtual flow metering on arbitrary well patterns for transient multiphase prediction driven by mechanistic and data model