Submerged sewage pumps
Catch ragging early. Avoid pollution incidents. Reduce out-of-hours call-outs.
SAM4 reads the motor at the cabinet, so the pump stays in the well.
Submerged sewage pumps fail underwater, in places direct sensors can't reach economically. SAM4 reads electrical and load signatures from the motor control cabinet and flags ragging, blockage, air lock, phase issues, and sustained load-pattern anomalies before the pump trips. Move sewage pump maintenance from reactive call-out to planned intervention.
Pumping stations are now a board issue
The Water (Special Measures) Act 2025 has made pollution a governance issue. Submerged pumping stations sit at the centre of that exposure: critical, distributed, and largely unmonitored.
The blockage problem
Rags, wet wipes, and debris cause the majority of sewage pump failures. Ragging incidents have risen sharply as disposable wipe use has grown. Progressive rag wrapping goes undetected by telemetry until the motor stalls or trips.
The telemetry lag
SCADA and telemetry do their job: levels, flow, run status, alarms. They catch faults at the moment of failure, when the pump trips or the wet well overflows. Slow rag build, developing cavitation, and gradual bearing wear give no telemetry signal until they cause the event. Telemetry is a lagging indicator. With personal director liability and multi-million pound pollution fines now on the table, lagging is not enough.
The access barrier
Submerged sewage pumps sit 3-10 metres below ground in sealed wet wells. Mounting and maintaining vibration sensors there is impractical at fleet scale. The only way to assess condition is to pull the pump: a confined-space operation costing £5,000 to £15,000 per event.
“Critical in our monitoring of hard to reach assets such as submersible pumps.”
Reading the pump from above ground
The motor cable already carries the diagnostic signal. SAM4 reads at the cabinet, classifies the fault, and routes it through expert review before it reaches you.
Many mechanical, electrical, and hydraulic changes on a submerged sewage pump create measurable changes in current, voltage, or load pattern. Ragging shifts the load profile. A failing impeller modulates the torque. Phase imbalance distorts the current waveform. Air lock changes the operating point.
These changes propagate up the motor cable to the motor control cabinet. SAM4 reads the three current and three voltage signals at the cabinet, sampled at the rate needed to resolve the fault frequencies of interest. No sensor enters the wet well. No personnel access is needed. The pump runs as it always has.
For submerged sewage pumps, this is the only practical path. The asset is sealed. The wet well is a confined space. Mounting and maintaining vibration sensors at scale is uneconomic, and on the worst-affected pumps it isn't possible at all. The cable is already there.
Reading the cabinet. A clogging event detected on a wet-well pump. Load-pattern anomaly highlighted, fault classified with evidence level, action recommended.
Signal flagged
Expert review
Fault classified
Action recommended
What SAM4 detects on this asset, and where it doesn't fit
One table. Each fault class appears once with its signal path, the strength of field evidence on this asset class, and the recommended use of SAM4. The signal path is defined by physics. The recommendation reflects what we suggest you act on.
| Fault class | Signal path | Field evidence on this asset | Use SAM4 as |
|---|---|---|---|
| Clogging and ragging | Indirect / load. Debris build-up shifts the pump load before trip. | 1,213 scored events. 98.5% recall, 2.0% false-alert share. | Primary monitoring |
| Phase loss and voltage imbalance | Direct / electrical. Resolved at the cabinet from current and voltage symmetry. | 30+ scored events. No misses observed in the reviewed sample. | Primary monitoring |
| Sustained load-pattern anomalies | Indirect / load. Process-driven changes resolve through the motor current. | Detected consistently across the cohort. | Primary monitoring |
| Mechanical unbalance | Load signature + 1x running speed. Reaches motor current through the rotor. | Small sample reviewed. No misses observed. | Primary monitoring |
| Repeated abnormal duty cycles | Trend / infer. Pattern across operating cycles, not a single event. | Established on this asset class. | Primary monitoring |
| Air lock | Load step-change. Operating point change under variable duty. | Cases reviewed. Pattern detected consistently. | Conditional |
| Impeller degradation or load shift | Long-window 1x harmonic trend. Detectable where the load actually shifts. | Cases reviewed. Cohort still small. | Conditional |
| Cavitation-like operating patterns | Load signature + current variance. Pattern visible. Severity not graded. | Cases reviewed. Useful as a flag, not a measurement. | Conditional |
| Bearing degradation | Indirect / late. Visible once degradation reaches the motor current. Attenuated through the submerged assembly. | Stable runtime helps; intermittent duty thins the signal. Vibration on accessible critical pumps remains the better tool for raceway-level diagnosis. | Late-stage detection |
| Seal-related anomalies | Indirect / late. No reliable pre-leak signature. Detectable once degradation affects load or current. | Seal anomalies produce no torque modulation, so ESA cannot see the leak directly. SAM4 catches the root causes (misalignment, unbalance, cavitation) and cascaded effects. | Late-stage detection |
| Structural defects | Outside envelope. No reliable electrical or load expression. | Not claimed. | Use other methods |
| Cavitation severity grading | Outside envelope. Pattern visible. Severity calibration is out of scope. | Use hydraulic instrumentation if grading is required. | Use other methods |
Where ESA fits in your monitoring architecture
On most submerged sewage pump fleets, SCADA covers process telemetry but no system monitors asset health inside the wet well. ESA fills that gap. It runs alongside what you already have, not in place of it.
