How a Water Utility Group Achieves Cost Reduction and Efficiency Improvement Through Drainage Pipe Network Monitoring

I. Project Background: Compounded Capacity Constraints and Waterlogging Risks of Old Town Drainage Networks

The project was implemented in the central urban area of a certain city, which has a large built-up area and underground drainage networks constructed over a long period. Within the project scope, the total length of drainage pipes is approximately 1,240 km, including approximately 680 km of stormwater pipes and channels, 360 km of combined sewers, and 200 km of sewage pipes. Some pipe sections in the old town have relatively small diameters, and a gap has emerged between design standards and recent heavy rainfall events. Particularly in combined sewer areas, rainwater, sewage, and surface runoff enter the network simultaneously during rainfall, easily causing rapid local water level rises and downstream outfall backwater effects.

Traditional drainage management relied mainly on manual patrols and post-event emergency response, making it difficult to obtain real-time operational data such as water levels and flows.

According to flood-season statistics from 2021 to 2023, the project area historically had 17 waterlogging-prone points. During a single heavy rainfall event, the maximum road surface water depth once reached 68 cm, traffic disruption on some road sections exceeded 2 hours, and drainage complaints from businesses and residents along affected routes were concentrated. Therefore, in drainage network renovation and flood-prevention capacity enhancement projects, local authorities promoted online water level and flow monitoring as a foundational preliminary engineering task.

II. Technical Challenges: Synchronizing Accurate Water Level, Flow, and Flood-Prevention Thresholds under Complex Operating Conditions

The primary challenge in drainage network monitoring lies in the complexity of operating conditions for water level measurement. The project area includes buried pipes, open channels, box culverts, and manholes, with some pipe sections having deposits of sludge, floating debris, grease, and trash. Traditional submersible level transmitters are prone to reading drift when covered by sludge, installed in narrow shafts, or exposed to large water surface fluctuations. The project requires water level monitoring to cover a range of 0 to 10 m, with key nodes achieving an accuracy of ±0.25% FS and a resolution of 1 mm, and stable operation under humid, high-temperature, and short-term immersion conditions.

Flow monitoring challenges are more pronounced. Flow in drainage networks is not always full-pipe. At the beginning of rainfall, only low-velocity sewage or rainwater may be present, while during storm peaks, flow may approach full-pipe conditions or even overflow. The project involves pipe sections with diameters ranging from DN300 to DN2000, some of which are combined sewers with a wide velocity variation range. Partially full pipe electromagnetic flowmeters are required to cover a velocity range of 0.01 to 5 m/s with accuracy controlled between ±2% and ±3%.

The challenge in flood prevention warning lies in the coupled relationships among water level, flow, rainfall, river water level, and pump station operation. Relying solely on a single water level threshold for alarms is prone to false alarms or missed alarms. For example, when downstream river water levels rise, discharge outlets are affected by backwater, and pipe section water levels increase, but actual waterlogging risk does not necessarily rise synchronously. The project team needed to establish multi-source data associations in drainage monitoring to avoid situations such as 'device alarm without actual site risk' or 'site waterlogging without timely system response'.

III. Solution: Building an Integrated Water Level and Flow Monitoring and Flood Prevention Warning System

In device selection, the project specified clear parameters. Submersible level transmitters use stainless steel encapsulation, IP68 protection, temperature compensation, and anti-clogging design. Radar level transmitters are used at open channels with significant water surface fluctuations, offering ±3 mm accuracy. Partially full pipe electromagnetic flowmeters are configured with multiple measuring electrodes suitable for DN300 to DN1200 pipe sections. Rain gauges have a resolution of 0.2 mm, and water depth monitors can identify waterlogging risk above 15 cm.

For warning models, the project established four categories of thresholds: water level, flow, rainfall, and standing water. A yellow warning is triggered when water levels reach 70% of pipe diameter, orange at 85%, and red when full-pipe or overflow risk rises significantly. Standing water thresholds are set at 15 cm (yellow), 30 cm (orange), and 50 cm (red). The platform combines multi-source data trends to calculate waterlogging risk levels, providing lead time for flood prevention dispatching.

