Urban Pipeline Network Project Practice: Drainage Pipeline Monitoring Application Case

Urban Pipeline Network Project Practice: Drainage Pipeline Monitoring Application Case

In recent years, with the increasing frequency of extreme rainfall events, urban drainage systems have faced continuously rising pressure. Traditional drainage management has relied more on manual patrols, post-event disposal, and experience-based judgment, making it difficult to accurately grasp changes in pipe network water levels, flow rates, and flood-prone spots before rainfall arrives.

A drainage pipe network monitoring project in the old urban area of a certain city was implemented against this background. The project focuses on key indicators including drainage, water level, flow rate, and flood prevention, and deploys monitoring devices on trunk pipes, branch pipes, pump station forebays, underpasses, and low-lying road sections, forming a sensing network covering a catchment area of approximately 136 square kilometers. Through continuous operation and validation across multiple rainfall events, the project has preliminarily achieved the transition from “post-flooding disposal” to “pre-rainfall early warning, in-rainfall dispatching, and post-rainfall assessment.”

Project Background: The Contradiction between Drainage Capacity in Old Urban Areas and Extreme Rainfall

The project area is a typical old urban area, with a built-up area of approximately 136 square kilometers and a permanent population of about 820,000. The drainage pipe network in this area was built over a long period. Early pipelines were mainly combined sewers and direct stormwater discharge; since 2000, stormwater-sewage separation retrofitting has been gradually carried out. However, constrained by road space, underground utilities, building foundations, and the existing drainage system, some areas still have residual combined sewer systems, undersized pipes, and poor outlet conditions.

"Before the project was launched, local waterlogging incidents occurred multiple times. During a heavy rainfall event in 2023, the maximum hourly rainfall reached 62 mm. The waterlogging depth in an underpass exceeded 45 cm, interrupting traffic for about 3 hours."

Post-event analysis found that there was a lack of real-time data on internal water level changes. Pump station dispatching mainly relied on on-site duty personnel and manual judgment. Based on this, the project construction objectives were clearly defined as establishing a drainage pipe network monitoring network for water level, flow rate, rainfall, and waterlogging; forming flood prevention early warning and dispatch support capabilities; and improving the operation and management efficiency of the drainage system.

Technical Challenges: Multi-Source Data Governance

Drainage pipe network monitoring is not simply the installation of a few level gauges or flow meters. Its core difficulty lies in the complex operating state of the pipe network, strong data correlation, and harsh field environments.

Solution: A Sensing, Analysis, and Dispatch System

The overall project scheme is divided into three parts: sensing layer, data layer, and application layer.

Data Layer Strategy: The system uploads data every 5 minutes normally, switching to 1-minute sampling during rainfall or orange-and-above warnings. To reduce false alarms, a sliding time window and trend judgment algorithm is introduced. For example, a higher-level warning is triggered only when water level continuously rises by more than 8% of the pipe diameter within 10 minutes AND upstream rainfall intensity exceeds 10 mm/h.

Application Layer Functions:

• Operational Situational Awareness: Displays pipe network status, identifying high water levels, flow velocities, and abnormal overflows.
• Flood Prevention Early Warning & Dispatch Assistance: Generates pump/gate dispatch suggestions based on forecasts, real-time data, and capacity.
• Post-Event Review: Quantifies warning time, response time, and drainage effectiveness for each rainfall event.

Implementation Process: From Survey to Commissioning

The project launched field surveys in March 2024. After review, the initial design of 158 points was adjusted to a total scale of about 300 devices (126 water level, 92 waterlogging, 38 flow). Equipment installation was concentrated from May to July 2024, often conducted at night with segmented traffic diversion to minimize impact. Approximately 420 confined-space well entries were completed with no safety incidents.

During the trial operation (July 2024 – January 2025), the system collected over 1.32 million water level data points and issued 368 warnings. After multiple rounds of threshold adjustment, the effective alarm rate increased from 64% to 78%, and the false alarm rate decreased from 36% to 22%.

Operational Data and Results

On July 18, 2024, a heavy rainfall event occurred (max hourly rainfall 68 mm). Within 30 minutes, the system detected rapid water level rise in the upstream catchment. A radar gauge showed water level rising from 42% to 78% of pipe diameter. The platform issued an orange warning about 18 minutes before waterlogging occurred.

Another effect is that pump station dispatching shifted from “experience-based” to “data-based.” In August 2024, the system judged a pump station's rising forebay level as downstream outlet backwater (since downstream flow didn't increase) and recommended stopping some pumps. This reduced ineffective operation by ~2.5 hours and saved electricity consumption by ~18%.

In flood prevention management, all 12 key risk spots achieved linked monitoring of water level, waterlogging, and video. As of January 2025, the project had cumulatively identified 17 severely silted pipe sections and 9 pipe sections with abnormal water levels, incorporating them into maintenance plans.

Case Review: Experience and Limitations

While successful, the project exposed common limitations in engineering implementation:

• Equipment Accuracy vs. Data Reliability: Radar gauges can be affected by well walls/edges in narrow spaces. Ultrasonic meters struggle with high sediment/floating objects. Regular patrol, dredging, and calibration mechanisms are essential.
• Threshold Optimization is Continuous: Initial rules relying solely on pipe diameter ratio caused issues. Differentiated thresholds using “water level + flow + rainfall + waterlogging depth” improved effective alarm rates but required ongoing tuning.
• Complex Construction Conditions: Field factors like utility conflicts and traffic permits caused delays and relocations. Fine-grained field surveys regarding well space, slope, and communication coverage are critical to avoid idealized designs.

Outlook: From Monitoring to Resilient Drainage Management

The core value of drainage pipe network monitoring is organizing data into an operable management closed loop. Future governance may deepen along three directions:

Overall, this project demonstrates that multi-source monitoring can significantly improve flood prevention capabilities without immediate large-scale civil works. It provides a practical sample for transforming urban drainage systems from passive emergency response to proactive, resilient governance.