Steam Trap Monitoring
Steam Trap Monitoring (STM) gives you continuous, automated visibility into the health of every steam trap in your facility. Batteryless Eversensors attach directly to each trap, measure condensate and steam temperatures every ~60 seconds, and send that data wirelessly through an on-site gateway to the cloud. The platform runs analytics against those readings — using a combination of ML-derived thresholds and physics-based algorithms — to determine each trap's condition and write the result back where your team can act on it.
No manual inspection schedule is required to maintain a live health picture of your trap population.
How Detection Works
Each reading follows a fixed path from sensor to state:
- Dual temperature probes on the Eversensor measure condensate temperature and steam temperature independently, once per ~60 seconds.
- Eversensor transmits readings wirelessly to the nearest Evergateway on site.
- Gateway forwards readings over MQTT to the cloud ingestion pipeline.
- Analytics pipeline runs steam state analysis and failure detection algorithms against the new data.
- State is written back to the trap record via the Public API.
- Notifications and webhooks fire if the trap's condition changed.
The full round-trip from a new reading to an updated condition is typically under a few minutes.
What the Platform Detects
| Condition | What It Signals |
|---|---|
| Good | Trap is operating normally |
| Blowthrough | Trap has failed open; live steam is passing continuously |
| Intermittent Blowthrough | Trap is intermittently passing steam (Light, Medium, or Heavy severity) |
| Leaking By | Trap is not fully closing; partial steam loss when the trap should be sealed |
| Unexpected Cold Condition | Steam stopped during a period when the trap should be operational |
| Failed Cold | Trap has been confirmed failed in the closed position |
State changes — not individual readings — trigger notifications and webhooks. A trap already in Blowthrough does not generate a duplicate alert on the next reading.
Supported Trap Operation Types
The operation type you assign to each trap controls which analytics apply to it:
| Operation Type | Description |
|---|---|
| Continuous | Trap is expected to pass steam at all times; unexpected cold detection is always active |
| Seasonal | Trap is operational only during a configured date window; unexpected cold detection is active within the window only |
| Intermittent | Trap cycles on and off by design; unexpected cold detection is suppressed |
See Steam Trap Configuration for how to set the operation type and other trap metadata.
Steam State Detection
The platform tracks whether steam is flowing through each trap at any given time. Two detection techniques run in parallel and their confidence scores are combined to produce a final on/off determination:
- Threshold method: Steam On at 90°C, Steam Off at 60°C (both configurable), with hysteresis to avoid rapid toggling.
- Floating min/max method: Tracks a temperature envelope over time to detect on/off transitions without relying on fixed thresholds.
The result of this analysis feeds directly into unexpected cold detection and other failure-state algorithms.
Data Retention
| Data | Retention |
|---|---|
| Raw sensor readings | ~1.5 years (548 days) |
| Sensor aggregates | ~6 months (180 days) |
| Events and trap states | Indefinite |
See Also
- Sensor Installation & Field Troubleshooting — mounting an Eversensor and diagnosing hardware issues
- Trap States — complete reference for all trap condition values and recommended actions
- Unexpected Cold Events — how the platform detects and suppresses unplanned steam loss
- Seasonal Configuration — configuring date windows for seasonal traps
- Cost of Steam Analytics — financial impact of failing traps
- Shoplogix IMS Dashboard — where to view trap condition summaries