IV. Implementation Process: From Site Layout and Construction Renovation to Joint Commissioning and Acceptance

The project implementation began in April 2023, trial operation was completed in January 2024, and flood-season acceptance was passed in June 2024, with a total duration of approximately 10 months. During the preliminary site layout phase, the project team combined drainage network GIS data, historical waterlogging records, and pump station operation logs to select monitoring sites. Principles included placing water level and rainfall monitoring at catchment unit boundaries, flow monitoring at bottleneck sections, and joint monitoring at pump station forebays and outlets.

The construction phase adopted a combination of trenchless and local excavation methods. For level transmitter installation inside manholes, in-well fixing brackets and anti-drop chains were used. Dredging and shaft tidying were carried out simultaneously, completing dredging at approximately 340 key nodes to ensure basic conditions for accurate measurement.

During joint commissioning, the project team calibrated 260 sets of water level devices and 70 sets of flow devices point by point using standard staff gauges, pressure calibrators, and portable ultrasonic flowmeters. Points with errors exceeding ±2 mm were reinstalled or corrected. During trial operation, the project connected more than 450,000 data points, completed three rounds of storm simulation, and one round of practical dispatching drills.

V. Operational Data: Water Level and Flow Acquisition Quality and Flood Prevention Warning Indicators

Within 12 months after formal system operation, the validity rate of water level and flow monitoring data reached 99.2%. Water level data anomalies mainly resulted from device obstruction and sedimentation, addressed through inspections and threshold corrections. For flow data, partially full pipe electromagnetic flowmeters showed slightly higher errors under low-velocity conditions; the project controlled cumulative errors within ±3.5% by adding water level linkage judgment and filtering algorithms.

Regarding warning indicators, the platform established three-level flood prevention warnings (yellow, orange, red). After the system went online, warning accuracy improved from 76% to 88%, and the false alarm rate decreased by 42%, mainly due to combined judgment using multi-source data. In terms of drainage dispatching effectiveness, the average advance time for pump station startup increased by about 12 minutes compared to traditional manual judgment, reducing overflow risks caused by rapid water level rises.

VI. Case Study: Water Level and Flow Monitoring and Flood Prevention Dispatching during a Heavy Rainfall Event

On July 11, 2024, a heavy rainfall event occurred with cumulative rainfall reaching 102 mm in three hours. Platform water level monitoring showed that the water level in the old town combined sewer trunk pipe rose from 1.2 m to 2.1 m within 30 minutes. Simultaneously, pump station forebay flow monitoring showed instantaneous inflow reaching 8,600 m³/h, more than six times the normal dry-weather flow. Combining rainfall, water level, flow, and downstream river water level, the platform judged an orange flood prevention risk and issued a warning about 26 minutes before standing water reached 30 cm.

After the warning, the drainage dispatching center activated response, recommending the opening of 8 pump units across 3 stations and closing 2 gates susceptible to backwater. On-site water depth monitoring showed peak standing water in the underpass was 31 cm, below the red threshold. The maximum road surface water depth was 29 cm, about 39 cm less than a comparable rainfall event in 2023. The recession time at waterlogging points was about 1 hour and 20 minutes, approximately 45% shorter than the historical average.

Project acceptance statistics show that waterlogging-prone points decreased from 17 to 4, the maximum single-event standing water depth decreased from 68 cm to 29 cm, and events with traffic disruption exceeding 2 hours dropped to zero.

VII. Conclusion and Outlook: Extending Drainage Network Monitoring toward Resilient Governance

This engineering case demonstrates that drainage network water level and flow monitoring is not simply device deployment, but a comprehensive governance project involving drainage system structure, rainfall patterns, pipe network hydraulic conditions, and flood prevention dispatching processes. Through 260 sets of water level devices, 70 sets of flow devices, 12 sets of rainfall devices, and 36 sets of standing water devices, the project preliminarily achieved online monitoring of key nodes in a 1,240 km drainage network.

As a next step, local authorities plan to advance Phase II of drainage network monitoring, focusing on supplementing river water level, pump station operation, water quality monitoring, and pipe network structural defect detection data. The platform will combine digital twins and pipe network hydraulic models to further improve prediction capabilities, extending warning lead time from the current 18–35 minutes to more than 60 minutes. By sharing drainage monitoring data with transportation, meteorological, and emergency systems, urban flood prevention will shift from passive emergency response to proactive dispatching and precise handling, developing toward resilience, refinement, and intelligence.