LEUNG S S, Machine‑learning‑based fresh water service reservoir outflow data logging system and virtual outlet flowmeter, HKIE Transactions, Vol. 33, No. 1 (Regular Issue), Article THIE-2025-0030.R1, 2026, 10.33430/V33N1THIE‑2025‑0030
This paper presents the implementation of a Virtual Outlet Flowmeter (VOF) and a Service Reservoir Outflow Data Logging System (SRODLS) in a Fresh Water Service Reservoir (FWSR). A VOF offers a cost‑effective solution for estimating outlet flow rates without physical meters. SRODLS improves the data quality by providing continuous logging and remote access to critical data, facilitating efficient fault detection and troubleshooting. The paper also explores machine learning anomaly detection systems to identify data inconsistencies and potential issues in water flow data. By leveraging historical data and advanced algorithms, the system can detect unusual patterns that may indicate problems such as leaks or infrastructure failures. The application of machine learning in forecasting and anomaly detection is seen as a significant step forward in water resource management. The results show that the implementation of VOFs and SRODLS in pilot FWSRs has been successful, contributing to better planning and management of water resources with minimal operational impact. The paper concludes with the potential for future developments, including the analysis of minimum night flow and the creation of APIs for hydraulic modelling, to further enhance water loss detection and resource management.
Keywords:
water supply; data analytics; virtual metering; machine learning; water resources, control automation and instrumentation
Reference List:
[1] Bikmukhametov T and Jäschke J (2020). First Principles and Machine Learning Virtual Flow Metering꞉ A Literature Review, Journal of Petroleum Science and Engineering,
[2] Taylor S.J. and Letham B. (2017). Forecasting at Scale, PeerJ Preprints.
[3] Sulaiman R.B., G. Hariprasath & P. Dhinakaran & Köse.U (2023). Time‑series Forecasting of Web Traffic Using Prophet Machine Learning Model. 1. 161‑177.
[4] Loizou, K. and Koutroulis, E. (2016). Water level sensing꞉ State of the art review and performance evaluation of a low‑cost measurement system Volume 89, 2016, Pages 204‑214,
[5] Baker, R.C. (2016). Flow Measurement Handbook꞉ Industrial Designs, Operating Principles, Performance, and Applications. 2nd ed. Cambridge University Press.
[6] Shercliff, J.A. (2022). Electromagnetic Flow Measurement꞉ Theory and Practice. 3rd ed. Cambridge University Press.
[7] Wu JP, Xu KJ, Xu W, Yu XL. (2018) Transient process based electromagnetic flow measurement methods and implementation. Rev Sci Instrum;89
[8] B. Sun, S. Chen, Q. Liu et al. (2020), Review of sewage flow measuring instruments, Ain Shams Engineering Journal
[9] G. Fowles, W.H. Boyes (2010), Chapter 6 ‑ Measurement of Flow, Instrumentation Reference Book (Fourth Edition), Butterworth‑Heinemann, Pages 31‑68
[10] Shenoy, V and Krishnan Venkata, S. (2025), An In‑Depth Analysis of Liquid Level Measurement Techniques and Performance Evaluation Using Computational Fluid Dynamics, Journal of Sensors, 2025, 4412250, 34 pages
[11] Nikolai A.(2018), A Machine Learning Approach for Virtual Flow Metering and Forecasting, IFACPapersOnLine,
Volume 51, Issue 8, Pages 191‑196
[12] Udhbhav N., et al (2024), Comparative Analysis for Water Demand Forecasting using LSTM, Prophet,
ARIMA & SARIMA, Proceedings of the 3rd International Conference on Optimization Techniques in the Field of Engineering (ICOFE‑2024)
