Application of Temperature Sensors in Centrifugal Pumps

Why Centrifugal Pumps need Temperature Sensor Monitoring

The core issue is that centrifugal pumps failures cause above 35% of unplanned downtime in petrochemicals; bearing overheating accounts for 21% of faults.

The value of temperature sensors monitoring is that 72 hours early warning avoids unplanned downtime; each critical centrifugal pump outage costs hundreds of thousands to millions dollars per day.

There are four key monitoring points which are bearing temperature, seal temperature, casing temperature and motor winding temperature.

Thermal Failure Mechanisms of Centrifugal Pumps

1.Bearing Thermal Failure

The bearing is the highest priority measurement point for temperature monitoring in centrifugal pumps, as bearing failures account for 21% of pump failures, and temperature sensors will monitor the temperature parameter in the early stages of failure.

2.Mechanical Seal Thermal Failure

The temperature sensor will monitor the failure temperature threshold of mechanical seals which is crucial and more subtle.

High temperature pumps should be equipped with Pt100 temperature sensors in the sealing flushing liquid circuit to monitor the temperature difference between the inlet and outlet of the flushing liquid. An increase in temperature difference indicates that the sealing end face is heating up more intensely, which is an early signal that the seal is about to fail. Double-end face sealing systems should be equipped with sealing chamber temperature sensors.

3 Thermal Expansion and Minimum Flow

The temperature sensor monitoring can serve as a redundant criterion for small flow protection, when the outlet temperature is more than 15°C higher than the inlet temperature, forced load reduction or shutdown should be implemented.

Summary

The temperature sensor is the earliest measurable signal of bearing, seal and rotor failure.

The value of temperature sensor monitoring lies not in "knowing the current temperature" but in "predicting what it will become"，from passive alarm to proactive prediction, this is the technical foundation of the shift from scheduled to predictive maintenance.