Background: Why Sewer Flow Monitoring Is Hard
Urban sewer networks are among the most hostile environments for flow measurement. Gravity flow oscillates between partially full and surcharged conditions, flow ranges span more than two orders of magnitude, and the pipe interior combines H₂S corrosion, grease deposits and sediment. Submersible Doppler area-velocity meters have been the workhorse for years, yet utilities routinely report first-year failure or drift rates above 30% due to transducer fouling, burial in sediment and cable degradation. As 24 GHz Doppler surface-velocity radars and 80 GHz FMCW level radars have fallen in price, non-contact radar has become one of the mainstream approaches for sewer flow monitoring and urban-lifeline safety programmes.
Technical Analysis: Surface Velocimetry and Flow Computation
- Principle and specifications. A 24 GHz Doppler radar illuminates the water surface at a 30°–45° incidence angle and derives surface velocity from the frequency shift of the backscattered signal. Typical range is 0.1–15 m/s with an accuracy of about ±1% of reading ±0.01 m/s. The sensor mounts at the pipe crown or inside the manhole cover and never contacts the medium — immunity to corrosion and fouling comes by design.
- Flow computation. Discharge follows the area-velocity method: Q = V × A. The wetted area A is computed from the stage measured by an 80 GHz FMCW level radar (±3 mm class) combined with the pipe geometry (circular sections DN300–DN2000). Surface velocity is converted to cross-section mean velocity through a coefficient k, typically 0.85–0.92 in circular pipes, which must be calibrated in the field.
- Dominant error sources. (1) Uncertainty in the surface-to-mean velocity ratio is the largest error term; calibration against ADCP or ultrasonic transit-time measurements is recommended, consistent with the velocity-area framework of ISO 748. (2) In the low-flow dead zone below roughly 5 cm depth, bed roughness disturbs the surface flow and radar readings should be flagged as degraded. (3) Under surcharged conditions surface velocimetry is invalid — level rate-of-change and pressure data must be used to flag the regime. (4) Foam and floating debris generate false echoes; echo-quality filtering is mandatory.
- Route comparison. Submersible Doppler offers better low-flow accuracy at the cost of maintenance; in-line electromagnetic meters (ISO 4064 / OIML R49) are accurate but require full-bore flow and significant civil works. Radar surface velocimetry is maintenance-free but depends on section calibration. In practice the three routes are combined per reach conditions.
Implementation: Installation, Calibration and Telemetry
- Siting. Choose a straight reach of at least 5 pipe diameters with uniform flow, away from drops, sharp bends and backwater effects. Keep the beam incidence at 30°–45°, and ensure the distance from antenna to maximum water level sits within the level-meter range with at least 20% headroom.
- Commissioning calibration. Perform field comparison at no fewer than three flow stages (low, medium, high) to build a level-velocity (H-V) rating, and cross-check against Manning's equation Q = (1/n)·A·R^(2/3)·S^(1/2). If the two diverge by more than 15%, re-survey the cross-section geometry and roughness n.
- Telemetry. An IoT RTU acquires discharge, stage and surface velocity, reporting every 15 minutes normally, intensifying to 5 minutes during rainfall, with event-triggered reports when the level gradient exceeds a threshold. Power comes from battery plus solar panel; transport uses 4G/NB-IoT. Enclosures are IP68 per IEC 60529 and electronics qualified to IEC 60068 for pit-mounted service.
- Maintenance. Inspect antenna window contamination, bracket corrosion and manhole-cover sealing quarterly; verify anti-theft fixings before the flood season; re-survey the pipe section and re-check the k coefficient annually after the wet season.
Conclusion
Non-contact radar flow measurement trades one-time calibration effort for long-term, maintenance-free operation. For urban drainage networks, a radar surface-velocity plus radar-level area-velocity scheme, backed by field calibration and echo-quality control, sustains above 95% data availability — a dependable data foundation for CSO pollution control, flood dispatching and urban lifeline monitoring.
