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

Introduction: The "Nerve Endings" of Urban Flood Prevention

In recent years, the frequency of extreme heavy rainfall events has placed immense pressure on urban drainage systems. Traditional management approaches—relying on manual inspections, experience-based judgment, and post-disaster emergency responses—are no longer adequate for high-density urban areas. As the core channel for stormwater discharge, the ability to perceive the operational status of the drainage network, warn of anomalies in advance, and coordinate dispatch is critical.

Based on industry research by Yicai Technology, this article reviews the implementation of an online monitoring and flood prevention warning project in the eastern district of a city in the Yangtze River Delta. This case demonstrates how data on liquid level, flow rate, rainfall, and ponding supports refined management and decision-making.

During the 2024 flood season, the system successfully handled 12 heavy rainfall events (max intensity 68mm/hr). Key results included:

• Significant ponding points decreased from 19 to 7.
• Average water recession time shortened from 58 to 27 minutes.
• Emergency dispatch response time reduced from 22 to 8 minutes.

Project Background: Pressures on Old Urban Systems

The project area is a typical old urban district with complex underground conditions. Variations in pipe materials, diameters, and burial depths create intricate drainage paths. With a road hardening rate of ~78%, runoff forms quickly, leading to ponding in low-lying areas when rainfall exceeds discharge capacity.

Prior to the project, management relied on manual patrols and hotline feedback. During the 2022 flood season, there were 19 significant ponding points, causing traffic disruptions of up to 3.5 hours and generating 127 emergency work orders. The average time to manually detect anomalies was about 40 minutes, making proactive dispatch nearly impossible.

"Sedimentation in older pipes accounted for 15% to 25% of the pipe diameter, significantly reducing flow capacity and causing premature high liquid level states."

Technical Challenges: Complex Underground Environments

Drainage network monitoring requires overcoming five major technical hurdles:

Solution: A Five-Layer System Architecture

The project adopts a comprehensive "Perception-Transmission-Platform-Model-Dispatch" approach.

1. Perception Layer

• Liquid Level: Radar/Ultrasonic gauges (IP68, ±0.5% accuracy). Radar preferred for foam/vapor resistance.
• Flow: Non-full-pipe scheme combining level, velocity, and cross-section (DN300-DN2200).
• Ponding: Electronic staff gauges/radar meters at underpasses and low-lying intersections.

2. Transmission & Platform

Data collection intervals adapt dynamically: 5 mins (dry) → 1 min (rain) → 30 secs (critical risk). The platform integrates GIS, video, and maintenance data, processing ~186,000 entries daily.

3. Model & Dispatch

A hydraulic model predicts liquid level trends 15-30 minutes in advance. Warnings are categorized into Blue, Yellow, Orange, and Red levels, automatically pushing alerts to pump operators and maintenance teams via SMS/App.

Implementation: From Survey to Operation

The project followed a rigorous five-step process:

1. Survey (Nov-Dec 2023): Identified key nodes based on CCTV, historical data, and temporary flow monitoring.
2. Installation (Jan-Feb 2024): Night operations to minimize traffic impact. Average install time: 45 mins/point.
3. Debugging (Mar 2024): Integrated data streams and tested alarm triggers. Resolved 38 initial anomalies.
4. Trial Run (Apr-May 2024): Validated system performance against spring rains. Adjusted filtering parameters for debris interference.
5. Acceptance (Jun 2024): Equipment online rate reached 99.2%. Alarm push time <30 seconds.

Typical Scenario: Heavy Rainfall Response

On July 18, 2024, the district faced its heaviest rain (68mm/hr max intensity). The system demonstrated its value through early intervention:

By detecting upstream trends 18 minutes before obvious surface ponding, the system allowed for advance pump dispatch and traffic control. Compared to a similar 2022 event (90-min disruption), this event saw only minor, short-lived ponding.

Results: Efficiency Gains and Data-Driven Management

After the 2024 flood season, the project delivered measurable improvements:

Maintenance efficiency also improved. Data-driven analysis identified 23 sections with sustained high levels and 18 suspected blockages. Targeted cleaning recovered flow capacity by ~12% and reduced unnecessary patrols by 32%.

Conclusion: From Monitoring to Smart Prevention

This case study highlights that successful drainage monitoring relies on four pillars:

1. Risk-Oriented Layout: Focus on historical ponding points and bottleneck sections, not just even distribution.
2. Equipment Reliability: Sensors must withstand harsh underground environments (foam, vapor, debris).
3. Data Governance: Raw data must be cleaned and correlated to avoid "alarm fatigue."
4. Dispatch Integration: Data must trigger actionable workflows (pump control, traffic closure).

Looking forward, the integration of digital twins and AI will shift urban flood prevention from "response during rain" to "prevention before rain," making drainage networks a critical component of smart city infrastructure.