Process layer
SCADA on flow, head, level, and kWh. Pollution sensors downstream. Process control on the VFDs. These tell you when the system stops performing. They don't tell you which pump is degrading, or why.
Asset-health layer
Continuous monitoring on every pump, read from the motor control cabinet. Each detection passes through expert review and arrives with an evidence tier showing the strength of the supporting field evidence. No wet well access. No extra sensors on the asset.
Specialised layer
Pre-leak seal moisture, very early-stage bearing pitting, water chemistry and process-quality monitoring. These sit outside what the electrical signature can resolve. Use seal-leak sensors for moisture. Use vibration spot-checks if your fleet warrants it.
ESA is not a replacement. It covers the continuous monitoring gap on submerged assets that vibration cannot reach at fleet scale, and reads the same signal that drives the pump.
Pump performance curves from electrical data
SAM4 calculates instantaneous head and flow using affinity laws and the pump's reference curve. The result is a real-time performance curve showing where each sewage pump operates relative to its best efficiency point. A pump drifting left of BEP signals cavitation risk. A curve shifting downward indicates impeller fouling, ragging, or wear ring degradation. The same view used to detect faults reveals where energy is being wasted.
Continuous BEP tracking
Every pump's operating point is tracked against its design curve. Deviation from BEP triggers alerts for cavitation risk, oversized duty, or fouling that shifts the curve. No manual test required, no flow meter needed.
Energy efficiency baseline
SAM4 benchmarks each pump's energy consumption against its theoretical optimum. The gap between actual and optimal consumption is your recoverable energy waste. Quantified per pump, per day. Especially relevant on stations where motors are sized for peak inflow but run far below design most of the time.
One sensor. Two value streams. The same electrical signal that detects faults reveals where pumps waste energy.
Explore Energy Optimisation →Real detections on submerged and submersible pumps
Two real events from water utility deployments. One prevented a pollution incident. One identified recoverable energy waste. Both follow the same evidence chain: signal flagged, fault hypothesised, action recommended, customer inspection, outcome confirmed.
How Yorkshire Water saved £390k in potential fines
Yorkshire Water is committed to building robust and resilient clean and wastewater networks for the future in the run up to 2050 and beyond. This is the basis
Southern Water’s success story: preventing three failures, saving £748K, and ensuring operational resilience
Southern Water is committed to reducing pollution incidents and improving infrastructure resilience. As part of this effort, SAM4 was deployed across 637
Improving submersible pump efficiency at Sabesp
Founded in 1973, Sabesp (Companhia de Saneamento Básico do Estado de São Paulo) is one of the largest sanitation companies in the world. It provides water and
“SAM4 delivers clear and actionable insights that enable our teams to make swift, informed decisions.”
Under 60 minutes. No wet well entry required
1. Open the motor control cabinet
SAM4 installs at the MCC, the same panel your electricians already access. No wet well entry. No confined space permit. No interruption to the wet-well environment.
2. Clip sensors onto motor supply cables
Current and voltage sensors clip directly onto existing motor cabling. Installation requires a brief motor de-energisation while sensors are fitted, typically scheduled with operations. No wiring changes.
3. Connect and commission
The SAM4 gateway connects via cellular (4G/LTE). No dependency on your IT network. Monitoring starts immediately, with first diagnostic results within 48 hours.
Methodology and validation detail
Reviewed evidence from the 12 months ending 2026-05-01. Exact counts and exclusions are in the validation report.
Review window
Window: 12 months ending 2026-05-01.
Every alert in the window is reviewed by a Samotics analyst and scored as: fault detected, fault missed, false communication, inconclusive, or not applicable. Inconclusive and not-applicable cases are excluded from both metrics.
Recall and false-alert share
Of all confirmed fault events in the review window, how many did SAM4 detect before the event was resolved or failed?
Recall on confirmed fault events = TP / (TP + FN) False-alert share after review = FP / (TP + FP) TP = confirmed fault events SAM4 flagged before resolution FN = confirmed fault events SAM4 missed in the window FP = reviewed alerts rejected as false after expert review
Move sewage pump maintenance from reactive to scheduled.
Bring your asset list, drive configuration, and dominant failure modes. We will tell you what fits.