Online leak diagnosis in pipelines using an EKF-based and steady-state mixed approach. (December 2018)

• Record Type: Journal Article
• Authors: Santos-Ruiz, I.; Bermúdez, J.R.; López-Estrada, F.R.; Puig, V.; Torres, L.; Delgado-Aguiñaga, J.A.
• Is Part Of: Control engineering practice. Volume 81(2018)
• Page Start: 55
• Page End: 64
• Publication Date: 2018-12
• Subjects: Pipelines -- Leak diagnosis -- Extended Kalman Filter -- Data fusion -- Steady state -- Real-time; Automatic control -- Periodicals; 629.89
• DOI: 10.1016/j.conengprac.2018.09.006
• Languages: English
• ISSNs: 0967-0661

Abstract

Abstract: This paper proposes a methodology for leak detection and isolation (LDI) in pipelines based on data fusion from two approaches: a steady-state estimation and an Extended Kalman Filter (EKF). The proposed method considers only pressure head and flow rate measurements at the pipeline ends, which contain intrinsic sensor and process noise. The LDI system is tested in real-time by using an USB data acquisition device that is implemented in MATLAB environment. The effectiveness of the method is analyzed by considering: online detection, location as well as quantification of non-concurrent leaks at different positions. The leak estimation error average is less than 1% of the flow rate and less than 3% in the leakage position. Furthermore, the incorporation of a steady-state estimation shows that the solution of the LDI problem has improved significantly with respect to the one that only considers the EKF estimation. An experimental analysis was also performed on the effectiveness of the proposed approach for different sampling rates and for different leakage positions. Highlights: This paper focuses on improving some practical limitations in the real-time implementation of an Extended Kalman Filter (EKF) for leak detection and isolation (LDI) in pipelines. Results of the diagnosis program tests in an experimental pipeline are included. The EKF was tuned according to the uncertainty of the available sensors. The proposed LDI scheme considers the incorporation of the steady-state solution of the dynamic model into EKF outputs to improve the estimates of the leak parameters. The leak-detection system considers different reliability regions according to the sample time.