Background: Why Sewer Networks Are So Hard to Instrument
Sewer networks are a critical part of urban lifeline infrastructure, yet they remain one of the least instrumented assets in most cities. Typical monitoring density is below one station per 10 km of sewer — far short of what flood dispatching and CSO (combined sewer overflow) control require. The constraints are structural: manholes are humid, corrosive (H₂S), and attenuate radio signals, while conventional monitoring solutions that require excavation and grid power simply do not scale economically.
The Four Engineering Challenges
- Power. No grid power exists inside a manhole. Node lifetime is dictated by battery capacity and reporting duty cycle, so sleep current must be kept in the tens of microamps and reporting intervals must be adaptive.
- Communications. RF penetration loss through manhole covers commonly reaches 20–30 dB. A reliable uplink demands either an external antenna or link-budget-friendly LPWAN/cellular technologies such as NB-IoT or LTE Cat.1.
- Sensing. Ultrasonic level sensors drift under condensation and foam; submerged pressure sensors suffer zero-point drift once buried in silt. 80 GHz radar level sensors are immune to humidity and corrosive gas, with typical accuracy of ±3 mm, and have become the mainstream choice for manhole level monitoring. For sewer flow monitoring, the area-velocity method — Doppler or non-contact radar surface velocimetry — covers non-full pipe conditions.
- O&M. Any manhole entry is a confined-space operation with significant cost. Equipment should be engineered for five-year maintenance-free service, with ingress protection of at least IP68 per IEC 60529.
A Tiered, Low-Cost Sensing Architecture
A proven engineering approach is to classify monitoring points into three tiers and equip each tier differently, instead of applying one high-spec design everywhere:
- Tier 1 – critical nodes (outfalls, siphons, pump-station inlets, flood-prone points): 80 GHz radar level sensor + Doppler velocity sensor + IoT RTU, reporting every 5–15 minutes, delivering full level–flow datasets for CSO accounting and flood dispatch.
- Tier 2 – trunk sewers: radar level sensor + low-power RTU, reporting every 30–60 minutes with event-triggered uplift during storms, used to assess pipe surcharging and locate bottleneck reaches.
- Tier 3 – branch sewers and CSO weirs: low-cost level-only nodes over NB-IoT/LoRaWAN sending threshold alarms plus periodic heartbeats, cutting per-point cost to one third to one fifth of Tier 1 and enabling grid-like coverage density.
Selection guidance: use radar sensors with a beam angle of 4° or less to suppress manhole-wall echoes; require IP68 minimum; verify EMC immunity to EN 61326-1 for industrial environments; and specify 3–5 years of battery life at 15-minute reporting intervals. The RTU should support store-and-forward and edge-side data quality checks so that transient foam or wave spikes never reach the operations platform as false alarms.
Conclusion
In urban drainage monitoring, coverage and data continuity matter more than peak single-point accuracy. A tiered architecture built on low-cost sensing and battery-powered IoT RTU telemetry delivers high-density sewer coverage within a controlled budget, creating the data foundation for flood dispatching, CSO control and network health assessment. Start at critical nodes, validate data quality and O&M workflows, then expand along trunk sewers in phases.
